rust/tantivy
1
0
Fork 0
mirror of https://github.com/quickwit-oss/tantivy.git synced 2025-12-28 13:02:55 +00:00
Code Issues Packages Projects Releases Wiki Activity

Compare commits

base: rust:issue/922
Branches Tags
rust:main rust:stuhood.upstream-erased-sort-computer rust:trinity.pointard/ff-query rust:bucket_id_agg rust:paradedb/fix-agg-validation rust:flatheadmill-geo rust:paradedb-paradedb/fix-overflow-issue rust:low_card_optimization rust:paul.masurel/clippy rust:fix_size_hint rust:qw-airmail-20250916 rust:unit-test-aggregation-sum-other rust:clippy-and-cleanups rust:trinity.pointard/opti-sstable-sorted-ords-to-term rust:release_0.24 rust:dependabot/cargo/rand-0.9.1 rust:dependabot/cargo/lru-0.14.0 rust:dependabot/cargo/fs4-0.13.1 rust:tagged-user-input-ast rust:1686a/parse-agg-res rust:dependabot/cargo/rand_distr-0.5.1 rust:issue/2411-adding-panic-handler-to-rayon-merge-thread-pool rust:dependabot/cargo/nom-8 rust:upgrade_deps rust:test_parser rust:dependabot/github_actions/codecov/codecov-action-5 rust:binggan-0.12.0 rust:agg_format rust:path_collector rust:test_order_bug rust:test_columnar_multi rust:columnar_optionalmask_in_multivalues rust:prefix-phrase-query-optim rust:tantivy-with-oneshot rust:update_example rust:store-reader-enumerate rust:document-deserialize-seed rust:check_doc_mem rust:experimental-do-not-merge rust:expdonotmerge rust:cherry-pick-vec-instead-of-btreemap rust:validate_length rust:raphael_bench_avx_filter rust:raphael_optim_memset2 rust:query-field-name-avoid-alloc rust:check_bug rust:missing_test rust:list_fields rust:block-cache-control rust:trinity--multilayer-sstable rust:tracing-tantivy2 rust:tracing-tantivy rust:index_writer_err rust:default_fast_field_tokenizer rust:fmassot/add-preserve-original-ascii-folding rust:numeric_with_str_col rust:straightforward-simplify rust:fmassot/fake-test-branch rust:fmassot/dynamic-tokenizer-poc-follow-up rust:fmassot/add-box-token-filter-and-refactor rust:dynamic-tokenizer-proof-of-concept rust:fmassot/add-box-tokenizer-and-token-filter rust:tokenizer-api-0.1.1 rust:addconversion rust:removed-extra-enum-alternative-for-datetime-precision rust:issue/1981-reproduce-bm25-bug rust:tokenizer_docs rust:merge-proptests rust:columnar-todo rust:quickwit-0.5 rust:remove-byteorder rust:remove_dyn rust:pr/alexcole/1855-inject-bm25-statistics rust:evance-expose-build-query-from-user-input-ast rust:missing-specialized-method-arcdyncolumn rust:fast-u64-range-query rust:0.19.2 rust:columnar-cli2 rust:trinity--pub-term-as-slice rust:typed-column3 rust:columnar-merge3 rust:columnar-merge-exploring-fn rust:use_columnar rust:typed-column2 rust:term_agg_bucket_limit rust:range rust:typed-column rust:columnar-main rust:forced-types rust:u128-columnar rust:use_stats rust:truncation_comment rust:0.19.1 rust:trinity/yoke-documents rust:column-handle rust:columnar-cli rust:columnar rust:revert-1711-sparse_dense_index rust:commit-changes rust:sparse_dense_index_headroom rust:optional_column rust:no-column-for-multivalued rust:fmassot--bench-hdfs-with-array rust:bugfix-position-broken rust:float rust:simplify rust:debug_time rust:sparse_codecs rust:fix_estimate rust:column-reader rust:column-reader-b rust:columnreader-simple rust:col-trait-refact rust:ip_fastfield rust:column-trait-2 rust:refact-dyn-column rust:column-trait rust:refact-codec-trait rust:pr/PSeitz/1485-1 rust:fastfieldcodec rust:termmap_per_field rust:fastfield-codecs-tests rust:rayon-indexing-writers rust:issue/1251 rust:index-bench rust:fix_open_bytes_index rust:warming rust:revert-1222-issue/1195 rust:add-random-bench-for-fast-field rust:wasm-friendly-exploratory rust:revert-1159-patch-1 rust:debug-position-panic-0.15.3 rust:fix-issue-1111 rust:remove_rand rust:unused-deps rust:hfb/phrase_match_slop rust:0.15.3 rust:expose-min-max-on-multi-fast-field-reader rust:revert-1094-expose-min-max-on-multi-fast-field-reader rust:hotfix-1088 rust:owned-bytes-instead-of-rawdoc rust:rihardsk-date-ranges rust:issue/997 rust:0.13.3 rust:fail-fast-store-checkpoint rust:debugging-store rust:issue/reproduce-897 rust:issue/969-reproduce rust:troublescooter-static rust:intermediate-mergeable-troublescooter rust:barrotsteindev-filter-collector-tpredicatevalue rust:issue/938b rust:issue/938 rust:barrotsteindev-filter-collector rust:issue/922 rust:issue/896 rust:quickfix-explain rust:slog rust:limit-random-seeks rust:0.13.1 rust:issue/866b rust:bugfix-unittest-macos rust:githubactions rust:fieldnorm_readers rust:blockwand rust:nodeffeatfails rust:removedaliveiterator rust:robyoung-introduce-offset-to-top-docs rust:specialization-of-foreach-for-union-scorer rust:elshize-bmw rust:cutt rust:nrt rust:storewriter-out rust:segment_writer_in_ram2 rust:polling rust:audunhalland-query-boost rust:poll-meta rust:segment_writer_in_ram rust:bundle rust:petr-tik-n662_footer_is_incompatible_impl rust:race-condition-in-tests rust:hotfix-0.10.3-issue681 rust:remove-tru rust:kkoziara-pre-tokenized-text rust:issue/681 rust:issue/680 rust:refactoring/field rust:updating/uuid rust:fix-bench rust:hotfix-656 rust:readd-macos-ci rust:incremental-search rust:crate-division rust:streamer-with-state rust:criterion rust:python-binding-changes rust:bump-version rust:gh-pages rust:issue/554 rust:perf/experimental-union-for_each rust:issue/526b rust:softcommits rust:0.9.1-hotfix rust:revert-521-fix-non-english-stemmers rust:issue/500-hotfix rust:hotfix/0.8.2 rust:issue/483 rust:bug/merge-deted rust:issue/407 rust:issue/457 rust:segment_fruits rust:doc rust:dds/lenient rust:issue/webasm rust:issue/396 rust:broken_collector_with_result rust:dss/automaton-weight-suggestion rust:wasm rust:common-crawl rust:tantivy-imhotep
rust:0.25.0 rust:0.22.1 rust:0.24.2 rust:0.24.1 rust:tantivy-query-grammar-v0.24.0 rust:tantivy-columnar-v0.5.0 rust:tantivy-bitpacker-v0.8.0 rust:ownedbytes-v0.9.0 rust:tantivy-stacker-v0.5.0 rust:tantivy-sstable-v0.5.0 rust:tantivy-common-v0.9.0 rust:tantivy-tokenizer-api-v0.5.0 rust:0.24 rust:qw-airmail-20250224-hotfix-merge-panic rust:qw-airmail-20250219-hotfix-merge-panic rust:qw-airmail-202406 rust:qw-airmail-20240611 rust:qw-airmail-20240606 rust:quickwit-rev rust:0.22.0 rust:tantivy-stacker-v0.3.0 rust:ownedbytes-v0.7.0 rust:tantivy-tokenizer-api-v0.3.0 rust:tantivy-columnar-v0.3.0 rust:tantivy-common-v0.7.0 rust:tantivy-query-grammar-v0.22.0 rust:tantivy-sstable-v0.3.0 rust:tantivy-bitpacker-v0.6.0 rust:0.21.1 rust:0.21 rust:0.20.2 rust:0.20.1 rust:0.20 rust:quickwit-0.5-rev rust:0.19.2 rust:0.19.1 rust:0.19 rust:0.18.1 rust:0.18 rust:0.17 rust:0.16.1 rust:0.15.3 rust:0.15.2 rust:0.15.1 rust:0.15 rust:0.14 rust:0.13.3 rust:0.13.2 rust:0.13.1 rust:0.13 rust:0.12 rust:0.11.3 rust:0.11.1 rust:0.11 rust:0.10.3 rust:0.1 rust:0.10.2 rust:0.10.1 rust:0.10.0 rust:0.9.1 rust:0.9 rust:0.8.2 rust:0.8.2b rust:0.8.1 rust:0.8.0 rust:0.7.2 rust:0.7.1 rust:0.7.0 rust:0.6.1 rust:0.6.0 rust:0.5.2 rust:0.5.1 rust:0.5.0 rust:0.4.3 rust:0.4.2 rust:0.4.0 rust:0.3.1 rust:0.3.0 rust:0.2.0
..
compare: rust:nrt
Branches Tags
rust:stuhood.upstream-erased-sort-computer rust:main rust:trinity.pointard/ff-query rust:bucket_id_agg rust:paradedb/fix-agg-validation rust:flatheadmill-geo rust:paradedb-paradedb/fix-overflow-issue rust:low_card_optimization rust:paul.masurel/clippy rust:fix_size_hint rust:qw-airmail-20250916 rust:unit-test-aggregation-sum-other rust:clippy-and-cleanups rust:trinity.pointard/opti-sstable-sorted-ords-to-term rust:release_0.24 rust:dependabot/cargo/rand-0.9.1 rust:dependabot/cargo/lru-0.14.0 rust:dependabot/cargo/fs4-0.13.1 rust:tagged-user-input-ast rust:1686a/parse-agg-res rust:dependabot/cargo/rand_distr-0.5.1 rust:issue/2411-adding-panic-handler-to-rayon-merge-thread-pool rust:dependabot/cargo/nom-8 rust:upgrade_deps rust:test_parser rust:dependabot/github_actions/codecov/codecov-action-5 rust:binggan-0.12.0 rust:agg_format rust:path_collector rust:test_order_bug rust:test_columnar_multi rust:columnar_optionalmask_in_multivalues rust:prefix-phrase-query-optim rust:tantivy-with-oneshot rust:update_example rust:store-reader-enumerate rust:document-deserialize-seed rust:check_doc_mem rust:experimental-do-not-merge rust:expdonotmerge rust:cherry-pick-vec-instead-of-btreemap rust:validate_length rust:raphael_bench_avx_filter rust:raphael_optim_memset2 rust:query-field-name-avoid-alloc rust:check_bug rust:missing_test rust:list_fields rust:block-cache-control rust:trinity--multilayer-sstable rust:tracing-tantivy2 rust:tracing-tantivy rust:index_writer_err rust:default_fast_field_tokenizer rust:fmassot/add-preserve-original-ascii-folding rust:numeric_with_str_col rust:straightforward-simplify rust:fmassot/fake-test-branch rust:fmassot/dynamic-tokenizer-poc-follow-up rust:fmassot/add-box-token-filter-and-refactor rust:dynamic-tokenizer-proof-of-concept rust:fmassot/add-box-tokenizer-and-token-filter rust:tokenizer-api-0.1.1 rust:addconversion rust:removed-extra-enum-alternative-for-datetime-precision rust:issue/1981-reproduce-bm25-bug rust:tokenizer_docs rust:merge-proptests rust:columnar-todo rust:quickwit-0.5 rust:remove-byteorder rust:remove_dyn rust:pr/alexcole/1855-inject-bm25-statistics rust:evance-expose-build-query-from-user-input-ast rust:missing-specialized-method-arcdyncolumn rust:fast-u64-range-query rust:0.19.2 rust:columnar-cli2 rust:trinity--pub-term-as-slice rust:typed-column3 rust:columnar-merge3 rust:columnar-merge-exploring-fn rust:use_columnar rust:typed-column2 rust:term_agg_bucket_limit rust:range rust:typed-column rust:columnar-main rust:forced-types rust:u128-columnar rust:use_stats rust:truncation_comment rust:0.19.1 rust:trinity/yoke-documents rust:column-handle rust:columnar-cli rust:columnar rust:revert-1711-sparse_dense_index rust:commit-changes rust:sparse_dense_index_headroom rust:optional_column rust:no-column-for-multivalued rust:fmassot--bench-hdfs-with-array rust:bugfix-position-broken rust:float rust:simplify rust:debug_time rust:sparse_codecs rust:fix_estimate rust:column-reader rust:column-reader-b rust:columnreader-simple rust:col-trait-refact rust:ip_fastfield rust:column-trait-2 rust:refact-dyn-column rust:column-trait rust:refact-codec-trait rust:pr/PSeitz/1485-1 rust:fastfieldcodec rust:termmap_per_field rust:fastfield-codecs-tests rust:rayon-indexing-writers rust:issue/1251 rust:index-bench rust:fix_open_bytes_index rust:warming rust:revert-1222-issue/1195 rust:add-random-bench-for-fast-field rust:wasm-friendly-exploratory rust:revert-1159-patch-1 rust:debug-position-panic-0.15.3 rust:fix-issue-1111 rust:remove_rand rust:unused-deps rust:hfb/phrase_match_slop rust:0.15.3 rust:expose-min-max-on-multi-fast-field-reader rust:revert-1094-expose-min-max-on-multi-fast-field-reader rust:hotfix-1088 rust:owned-bytes-instead-of-rawdoc rust:rihardsk-date-ranges rust:issue/997 rust:0.13.3 rust:fail-fast-store-checkpoint rust:debugging-store rust:issue/reproduce-897 rust:issue/969-reproduce rust:troublescooter-static rust:intermediate-mergeable-troublescooter rust:barrotsteindev-filter-collector-tpredicatevalue rust:issue/938b rust:issue/938 rust:barrotsteindev-filter-collector rust:issue/922 rust:issue/896 rust:quickfix-explain rust:slog rust:limit-random-seeks rust:0.13.1 rust:issue/866b rust:bugfix-unittest-macos rust:githubactions rust:fieldnorm_readers rust:blockwand rust:nodeffeatfails rust:removedaliveiterator rust:robyoung-introduce-offset-to-top-docs rust:specialization-of-foreach-for-union-scorer rust:elshize-bmw rust:cutt rust:nrt rust:storewriter-out rust:segment_writer_in_ram2 rust:polling rust:audunhalland-query-boost rust:poll-meta rust:segment_writer_in_ram rust:bundle rust:petr-tik-n662_footer_is_incompatible_impl rust:race-condition-in-tests rust:hotfix-0.10.3-issue681 rust:remove-tru rust:kkoziara-pre-tokenized-text rust:issue/681 rust:issue/680 rust:refactoring/field rust:updating/uuid rust:fix-bench rust:hotfix-656 rust:readd-macos-ci rust:incremental-search rust:crate-division rust:streamer-with-state rust:criterion rust:python-binding-changes rust:bump-version rust:gh-pages rust:issue/554 rust:perf/experimental-union-for_each rust:issue/526b rust:softcommits rust:0.9.1-hotfix rust:revert-521-fix-non-english-stemmers rust:issue/500-hotfix rust:hotfix/0.8.2 rust:issue/483 rust:bug/merge-deted rust:issue/407 rust:issue/457 rust:segment_fruits rust:doc rust:dds/lenient rust:issue/webasm rust:issue/396 rust:broken_collector_with_result rust:dss/automaton-weight-suggestion rust:wasm rust:common-crawl rust:tantivy-imhotep
rust:0.25.0 rust:0.22.1 rust:0.24.2 rust:0.24.1 rust:tantivy-query-grammar-v0.24.0 rust:tantivy-columnar-v0.5.0 rust:tantivy-bitpacker-v0.8.0 rust:ownedbytes-v0.9.0 rust:tantivy-stacker-v0.5.0 rust:tantivy-sstable-v0.5.0 rust:tantivy-common-v0.9.0 rust:tantivy-tokenizer-api-v0.5.0 rust:0.24 rust:qw-airmail-20250224-hotfix-merge-panic rust:qw-airmail-20250219-hotfix-merge-panic rust:qw-airmail-202406 rust:qw-airmail-20240611 rust:qw-airmail-20240606 rust:quickwit-rev rust:0.22.0 rust:tantivy-stacker-v0.3.0 rust:ownedbytes-v0.7.0 rust:tantivy-tokenizer-api-v0.3.0 rust:tantivy-columnar-v0.3.0 rust:tantivy-common-v0.7.0 rust:tantivy-query-grammar-v0.22.0 rust:tantivy-sstable-v0.3.0 rust:tantivy-bitpacker-v0.6.0 rust:0.21.1 rust:0.21 rust:0.20.2 rust:0.20.1 rust:0.20 rust:quickwit-0.5-rev rust:0.19.2 rust:0.19.1 rust:0.19 rust:0.18.1 rust:0.18 rust:0.17 rust:0.16.1 rust:0.15.3 rust:0.15.2 rust:0.15.1 rust:0.15 rust:0.14 rust:0.13.3 rust:0.13.2 rust:0.13.1 rust:0.13 rust:0.12 rust:0.11.3 rust:0.11.1 rust:0.11 rust:0.10.3 rust:0.1 rust:0.10.2 rust:0.10.1 rust:0.10.0 rust:0.9.1 rust:0.9 rust:0.8.2 rust:0.8.2b rust:0.8.1 rust:0.8.0 rust:0.7.2 rust:0.7.1 rust:0.7.0 rust:0.6.1 rust:0.6.0 rust:0.5.2 rust:0.5.1 rust:0.5.0 rust:0.4.3 rust:0.4.2 rust:0.4.0 rust:0.3.1 rust:0.3.0 rust:0.2.0

10 Commits
issue/922 ... nrt

Author SHA1 Message Date
Paul Masurel
9bc6a43917 persistor 2020-03-25 21:46:07 +09:00
Paul Masurel
e5bf41c1f6 moving merge inventory to the segment_manager 2020-03-25 10:05:01 +09:00
Paul Masurel
58d40ebf95 Add resource manager 2020-03-22 15:24:25 +09:00
Paul Masurel
f07634517f Merge branch 'nrt' of github.com:tantivy-search/tantivy into nrt 2020-03-17 20:57:22 +09:00
Paul Masurel
2edea107ef in ram segments 2020-03-13 18:44:34 +09:00
Paul Masurel
a6b5f4f5b5 in ram segments 2020-03-13 09:01:44 +09:00
Paul Masurel
7859ee4a39 NRT volatile 2020-03-11 18:55:09 +09:00
Paul Masurel
f06e116aae Update 2020-03-11 18:55:09 +09:00
Paul Masurel
ec32e0546a Added the concept of Volatile Segment. 2020-03-11 18:55:09 +09:00
Paul Masurel
0cd10e1197 Moving the StoreWriter out of the SegmentSerializer. 2020-03-11 18:55:09 +09:00
164 changed files with 6700 additions and 8547 deletions
Expand all files Collapse all files

2
.gitignore vendored
View File

@@ -1,5 +1,4 @@
tantivy.iml
proptest-regressions
*.swp
target
target/debug
@@ -12,4 +11,3 @@ cpp/simdcomp/bitpackingbenchmark
*.bk
.idea
trace.dat
cargo-timing*

50
CHANGELOG.md
View File

@@ -1,53 +1,3 @@
Tantivy 0.14.0
=========================
- Remove dependency to atomicwrites #833 .Implemented by @pmasurel upon suggestion and research from @asafigan).
- Migrated tantivy error from the now deprecated `failure` crate to `thiserror` #760. (@hirevo)
- API Change. Accessing the typed value off a `Schema::Value` now returns an Option instead of panicking if the type does not match.
- Large API Change in the Directory API. Tantivy used to assume that all files could be somehow memory mapped. After this change, Directory return a `FileSlice` that can be reduced and eventually read into an `OwnedBytes` object. Long and blocking io operation are still required by they do not span over the entire file.
- Added support for Brotli compression in the DocStore. (@ppodolsky)
- Added helper for building intersections and unions in BooleanQuery (@guilload)
- Bugfix in `Query::explain`
- Making it possible to opt out the generation of fieldnorms information for indexed fields. This change breaks compatibility as the meta.json file format is slightly changed. (#922, @pmasurel)
Tantivy 0.13.2
===================
Bugfix. Acquiring a facet reader on a segment that does not contain any
doc with this facet returns `None`. (#896)
Tantivy 0.13.1
===================
Made `Query` and `Collector` `Send + Sync`.
Updated misc dependency versions.
Tantivy 0.13.0
======================
Tantivy 0.13 introduce a change in the index format that will require
you to reindex your index (BlockWAND information are added in the skiplist).
The index size increase is minor as this information is only added for
full blocks.
If you have a massive index for which reindexing is not an option, please contact me
so that we can discuss possible solutions.
- Bugfix in `FuzzyTermQuery` not matching terms by prefix when it should (@Peachball)
- Relaxed constraints on the custom/tweak score functions. At the segment level, they can be mut, and they are not required to be Sync + Send.
- `MMapDirectory::open` does not return a `Result` anymore.
- Change in the DocSet and Scorer API. (@fulmicoton).
A freshly created DocSet point directly to their first doc. A sentinel value called TERMINATED marks the end of a DocSet.
`.advance()` returns the new DocId. `Scorer::skip(target)` has been replaced by `Scorer::seek(target)` and returns the resulting DocId.
As a result, iterating through DocSet now looks as follows
```rust
let mut doc = docset.doc();
while doc != TERMINATED {
// ...
doc = docset.advance();
}
```
The change made it possible to greatly simplify a lot of the docset's code.
- Misc internal optimization and introduction of the `Scorer::for_each_pruning` function. (@fulmicoton)
- Added an offset option to the Top(.*)Collectors. (@robyoung)
- Added Block WAND. Performance on TOP-K on term-unions should be greatly increased. (@fulmicoton, and special thanks
to the PISA team for answering all my questions!)
Tantivy 0.12.0
======================
- Removing static dispatch in tokenizers for simplicity. (#762)

58
Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy"
version = "0.14.0-dev"
version = "0.12.0"
authors = ["Paul Masurel <paul.masurel@gmail.com>"]
license = "MIT"
categories = ["database-implementations", "data-structures"]
@@ -13,40 +13,42 @@ keywords = ["search", "information", "retrieval"]
edition = "2018"
[dependencies]
base64 = "0.13"
byteorder = "1"
crc32fast = "1"
once_cell = "1"
regex ={version = "1", default-features = false, features = ["std"]}
tantivy-fst = "0.3"
base64 = "0.12.0"
byteorder = "1.0"
crc32fast = "1.2.0"
once_cell = "1.0"
regex ={version = "1.3.0", default-features = false, features = ["std"]}
tantivy-fst = "0.2.1"
memmap = {version = "0.7", optional=true}
lz4 = {version="1", optional=true}
brotli = {version="3.3.0", optional=true}
lz4 = {version="1.20", optional=true}
snap = "1"
tempfile = {version="3", optional=true}
atomicwrites = {version="0.2.2", optional=true}
tempfile = "3.0"
log = "0.4"
serde = {version="1", features=["derive"]}
serde_json = "1"
num_cpus = "1"
serde = {version="1.0", features=["derive"]}
serde_json = "1.0"
num_cpus = "1.2"
fs2={version="0.4", optional=true}
levenshtein_automata = "0.2"
levenshtein_automata = "0.1"
notify = {version="4", optional=true}
uuid = { version = "0.8", features = ["v4", "serde"] }
crossbeam = "0.8"
crossbeam = "0.7"
futures = {version = "0.3", features=["thread-pool"] }
tantivy-query-grammar = { version="0.14.0-dev", path="./query-grammar" }
stable_deref_trait = "1"
rust-stemmers = "1"
downcast-rs = "1"
owning_ref = "0.4"
stable_deref_trait = "1.0.0"
rust-stemmers = "1.2"
downcast-rs = { version="1.0" }
tantivy-query-grammar = { version="0.12", path="./query-grammar" }
bitpacking = {version="0.8", default-features = false, features=["bitpacker4x"]}
census = "0.4"
fnv = "1"
thiserror = "1.0"
htmlescape = "0.3"
fail = "0.4"
census = {path="../census"}
fnv = "1.0.6"
owned-read = "0.4"
failure = "0.1"
htmlescape = "0.3.1"
fail = "0.3"
murmurhash32 = "0.2"
chrono = "0.4"
smallvec = "1"
smallvec = "1.0"
rayon = "1"
[target.'cfg(windows)'.dependencies]
@@ -56,10 +58,9 @@ winapi = "0.3"
rand = "0.7"
maplit = "1"
matches = "0.1.8"
proptest = "0.10"
[dev-dependencies.fail]
version = "0.4"
version = "0.3"
features = ["failpoints"]
[profile.release]
@@ -73,8 +74,7 @@ overflow-checks = true
[features]
default = ["mmap"]
mmap = ["fs2", "tempfile", "memmap", "notify"]
brotli-compression = ["brotli"]
mmap = ["atomicwrites", "fs2", "memmap", "notify"]
lz4-compression = ["lz4"]
failpoints = ["fail/failpoints"]
unstable = [] # useful for benches.

8
README.md
View File

@@ -5,6 +5,7 @@
[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
[![Build status](https://ci.appveyor.com/api/projects/status/r7nb13kj23u8m9pj/branch/master?svg=true)](https://ci.appveyor.com/project/fulmicoton/tantivy/branch/master)
[![Crates.io](https://img.shields.io/crates/v/tantivy.svg)](https://crates.io/crates/tantivy)
[![Say Thanks!](https://img.shields.io/badge/Say%20Thanks-!-1EAEDB.svg)](https://saythanks.io/to/fulmicoton)
![Tantivy](https://tantivy-search.github.io/logo/tantivy-logo.png)
@@ -30,11 +31,12 @@ Tantivy is, in fact, strongly inspired by Lucene's design.
# Benchmark
Tantivy is typically faster than Lucene, but the results depend on
the nature of the queries in your workload.
The following [benchmark](https://tantivy-search.github.io/bench/) break downs
performance for different type of queries / collection.
Your mileage WILL vary depending on the nature of queries and their load.
# Features
- Full-text search
@@ -84,7 +86,7 @@ There are many ways to support this project.
- Help with documentation by asking questions or submitting PRs
- Contribute code (you can join [our Gitter](https://gitter.im/tantivy-search/tantivy))
- Talk about Tantivy around you
- [![Become a patron](https://c5.patreon.com/external/logo/become_a_patron_button.png)](https://www.patreon.com/fulmicoton)
- Drop a word on on [![Say Thanks!](https://img.shields.io/badge/Say%20Thanks-!-1EAEDB.svg)](https://saythanks.io/to/fulmicoton) or even [![Become a patron](https://c5.patreon.com/external/logo/become_a_patron_button.png)](https://www.patreon.com/fulmicoton)
# Contributing code

9
TODO.md Normal file
View File

@@ -0,0 +1,9 @@
- segment writer serializes if already flush
- flush if exceed max doc.
- unit tests
- reader from `IndexWriter`
- configurable limit to flush
- bundle directory
- add index worker when exceeds some doc limit
- flush before prepare commit.
- segment_writer should not receive a segment

2
appveyor.yml
View File

@@ -18,5 +18,5 @@ install:
build: false
test_script:
- REM SET RUST_LOG=tantivy,test & cargo test --all --verbose --no-default-features --features mmap
- REM SET RUST_LOG=tantivy,test & cargo test --verbose --no-default-features --features mmap
- REM SET RUST_BACKTRACE=1 & cargo build --examples

12
examples/basic_search.rs
View File

@@ -112,6 +112,18 @@ fn main() -> tantivy::Result<()> {
limbs and branches that arch over the pool"
));
index_writer.add_document(doc!(
title => "Of Mice and Men",
body => "A few miles south of Soledad, the Salinas River drops in close to the hillside \
bank and runs deep and green. The water is warm too, for it has slipped twinkling \
over the yellow sands in the sunlight before reaching the narrow pool. On one \
side of the river the golden foothill slopes curve up to the strong and rocky \
Gabilan Mountains, but on the valley side the water is lined with trees—willows \
fresh and green with every spring, carrying in their lower leaf junctures the \
debris of the winter’s flooding; and sycamores with mottled, white, recumbent \
limbs and branches that arch over the pool"
));
// Multivalued field just need to be repeated.
index_writer.add_document(doc!(
title => "Frankenstein",

4
examples/custom_collector.rs
View File

@@ -14,7 +14,7 @@ use tantivy::fastfield::FastFieldReader;
use tantivy::query::QueryParser;
use tantivy::schema::Field;
use tantivy::schema::{Schema, FAST, INDEXED, TEXT};
use tantivy::{doc, Index, Score, SegmentReader, TantivyError};
use tantivy::{doc, Index, SegmentReader, TantivyError};
#[derive(Default)]
struct Stats {
@@ -114,7 +114,7 @@ struct StatsSegmentCollector {
impl SegmentCollector for StatsSegmentCollector {
type Fruit = Option<Stats>;
fn collect(&mut self, doc: u32, _score: Score) {
fn collect(&mut self, doc: u32, _score: f32) {
let value = self.fast_field_reader.get(doc) as f64;
self.stats.count += 1;
self.stats.sum += value;

98
examples/faceted_search_with_tweaked_score.rs
View File

@@ -1,98 +0,0 @@
use std::collections::HashSet;
use tantivy::collector::TopDocs;
use tantivy::doc;
use tantivy::query::BooleanQuery;
use tantivy::schema::*;
use tantivy::{DocId, Index, Score, SegmentReader};
fn main() -> tantivy::Result<()> {
let mut schema_builder = Schema::builder();
let title = schema_builder.add_text_field("title", STORED);
let ingredient = schema_builder.add_facet_field("ingredient");
let schema = schema_builder.build();
let index = Index::create_in_ram(schema.clone());
let mut index_writer = index.writer(30_000_000)?;
index_writer.add_document(doc!(
title => "Fried egg",
ingredient => Facet::from("/ingredient/egg"),
ingredient => Facet::from("/ingredient/oil"),
));
index_writer.add_document(doc!(
title => "Scrambled egg",
ingredient => Facet::from("/ingredient/egg"),
ingredient => Facet::from("/ingredient/butter"),
ingredient => Facet::from("/ingredient/milk"),
ingredient => Facet::from("/ingredient/salt"),
));
index_writer.add_document(doc!(
title => "Egg rolls",
ingredient => Facet::from("/ingredient/egg"),
ingredient => Facet::from("/ingredient/garlic"),
ingredient => Facet::from("/ingredient/salt"),
ingredient => Facet::from("/ingredient/oil"),
ingredient => Facet::from("/ingredient/tortilla-wrap"),
ingredient => Facet::from("/ingredient/mushroom"),
));
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
{
let facets = vec![
Facet::from("/ingredient/egg"),
Facet::from("/ingredient/oil"),
Facet::from("/ingredient/garlic"),
Facet::from("/ingredient/mushroom"),
];
let query = BooleanQuery::new_multiterms_query(
facets
.iter()
.map(|key| Term::from_facet(ingredient, &key))
.collect(),
);
let top_docs_by_custom_score =
TopDocs::with_limit(2).tweak_score(move |segment_reader: &SegmentReader| {
let ingredient_reader = segment_reader.facet_reader(ingredient).unwrap();
let facet_dict = ingredient_reader.facet_dict();
let query_ords: HashSet<u64> = facets
.iter()
.filter_map(|key| facet_dict.term_ord(key.encoded_str()))
.collect();
let mut facet_ords_buffer: Vec<u64> = Vec::with_capacity(20);
move |doc: DocId, original_score: Score| {
ingredient_reader.facet_ords(doc, &mut facet_ords_buffer);
let missing_ingredients = facet_ords_buffer
.iter()
.filter(|ord| !query_ords.contains(ord))
.count();
let tweak = 1.0 / 4_f32.powi(missing_ingredients as i32);
original_score * tweak
}
});
let top_docs = searcher.search(&query, &top_docs_by_custom_score)?;
let titles: Vec<String> = top_docs
.iter()
.map(|(_, doc_id)| {
searcher
.doc(*doc_id)
.unwrap()
.get_first(title)
.unwrap()
.text()
.unwrap()
.to_owned()
})
.collect();
assert_eq!(titles, vec!["Fried egg", "Egg rolls"]);
}
Ok(())
}

25
examples/iterating_docs_and_positions.rs
View File

@@ -10,7 +10,7 @@
// ---
// Importing tantivy...
use tantivy::schema::*;
use tantivy::{doc, DocSet, Index, Postings, TERMINATED};
use tantivy::{doc, DocId, DocSet, Index, Postings};
fn main() -> tantivy::Result<()> {
// We first create a schema for the sake of the
@@ -45,7 +45,7 @@ fn main() -> tantivy::Result<()> {
// Inverted index stands for the combination of
// - the term dictionary
// - the inverted lists associated to each terms and their positions
let inverted_index = segment_reader.inverted_index(title)?;
let inverted_index = segment_reader.inverted_index(title);
// A `Term` is a text token associated with a field.
// Let's go through all docs containing the term `title:the` and access their position
@@ -58,15 +58,16 @@ fn main() -> tantivy::Result<()> {
// If you don't need all this information, you may get better performance by decompressing less
// information.
if let Some(mut segment_postings) =
inverted_index.read_postings(&term_the, IndexRecordOption::WithFreqsAndPositions)?
inverted_index.read_postings(&term_the, IndexRecordOption::WithFreqsAndPositions)
{
// this buffer will be used to request for positions
let mut positions: Vec<u32> = Vec::with_capacity(100);
let mut doc_id = segment_postings.doc();
while doc_id != TERMINATED {
while segment_postings.advance() {
// the number of time the term appears in the document.
let doc_id: DocId = segment_postings.doc(); //< do not try to access this before calling advance once.
// This MAY contains deleted documents as well.
if segment_reader.is_deleted(doc_id) {
doc_id = segment_postings.advance();
continue;
}
@@ -85,7 +86,6 @@ fn main() -> tantivy::Result<()> {
// Doc 2: TermFreq 1: [0]
// ```
println!("Doc {}: TermFreq {}: {:?}", doc_id, term_freq, positions);
doc_id = segment_postings.advance();
}
}
}
@@ -106,7 +106,7 @@ fn main() -> tantivy::Result<()> {
// Inverted index stands for the combination of
// - the term dictionary
// - the inverted lists associated to each terms and their positions
let inverted_index = segment_reader.inverted_index(title)?;
let inverted_index = segment_reader.inverted_index(title);
// This segment posting object is like a cursor over the documents matching the term.
// The `IndexRecordOption` arguments tells tantivy we will be interested in both term frequencies
@@ -115,18 +115,13 @@ fn main() -> tantivy::Result<()> {
// If you don't need all this information, you may get better performance by decompressing less
// information.
if let Some(mut block_segment_postings) =
inverted_index.read_block_postings(&term_the, IndexRecordOption::Basic)?
inverted_index.read_block_postings(&term_the, IndexRecordOption::Basic)
{
loop {
let docs = block_segment_postings.docs();
if docs.is_empty() {
break;
}
while block_segment_postings.advance() {
// Once again these docs MAY contains deleted documents as well.
let docs = block_segment_postings.docs();
// Prints `Docs [0, 2].`
println!("Docs {:?}", docs);
block_segment_postings.advance();
}
}
}

2
query-grammar/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy-query-grammar"
version = "0.14.0-dev"
version = "0.12.0"
authors = ["Paul Masurel <paul.masurel@gmail.com>"]
license = "MIT"
categories = ["database-implementations", "data-structures"]

46
query-grammar/src/occur.rs
View File

@@ -31,12 +31,22 @@ impl Occur {
/// Compose two occur values.
pub fn compose(left: Occur, right: Occur) -> Occur {
match (left, right) {
(Occur::Should, _) => right,
(Occur::Must, Occur::MustNot) => Occur::MustNot,
(Occur::Must, _) => Occur::Must,
(Occur::MustNot, Occur::MustNot) => Occur::Must,
(Occur::MustNot, _) => Occur::MustNot,
match left {
Occur::Should => right,
Occur::Must => {
if right == Occur::MustNot {
Occur::MustNot
} else {
Occur::Must
}
}
Occur::MustNot => {
if right == Occur::MustNot {
Occur::Must
} else {
Occur::MustNot
}
}
}
}
}
@@ -46,27 +56,3 @@ impl fmt::Display for Occur {
f.write_char(self.to_char())
}
}
#[cfg(test)]
mod test {
use crate::Occur;
#[test]
fn test_occur_compose() {
assert_eq!(Occur::compose(Occur::Should, Occur::Should), Occur::Should);
assert_eq!(Occur::compose(Occur::Should, Occur::Must), Occur::Must);
assert_eq!(
Occur::compose(Occur::Should, Occur::MustNot),
Occur::MustNot
);
assert_eq!(Occur::compose(Occur::Must, Occur::Should), Occur::Must);
assert_eq!(Occur::compose(Occur::Must, Occur::Must), Occur::Must);
assert_eq!(Occur::compose(Occur::Must, Occur::MustNot), Occur::MustNot);
assert_eq!(
Occur::compose(Occur::MustNot, Occur::Should),
Occur::MustNot
);
assert_eq!(Occur::compose(Occur::MustNot, Occur::Must), Occur::MustNot);
assert_eq!(Occur::compose(Occur::MustNot, Occur::MustNot), Occur::Must);
}
}

190
query-grammar/src/query_grammar.rs
View File

@@ -9,10 +9,8 @@ use combine::{
fn field<'a>() -> impl Parser<&'a str, Output = String> {
(
(letter().or(char('_'))),
many(satisfy(|c: char| {
c.is_alphanumeric() || c == '_' || c == '-'
})),
letter(),
many(satisfy(|c: char| c.is_alphanumeric() || c == '_')),
)
.skip(char(':'))
.map(|(s1, s2): (char, String)| format!("{}{}", s1, s2))
@@ -156,11 +154,17 @@ fn negate(expr: UserInputAST) -> UserInputAST {
expr.unary(Occur::MustNot)
}
fn must(expr: UserInputAST) -> UserInputAST {
expr.unary(Occur::Must)
}
fn leaf<'a>() -> impl Parser<&'a str, Output = UserInputAST> {
parser(|input| {
char('(')
.with(ast())
.skip(char(')'))
char('-')
.with(leaf())
.map(negate)
.or(char('+').with(leaf()).map(must))
.or(char('(').with(ast()).skip(char(')')))
.or(char('*').map(|_| UserInputAST::from(UserInputLeaf::All)))
.or(attempt(
string("NOT").skip(spaces1()).with(leaf()).map(negate),
@@ -172,17 +176,7 @@ fn leaf<'a>() -> impl Parser<&'a str, Output = UserInputAST> {
})
}
fn occur_symbol<'a>() -> impl Parser<&'a str, Output = Occur> {
char('-')
.map(|_| Occur::MustNot)
.or(char('+').map(|_| Occur::Must))
}
fn occur_leaf<'a>() -> impl Parser<&'a str, Output = (Option<Occur>, UserInputAST)> {
(optional(occur_symbol()), boosted_leaf())
}
fn positive_float_number<'a>() -> impl Parser<&'a str, Output = f64> {
fn positive_float_number<'a>() -> impl Parser<&'a str, Output = f32> {
(many1(digit()), optional((char('.'), many1(digit())))).map(
|(int_part, decimal_part_opt): (String, Option<(char, String)>)| {
let mut float_str = int_part;
@@ -190,18 +184,18 @@ fn positive_float_number<'a>() -> impl Parser<&'a str, Output = f64> {
float_str.push(chr);
float_str.push_str(&decimal_str);
}
float_str.parse::<f64>().unwrap()
float_str.parse::<f32>().unwrap()
},
)
}
fn boost<'a>() -> impl Parser<&'a str, Output = f64> {
fn boost<'a>() -> impl Parser<&'a str, Output = f32> {
(char('^'), positive_float_number()).map(|(_, boost)| boost)
}
fn boosted_leaf<'a>() -> impl Parser<&'a str, Output = UserInputAST> {
(leaf(), optional(boost())).map(|(leaf, boost_opt)| match boost_opt {
Some(boost) if (boost - 1.0).abs() > std::f64::EPSILON => {
Some(boost) if (boost - 1.0).abs() > std::f32::EPSILON => {
UserInputAST::Boost(Box::new(leaf), boost)
}
_ => leaf,
@@ -245,29 +239,21 @@ fn aggregate_binary_expressions(
}
}
fn operand_leaf<'a>() -> impl Parser<&'a str, Output = (BinaryOperand, UserInputAST)> {
(
pub fn ast<'a>() -> impl Parser<&'a str, Output = UserInputAST> {
let operand_leaf = (
binary_operand().skip(spaces()),
boosted_leaf().skip(spaces()),
)
}
pub fn ast<'a>() -> impl Parser<&'a str, Output = UserInputAST> {
let boolean_expr = (boosted_leaf().skip(spaces()), many1(operand_leaf()))
);
let boolean_expr = (boosted_leaf().skip(spaces().silent()), many1(operand_leaf))
.map(|(left, right)| aggregate_binary_expressions(left, right));
let whitespace_separated_leaves = many1(occur_leaf().skip(spaces().silent())).map(
|subqueries: Vec<(Option<Occur>, UserInputAST)>| {
let whitespace_separated_leaves =
many1(boosted_leaf().skip(spaces().silent())).map(|subqueries: Vec<UserInputAST>| {
if subqueries.len() == 1 {
let (occur_opt, ast) = subqueries.into_iter().next().unwrap();
match occur_opt.unwrap_or(Occur::Should) {
Occur::Must | Occur::Should => ast,
Occur::MustNot => UserInputAST::Clause(vec![(Some(Occur::MustNot), ast)]),
}
subqueries.into_iter().next().unwrap()
} else {
UserInputAST::Clause(subqueries.into_iter().collect())
}
},
);
});
let expr = attempt(boolean_expr).or(whitespace_separated_leaves);
spaces().with(expr).skip(spaces())
}
@@ -281,16 +267,14 @@ pub fn parse_to_ast<'a>() -> impl Parser<&'a str, Output = UserInputAST> {
#[cfg(test)]
mod test {
type TestParseResult = Result<(), StringStreamError>;
use super::*;
use combine::parser::Parser;
pub fn nearly_equals(a: f64, b: f64) -> bool {
pub fn nearly_equals(a: f32, b: f32) -> bool {
(a - b).abs() < 0.0005 * (a + b).abs()
}
fn assert_nearly_equals(expected: f64, val: f64) {
fn assert_nearly_equals(expected: f32, val: f32) {
assert!(
nearly_equals(val, expected),
"Got {}, expected {}.",
@@ -299,16 +283,9 @@ mod test {
);
}
#[test]
fn test_occur_symbol() -> TestParseResult {
assert_eq!(super::occur_symbol().parse("-")?, (Occur::MustNot, ""));
assert_eq!(super::occur_symbol().parse("+")?, (Occur::Must, ""));
Ok(())
}
#[test]
fn test_positive_float_number() {
fn valid_parse(float_str: &str, expected_val: f64, expected_remaining: &str) {
fn valid_parse(float_str: &str, expected_val: f32, expected_remaining: &str) {
let (val, remaining) = positive_float_number().parse(float_str).unwrap();
assert_eq!(remaining, expected_remaining);
assert_nearly_equals(val, expected_val);
@@ -316,9 +293,9 @@ mod test {
fn error_parse(float_str: &str) {
assert!(positive_float_number().parse(float_str).is_err());
}
valid_parse("1.0", 1.0, "");
valid_parse("1", 1.0, "");
valid_parse("0.234234 aaa", 0.234234f64, " aaa");
valid_parse("1.0", 1.0f32, "");
valid_parse("1", 1.0f32, "");
valid_parse("0.234234 aaa", 0.234234f32, " aaa");
error_parse(".3332");
error_parse("1.");
error_parse("-1.");
@@ -353,7 +330,7 @@ mod test {
"Err(UnexpectedParse)"
);
test_parse_query_to_ast_helper("NOTa", "\"NOTa\"");
test_parse_query_to_ast_helper("NOT a", "(-\"a\")");
test_parse_query_to_ast_helper("NOT a", "-(\"a\")");
}
#[test]
@@ -361,16 +338,16 @@ mod test {
assert!(parse_to_ast().parse("a^2^3").is_err());
assert!(parse_to_ast().parse("a^2^").is_err());
test_parse_query_to_ast_helper("a^3", "(\"a\")^3");
test_parse_query_to_ast_helper("a^3 b^2", "(*(\"a\")^3 *(\"b\")^2)");
test_parse_query_to_ast_helper("a^3 b^2", "((\"a\")^3 (\"b\")^2)");
test_parse_query_to_ast_helper("a^1", "\"a\"");
}
#[test]
fn test_parse_query_to_ast_binary_op() {
test_parse_query_to_ast_helper("a AND b", "(+\"a\" +\"b\")");
test_parse_query_to_ast_helper("a OR b", "(?\"a\" ?\"b\")");
test_parse_query_to_ast_helper("a OR b AND c", "(?\"a\" ?(+\"b\" +\"c\"))");
test_parse_query_to_ast_helper("a AND b AND c", "(+\"a\" +\"b\" +\"c\")");
test_parse_query_to_ast_helper("a AND b", "(+(\"a\") +(\"b\"))");
test_parse_query_to_ast_helper("a OR b", "(?(\"a\") ?(\"b\"))");
test_parse_query_to_ast_helper("a OR b AND c", "(?(\"a\") ?((+(\"b\") +(\"c\"))))");
test_parse_query_to_ast_helper("a AND b AND c", "(+(\"a\") +(\"b\") +(\"c\"))");
assert_eq!(
format!("{:?}", parse_to_ast().parse("a OR b aaa")),
"Err(UnexpectedParse)"
@@ -408,32 +385,6 @@ mod test {
test_parse_query_to_ast_helper("weight: <= 70.5", "weight:{\"*\" TO \"70.5\"]");
}
#[test]
fn test_occur_leaf() {
let ((occur, ast), _) = super::occur_leaf().parse("+abc").unwrap();
assert_eq!(occur, Some(Occur::Must));
assert_eq!(format!("{:?}", ast), "\"abc\"");
}
#[test]
fn test_field_name() -> TestParseResult {
assert_eq!(
super::field().parse("my-field-name:a")?,
("my-field-name".to_string(), "a")
);
assert_eq!(
super::field().parse("my_field_name:a")?,
("my_field_name".to_string(), "a")
);
assert!(super::field().parse(":a").is_err());
assert!(super::field().parse("-my_field:a").is_err());
assert_eq!(
super::field().parse("_my_field:a")?,
("_my_field".to_string(), "a")
);
Ok(())
}
#[test]
fn test_range_parser() {
// testing the range() parser separately
@@ -462,67 +413,32 @@ mod test {
fn test_parse_query_to_triming_spaces() {
test_parse_query_to_ast_helper(" abc", "\"abc\"");
test_parse_query_to_ast_helper("abc ", "\"abc\"");
test_parse_query_to_ast_helper("( a OR abc)", "(?\"a\" ?\"abc\")");
test_parse_query_to_ast_helper("(a OR abc)", "(?\"a\" ?\"abc\")");
test_parse_query_to_ast_helper("(a OR abc)", "(?\"a\" ?\"abc\")");
test_parse_query_to_ast_helper("a OR abc ", "(?\"a\" ?\"abc\")");
test_parse_query_to_ast_helper("(a OR abc )", "(?\"a\" ?\"abc\")");
test_parse_query_to_ast_helper("(a OR abc) ", "(?\"a\" ?\"abc\")");
test_parse_query_to_ast_helper("( a OR abc)", "(?(\"a\") ?(\"abc\"))");
test_parse_query_to_ast_helper("(a OR abc)", "(?(\"a\") ?(\"abc\"))");
test_parse_query_to_ast_helper("(a OR abc)", "(?(\"a\") ?(\"abc\"))");
test_parse_query_to_ast_helper("a OR abc ", "(?(\"a\") ?(\"abc\"))");
test_parse_query_to_ast_helper("(a OR abc )", "(?(\"a\") ?(\"abc\"))");
test_parse_query_to_ast_helper("(a OR abc) ", "(?(\"a\") ?(\"abc\"))");
}
#[test]
fn test_parse_query_single_term() {
fn test_parse_query_to_ast() {
test_parse_query_to_ast_helper("abc", "\"abc\"");
}
#[test]
fn test_parse_query_default_clause() {
test_parse_query_to_ast_helper("a b", "(*\"a\" *\"b\")");
}
#[test]
fn test_parse_query_must_default_clause() {
test_parse_query_to_ast_helper("+(a b)", "(*\"a\" *\"b\")");
}
#[test]
fn test_parse_query_must_single_term() {
test_parse_query_to_ast_helper("+d", "\"d\"");
}
#[test]
fn test_single_term_with_field() {
test_parse_query_to_ast_helper("a b", "(\"a\" \"b\")");
test_parse_query_to_ast_helper("+(a b)", "+((\"a\" \"b\"))");
test_parse_query_to_ast_helper("+d", "+(\"d\")");
test_parse_query_to_ast_helper("+(a b) +d", "(+((\"a\" \"b\")) +(\"d\"))");
test_parse_query_to_ast_helper("(+a +b) d", "((+(\"a\") +(\"b\")) \"d\")");
test_parse_query_to_ast_helper("(+a)", "+(\"a\")");
test_parse_query_to_ast_helper("(+a +b)", "(+(\"a\") +(\"b\"))");
test_parse_query_to_ast_helper("abc:toto", "abc:\"toto\"");
}
#[test]
fn test_single_term_with_float() {
test_parse_query_to_ast_helper("abc:1.1", "abc:\"1.1\"");
}
#[test]
fn test_must_clause() {
test_parse_query_to_ast_helper("(+a +b)", "(+\"a\" +\"b\")");
}
#[test]
fn test_parse_test_query_plus_a_b_plus_d() {
test_parse_query_to_ast_helper("+(a b) +d", "(+(*\"a\" *\"b\") +\"d\")");
}
#[test]
fn test_parse_test_query_other() {
test_parse_query_to_ast_helper("(+a +b) d", "(*(+\"a\" +\"b\") *\"d\")");
test_parse_query_to_ast_helper("+abc:toto", "abc:\"toto\"");
test_parse_query_to_ast_helper("(+abc:toto -titi)", "(+abc:\"toto\" -\"titi\")");
test_parse_query_to_ast_helper("-abc:toto", "(-abc:\"toto\")");
test_parse_query_to_ast_helper("abc:a b", "(*abc:\"a\" *\"b\")");
test_parse_query_to_ast_helper("+abc:toto", "+(abc:\"toto\")");
test_parse_query_to_ast_helper("(+abc:toto -titi)", "(+(abc:\"toto\") -(\"titi\"))");
test_parse_query_to_ast_helper("-abc:toto", "-(abc:\"toto\")");
test_parse_query_to_ast_helper("abc:a b", "(abc:\"a\" \"b\")");
test_parse_query_to_ast_helper("abc:\"a b\"", "abc:\"a b\"");
test_parse_query_to_ast_helper("foo:[1 TO 5]", "foo:[\"1\" TO \"5\"]");
}
#[test]
fn test_parse_query_with_range() {
test_parse_query_to_ast_helper("[1 TO 5]", "[\"1\" TO \"5\"]");
test_parse_query_to_ast_helper("foo:{a TO z}", "foo:{\"a\" TO \"z\"}");
test_parse_query_to_ast_helper("foo:[1 TO toto}", "foo:[\"1\" TO \"toto\"}");

32
query-grammar/src/user_input_ast.rs
View File

@@ -85,14 +85,15 @@ impl UserInputBound {
}
pub enum UserInputAST {
Clause(Vec<(Option<Occur>, UserInputAST)>),
Clause(Vec<UserInputAST>),
Unary(Occur, Box<UserInputAST>),
Leaf(Box<UserInputLeaf>),
Boost(Box<UserInputAST>, f64),
Boost(Box<UserInputAST>, f32),
}
impl UserInputAST {
pub fn unary(self, occur: Occur) -> UserInputAST {
UserInputAST::Clause(vec![(Some(occur), self)])
UserInputAST::Unary(occur, Box::new(self))
}
fn compose(occur: Occur, asts: Vec<UserInputAST>) -> UserInputAST {
@@ -103,7 +104,7 @@ impl UserInputAST {
} else {
UserInputAST::Clause(
asts.into_iter()
.map(|ast: UserInputAST| (Some(occur), ast))
.map(|ast: UserInputAST| ast.unary(occur))
.collect::<Vec<_>>(),
)
}
@@ -134,36 +135,25 @@ impl From<UserInputLeaf> for UserInputAST {
}
}
fn print_occur_ast(
occur_opt: Option<Occur>,
ast: &UserInputAST,
formatter: &mut fmt::Formatter,
) -> fmt::Result {
if let Some(occur) = occur_opt {
write!(formatter, "{}{:?}", occur, ast)?;
} else {
write!(formatter, "*{:?}", ast)?;
}
Ok(())
}
impl fmt::Debug for UserInputAST {
fn fmt(&self, formatter: &mut fmt::Formatter) -> fmt::Result {
fn fmt(&self, formatter: &mut fmt::Formatter<'_>) -> Result<(), fmt::Error> {
match *self {
UserInputAST::Clause(ref subqueries) => {
if subqueries.is_empty() {
write!(formatter, "<emptyclause>")?;
} else {
write!(formatter, "(")?;
print_occur_ast(subqueries[0].0, &subqueries[0].1, formatter)?;
write!(formatter, "{:?}", &subqueries[0])?;
for subquery in &subqueries[1..] {
write!(formatter, " ")?;
print_occur_ast(subquery.0, &subquery.1, formatter)?;
write!(formatter, " {:?}", subquery)?;
}
write!(formatter, ")")?;
}
Ok(())
}
UserInputAST::Unary(ref occur, ref subquery) => {
write!(formatter, "{}({:?})", occur, subquery)
}
UserInputAST::Leaf(ref subquery) => write!(formatter, "{:?}", subquery),
UserInputAST::Boost(ref leaf, boost) => write!(formatter, "({:?})^{}", leaf, boost),
}

8
src/collector/count_collector.rs
View File

@@ -96,18 +96,18 @@ mod tests {
}
{
let mut count_collector = SegmentCountCollector::default();
count_collector.collect(0u32, 1.0);
count_collector.collect(0u32, 1f32);
assert_eq!(count_collector.harvest(), 1);
}
{
let mut count_collector = SegmentCountCollector::default();
count_collector.collect(0u32, 1.0);
count_collector.collect(0u32, 1f32);
assert_eq!(count_collector.harvest(), 1);
}
{
let mut count_collector = SegmentCountCollector::default();
count_collector.collect(0u32, 1.0);
count_collector.collect(1u32, 1.0);
count_collector.collect(0u32, 1f32);
count_collector.collect(1u32, 1f32);
assert_eq!(count_collector.harvest(), 2);
}
}

16
src/collector/custom_score_top_collector.rs
View File

@@ -11,13 +11,13 @@ impl<TCustomScorer, TScore> CustomScoreTopCollector<TCustomScorer, TScore>
where
TScore: Clone + PartialOrd,
{
pub(crate) fn new(
pub fn new(
custom_scorer: TCustomScorer,
collector: TopCollector<TScore>,
limit: usize,
) -> CustomScoreTopCollector<TCustomScorer, TScore> {
CustomScoreTopCollector {
custom_scorer,
collector,
collector: TopCollector::with_limit(limit),
}
}
}
@@ -28,7 +28,7 @@ where
/// It is the segment local version of the [`CustomScorer`](./trait.CustomScorer.html).
pub trait CustomSegmentScorer<TScore>: 'static {
/// Computes the score of a specific `doc`.
fn score(&mut self, doc: DocId) -> TScore;
fn score(&self, doc: DocId) -> TScore;
}
/// `CustomScorer` makes it possible to define any kind of score.
@@ -46,7 +46,7 @@ pub trait CustomScorer<TScore>: Sync {
impl<TCustomScorer, TScore> Collector for CustomScoreTopCollector<TCustomScorer, TScore>
where
TCustomScorer: CustomScorer<TScore> + Send + Sync,
TCustomScorer: CustomScorer<TScore>,
TScore: 'static + PartialOrd + Clone + Send + Sync,
{
type Fruit = Vec<(TScore, DocAddress)>;
@@ -58,10 +58,10 @@ where
segment_local_id: u32,
segment_reader: &SegmentReader,
) -> crate::Result<Self::Child> {
let segment_scorer = self.custom_scorer.segment_scorer(segment_reader)?;
let segment_collector = self
.collector
.for_segment(segment_local_id, segment_reader)?;
let segment_scorer = self.custom_scorer.segment_scorer(segment_reader)?;
Ok(CustomScoreTopSegmentCollector {
segment_collector,
segment_scorer,
@@ -117,9 +117,9 @@ where
impl<F, TScore> CustomSegmentScorer<TScore> for F
where
F: 'static + FnMut(DocId) -> TScore,
F: 'static + Sync + Send + Fn(DocId) -> TScore,
{
fn score(&mut self, doc: DocId) -> TScore {
fn score(&self, doc: DocId) -> TScore {
(self)(doc)
}
}

61
src/collector/docset_collector.rs
View File

@@ -1,61 +0,0 @@
use std::collections::HashSet;
use crate::{DocAddress, DocId, Score};
use super::{Collector, SegmentCollector};
/// Collectors that returns the set of DocAddress that matches the query.
///
/// This collector is mostly useful for tests.
pub struct DocSetCollector;
impl Collector for DocSetCollector {
type Fruit = HashSet<DocAddress>;
type Child = DocSetChildCollector;
fn for_segment(
&self,
segment_local_id: crate::SegmentLocalId,
_segment: &crate::SegmentReader,
) -> crate::Result<Self::Child> {
Ok(DocSetChildCollector {
segment_local_id,
docs: HashSet::new(),
})
}
fn requires_scoring(&self) -> bool {
false
}
fn merge_fruits(
&self,
segment_fruits: Vec<(u32, HashSet<DocId>)>,
) -> crate::Result<Self::Fruit> {
let len: usize = segment_fruits.iter().map(|(_, docset)| docset.len()).sum();
let mut result = HashSet::with_capacity(len);
for (segment_local_id, docs) in segment_fruits {
for doc in docs {
result.insert(DocAddress(segment_local_id, doc));
}
}
Ok(result)
}
}
pub struct DocSetChildCollector {
segment_local_id: u32,
docs: HashSet<DocId>,
}
impl SegmentCollector for DocSetChildCollector {
type Fruit = (u32, HashSet<DocId>);
fn collect(&mut self, doc: crate::DocId, _score: Score) {
self.docs.insert(doc);
}
fn harvest(self) -> (u32, HashSet<DocId>) {
(self.segment_local_id, self.docs)
}
}

56
src/collector/facet_collector.rs
View File

@@ -1,5 +1,6 @@
use crate::collector::Collector;
use crate::collector::SegmentCollector;
use crate::docset::SkipResult;
use crate::fastfield::FacetReader;
use crate::schema::Facet;
use crate::schema::Field;
@@ -7,6 +8,7 @@ use crate::DocId;
use crate::Score;
use crate::SegmentLocalId;
use crate::SegmentReader;
use crate::TantivyError;
use std::cmp::Ordering;
use std::collections::btree_map;
use std::collections::BTreeMap;
@@ -186,11 +188,6 @@ pub struct FacetSegmentCollector {
collapse_facet_ords: Vec<u64>,
}
enum SkipResult {
Found,
NotFound,
}
fn skip<'a, I: Iterator<Item = &'a Facet>>(
target: &[u8],
collapse_it: &mut Peekable<I>,
@@ -200,14 +197,14 @@ fn skip<'a, I: Iterator<Item = &'a Facet>>(
Some(facet_bytes) => match facet_bytes.encoded_str().as_bytes().cmp(target) {
Ordering::Less => {}
Ordering::Greater => {
return SkipResult::NotFound;
return SkipResult::OverStep;
}
Ordering::Equal => {
return SkipResult::Found;
return SkipResult::Reached;
}
},
None => {
return SkipResult::NotFound;
return SkipResult::End;
}
}
collapse_it.next();
@@ -265,7 +262,10 @@ impl Collector for FacetCollector {
_: SegmentLocalId,
reader: &SegmentReader,
) -> crate::Result<FacetSegmentCollector> {
let facet_reader = reader.facet_reader(self.field)?;
let field_name = reader.schema().get_field_name(self.field);
let facet_reader = reader.facet_reader(self.field).ok_or_else(|| {
TantivyError::SchemaError(format!("Field {:?} is not a facet field.", field_name))
})?;
let mut collapse_mapping = Vec::new();
let mut counts = Vec::new();
@@ -281,7 +281,7 @@ impl Collector for FacetCollector {
// is positionned on a term that has not been processed yet.
let skip_result = skip(facet_streamer.key(), &mut collapse_facet_it);
match skip_result {
SkipResult::Found => {
SkipResult::Reached => {
// we reach a facet we decided to collapse.
let collapse_depth = facet_depth(facet_streamer.key());
let mut collapsed_id = 0;
@@ -301,7 +301,7 @@ impl Collector for FacetCollector {
}
break;
}
SkipResult::NotFound => {
SkipResult::End | SkipResult::OverStep => {
collapse_mapping.push(0);
if !facet_streamer.advance() {
break;
@@ -468,7 +468,7 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
let num_facets: usize = 3 * 4 * 5;
let facets: Vec<Facet> = (0..num_facets)
.map(|mut n| {
@@ -527,7 +527,7 @@ mod tests {
let facet_field = schema_builder.add_facet_field("facets");
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(
facet_field => Facet::from_text(&"/subjects/A/a"),
facet_field => Facet::from_text(&"/subjects/B/a"),
@@ -546,12 +546,12 @@ mod tests {
}
#[test]
fn test_doc_search_by_facet() -> crate::Result<()> {
fn test_doc_search_by_facet() {
let mut schema_builder = Schema::builder();
let facet_field = schema_builder.add_facet_field("facet");
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(
facet_field => Facet::from_text(&"/A/A"),
));
@@ -564,8 +564,8 @@ mod tests {
index_writer.add_document(doc!(
facet_field => Facet::from_text(&"/D/C/A"),
));
index_writer.commit()?;
let reader = index.reader()?;
index_writer.commit().unwrap();
let reader = index.reader().unwrap();
let searcher = reader.searcher();
assert_eq!(searcher.num_docs(), 4);
@@ -582,17 +582,17 @@ mod tests {
assert_eq!(count_facet("/A/C"), 1);
assert_eq!(count_facet("/A/C/A"), 1);
assert_eq!(count_facet("/C/A"), 0);
let query_parser = QueryParser::for_index(&index, vec![]);
{
let query = query_parser.parse_query("facet:/A/B")?;
assert_eq!(1, searcher.search(&query, &Count).unwrap());
let query_parser = QueryParser::for_index(&index, vec![]);
{
let query = query_parser.parse_query("facet:/A/B").unwrap();
assert_eq!(1, searcher.search(&query, &Count).unwrap());
}
{
let query = query_parser.parse_query("facet:/A").unwrap();
assert_eq!(3, searcher.search(&query, &Count).unwrap());
}
}
{
let query = query_parser.parse_query("facet:/A")?;
assert_eq!(3, searcher.search(&query, &Count)?);
}
Ok(())
}
#[test]
@@ -627,7 +627,7 @@ mod tests {
.collect();
docs[..].shuffle(&mut thread_rng());
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
for doc in docs {
index_writer.add_document(doc);
}
@@ -680,7 +680,7 @@ mod bench {
// 40425 docs
docs[..].shuffle(&mut thread_rng());
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
for doc in docs {
index_writer.add_document(doc);
}

127
src/collector/int_facet_collector.rs Normal file
View File

@@ -0,0 +1,127 @@
use std::cmp::Eq;
use std::collections::HashMap;
use std::hash::Hash;
use collector::Collector;
use fastfield::FastFieldReader;
use schema::Field;
use DocId;
use Result;
use Score;
use SegmentReader;
use SegmentLocalId;
/// Facet collector for i64/u64 fast field
pub struct IntFacetCollector<T>
where
T: FastFieldReader,
T::ValueType: Eq + Hash,
{
counters: HashMap<T::ValueType, u64>,
field: Field,
ff_reader: Option<T>,
}
impl<T> IntFacetCollector<T>
where
T: FastFieldReader,
T::ValueType: Eq + Hash,
{
/// Creates a new facet collector for aggregating a given field.
pub fn new(field: Field) -> IntFacetCollector<T> {
IntFacetCollector {
counters: HashMap::new(),
field: field,
ff_reader: None,
}
}
}
impl<T> Collector for IntFacetCollector<T>
where
T: FastFieldReader,
T::ValueType: Eq + Hash,
{
fn set_segment(&mut self, _: SegmentLocalId, reader: &SegmentReader) -> Result<()> {
self.ff_reader = Some(reader.get_fast_field_reader(self.field)?);
Ok(())
}
fn collect(&mut self, doc: DocId, _: Score) {
let val = self.ff_reader
.as_ref()
.expect(
"collect() was called before set_segment. \
This should never happen.",
)
.get(doc);
*(self.counters.entry(val).or_insert(0)) += 1;
}
}
#[cfg(test)]
mod tests {
use collector::{chain, IntFacetCollector};
use query::QueryParser;
use fastfield::{I64FastFieldReader, U64FastFieldReader};
use schema::{self, FAST, STRING};
use Index;
#[test]
// create 10 documents, set num field value to 0 or 1 for even/odd ones
// make sure we have facet counters correctly filled
fn test_facet_collector_results() {
let mut schema_builder = schema::Schema::builder();
let num_field_i64 = schema_builder.add_i64_field("num_i64", FAST);
let num_field_u64 = schema_builder.add_u64_field("num_u64", FAST);
let num_field_f64 = schema_builder.add_f64_field("num_f64", FAST);
let text_field = schema_builder.add_text_field("text", STRING);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema.clone());
{
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
{
for i in 0u64..10u64 {
index_writer.add_document(doc!(
num_field_i64 => ((i as i64) % 3i64) as i64,
num_field_u64 => (i % 2u64) as u64,
num_field_f64 => (i % 4u64) as f64,
text_field => "text"
));
}
}
assert_eq!(index_writer.commit().unwrap(), 10u64);
}
let searcher = index.reader().searcher();
let mut ffvf_i64: IntFacetCollector<I64FastFieldReader> = IntFacetCollector::new(num_field_i64);
let mut ffvf_u64: IntFacetCollector<U64FastFieldReader> = IntFacetCollector::new(num_field_u64);
let mut ffvf_f64: IntFacetCollector<F64FastFieldReader> = IntFacetCollector::new(num_field_f64);
{
// perform the query
let mut facet_collectors = chain().push(&mut ffvf_i64).push(&mut ffvf_u64).push(&mut ffvf_f64);
let mut query_parser = QueryParser::for_index(index, vec![text_field]);
let query = query_parser.parse_query("text:text").unwrap();
query.search(&searcher, &mut facet_collectors).unwrap();
}
assert_eq!(ffvf_u64.counters[&0], 5);
assert_eq!(ffvf_u64.counters[&1], 5);
assert_eq!(ffvf_i64.counters[&0], 4);
assert_eq!(ffvf_i64.counters[&1], 3);
assert_eq!(ffvf_f64.counters[&0.0], 3);
assert_eq!(ffvf_f64.counters[&2.0], 2);
}
}

50
src/collector/mod.rs
View File

@@ -109,10 +109,6 @@ pub use self::tweak_score_top_collector::{ScoreSegmentTweaker, ScoreTweaker};
mod facet_collector;
pub use self::facet_collector::FacetCollector;
use crate::query::Weight;
mod docset_collector;
pub use self::docset_collector::DocSetCollector;
/// `Fruit` is the type for the result of our collection.
/// e.g. `usize` for the `Count` collector.
@@ -136,13 +132,13 @@ impl<T> Fruit for T where T: Send + downcast_rs::Downcast {}
/// The collection logic itself is in the `SegmentCollector`.
///
/// Segments are not guaranteed to be visited in any specific order.
pub trait Collector: Sync + Send {
pub trait Collector: Sync {
/// `Fruit` is the type for the result of our collection.
/// e.g. `usize` for the `Count` collector.
type Fruit: Fruit;
/// Type of the `SegmentCollector` associated to this collector.
type Child: SegmentCollector;
type Child: SegmentCollector<Fruit = Self::Fruit>;
/// `set_segment` is called before beginning to enumerate
/// on this segment.
@@ -157,33 +153,7 @@ pub trait Collector: Sync + Send {
/// Combines the fruit associated to the collection of each segments
/// into one fruit.
fn merge_fruits(
&self,
segment_fruits: Vec<<Self::Child as SegmentCollector>::Fruit>,
) -> crate::Result<Self::Fruit>;
/// Created a segment collector and
fn collect_segment(
&self,
weight: &dyn Weight,
segment_ord: u32,
reader: &SegmentReader,
) -> crate::Result<<Self::Child as SegmentCollector>::Fruit> {
let mut segment_collector = self.for_segment(segment_ord as u32, reader)?;
if let Some(delete_bitset) = reader.delete_bitset() {
weight.for_each(reader, &mut |doc, score| {
if delete_bitset.is_alive(doc) {
segment_collector.collect(doc, score);
}
})?;
} else {
weight.for_each(reader, &mut |doc, score| {
segment_collector.collect(doc, score);
})?;
}
Ok(segment_collector.harvest())
}
fn merge_fruits(&self, segment_fruits: Vec<Self::Fruit>) -> crate::Result<Self::Fruit>;
}
/// The `SegmentCollector` is the trait in charge of defining the
@@ -230,11 +200,11 @@ where
fn merge_fruits(
&self,
segment_fruits: Vec<<Self::Child as SegmentCollector>::Fruit>,
children: Vec<(Left::Fruit, Right::Fruit)>,
) -> crate::Result<(Left::Fruit, Right::Fruit)> {
let mut left_fruits = vec![];
let mut right_fruits = vec![];
for (left_fruit, right_fruit) in segment_fruits {
for (left_fruit, right_fruit) in children {
left_fruits.push(left_fruit);
right_fruits.push(right_fruit);
}
@@ -288,10 +258,7 @@ where
self.0.requires_scoring() || self.1.requires_scoring() || self.2.requires_scoring()
}
fn merge_fruits(
&self,
children: Vec<<Self::Child as SegmentCollector>::Fruit>,
) -> crate::Result<Self::Fruit> {
fn merge_fruits(&self, children: Vec<Self::Fruit>) -> crate::Result<Self::Fruit> {
let mut one_fruits = vec![];
let mut two_fruits = vec![];
let mut three_fruits = vec![];
@@ -358,10 +325,7 @@ where
|| self.3.requires_scoring()
}
fn merge_fruits(
&self,
children: Vec<<Self::Child as SegmentCollector>::Fruit>,
) -> crate::Result<Self::Fruit> {
fn merge_fruits(&self, children: Vec<Self::Fruit>) -> crate::Result<Self::Fruit> {
let mut one_fruits = vec![];
let mut two_fruits = vec![];
let mut three_fruits = vec![];

14
src/collector/multi_collector.rs
View File

@@ -34,13 +34,13 @@ impl<TCollector: Collector> Collector for CollectorWrapper<TCollector> {
fn merge_fruits(
&self,
children: Vec<<Self::Child as SegmentCollector>::Fruit>,
children: Vec<<Self as Collector>::Fruit>,
) -> crate::Result<Box<dyn Fruit>> {
let typed_fruit: Vec<<TCollector::Child as SegmentCollector>::Fruit> = children
let typed_fruit: Vec<TCollector::Fruit> = children
.into_iter()
.map(|untyped_fruit| {
untyped_fruit
.downcast::<<TCollector::Child as SegmentCollector>::Fruit>()
.downcast::<TCollector::Fruit>()
.map(|boxed_but_typed| *boxed_but_typed)
.map_err(|_| {
TantivyError::InvalidArgument("Failed to cast child fruit.".to_string())
@@ -55,7 +55,7 @@ impl<TCollector: Collector> Collector for CollectorWrapper<TCollector> {
impl SegmentCollector for Box<dyn BoxableSegmentCollector> {
type Fruit = Box<dyn Fruit>;
fn collect(&mut self, doc: u32, score: Score) {
fn collect(&mut self, doc: u32, score: f32) {
self.as_mut().collect(doc, score);
}
@@ -65,7 +65,7 @@ impl SegmentCollector for Box<dyn BoxableSegmentCollector> {
}
pub trait BoxableSegmentCollector {
fn collect(&mut self, doc: u32, score: Score);
fn collect(&mut self, doc: u32, score: f32);
fn harvest_from_box(self: Box<Self>) -> Box<dyn Fruit>;
}
@@ -74,7 +74,7 @@ pub struct SegmentCollectorWrapper<TSegmentCollector: SegmentCollector>(TSegment
impl<TSegmentCollector: SegmentCollector> BoxableSegmentCollector
for SegmentCollectorWrapper<TSegmentCollector>
{
fn collect(&mut self, doc: u32, score: Score) {
fn collect(&mut self, doc: u32, score: f32) {
self.0.collect(doc, score);
}
@@ -259,7 +259,7 @@ mod tests {
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(text=>"abc"));
index_writer.add_document(doc!(text=>"abc abc abc"));
index_writer.add_document(doc!(text=>"abc abc"));

13
src/collector/tests.rs
View File

@@ -185,15 +185,12 @@ impl Collector for BytesFastFieldTestCollector {
_segment_local_id: u32,
segment_reader: &SegmentReader,
) -> crate::Result<BytesFastFieldSegmentCollector> {
let reader = segment_reader
.fast_fields()
.bytes(self.field)
.ok_or_else(|| {
crate::TantivyError::InvalidArgument("Field is not a bytes fast field.".to_string())
})?;
Ok(BytesFastFieldSegmentCollector {
vals: Vec::new(),
reader,
reader: segment_reader
.fast_fields()
.bytes(self.field)
.expect("Field is not a bytes fast field."),
})
}
@@ -209,7 +206,7 @@ impl Collector for BytesFastFieldTestCollector {
impl SegmentCollector for BytesFastFieldSegmentCollector {
type Fruit = Vec<u8>;
fn collect(&mut self, doc: u32, _score: Score) {
fn collect(&mut self, doc: u32, _score: f32) {
let data = self.reader.get_bytes(doc);
self.vals.extend(data);
}

86
src/collector/top_collector.rs
View File

@@ -18,9 +18,9 @@ use std::collections::BinaryHeap;
/// Two elements are equal if their feature is equal, and regardless of whether `doc`
/// is equal. This should be perfectly fine for this usage, but let's make sure this
/// struct is never public.
pub(crate) struct ComparableDoc<T, D> {
pub feature: T,
pub doc: D,
struct ComparableDoc<T, D> {
feature: T,
doc: D,
}
impl<T: PartialOrd, D: PartialOrd> PartialOrd for ComparableDoc<T, D> {
@@ -56,8 +56,7 @@ impl<T: PartialOrd, D: PartialOrd> PartialEq for ComparableDoc<T, D> {
impl<T: PartialOrd, D: PartialOrd> Eq for ComparableDoc<T, D> {}
pub(crate) struct TopCollector<T> {
pub limit: usize,
pub offset: usize,
limit: usize,
_marker: PhantomData<T>,
}
@@ -73,20 +72,14 @@ where
if limit < 1 {
panic!("Limit must be strictly greater than 0.");
}
Self {
TopCollector {
limit,
offset: 0,
_marker: PhantomData,
}
}
/// Skip the first "offset" documents when collecting.
///
/// This is equivalent to `OFFSET` in MySQL or PostgreSQL and `start` in
/// Lucene's TopDocsCollector.
pub fn and_offset(mut self, offset: usize) -> TopCollector<T> {
self.offset = offset;
self
pub fn limit(&self) -> usize {
self.limit
}
pub fn merge_fruits(
@@ -99,7 +92,7 @@ where
let mut top_collector = BinaryHeap::new();
for child_fruit in children {
for (feature, doc) in child_fruit {
if top_collector.len() < (self.limit + self.offset) {
if top_collector.len() < self.limit {
top_collector.push(ComparableDoc { feature, doc });
} else if let Some(mut head) = top_collector.peek_mut() {
if head.feature < feature {
@@ -111,7 +104,6 @@ where
Ok(top_collector
.into_sorted_vec()
.into_iter()
.skip(self.offset)
.map(|cdoc| (cdoc.feature, cdoc.doc))
.collect())
}
@@ -121,23 +113,7 @@ where
segment_id: SegmentLocalId,
_: &SegmentReader,
) -> crate::Result<TopSegmentCollector<F>> {
Ok(TopSegmentCollector::new(
segment_id,
self.limit + self.offset,
))
}
/// Create a new TopCollector with the same limit and offset.
///
/// Ideally we would use Into but the blanket implementation seems to cause the Scorer traits
/// to fail.
#[doc(hidden)]
pub(crate) fn into_tscore<TScore: PartialOrd + Clone>(self) -> TopCollector<TScore> {
TopCollector {
limit: self.limit,
offset: self.offset,
_marker: PhantomData,
}
Ok(TopSegmentCollector::new(segment_id, self.limit))
}
}
@@ -211,7 +187,7 @@ impl<T: PartialOrd + Clone> TopSegmentCollector<T> {
#[cfg(test)]
mod tests {
use super::{TopCollector, TopSegmentCollector};
use super::TopSegmentCollector;
use crate::DocAddress;
#[test]
@@ -272,48 +248,6 @@ mod tests {
top_collector_limit_3.harvest()[..2].to_vec(),
);
}
#[test]
fn test_top_collector_with_limit_and_offset() {
let collector = TopCollector::with_limit(2).and_offset(1);
let results = collector
.merge_fruits(vec![vec![
(0.9, DocAddress(0, 1)),
(0.8, DocAddress(0, 2)),
(0.7, DocAddress(0, 3)),
(0.6, DocAddress(0, 4)),
(0.5, DocAddress(0, 5)),
]])
.unwrap();
assert_eq!(
results,
vec![(0.8, DocAddress(0, 2)), (0.7, DocAddress(0, 3)),]
);
}
#[test]
fn test_top_collector_with_limit_larger_than_set_and_offset() {
let collector = TopCollector::with_limit(2).and_offset(1);
let results = collector
.merge_fruits(vec![vec![(0.9, DocAddress(0, 1)), (0.8, DocAddress(0, 2))]])
.unwrap();
assert_eq!(results, vec![(0.8, DocAddress(0, 2)),]);
}
#[test]
fn test_top_collector_with_limit_and_offset_larger_than_set() {
let collector = TopCollector::with_limit(2).and_offset(20);
let results = collector
.merge_fruits(vec![vec![(0.9, DocAddress(0, 1)), (0.8, DocAddress(0, 2))]])
.unwrap();
assert_eq!(results, vec![]);
}
}
#[cfg(all(test, feature = "unstable"))]

577
src/collector/top_score_collector.rs
View File

@@ -1,82 +1,19 @@
use super::Collector;
use crate::collector::top_collector::{ComparableDoc, TopCollector};
use crate::collector::custom_score_top_collector::CustomScoreTopCollector;
use crate::collector::top_collector::TopCollector;
use crate::collector::top_collector::TopSegmentCollector;
use crate::collector::tweak_score_top_collector::TweakedScoreTopCollector;
use crate::collector::{
CustomScorer, CustomSegmentScorer, ScoreSegmentTweaker, ScoreTweaker, SegmentCollector,
};
use crate::fastfield::FastFieldReader;
use crate::query::Weight;
use crate::schema::Field;
use crate::DocAddress;
use crate::DocId;
use crate::Score;
use crate::SegmentLocalId;
use crate::SegmentReader;
use crate::{collector::custom_score_top_collector::CustomScoreTopCollector, fastfield::FastValue};
use crate::{collector::top_collector::TopSegmentCollector, TantivyError};
use std::fmt;
use std::{collections::BinaryHeap, marker::PhantomData};
struct FastFieldConvertCollector<
TCollector: Collector<Fruit = Vec<(u64, DocAddress)>>,
TFastValue: FastValue,
> {
pub collector: TCollector,
pub field: Field,
pub fast_value: std::marker::PhantomData<TFastValue>,
}
impl<TCollector, TFastValue> Collector for FastFieldConvertCollector<TCollector, TFastValue>
where
TCollector: Collector<Fruit = Vec<(u64, DocAddress)>>,
TFastValue: FastValue + 'static,
{
type Fruit = Vec<(TFastValue, DocAddress)>;
type Child = TCollector::Child;
fn for_segment(
&self,
segment_local_id: crate::SegmentLocalId,
segment: &SegmentReader,
) -> crate::Result<Self::Child> {
let schema = segment.schema();
let field_entry = schema.get_field_entry(self.field);
if !field_entry.is_fast() {
return Err(TantivyError::SchemaError(format!(
"Field {:?} is not a fast field.",
field_entry.name()
)));
}
let schema_type = TFastValue::to_type();
let requested_type = field_entry.field_type().value_type();
if schema_type != requested_type {
return Err(TantivyError::SchemaError(format!(
"Field {:?} is of type {:?}!={:?}",
field_entry.name(),
schema_type,
requested_type
)));
}
self.collector.for_segment(segment_local_id, segment)
}
fn requires_scoring(&self) -> bool {
self.collector.requires_scoring()
}
fn merge_fruits(
&self,
segment_fruits: Vec<<Self::Child as SegmentCollector>::Fruit>,
) -> crate::Result<Self::Fruit> {
let raw_result = self.collector.merge_fruits(segment_fruits)?;
let transformed_result = raw_result
.into_iter()
.map(|(score, doc_address)| (TFastValue::from_u64(score), doc_address))
.collect::<Vec<_>>();
Ok(transformed_result)
}
}
/// The `TopDocs` collector keeps track of the top `K` documents
/// sorted by their score.
@@ -99,7 +36,7 @@ where
/// let schema = schema_builder.build();
/// let index = Index::create_in_ram(schema);
///
/// let mut index_writer = index.writer_with_num_threads(1, 10_000_000).unwrap();
/// let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
/// index_writer.add_document(doc!(title => "The Name of the Wind"));
/// index_writer.add_document(doc!(title => "The Diary of Muadib"));
/// index_writer.add_document(doc!(title => "A Dairy Cow"));
@@ -113,18 +50,14 @@ where
/// let query = query_parser.parse_query("diary").unwrap();
/// let top_docs = searcher.search(&query, &TopDocs::with_limit(2)).unwrap();
///
/// assert_eq!(top_docs[0].1, DocAddress(0, 1));
/// assert_eq!(top_docs[1].1, DocAddress(0, 3));
/// assert_eq!(&top_docs[0], &(0.7261542, DocAddress(0, 1)));
/// assert_eq!(&top_docs[1], &(0.6099695, DocAddress(0, 3)));
/// ```
pub struct TopDocs(TopCollector<Score>);
impl fmt::Debug for TopDocs {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
write!(
f,
"TopDocs(limit={}, offset={})",
self.0.limit, self.0.offset
)
write!(f, "TopDocs({})", self.0.limit())
}
}
@@ -133,8 +66,8 @@ struct ScorerByFastFieldReader {
}
impl CustomSegmentScorer<u64> for ScorerByFastFieldReader {
fn score(&mut self, doc: DocId) -> u64 {
self.ff_reader.get(doc)
fn score(&self, doc: DocId) -> u64 {
self.ff_reader.get_u64(u64::from(doc))
}
}
@@ -148,10 +81,10 @@ impl CustomScorer<u64> for ScorerByField {
fn segment_scorer(&self, segment_reader: &SegmentReader) -> crate::Result<Self::Child> {
let ff_reader = segment_reader
.fast_fields()
.u64_lenient(self.field)
.u64(self.field)
.ok_or_else(|| {
crate::TantivyError::SchemaError(format!(
"Field requested ({:?}) is not a fast field.",
"Field requested ({:?}) is not a i64/u64 fast field.",
self.field
))
})?;
@@ -168,57 +101,8 @@ impl TopDocs {
TopDocs(TopCollector::with_limit(limit))
}
/// Skip the first "offset" documents when collecting.
///
/// This is equivalent to `OFFSET` in MySQL or PostgreSQL and `start` in
/// Lucene's TopDocsCollector.
///
/// # Example
///
/// ```rust
/// use tantivy::collector::TopDocs;
/// use tantivy::query::QueryParser;
/// use tantivy::schema::{Schema, TEXT};
/// use tantivy::{doc, DocAddress, Index};
///
/// let mut schema_builder = Schema::builder();
/// let title = schema_builder.add_text_field("title", TEXT);
/// let schema = schema_builder.build();
/// let index = Index::create_in_ram(schema);
///
/// let mut index_writer = index.writer_with_num_threads(1, 10_000_000).unwrap();
/// index_writer.add_document(doc!(title => "The Name of the Wind"));
/// index_writer.add_document(doc!(title => "The Diary of Muadib"));
/// index_writer.add_document(doc!(title => "A Dairy Cow"));
/// index_writer.add_document(doc!(title => "The Diary of a Young Girl"));
/// index_writer.add_document(doc!(title => "The Diary of Lena Mukhina"));
/// assert!(index_writer.commit().is_ok());
///
/// let reader = index.reader().unwrap();
/// let searcher = reader.searcher();
///
/// let query_parser = QueryParser::for_index(&index, vec![title]);
/// let query = query_parser.parse_query("diary").unwrap();
/// let top_docs = searcher.search(&query, &TopDocs::with_limit(2).and_offset(1)).unwrap();
///
/// assert_eq!(top_docs.len(), 2);
/// assert_eq!(top_docs[0].1, DocAddress(0, 4));
/// assert_eq!(top_docs[1].1, DocAddress(0, 3));
/// ```
pub fn and_offset(self, offset: usize) -> TopDocs {
TopDocs(self.0.and_offset(offset))
}
/// Set top-K to rank documents by a given fast field.
///
/// If the field is not a fast or does not exist, this method returns successfully (it is not aware of any schema).
/// An error will be returned at the moment of search.
///
/// If the field is a FAST field but not a u64 field, search will return successfully but it will return
/// returns a monotonic u64-representation (ie. the order is still correct) of the requested field type.
///
/// # Example
///
/// ```rust
/// # use tantivy::schema::{Schema, FAST, TEXT};
/// # use tantivy::{doc, Index, DocAddress};
@@ -234,13 +118,13 @@ impl TopDocs {
/// # let schema = schema_builder.build();
/// #
/// # let index = Index::create_in_ram(schema);
/// # let mut index_writer = index.writer_with_num_threads(1, 10_000_000)?;
/// # let mut index_writer = index.writer_with_num_threads(1, 3_000_000)?;
/// # index_writer.add_document(doc!(title => "The Name of the Wind", rating => 92u64));
/// # index_writer.add_document(doc!(title => "The Diary of Muadib", rating => 97u64));
/// # index_writer.add_document(doc!(title => "A Dairy Cow", rating => 63u64));
/// # index_writer.add_document(doc!(title => "The Diary of a Young Girl", rating => 80u64));
/// # assert!(index_writer.commit().is_ok());
/// # let reader = index.reader()?;
/// # let reader = index.reader().unwrap();
/// # let query = QueryParser::for_index(&index, vec![title]).parse_query("diary")?;
/// # let top_docs = docs_sorted_by_rating(&reader.searcher(), &query, rating)?;
/// # assert_eq!(top_docs,
@@ -248,20 +132,25 @@ impl TopDocs {
/// # (80u64, DocAddress(0u32, 3))]);
/// # Ok(())
/// # }
///
///
/// /// Searches the document matching the given query, and
/// /// collects the top 10 documents, order by the u64-`field`
/// /// given in argument.
/// ///
/// /// `field` is required to be a FAST field.
/// fn docs_sorted_by_rating(searcher: &Searcher,
/// query: &dyn Query,
/// rating_field: Field)
/// sort_by_field: Field)
/// -> tantivy::Result<Vec<(u64, DocAddress)>> {
///
/// // This is where we build our topdocs collector
/// //
/// // Note the `rating_field` needs to be a FAST field here.
/// let top_books_by_rating = TopDocs
/// // Note the generics parameter that needs to match the
/// // type `sort_by_field`.
/// let top_docs_by_rating = TopDocs
/// ::with_limit(10)
/// .order_by_u64_field(rating_field);
/// .order_by_u64_field(sort_by_field);
///
/// // ... and here are our documents. Note this is a simple vec.
/// // The `u64` in the pair is the value of our fast field for
@@ -271,105 +160,21 @@ impl TopDocs {
/// // length of 10, or less if not enough documents matched the
/// // query.
/// let resulting_docs: Vec<(u64, DocAddress)> =
/// searcher.search(query, &top_books_by_rating)?;
/// searcher.search(query, &top_docs_by_rating)?;
///
/// Ok(resulting_docs)
/// }
/// ```
///
/// # See also
///
/// To confortably work with `u64`s, `i64`s, `f64`s, or `date`s, please refer to
/// [.order_by_fast_field(...)](#method.order_by_fast_field) method.
/// # Panics
///
/// May panic if the field requested is not a fast field.
///
pub fn order_by_u64_field(
self,
field: Field,
) -> impl Collector<Fruit = Vec<(u64, DocAddress)>> {
CustomScoreTopCollector::new(ScorerByField { field }, self.0.into_tscore())
}
/// Set top-K to rank documents by a given fast field.
///
/// If the field is not a fast field, or its field type does not match the generic type, this method does not panic,
/// but an explicit error will be returned at the moment of collection.
///
/// Note that this method is a generic. The requested fast field type will be often
/// inferred in your code by the rust compiler.
///
/// Implementation-wise, for performance reason, tantivy will manipulate the u64 representation of your fast
/// field until the last moment.
///
/// # Example
///
/// ```rust
/// # use tantivy::schema::{Schema, FAST, TEXT};
/// # use tantivy::{doc, Index, DocAddress};
/// # use tantivy::query::{Query, AllQuery};
/// use tantivy::Searcher;
/// use tantivy::collector::TopDocs;
/// use tantivy::schema::Field;
///
/// # fn main() -> tantivy::Result<()> {
/// # let mut schema_builder = Schema::builder();
/// # let title = schema_builder.add_text_field("company", TEXT);
/// # let rating = schema_builder.add_i64_field("revenue", FAST);
/// # let schema = schema_builder.build();
/// #
/// # let index = Index::create_in_ram(schema);
/// # let mut index_writer = index.writer_with_num_threads(1, 10_000_000)?;
/// # index_writer.add_document(doc!(title => "MadCow Inc.", rating => 92_000_000i64));
/// # index_writer.add_document(doc!(title => "Zozo Cow KKK", rating => 119_000_000i64));
/// # index_writer.add_document(doc!(title => "Declining Cow", rating => -63_000_000i64));
/// # assert!(index_writer.commit().is_ok());
/// # let reader = index.reader()?;
/// # let top_docs = docs_sorted_by_revenue(&reader.searcher(), &AllQuery, rating)?;
/// # assert_eq!(top_docs,
/// # vec![(119_000_000i64, DocAddress(0, 1)),
/// # (92_000_000i64, DocAddress(0, 0))]);
/// # Ok(())
/// # }
/// /// Searches the document matching the given query, and
/// /// collects the top 10 documents, order by the u64-`field`
/// /// given in argument.
/// fn docs_sorted_by_revenue(searcher: &Searcher,
/// query: &dyn Query,
/// revenue_field: Field)
/// -> tantivy::Result<Vec<(i64, DocAddress)>> {
///
/// // This is where we build our topdocs collector
/// //
/// // Note the generics parameter that needs to match the
/// // type `sort_by_field`. revenue_field here is a FAST i64 field.
/// let top_company_by_revenue = TopDocs
/// ::with_limit(2)
/// .order_by_fast_field(revenue_field);
///
/// // ... and here are our documents. Note this is a simple vec.
/// // The `i64` in the pair is the value of our fast field for
/// // each documents.
/// //
/// // The vec is sorted decreasingly by `sort_by_field`, and has a
/// // length of 10, or less if not enough documents matched the
/// // query.
/// let resulting_docs: Vec<(i64, DocAddress)> =
/// searcher.search(query, &top_company_by_revenue)?;
///
/// Ok(resulting_docs)
/// }
/// ```
pub fn order_by_fast_field<TFastValue>(
self,
fast_field: Field,
) -> impl Collector<Fruit = Vec<(TFastValue, DocAddress)>>
where
TFastValue: FastValue + 'static,
{
let u64_collector = self.order_by_u64_field(fast_field);
FastFieldConvertCollector {
collector: u64_collector,
field: fast_field,
fast_value: PhantomData,
}
self.custom_score(ScorerByField { field })
}
/// Ranks the documents using a custom score.
@@ -414,7 +219,7 @@ impl TopDocs {
/// fn create_index() -> tantivy::Result<Index> {
/// let schema = create_schema();
/// let index = Index::create_in_ram(schema);
/// let mut index_writer = index.writer_with_num_threads(1, 10_000_000)?;
/// let mut index_writer = index.writer_with_num_threads(1, 3_000_000)?;
/// let product_name = index.schema().get_field("product_name").unwrap();
/// let popularity: Field = index.schema().get_field("popularity").unwrap();
/// index_writer.add_document(doc!(product_name => "The Diary of Muadib", popularity => 1u64));
@@ -453,7 +258,7 @@ impl TopDocs {
/// let popularity: u64 = popularity_reader.get(doc);
/// // Well.. For the sake of the example we use a simple logarithm
/// // function.
/// let popularity_boost_score = ((2u64 + popularity) as Score).log2();
/// let popularity_boost_score = ((2u64 + popularity) as f32).log2();
/// popularity_boost_score * original_score
/// }
/// });
@@ -474,9 +279,9 @@ impl TopDocs {
where
TScore: 'static + Send + Sync + Clone + PartialOrd,
TScoreSegmentTweaker: ScoreSegmentTweaker<TScore> + 'static,
TScoreTweaker: ScoreTweaker<TScore, Child = TScoreSegmentTweaker> + Send + Sync,
TScoreTweaker: ScoreTweaker<TScore, Child = TScoreSegmentTweaker>,
{
TweakedScoreTopCollector::new(score_tweaker, self.0.into_tscore())
TweakedScoreTopCollector::new(score_tweaker, self.0.limit())
}
/// Ranks the documents using a custom score.
@@ -521,7 +326,7 @@ impl TopDocs {
/// # fn main() -> tantivy::Result<()> {
/// # let schema = create_schema();
/// # let index = Index::create_in_ram(schema);
/// # let mut index_writer = index.writer_with_num_threads(1, 10_000_000)?;
/// # let mut index_writer = index.writer_with_num_threads(1, 3_000_000)?;
/// # let product_name = index.schema().get_field("product_name").unwrap();
/// #
/// let popularity: Field = index.schema().get_field("popularity").unwrap();
@@ -588,9 +393,9 @@ impl TopDocs {
where
TScore: 'static + Send + Sync + Clone + PartialOrd,
TCustomSegmentScorer: CustomSegmentScorer<TScore> + 'static,
TCustomScorer: CustomScorer<TScore, Child = TCustomSegmentScorer> + Send + Sync,
TCustomScorer: CustomScorer<TScore, Child = TCustomSegmentScorer>,
{
CustomScoreTopCollector::new(custom_score, self.0.into_tscore())
CustomScoreTopCollector::new(custom_score, self.0.limit())
}
}
@@ -618,64 +423,6 @@ impl Collector for TopDocs {
) -> crate::Result<Self::Fruit> {
self.0.merge_fruits(child_fruits)
}
fn collect_segment(
&self,
weight: &dyn Weight,
segment_ord: u32,
reader: &SegmentReader,
) -> crate::Result<<Self::Child as SegmentCollector>::Fruit> {
let heap_len = self.0.limit + self.0.offset;
let mut heap: BinaryHeap<ComparableDoc<Score, DocId>> = BinaryHeap::with_capacity(heap_len);
if let Some(delete_bitset) = reader.delete_bitset() {
let mut threshold = Score::MIN;
weight.for_each_pruning(threshold, reader, &mut |doc, score| {
if delete_bitset.is_deleted(doc) {
return threshold;
}
let heap_item = ComparableDoc {
feature: score,
doc,
};
if heap.len() < heap_len {
heap.push(heap_item);
if heap.len() == heap_len {
threshold = heap.peek().map(|el| el.feature).unwrap_or(Score::MIN);
}
return threshold;
}
*heap.peek_mut().unwrap() = heap_item;
threshold = heap.peek().map(|el| el.feature).unwrap_or(Score::MIN);
threshold
})?;
} else {
weight.for_each_pruning(Score::MIN, reader, &mut |doc, score| {
let heap_item = ComparableDoc {
feature: score,
doc,
};
if heap.len() < heap_len {
heap.push(heap_item);
// TODO the threshold is suboptimal for heap.len == heap_len
if heap.len() == heap_len {
return heap.peek().map(|el| el.feature).unwrap_or(Score::MIN);
} else {
return Score::MIN;
}
}
*heap.peek_mut().unwrap() = heap_item;
heap.peek().map(|el| el.feature).unwrap_or(Score::MIN)
})?;
}
let fruit = heap
.into_sorted_vec()
.into_iter()
.map(|cid| (cid.feature, DocAddress(segment_ord, cid.doc)))
.collect();
Ok(fruit)
}
}
/// Segment Collector associated to `TopDocs`.
@@ -685,7 +432,7 @@ impl SegmentCollector for TopScoreSegmentCollector {
type Fruit = Vec<(Score, DocAddress)>;
fn collect(&mut self, doc: DocId, score: Score) {
self.0.collect(doc, score);
self.0.collect(doc, score)
}
fn harvest(self) -> Vec<(Score, DocAddress)> {
@@ -699,10 +446,10 @@ mod tests {
use crate::collector::Collector;
use crate::query::{AllQuery, Query, QueryParser};
use crate::schema::{Field, Schema, FAST, STORED, TEXT};
use crate::DocAddress;
use crate::Index;
use crate::IndexWriter;
use crate::Score;
use crate::{DocAddress, DocId, SegmentReader};
fn make_index() -> Index {
let mut schema_builder = Schema::builder();
@@ -711,7 +458,7 @@ mod tests {
let index = Index::create_in_ram(schema);
{
// writing the segment
let mut index_writer = index.writer_with_num_threads(1, 10_000_000).unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(text_field=>"Hello happy tax payer."));
index_writer.add_document(doc!(text_field=>"Droopy says hello happy tax payer"));
index_writer.add_document(doc!(text_field=>"I like Droopy"));
@@ -720,13 +467,6 @@ mod tests {
index
}
fn assert_results_equals(results: &[(Score, DocAddress)], expected: &[(Score, DocAddress)]) {
for (result, expected) in results.iter().zip(expected.iter()) {
assert_eq!(result.1, expected.1);
crate::assert_nearly_equals!(result.0, expected.0);
}
}
#[test]
fn test_top_collector_not_at_capacity() {
let index = make_index();
@@ -739,31 +479,16 @@ mod tests {
.searcher()
.search(&text_query, &TopDocs::with_limit(4))
.unwrap();
assert_results_equals(
&score_docs,
&[
assert_eq!(
score_docs,
vec![
(0.81221175, DocAddress(0u32, 1)),
(0.5376842, DocAddress(0u32, 2)),
(0.48527452, DocAddress(0, 0)),
],
(0.48527452, DocAddress(0, 0))
]
);
}
#[test]
fn test_top_collector_not_at_capacity_with_offset() {
let index = make_index();
let field = index.schema().get_field("text").unwrap();
let query_parser = QueryParser::for_index(&index, vec![field]);
let text_query = query_parser.parse_query("droopy tax").unwrap();
let score_docs: Vec<(Score, DocAddress)> = index
.reader()
.unwrap()
.searcher()
.search(&text_query, &TopDocs::with_limit(4).and_offset(2))
.unwrap();
assert_results_equals(&score_docs[..], &[(0.48527452, DocAddress(0, 0))]);
}
#[test]
fn test_top_collector_at_capacity() {
let index = make_index();
@@ -776,33 +501,12 @@ mod tests {
.searcher()
.search(&text_query, &TopDocs::with_limit(2))
.unwrap();
assert_results_equals(
&score_docs,
&[
assert_eq!(
score_docs,
vec![
(0.81221175, DocAddress(0u32, 1)),
(0.5376842, DocAddress(0u32, 2)),
],
);
}
#[test]
fn test_top_collector_at_capacity_with_offset() {
let index = make_index();
let field = index.schema().get_field("text").unwrap();
let query_parser = QueryParser::for_index(&index, vec![field]);
let text_query = query_parser.parse_query("droopy tax").unwrap();
let score_docs: Vec<(Score, DocAddress)> = index
.reader()
.unwrap()
.searcher()
.search(&text_query, &TopDocs::with_limit(2).and_offset(1))
.unwrap();
assert_results_equals(
&score_docs[..],
&[
(0.5376842, DocAddress(0u32, 2)),
(0.48527452, DocAddress(0, 0)),
],
]
);
}
@@ -863,8 +567,8 @@ mod tests {
let top_collector = TopDocs::with_limit(4).order_by_u64_field(size);
let top_docs: Vec<(u64, DocAddress)> = searcher.search(&query, &top_collector).unwrap();
assert_eq!(
&top_docs[..],
&[
top_docs,
vec![
(64, DocAddress(0, 1)),
(16, DocAddress(0, 2)),
(12, DocAddress(0, 0))
@@ -872,94 +576,6 @@ mod tests {
);
}
#[test]
fn test_top_field_collector_datetime() -> crate::Result<()> {
use std::str::FromStr;
let mut schema_builder = Schema::builder();
let name = schema_builder.add_text_field("name", TEXT);
let birthday = schema_builder.add_date_field("birthday", FAST);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
let pr_birthday = crate::DateTime::from_str("1898-04-09T00:00:00+00:00")?;
index_writer.add_document(doc!(
name => "Paul Robeson",
birthday => pr_birthday
));
let mr_birthday = crate::DateTime::from_str("1947-11-08T00:00:00+00:00")?;
index_writer.add_document(doc!(
name => "Minnie Riperton",
birthday => mr_birthday
));
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let top_collector = TopDocs::with_limit(3).order_by_fast_field(birthday);
let top_docs: Vec<(crate::DateTime, DocAddress)> =
searcher.search(&AllQuery, &top_collector)?;
assert_eq!(
&top_docs[..],
&[
(mr_birthday, DocAddress(0, 1)),
(pr_birthday, DocAddress(0, 0)),
]
);
Ok(())
}
#[test]
fn test_top_field_collector_i64() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let city = schema_builder.add_text_field("city", TEXT);
let altitude = schema_builder.add_i64_field("altitude", FAST);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(
city => "georgetown",
altitude => -1i64,
));
index_writer.add_document(doc!(
city => "tokyo",
altitude => 40i64,
));
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let top_collector = TopDocs::with_limit(3).order_by_fast_field(altitude);
let top_docs: Vec<(i64, DocAddress)> = searcher.search(&AllQuery, &top_collector)?;
assert_eq!(
&top_docs[..],
&[(40i64, DocAddress(0, 1)), (-1i64, DocAddress(0, 0)),]
);
Ok(())
}
#[test]
fn test_top_field_collector_f64() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let city = schema_builder.add_text_field("city", TEXT);
let altitude = schema_builder.add_f64_field("altitude", FAST);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(
city => "georgetown",
altitude => -1.0f64,
));
index_writer.add_document(doc!(
city => "tokyo",
altitude => 40f64,
));
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let top_collector = TopDocs::with_limit(3).order_by_fast_field(altitude);
let top_docs: Vec<(f64, DocAddress)> = searcher.search(&AllQuery, &top_collector)?;
assert_eq!(
&top_docs[..],
&[(40f64, DocAddress(0, 1)), (-1.0f64, DocAddress(0, 0)),]
);
Ok(())
}
#[test]
#[should_panic]
fn test_field_does_not_exist() {
@@ -982,85 +598,29 @@ mod tests {
}
#[test]
fn test_field_not_fast_field() -> crate::Result<()> {
fn test_field_not_fast_field() {
let mut schema_builder = Schema::builder();
let title = schema_builder.add_text_field(TITLE, TEXT);
let size = schema_builder.add_u64_field(SIZE, STORED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(size=>1u64));
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let (index, _) = index("beer", title, schema, |index_writer| {
index_writer.add_document(doc!(
title => "bottle of beer",
size => 12u64,
));
});
let searcher = index.reader().unwrap().searcher();
let segment = searcher.segment_reader(0);
let top_collector = TopDocs::with_limit(4).order_by_u64_field(size);
let err = top_collector.for_segment(0, segment).err().unwrap();
assert!(
matches!(err, crate::TantivyError::SchemaError(msg) if msg == "Field requested (Field(0)) is not a fast field.")
);
Ok(())
}
#[test]
fn test_field_wrong_type() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let size = schema_builder.add_u64_field(SIZE, STORED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(size=>1u64));
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let segment = searcher.segment_reader(0);
let top_collector = TopDocs::with_limit(4).order_by_fast_field::<i64>(size);
let err = top_collector.for_segment(0, segment).err().unwrap();
assert!(
matches!(err, crate::TantivyError::SchemaError(msg) if msg == "Field \"size\" is not a fast field.")
);
Ok(())
}
#[test]
fn test_tweak_score_top_collector_with_offset() {
let index = make_index();
let field = index.schema().get_field("text").unwrap();
let query_parser = QueryParser::for_index(&index, vec![field]);
let text_query = query_parser.parse_query("droopy tax").unwrap();
let collector = TopDocs::with_limit(2).and_offset(1).tweak_score(
move |_segment_reader: &SegmentReader| move |doc: DocId, _original_score: Score| doc,
);
let score_docs: Vec<(u32, DocAddress)> = index
.reader()
.unwrap()
.searcher()
.search(&text_query, &collector)
.unwrap();
assert_eq!(
score_docs,
vec![(1, DocAddress(0, 1)), (0, DocAddress(0, 0)),]
);
}
#[test]
fn test_custom_score_top_collector_with_offset() {
let index = make_index();
let field = index.schema().get_field("text").unwrap();
let query_parser = QueryParser::for_index(&index, vec![field]);
let text_query = query_parser.parse_query("droopy tax").unwrap();
let collector = TopDocs::with_limit(2)
.and_offset(1)
.custom_score(move |_segment_reader: &SegmentReader| move |doc: DocId| doc);
let score_docs: Vec<(u32, DocAddress)> = index
.reader()
.unwrap()
.searcher()
.search(&text_query, &collector)
.unwrap();
assert_eq!(
score_docs,
vec![(1, DocAddress(0, 1)), (0, DocAddress(0, 0)),]
);
let err = top_collector.for_segment(0, segment);
if let Err(crate::TantivyError::SchemaError(msg)) = err {
assert_eq!(
msg,
"Field requested (Field(1)) is not a i64/u64 fast field."
);
} else {
assert!(false);
}
}
fn index(
@@ -1070,7 +630,8 @@ mod tests {
mut doc_adder: impl FnMut(&mut IndexWriter) -> (),
) -> (Index, Box<dyn Query>) {
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_with_num_threads(1, 10_000_000).unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
doc_adder(&mut index_writer);
index_writer.commit().unwrap();
let query_parser = QueryParser::for_index(&index, vec![query_field]);

12
src/collector/tweak_score_top_collector.rs
View File

@@ -14,11 +14,11 @@ where
{
pub fn new(
score_tweaker: TScoreTweaker,
collector: TopCollector<TScore>,
limit: usize,
) -> TweakedScoreTopCollector<TScoreTweaker, TScore> {
TweakedScoreTopCollector {
score_tweaker,
collector,
collector: TopCollector::with_limit(limit),
}
}
}
@@ -29,7 +29,7 @@ where
/// It is the segment local version of the [`ScoreTweaker`](./trait.ScoreTweaker.html).
pub trait ScoreSegmentTweaker<TScore>: 'static {
/// Tweak the given `score` for the document `doc`.
fn score(&mut self, doc: DocId, score: Score) -> TScore;
fn score(&self, doc: DocId, score: Score) -> TScore;
}
/// `ScoreTweaker` makes it possible to tweak the score
@@ -49,7 +49,7 @@ pub trait ScoreTweaker<TScore>: Sync {
impl<TScoreTweaker, TScore> Collector for TweakedScoreTopCollector<TScoreTweaker, TScore>
where
TScoreTweaker: ScoreTweaker<TScore> + Send + Sync,
TScoreTweaker: ScoreTweaker<TScore>,
TScore: 'static + PartialOrd + Clone + Send + Sync,
{
type Fruit = Vec<(TScore, DocAddress)>;
@@ -121,9 +121,9 @@ where
impl<F, TScore> ScoreSegmentTweaker<TScore> for F
where
F: 'static + FnMut(DocId, Score) -> TScore,
F: 'static + Sync + Send + Fn(DocId, Score) -> TScore,
{
fn score(&mut self, doc: DocId, score: Score) -> TScore {
fn score(&self, doc: DocId, score: Score) -> TScore {
(self)(doc, score)
}
}

24
src/common/bitpacker.rs
View File

@@ -1,7 +1,6 @@
use byteorder::{ByteOrder, LittleEndian, WriteBytesExt};
use std::io;
use crate::directory::OwnedBytes;
use std::ops::Deref;
pub(crate) struct BitPacker {
mini_buffer: u64,
@@ -61,14 +60,20 @@ impl BitPacker {
}
#[derive(Clone)]
pub struct BitUnpacker {
pub struct BitUnpacker<Data>
where
Data: Deref<Target = [u8]>,
{
num_bits: u64,
mask: u64,
data: OwnedBytes,
data: Data,
}
impl BitUnpacker {
pub fn new(data: OwnedBytes, num_bits: u8) -> BitUnpacker {
impl<Data> BitUnpacker<Data>
where
Data: Deref<Target = [u8]>,
{
pub fn new(data: Data, num_bits: u8) -> BitUnpacker<Data> {
let mask: u64 = if num_bits == 64 {
!0u64
} else {
@@ -85,7 +90,7 @@ impl BitUnpacker {
if self.num_bits == 0 {
return 0u64;
}
let data: &[u8] = self.data.as_slice();
let data: &[u8] = &*self.data;
let num_bits = self.num_bits;
let mask = self.mask;
let addr_in_bits = idx * num_bits;
@@ -104,9 +109,8 @@ impl BitUnpacker {
#[cfg(test)]
mod test {
use super::{BitPacker, BitUnpacker};
use crate::directory::OwnedBytes;
fn create_fastfield_bitpacker(len: usize, num_bits: u8) -> (BitUnpacker, Vec<u64>) {
fn create_fastfield_bitpacker(len: usize, num_bits: u8) -> (BitUnpacker<Vec<u8>>, Vec<u64>) {
let mut data = Vec::new();
let mut bitpacker = BitPacker::new();
let max_val: u64 = (1u64 << num_bits as u64) - 1u64;
@@ -118,7 +122,7 @@ mod test {
}
bitpacker.close(&mut data).unwrap();
assert_eq!(data.len(), ((num_bits as usize) * len + 7) / 8 + 7);
let bitunpacker = BitUnpacker::new(OwnedBytes::new(data), num_bits);
let bitunpacker = BitUnpacker::new(data, num_bits);
(bitunpacker, vals)
}

29
src/common/bitset.rs
View File

@@ -33,10 +33,6 @@ impl TinySet {
TinySet(0u64)
}
pub fn clear(&mut self) {
self.0 = 0u64;
}
/// Returns the complement of the set in `[0, 64[`.
fn complement(self) -> TinySet {
TinySet(!self.0)
@@ -47,11 +43,6 @@ impl TinySet {
!self.intersect(TinySet::singleton(el)).is_empty()
}
/// Returns the number of elements in the TinySet.
pub fn len(self) -> u32 {
self.0.count_ones()
}
/// Returns the intersection of `self` and `other`
pub fn intersect(self, other: TinySet) -> TinySet {
TinySet(self.0 & other.0)
@@ -118,12 +109,22 @@ impl TinySet {
pub fn range_greater_or_equal(from_included: u32) -> TinySet {
TinySet::range_lower(from_included).complement()
}
pub fn clear(&mut self) {
self.0 = 0u64;
}
pub fn len(self) -> u32 {
self.0.count_ones()
}
}
#[derive(Clone)]
pub struct BitSet {
tinysets: Box<[TinySet]>,
len: usize,
len: usize, //< Technically it should be u32, but we
// count multiple inserts.
// `usize` guards us from overflow.
max_value: u32,
}
@@ -203,7 +204,7 @@ mod tests {
use super::BitSet;
use super::TinySet;
use crate::docset::{DocSet, TERMINATED};
use crate::docset::DocSet;
use crate::query::BitSetDocSet;
use crate::tests;
use crate::tests::generate_nonunique_unsorted;
@@ -277,13 +278,11 @@ mod tests {
}
assert_eq!(btreeset.len(), bitset.len());
let mut bitset_docset = BitSetDocSet::from(bitset);
let mut remaining = true;
for el in btreeset.into_iter() {
assert!(remaining);
bitset_docset.advance();
assert_eq!(bitset_docset.doc(), el);
remaining = bitset_docset.advance() != TERMINATED;
}
assert!(!remaining);
assert!(!bitset_docset.advance());
}
#[test]

69
src/common/composite_file.rs
View File

@@ -1,15 +1,14 @@
use crate::common::BinarySerializable;
use crate::common::CountingWriter;
use crate::common::VInt;
use crate::directory::FileSlice;
use crate::directory::ReadOnlySource;
use crate::directory::{TerminatingWrite, WritePtr};
use crate::schema::Field;
use crate::space_usage::FieldUsage;
use crate::space_usage::PerFieldSpaceUsage;
use std::collections::HashMap;
use std::io::{self, Read, Write};
use super::HasLen;
use std::io::Write;
use std::io::{self, Read};
#[derive(Eq, PartialEq, Hash, Copy, Ord, PartialOrd, Clone, Debug)]
pub struct FileAddr {
@@ -104,26 +103,25 @@ impl<W: TerminatingWrite + Write> CompositeWrite<W> {
/// for each field.
#[derive(Clone)]
pub struct CompositeFile {
data: FileSlice,
data: ReadOnlySource,
offsets_index: HashMap<FileAddr, (usize, usize)>,
}
impl CompositeFile {
/// Opens a composite file stored in a given
/// `FileSlice`.
pub fn open(data: &FileSlice) -> io::Result<CompositeFile> {
/// `ReadOnlySource`.
pub fn open(data: &ReadOnlySource) -> io::Result<CompositeFile> {
let end = data.len();
let footer_len_data = data.slice_from(end - 4).read_bytes()?;
let footer_len_data = data.slice_from(end - 4);
let footer_len = u32::deserialize(&mut footer_len_data.as_slice())? as usize;
let footer_start = end - 4 - footer_len;
let footer_data = data
.slice(footer_start, footer_start + footer_len)
.read_bytes()?;
let footer_data = data.slice(footer_start, footer_start + footer_len);
let mut footer_buffer = footer_data.as_slice();
let num_fields = VInt::deserialize(&mut footer_buffer)?.0 as usize;
let mut file_addrs = vec![];
let mut offsets = vec![];
let mut field_index = HashMap::new();
let mut offset = 0;
@@ -152,19 +150,19 @@ impl CompositeFile {
pub fn empty() -> CompositeFile {
CompositeFile {
offsets_index: HashMap::new(),
data: FileSlice::empty(),
data: ReadOnlySource::empty(),
}
}
/// Returns the `FileSlice` associated
/// Returns the `ReadOnlySource` associated
/// to a given `Field` and stored in a `CompositeFile`.
pub fn open_read(&self, field: Field) -> Option<FileSlice> {
pub fn open_read(&self, field: Field) -> Option<ReadOnlySource> {
self.open_read_with_idx(field, 0)
}
/// Returns the `FileSlice` associated
/// Returns the `ReadOnlySource` associated
/// to a given `Field` and stored in a `CompositeFile`.
pub fn open_read_with_idx(&self, field: Field, idx: usize) -> Option<FileSlice> {
pub fn open_read_with_idx(&self, field: Field, idx: usize) -> Option<ReadOnlySource> {
self.offsets_index
.get(&FileAddr { field, idx })
.map(|&(from, to)| self.data.slice(from, to))
@@ -192,46 +190,49 @@ mod test {
use crate::schema::Field;
use std::io::Write;
use std::path::Path;
#[test]
fn test_composite_file() -> crate::Result<()> {
fn test_composite_file() {
let path = Path::new("test_path");
let directory = RAMDirectory::create();
let mut directory = RAMDirectory::default();
{
let w = directory.open_write(path).unwrap();
let mut composite_write = CompositeWrite::wrap(w);
let mut write_0 = composite_write.for_field(Field::from_field_id(0u32));
VInt(32431123u64).serialize(&mut write_0)?;
write_0.flush()?;
let mut write_4 = composite_write.for_field(Field::from_field_id(4u32));
VInt(2).serialize(&mut write_4)?;
write_4.flush()?;
composite_write.close()?;
{
let mut write_0 = composite_write.for_field(Field::from_field_id(0u32));
VInt(32431123u64).serialize(&mut write_0).unwrap();
write_0.flush().unwrap();
}
{
let mut write_4 = composite_write.for_field(Field::from_field_id(4u32));
VInt(2).serialize(&mut write_4).unwrap();
write_4.flush().unwrap();
}
composite_write.close().unwrap();
}
{
let r = directory.open_read(path)?;
let composite_file = CompositeFile::open(&r)?;
let r = directory.open_read(path).unwrap();
let composite_file = CompositeFile::open(&r).unwrap();
{
let file0 = composite_file
.open_read(Field::from_field_id(0u32))
.unwrap()
.read_bytes()?;
.unwrap();
let mut file0_buf = file0.as_slice();
let payload_0 = VInt::deserialize(&mut file0_buf)?.0;
let payload_0 = VInt::deserialize(&mut file0_buf).unwrap().0;
assert_eq!(file0_buf.len(), 0);
assert_eq!(payload_0, 32431123u64);
}
{
let file4 = composite_file
.open_read(Field::from_field_id(4u32))
.unwrap()
.read_bytes()?;
.unwrap();
let mut file4_buf = file4.as_slice();
let payload_4 = VInt::deserialize(&mut file4_buf)?.0;
let payload_4 = VInt::deserialize(&mut file4_buf).unwrap().0;
assert_eq!(file4_buf.len(), 0);
assert_eq!(payload_4, 2u64);
}
}
Ok(())
}
}

6
src/common/mod.rs
View File

@@ -2,6 +2,7 @@ pub mod bitpacker;
mod bitset;
mod composite_file;
mod counting_writer;
mod mutable_enum;
mod serialize;
mod vint;
@@ -9,10 +10,9 @@ pub use self::bitset::BitSet;
pub(crate) use self::bitset::TinySet;
pub(crate) use self::composite_file::{CompositeFile, CompositeWrite};
pub use self::counting_writer::CountingWriter;
pub(crate) use self::mutable_enum::MutableEnum;
pub use self::serialize::{BinarySerializable, FixedSize};
pub use self::vint::{
read_u32_vint, read_u32_vint_no_advance, serialize_vint_u32, write_u32_vint, VInt,
};
pub use self::vint::{read_u32_vint, serialize_vint_u32, write_u32_vint, VInt};
pub use byteorder::LittleEndian as Endianness;
/// Segment's max doc must be `< MAX_DOC_LIMIT`.

37
src/common/mutable_enum.rs Normal file
View File

@@ -0,0 +1,37 @@
use std::ops::{Deref, DerefMut};
pub(crate) struct MutableEnum<T>(Option<T>);
impl<T> MutableEnum<T> {
pub fn wrap(val: T) -> Self {
MutableEnum(Some(val))
}
pub fn into(self) -> T {
self.0.unwrap()
}
}
impl<T> MutableEnum<T> {
pub fn map_mutate<E, F>(&mut self, transformation: F) -> Result<(), E>
where
F: FnOnce(T) -> Result<T, E>,
{
self.0 = self.0.take().map(transformation).transpose()?;
Ok(())
}
}
impl<T> Deref for MutableEnum<T> {
type Target = T;
fn deref(&self) -> &Self::Target {
self.0.as_ref().unwrap()
}
}
impl<T> DerefMut for MutableEnum<T> {
fn deref_mut(&mut self) -> &mut Self::Target {
self.0.as_mut().unwrap()
}
}

13
src/common/serialize.rs
View File

@@ -89,19 +89,6 @@ impl FixedSize for u64 {
const SIZE_IN_BYTES: usize = 8;
}
impl BinarySerializable for f32 {
fn serialize<W: Write>(&self, writer: &mut W) -> io::Result<()> {
writer.write_f32::<Endianness>(*self)
}
fn deserialize<R: Read>(reader: &mut R) -> io::Result<Self> {
reader.read_f32::<Endianness>()
}
}
impl FixedSize for f32 {
const SIZE_IN_BYTES: usize = 4;
}
impl BinarySerializable for i64 {
fn serialize<W: Write>(&self, writer: &mut W) -> io::Result<()> {
writer.write_i64::<Endianness>(*self)

38
src/common/vint.rs
View File

@@ -5,12 +5,12 @@ use std::io::Read;
use std::io::Write;
/// Wrapper over a `u64` that serializes as a variable int.
#[derive(Clone, Copy, Debug, Eq, PartialEq)]
#[derive(Debug, Eq, PartialEq)]
pub struct VInt(pub u64);
const STOP_BIT: u8 = 128;
pub fn serialize_vint_u32(val: u32, buf: &mut [u8; 8]) -> &[u8] {
pub fn serialize_vint_u32(val: u32) -> (u64, usize) {
const START_2: u64 = 1 << 7;
const START_3: u64 = 1 << 14;
const START_4: u64 = 1 << 21;
@@ -29,7 +29,7 @@ pub fn serialize_vint_u32(val: u32, buf: &mut [u8; 8]) -> &[u8] {
let val = u64::from(val);
const STOP_BIT: u64 = 128u64;
let (res, num_bytes) = match val {
match val {
0..=STOP_1 => (val | STOP_BIT, 1),
START_2..=STOP_2 => (
(val & MASK_1) | ((val & MASK_2) << 1) | (STOP_BIT << (8)),
@@ -56,9 +56,7 @@ pub fn serialize_vint_u32(val: u32, buf: &mut [u8; 8]) -> &[u8] {
| (STOP_BIT << (8 * 4)),
5,
),
};
LittleEndian::write_u64(&mut buf[..], res);
&buf[0..num_bytes]
}
}
/// Returns the number of bytes covered by a
@@ -87,26 +85,23 @@ fn vint_len(data: &[u8]) -> usize {
/// If the buffer does not start by a valid
/// vint payload
pub fn read_u32_vint(data: &mut &[u8]) -> u32 {
let (result, vlen) = read_u32_vint_no_advance(*data);
*data = &data[vlen..];
result
}
pub fn read_u32_vint_no_advance(data: &[u8]) -> (u32, usize) {
let vlen = vint_len(data);
let vlen = vint_len(*data);
let mut result = 0u32;
let mut shift = 0u64;
for &b in &data[..vlen] {
result |= u32::from(b & 127u8) << shift;
shift += 7;
}
(result, vlen)
*data = &data[vlen..];
result
}
/// Write a `u32` as a vint payload.
pub fn write_u32_vint<W: io::Write>(val: u32, writer: &mut W) -> io::Result<()> {
let mut buf = [0u8; 8];
let data = serialize_vint_u32(val, &mut buf);
writer.write_all(&data)
let (val, num_bytes) = serialize_vint_u32(val);
let mut buffer = [0u8; 8];
LittleEndian::write_u64(&mut buffer, val);
writer.write_all(&buffer[..num_bytes])
}
impl VInt {
@@ -177,6 +172,7 @@ mod tests {
use super::serialize_vint_u32;
use super::VInt;
use crate::common::BinarySerializable;
use byteorder::{ByteOrder, LittleEndian};
fn aux_test_vint(val: u64) {
let mut v = [14u8; 10];
@@ -212,10 +208,12 @@ mod tests {
fn aux_test_serialize_vint_u32(val: u32) {
let mut buffer = [0u8; 10];
let mut buffer2 = [0u8; 8];
let mut buffer2 = [0u8; 10];
let len_vint = VInt(val as u64).serialize_into(&mut buffer);
let res2 = serialize_vint_u32(val, &mut buffer2);
assert_eq!(&buffer[..len_vint], res2, "array wrong for {}", val);
let (vint, len) = serialize_vint_u32(val);
assert_eq!(len, len_vint, "len wrong for val {}", val);
LittleEndian::write_u64(&mut buffer2, vint);
assert_eq!(&buffer[..len], &buffer2[..len], "array wrong for {}", val);
}
#[test]

128
src/core/index.rs
View File

@@ -12,8 +12,8 @@ use crate::directory::INDEX_WRITER_LOCK;
use crate::directory::{Directory, RAMDirectory};
use crate::error::DataCorruption;
use crate::error::TantivyError;
use crate::indexer::index_writer::HEAP_SIZE_MIN;
use crate::indexer::segment_updater::save_new_metas;
use crate::indexer::IndexWriterConfig;
use crate::reader::IndexReader;
use crate::reader::IndexReaderBuilder;
use crate::schema::Field;
@@ -21,12 +21,12 @@ use crate::schema::FieldType;
use crate::schema::Schema;
use crate::tokenizer::{TextAnalyzer, TokenizerManager};
use crate::IndexWriter;
use num_cpus;
use std::borrow::BorrowMut;
use std::collections::HashSet;
use std::fmt;
#[cfg(feature = "mmap")]
use std::path::Path;
use std::path::PathBuf;
use std::path::{Path, PathBuf};
use std::sync::Arc;
fn load_metas(
@@ -56,9 +56,7 @@ pub struct Index {
}
impl Index {
/// Examines the directory to see if it contains an index.
///
/// Effectively, it only checks for the presence of the `meta.json` file.
/// Examines the director to see if it contains an index
pub fn exists<Dir: Directory>(dir: &Dir) -> bool {
dir.exists(&META_FILEPATH)
}
@@ -141,9 +139,7 @@ impl Index {
Index::create(mmap_directory, schema)
}
/// Creates a new index given an implementation of the trait `Directory`.
///
/// If a directory previously existed, it will be erased.
/// Creates a new index given an implementation of the trait `Directory`
pub fn create<Dir: Directory>(dir: Dir, schema: Schema) -> crate::Result<Index> {
let directory = ManagedDirectory::wrap(dir)?;
Index::from_directory(directory, schema)
@@ -152,8 +148,8 @@ impl Index {
/// Create a new index from a directory.
///
/// This will overwrite existing meta.json
fn from_directory(directory: ManagedDirectory, schema: Schema) -> crate::Result<Index> {
save_new_metas(schema.clone(), &directory)?;
fn from_directory(mut directory: ManagedDirectory, schema: Schema) -> crate::Result<Index> {
save_new_metas(schema.clone(), directory.borrow_mut())?;
let metas = IndexMeta::with_schema(schema);
Index::create_from_metas(directory, &metas, SegmentMetaInventory::default())
}
@@ -277,8 +273,40 @@ impl Index {
pub fn writer_with_num_threads(
&self,
num_threads: usize,
overall_heap_size_in_bytes: usize,
overall_heap_size_in_bytes: u64,
) -> crate::Result<IndexWriter> {
let config = IndexWriterConfig {
max_indexing_threads: num_threads,
memory_budget: overall_heap_size_in_bytes,
..Default::default()
};
self.writer_from_config(config)
}
/// Creates a multithreaded writer
///
/// Tantivy will automatically define the number of threads to use.
/// `overall_heap_size_in_bytes` is the total target memory usage that will be split
/// between a given number of threads.
///
/// # Errors
/// If the lockfile already exists, returns `Error::FileAlreadyExists`.
/// # Panics
/// If the heap size per thread is too small, panics.
pub fn writer(&self, overall_heap_size_in_bytes: u64) -> crate::Result<IndexWriter> {
let config = IndexWriterConfig {
max_indexing_threads: num_cpus::get(),
memory_budget: overall_heap_size_in_bytes,
..Default::default()
};
self.writer_from_config(config)
}
/// Creates a new writer with a given configuration.
///
/// See [`IndexWriterConfig`](./struct.IndexWriterConfig.html) for more information.
pub fn writer_from_config(&self, mut config: IndexWriterConfig) -> crate::Result<IndexWriter> {
config.validate()?;
let directory_lock = self
.directory
.acquire_lock(&INDEX_WRITER_LOCK)
@@ -294,41 +322,7 @@ impl Index {
),
)
})?;
let heap_size_in_bytes_per_thread = overall_heap_size_in_bytes / num_threads;
IndexWriter::new(
self,
num_threads,
heap_size_in_bytes_per_thread,
directory_lock,
)
}
/// Helper to create an index writer for tests.
///
/// That index writer only simply has a single thread and a heap of 5 MB.
/// Using a single thread gives us a deterministic allocation of DocId.
#[cfg(test)]
pub fn writer_for_tests(&self) -> crate::Result<IndexWriter> {
self.writer_with_num_threads(1, 10_000_000)
}
/// Creates a multithreaded writer
///
/// Tantivy will automatically define the number of threads to use.
/// `overall_heap_size_in_bytes` is the total target memory usage that will be split
/// between a given number of threads.
///
/// # Errors
/// If the lockfile already exists, returns `Error::FileAlreadyExists`.
/// # Panics
/// If the heap size per thread is too small, panics.
pub fn writer(&self, overall_heap_size_in_bytes: usize) -> crate::Result<IndexWriter> {
let mut num_threads = num_cpus::get();
let heap_size_in_bytes_per_thread = overall_heap_size_in_bytes / num_threads;
if heap_size_in_bytes_per_thread < HEAP_SIZE_MIN {
num_threads = (overall_heap_size_in_bytes / HEAP_SIZE_MIN).max(1);
}
self.writer_with_num_threads(num_threads, overall_heap_size_in_bytes)
IndexWriter::new(self, config, directory_lock)
}
/// Accessor to the index schema
@@ -482,7 +476,7 @@ mod tests {
.try_into()
.unwrap();
assert_eq!(reader.searcher().num_docs(), 0);
test_index_on_commit_reload_policy_aux(field, &index, &reader);
test_index_on_commit_reload_policy_aux(field, index.clone(), &index, &reader);
}
#[cfg(feature = "mmap")]
@@ -506,7 +500,7 @@ mod tests {
.try_into()
.unwrap();
assert_eq!(reader.searcher().num_docs(), 0);
test_index_on_commit_reload_policy_aux(field, &index, &reader);
test_index_on_commit_reload_policy_aux(field, index.clone(), &index, &reader);
}
#[test]
@@ -514,7 +508,7 @@ mod tests {
let schema = throw_away_schema();
let field = schema.get_field("num_likes").unwrap();
let mut index = Index::create_from_tempdir(schema).unwrap();
let mut writer = index.writer_for_tests().unwrap();
let mut writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
writer.commit().unwrap();
let reader = index
.reader_builder()
@@ -548,36 +542,30 @@ mod tests {
.try_into()
.unwrap();
assert_eq!(reader.searcher().num_docs(), 0);
test_index_on_commit_reload_policy_aux(field, &write_index, &reader);
test_index_on_commit_reload_policy_aux(field, read_index, &write_index, &reader);
}
}
fn test_index_on_commit_reload_policy_aux(field: Field, index: &Index, reader: &IndexReader) {
let mut reader_index = reader.index();
fn test_index_on_commit_reload_policy_aux(
field: Field,
mut reader_index: Index,
index: &Index,
reader: &IndexReader,
) {
let (sender, receiver) = crossbeam::channel::unbounded();
let _watch_handle = reader_index.directory_mut().watch(Box::new(move || {
let _ = sender.send(());
}));
let mut writer = index.writer_for_tests().unwrap();
let mut writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
assert_eq!(reader.searcher().num_docs(), 0);
writer.add_document(doc!(field=>1u64));
writer.commit().unwrap();
// We need a loop here because it is possible for notify to send more than
// one modify event. It was observed on CI on MacOS.
loop {
assert!(receiver.recv().is_ok());
if reader.searcher().num_docs() == 1 {
break;
}
}
assert!(receiver.recv().is_ok());
assert_eq!(reader.searcher().num_docs(), 1);
writer.add_document(doc!(field=>2u64));
writer.commit().unwrap();
// ... Same as above
loop {
assert!(receiver.recv().is_ok());
if reader.searcher().num_docs() == 2 {
break;
}
}
assert!(receiver.recv().is_ok());
assert_eq!(reader.searcher().num_docs(), 2);
}
// This test will not pass on windows, because windows

6
src/core/index_meta.rs
View File

@@ -3,7 +3,9 @@ use crate::core::SegmentId;
use crate::schema::Schema;
use crate::Opstamp;
use census::{Inventory, TrackedObject};
use serde;
use serde::{Deserialize, Serialize};
use serde_json;
use std::collections::HashSet;
use std::fmt;
use std::path::PathBuf;
@@ -213,7 +215,7 @@ pub struct IndexMeta {
#[serde(skip_serializing_if = "Option::is_none")]
/// Payload associated to the last commit.
///
/// Upon commit, clients can optionally add a small `String` payload to their commit
/// Upon commit, clients can optionally add a small `Striing` payload to their commit
/// to help identify this commit.
/// This payload is entirely unused by tantivy.
pub payload: Option<String>,
@@ -301,7 +303,7 @@ mod tests {
let json = serde_json::ser::to_string(&index_metas).expect("serialization failed");
assert_eq!(
json,
r#"{"segments":[],"schema":[{"name":"text","type":"text","options":{"indexing":{"record":"position","tokenizer":"default","fieldnorms":true},"stored":false}}],"opstamp":0}"#
r#"{"segments":[],"schema":[{"name":"text","type":"text","options":{"indexing":{"record":"position","tokenizer":"default"},"stored":false}}],"opstamp":0}"#
);
}
}

106
src/core/inverted_index_reader.rs
View File

@@ -1,13 +1,13 @@
use std::io;
use crate::common::BinarySerializable;
use crate::directory::FileSlice;
use crate::directory::ReadOnlySource;
use crate::positions::PositionReader;
use crate::postings::TermInfo;
use crate::postings::{BlockSegmentPostings, SegmentPostings};
use crate::schema::FieldType;
use crate::schema::IndexRecordOption;
use crate::schema::Term;
use crate::termdict::TermDictionary;
use owned_read::OwnedRead;
/// The inverted index reader is in charge of accessing
/// the inverted index associated to a specific field.
@@ -16,7 +16,7 @@ use crate::termdict::TermDictionary;
///
/// It is safe to delete the segment associated to
/// an `InvertedIndexReader`. As long as it is open,
/// the `FileSlice` it is relying on should
/// the `ReadOnlySource` it is relying on should
/// stay available.
///
///
@@ -24,9 +24,9 @@ use crate::termdict::TermDictionary;
/// the `SegmentReader`'s [`.inverted_index(...)`] method
pub struct InvertedIndexReader {
termdict: TermDictionary,
postings_file_slice: FileSlice,
positions_file_slice: FileSlice,
positions_idx_file_slice: FileSlice,
postings_source: ReadOnlySource,
positions_source: ReadOnlySource,
positions_idx_source: ReadOnlySource,
record_option: IndexRecordOption,
total_num_tokens: u64,
}
@@ -35,31 +35,35 @@ impl InvertedIndexReader {
#[cfg_attr(feature = "cargo-clippy", allow(clippy::needless_pass_by_value))] // for symmetry
pub(crate) fn new(
termdict: TermDictionary,
postings_file_slice: FileSlice,
positions_file_slice: FileSlice,
positions_idx_file_slice: FileSlice,
postings_source: ReadOnlySource,
positions_source: ReadOnlySource,
positions_idx_source: ReadOnlySource,
record_option: IndexRecordOption,
) -> io::Result<InvertedIndexReader> {
let (total_num_tokens_slice, postings_body) = postings_file_slice.split(8);
let total_num_tokens = u64::deserialize(&mut total_num_tokens_slice.read_bytes()?)?;
Ok(InvertedIndexReader {
) -> InvertedIndexReader {
let total_num_tokens_data = postings_source.slice(0, 8);
let mut total_num_tokens_cursor = total_num_tokens_data.as_slice();
let total_num_tokens = u64::deserialize(&mut total_num_tokens_cursor).unwrap_or(0u64);
InvertedIndexReader {
termdict,
postings_file_slice: postings_body,
positions_file_slice,
positions_idx_file_slice,
postings_source: postings_source.slice_from(8),
positions_source,
positions_idx_source,
record_option,
total_num_tokens,
})
}
}
/// Creates an empty `InvertedIndexReader` object, which
/// contains no terms at all.
pub fn empty(record_option: IndexRecordOption) -> InvertedIndexReader {
pub fn empty(field_type: &FieldType) -> InvertedIndexReader {
let record_option = field_type
.get_index_record_option()
.unwrap_or(IndexRecordOption::Basic);
InvertedIndexReader {
termdict: TermDictionary::empty(),
postings_file_slice: FileSlice::empty(),
positions_file_slice: FileSlice::empty(),
positions_idx_file_slice: FileSlice::empty(),
postings_source: ReadOnlySource::empty(),
positions_source: ReadOnlySource::empty(),
positions_idx_source: ReadOnlySource::empty(),
record_option,
total_num_tokens: 0u64,
}
@@ -89,12 +93,12 @@ impl InvertedIndexReader {
&self,
term_info: &TermInfo,
block_postings: &mut BlockSegmentPostings,
) -> io::Result<()> {
let postings_slice = self
.postings_file_slice
.slice_from(term_info.postings_offset as usize);
block_postings.reset(term_info.doc_freq, postings_slice.read_bytes()?);
Ok(())
) {
let offset = term_info.postings_offset as usize;
let end_source = self.postings_source.len();
let postings_slice = self.postings_source.slice(offset, end_source);
let postings_reader = OwnedRead::new(postings_slice);
block_postings.reset(term_info.doc_freq, postings_reader);
}
/// Returns a block postings given a `Term`.
@@ -105,11 +109,9 @@ impl InvertedIndexReader {
&self,
term: &Term,
option: IndexRecordOption,
) -> io::Result<Option<BlockSegmentPostings>> {
Ok(self
.get_term_info(term)
) -> Option<BlockSegmentPostings> {
self.get_term_info(term)
.map(move |term_info| self.read_block_postings_from_terminfo(&term_info, option))
.transpose()?)
}
/// Returns a block postings given a `term_info`.
@@ -120,12 +122,12 @@ impl InvertedIndexReader {
&self,
term_info: &TermInfo,
requested_option: IndexRecordOption,
) -> io::Result<BlockSegmentPostings> {
) -> BlockSegmentPostings {
let offset = term_info.postings_offset as usize;
let postings_data = self.postings_file_slice.slice_from(offset);
BlockSegmentPostings::open(
let postings_data = self.postings_source.slice_from(offset);
BlockSegmentPostings::from_data(
term_info.doc_freq,
postings_data,
OwnedRead::new(postings_data),
self.record_option,
requested_option,
)
@@ -139,23 +141,20 @@ impl InvertedIndexReader {
&self,
term_info: &TermInfo,
option: IndexRecordOption,
) -> io::Result<SegmentPostings> {
let block_postings = self.read_block_postings_from_terminfo(term_info, option)?;
) -> SegmentPostings {
let block_postings = self.read_block_postings_from_terminfo(term_info, option);
let position_stream = {
if option.has_positions() {
let position_reader = self.positions_file_slice.clone();
let skip_reader = self.positions_idx_file_slice.clone();
let position_reader = self.positions_source.clone();
let skip_reader = self.positions_idx_source.clone();
let position_reader =
PositionReader::new(position_reader, skip_reader, term_info.positions_idx)?;
PositionReader::new(position_reader, skip_reader, term_info.positions_idx);
Some(position_reader)
} else {
None
}
};
Ok(SegmentPostings::from_block_postings(
block_postings,
position_stream,
))
SegmentPostings::from_block_postings(block_postings, position_stream)
}
/// Returns the total number of tokens recorded for all documents
@@ -174,31 +173,24 @@ impl InvertedIndexReader {
/// For instance, requesting `IndexRecordOption::Freq` for a
/// `TextIndexingOptions` that does not index position will return a `SegmentPostings`
/// with `DocId`s and frequencies.
pub fn read_postings(
&self,
term: &Term,
option: IndexRecordOption,
) -> io::Result<Option<SegmentPostings>> {
pub fn read_postings(&self, term: &Term, option: IndexRecordOption) -> Option<SegmentPostings> {
self.get_term_info(term)
.map(move |term_info| self.read_postings_from_terminfo(&term_info, option))
.transpose()
}
pub(crate) fn read_postings_no_deletes(
&self,
term: &Term,
option: IndexRecordOption,
) -> io::Result<Option<SegmentPostings>> {
) -> Option<SegmentPostings> {
self.get_term_info(term)
.map(|term_info| self.read_postings_from_terminfo(&term_info, option))
.transpose()
}
/// Returns the number of documents containing the term.
pub fn doc_freq(&self, term: &Term) -> io::Result<u32> {
Ok(self
.get_term_info(term)
pub fn doc_freq(&self, term: &Term) -> u32 {
self.get_term_info(term)
.map(|term_info| term_info.doc_freq)
.unwrap_or(0u32))
.unwrap_or(0u32)
}
}

73
src/core/searcher.rs
View File

@@ -1,8 +1,11 @@
use crate::collector::Collector;
use crate::collector::SegmentCollector;
use crate::core::Executor;
use crate::core::InvertedIndexReader;
use crate::core::SegmentReader;
use crate::query::Query;
use crate::query::Scorer;
use crate::query::Weight;
use crate::schema::Document;
use crate::schema::Schema;
use crate::schema::{Field, Term};
@@ -11,8 +14,28 @@ use crate::store::StoreReader;
use crate::termdict::TermMerger;
use crate::DocAddress;
use crate::Index;
use std::fmt;
use std::sync::Arc;
use std::{fmt, io};
fn collect_segment<C: Collector>(
collector: &C,
weight: &dyn Weight,
segment_ord: u32,
segment_reader: &SegmentReader,
) -> crate::Result<C::Fruit> {
let mut scorer = weight.scorer(segment_reader, 1.0f32)?;
let mut segment_collector = collector.for_segment(segment_ord as u32, segment_reader)?;
if let Some(delete_bitset) = segment_reader.delete_bitset() {
scorer.for_each(&mut |doc, score| {
if delete_bitset.is_alive(doc) {
segment_collector.collect(doc, score);
}
});
} else {
scorer.for_each(&mut |doc, score| segment_collector.collect(doc, score));
}
Ok(segment_collector.harvest())
}
/// Holds a list of `SegmentReader`s ready for search.
///
@@ -32,17 +55,17 @@ impl Searcher {
schema: Schema,
index: Index,
segment_readers: Vec<SegmentReader>,
) -> io::Result<Searcher> {
let store_readers: Vec<StoreReader> = segment_readers
) -> Searcher {
let store_readers = segment_readers
.iter()
.map(SegmentReader::get_store_reader)
.collect::<io::Result<Vec<_>>>()?;
Ok(Searcher {
.collect();
Searcher {
schema,
index,
segment_readers,
store_readers,
})
}
}
/// Returns the `Index` associated to the `Searcher`
@@ -75,14 +98,13 @@ impl Searcher {
/// Return the overall number of documents containing
/// the given term.
pub fn doc_freq(&self, term: &Term) -> crate::Result<u64> {
let mut total_doc_freq = 0;
for segment_reader in &self.segment_readers {
let inverted_index = segment_reader.inverted_index(term.field())?;
let doc_freq = inverted_index.doc_freq(term)?;
total_doc_freq += u64::from(doc_freq);
}
Ok(total_doc_freq)
pub fn doc_freq(&self, term: &Term) -> u64 {
self.segment_readers
.iter()
.map(|segment_reader| {
u64::from(segment_reader.inverted_index(term.field()).doc_freq(term))
})
.sum::<u64>()
}
/// Return the list of segment readers
@@ -141,7 +163,12 @@ impl Searcher {
let segment_readers = self.segment_readers();
let fruits = executor.map(
|(segment_ord, segment_reader)| {
collector.collect_segment(weight.as_ref(), segment_ord as u32, segment_reader)
collect_segment(
collector,
weight.as_ref(),
segment_ord as u32,
segment_reader,
)
},
segment_readers.iter().enumerate(),
)?;
@@ -149,22 +176,22 @@ impl Searcher {
}
/// Return the field searcher associated to a `Field`.
pub fn field(&self, field: Field) -> crate::Result<FieldSearcher> {
let inv_index_readers: Vec<Arc<InvertedIndexReader>> = self
pub fn field(&self, field: Field) -> FieldSearcher {
let inv_index_readers = self
.segment_readers
.iter()
.map(|segment_reader| segment_reader.inverted_index(field))
.collect::<crate::Result<Vec<_>>>()?;
Ok(FieldSearcher::new(inv_index_readers))
.collect::<Vec<_>>();
FieldSearcher::new(inv_index_readers)
}
/// Summarize total space usage of this searcher.
pub fn space_usage(&self) -> io::Result<SearcherSpaceUsage> {
pub fn space_usage(&self) -> SearcherSpaceUsage {
let mut space_usage = SearcherSpaceUsage::new();
for segment_reader in &self.segment_readers {
space_usage.add_segment(segment_reader.space_usage()?);
for segment_reader in self.segment_readers.iter() {
space_usage.add_segment(segment_reader.space_usage());
}
Ok(space_usage)
space_usage
}
}

136
src/core/segment.rs
View File

@@ -3,41 +3,140 @@ use crate::core::Index;
use crate::core::SegmentId;
use crate::core::SegmentMeta;
use crate::directory::error::{OpenReadError, OpenWriteError};
use crate::directory::Directory;
use crate::directory::{FileSlice, WritePtr};
use crate::directory::{Directory, ManagedDirectory, RAMDirectory};
use crate::directory::{ReadOnlySource, WritePtr};
use crate::indexer::segment_serializer::SegmentSerializer;
use crate::schema::Schema;
use crate::Opstamp;
use std::fmt;
use std::ops::{Deref, DerefMut};
use std::path::PathBuf;
use crate::indexer::{ResourceManager};
#[derive(Clone)]
pub(crate) enum SegmentDirectory {
Persisted(ManagedDirectory),
Volatile {
directory: RAMDirectory,
},
}
impl SegmentDirectory {
fn new_volatile(memory_manager: ResourceManager) -> SegmentDirectory {
SegmentDirectory::Volatile {
directory: RAMDirectory::create_with_memory_manager(memory_manager),
}
}
}
impl fmt::Debug for SegmentDirectory {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
match self {
SegmentDirectory::Volatile { .. } => write!(f, "volatile")?,
SegmentDirectory::Persisted(dir) => write!(f, "Persisted({:?})", dir)?,
}
Ok(())
}
}
impl From<ManagedDirectory> for SegmentDirectory {
fn from(directory: ManagedDirectory) -> Self {
SegmentDirectory::Persisted(directory)
}
}
impl Deref for SegmentDirectory {
type Target = dyn Directory;
fn deref(&self) -> &Self::Target {
match self {
SegmentDirectory::Volatile {
directory, ..
} => directory,
SegmentDirectory::Persisted(dir) => dir,
}
}
}
impl DerefMut for SegmentDirectory {
fn deref_mut(&mut self) -> &mut Self::Target {
match self {
SegmentDirectory::Volatile { directory, .. } => directory,
SegmentDirectory::Persisted(dir) => dir,
}
}
}
/// A segment is a piece of the index.
#[derive(Clone)]
pub struct Segment {
index: Index,
schema: Schema,
meta: SegmentMeta,
directory: SegmentDirectory,
}
impl fmt::Debug for Segment {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
write!(f, "Segment({:?})", self.id().uuid_string())
write!(
f,
"Segment(id={:?}, directory={:?})",
self.id().uuid_string(),
self.directory
)
}
}
impl Segment {
/// Creates a new segment given an `Index` and a `SegmentId`
pub(crate) fn for_index(index: Index, meta: SegmentMeta) -> Segment {
Segment { index, meta }
pub(crate) fn new_persisted(
meta: SegmentMeta,
directory: ManagedDirectory,
schema: Schema,
) -> Segment {
Segment {
meta,
schema,
directory: SegmentDirectory::from(directory),
}
}
/// Returns the index the segment belongs to.
pub fn index(&self) -> &Index {
&self.index
/// Creates a new segment that embeds its own `RAMDirectory`.
///
/// That segment is entirely dissociated from the index directory.
/// It will be persisted by a background thread in charge of IO.
pub fn new_volatile(meta: SegmentMeta, schema: Schema, resource_manager: ResourceManager) -> Segment {
Segment {
schema,
meta,
directory: SegmentDirectory::new_volatile(resource_manager),
}
}
/// Creates a new segment given an `Index` and a `SegmentId`
pub(crate) fn for_index(index: Index, meta: SegmentMeta) -> Segment {
let segment_directory = index.directory().clone();
Segment::new_persisted(meta, segment_directory, index.schema())
}
/// Persists a given `Segment` to a directory.
pub fn persist(&mut self, mut dest_directory: ManagedDirectory) -> crate::Result<()> {
if let SegmentDirectory::Persisted(_) = self.directory {
// this segment is already persisted.
return Ok(());
}
if let SegmentDirectory::Volatile { directory, ..} = &self.directory {
directory.persist(&mut dest_directory)?;
}
self.directory = SegmentDirectory::Persisted(dest_directory);
Ok(())
}
pub fn into_volatile(&self, memory_manager: ResourceManager) -> Segment {
Segment::new_volatile(self.meta.clone(), self.schema.clone(), memory_manager)
}
/// Returns our index's schema.
pub fn schema(&self) -> Schema {
self.index.schema()
self.schema.clone()
}
/// Returns the segment meta-information
@@ -51,16 +150,18 @@ impl Segment {
/// as we finalize a fresh new segment.
pub(crate) fn with_max_doc(self, max_doc: u32) -> Segment {
Segment {
index: self.index,
schema: self.schema,
meta: self.meta.with_max_doc(max_doc),
directory: self.directory,
}
}
#[doc(hidden)]
pub fn with_delete_meta(self, num_deleted_docs: u32, opstamp: Opstamp) -> Segment {
Segment {
index: self.index,
schema: self.schema,
meta: self.meta.with_delete_meta(num_deleted_docs, opstamp),
directory: self.directory,
}
}
@@ -78,16 +179,17 @@ impl Segment {
}
/// Open one of the component file for a *regular* read.
pub fn open_read(&self, component: SegmentComponent) -> Result<FileSlice, OpenReadError> {
pub fn open_read(&self, component: SegmentComponent) -> Result<ReadOnlySource, OpenReadError> {
let path = self.relative_path(component);
self.index.directory().open_read(&path)
let source = self.directory.open_read(&path)?;
Ok(source)
}
/// Open one of the component file for *regular* write.
pub fn open_write(&mut self, component: SegmentComponent) -> Result<WritePtr, OpenWriteError> {
let path = self.relative_path(component);
let write = self.index.directory_mut().open_write(&path)?;
Ok(write)
let wrt = self.directory.open_write(&path)?;
Ok(wrt)
}
}

2
src/core/segment_component.rs
View File

@@ -20,7 +20,7 @@ pub enum SegmentComponent {
/// Dictionary associating `Term`s to `TermInfo`s which is
/// simply an address into the `postings` file and the `positions` file.
TERMS,
/// Row-oriented, compressed storage of the documents.
/// Row-oriented, LZ4-compressed storage of the documents.
/// Accessing a document from the store is relatively slow, as it
/// requires to decompress the entire block it belongs to.
STORE,

205
src/core/segment_reader.rs
View File

@@ -1,26 +1,26 @@
use crate::common::CompositeFile;
use crate::common::HasLen;
use crate::core::InvertedIndexReader;
use crate::core::Segment;
use crate::core::SegmentComponent;
use crate::core::SegmentId;
use crate::directory::FileSlice;
use crate::directory::ReadOnlySource;
use crate::fastfield::DeleteBitSet;
use crate::fastfield::FacetReader;
use crate::fastfield::FastFieldReaders;
use crate::fieldnorm::{FieldNormReader, FieldNormReaders};
use crate::fieldnorm::FieldNormReader;
use crate::schema::Field;
use crate::schema::FieldType;
use crate::schema::Schema;
use crate::schema::{Field, IndexRecordOption};
use crate::space_usage::SegmentSpaceUsage;
use crate::store::StoreReader;
use crate::termdict::TermDictionary;
use crate::DocId;
use crate::{common::CompositeFile, error::DataCorruption};
use fail::fail_point;
use std::collections::HashMap;
use std::fmt;
use std::sync::Arc;
use std::sync::RwLock;
use std::{collections::HashMap, io};
/// Entry point to access all of the datastructures of the `Segment`
///
@@ -48,9 +48,9 @@ pub struct SegmentReader {
positions_composite: CompositeFile,
positions_idx_composite: CompositeFile,
fast_fields_readers: Arc<FastFieldReaders>,
fieldnorm_readers: FieldNormReaders,
fieldnorms_composite: CompositeFile,
store_file: FileSlice,
store_source: ReadOnlySource,
delete_bitset_opt: Option<DeleteBitSet>,
schema: Schema,
}
@@ -106,26 +106,16 @@ impl SegmentReader {
}
/// Accessor to the `FacetReader` associated to a given `Field`.
pub fn facet_reader(&self, field: Field) -> crate::Result<FacetReader> {
pub fn facet_reader(&self, field: Field) -> Option<FacetReader> {
let field_entry = self.schema.get_field_entry(field);
if field_entry.field_type() != &FieldType::HierarchicalFacet {
return Err(crate::TantivyError::InvalidArgument(format!(
"Field {:?} is not a facet field.",
field_entry.name()
)));
return None;
}
let term_ords_reader = self.fast_fields().u64s(field).ok_or_else(|| {
DataCorruption::comment_only(format!(
"Cannot find data for hierarchical facet {:?}",
field_entry.name()
))
})?;
let termdict = self
.termdict_composite
.open_read(field)
.map(TermDictionary::open)
.unwrap_or_else(|| Ok(TermDictionary::empty()))?;
Ok(FacetReader::new(term_ords_reader, termdict))
let term_ords_reader = self.fast_fields().u64s(field)?;
let termdict_source = self.termdict_composite.open_read(field)?;
let termdict = TermDictionary::from_source(&termdict_source);
let facet_reader = FacetReader::new(term_ords_reader, termdict);
Some(facet_reader)
}
/// Accessor to the segment's `Field norms`'s reader.
@@ -135,45 +125,47 @@ impl SegmentReader {
///
/// They are simply stored as a fast field, serialized in
/// the `.fieldnorm` file of the segment.
pub fn get_fieldnorms_reader(&self, field: Field) -> crate::Result<FieldNormReader> {
self.fieldnorm_readers.get_field(field)?.ok_or_else(|| {
pub fn get_fieldnorms_reader(&self, field: Field) -> FieldNormReader {
if let Some(fieldnorm_source) = self.fieldnorms_composite.open_read(field) {
FieldNormReader::open(fieldnorm_source)
} else {
let field_name = self.schema.get_field_name(field);
let err_msg = format!(
"Field norm not found for field {:?}. Was it marked as indexed during indexing?",
"Field norm not found for field {:?}. Was it market as indexed during indexing.",
field_name
);
crate::TantivyError::SchemaError(err_msg)
})
panic!(err_msg);
}
}
/// Accessor to the segment's `StoreReader`.
pub fn get_store_reader(&self) -> io::Result<StoreReader> {
StoreReader::open(self.store_file.clone())
pub fn get_store_reader(&self) -> StoreReader {
StoreReader::from_source(self.store_source.clone())
}
/// Open a new segment for reading.
pub fn open(segment: &Segment) -> crate::Result<SegmentReader> {
let termdict_file = segment.open_read(SegmentComponent::TERMS)?;
let termdict_composite = CompositeFile::open(&termdict_file)?;
let termdict_source = segment.open_read(SegmentComponent::TERMS)?;
let termdict_composite = CompositeFile::open(&termdict_source)?;
let store_file = segment.open_read(SegmentComponent::STORE)?;
let store_source = segment.open_read(SegmentComponent::STORE)?;
fail_point!("SegmentReader::open#middle");
let postings_file = segment.open_read(SegmentComponent::POSTINGS)?;
let postings_composite = CompositeFile::open(&postings_file)?;
let postings_source = segment.open_read(SegmentComponent::POSTINGS)?;
let postings_composite = CompositeFile::open(&postings_source)?;
let positions_composite = {
if let Ok(positions_file) = segment.open_read(SegmentComponent::POSITIONS) {
CompositeFile::open(&positions_file)?
if let Ok(source) = segment.open_read(SegmentComponent::POSITIONS) {
CompositeFile::open(&source)?
} else {
CompositeFile::empty()
}
};
let positions_idx_composite = {
if let Ok(positions_skip_file) = segment.open_read(SegmentComponent::POSITIONSSKIP) {
CompositeFile::open(&positions_skip_file)?
if let Ok(source) = segment.open_read(SegmentComponent::POSITIONSSKIP) {
CompositeFile::open(&source)?
} else {
CompositeFile::empty()
}
@@ -186,27 +178,26 @@ impl SegmentReader {
let fast_field_readers =
Arc::new(FastFieldReaders::load_all(&schema, &fast_fields_composite)?);
let fieldnorm_data = segment.open_read(SegmentComponent::FIELDNORMS)?;
let fieldnorm_readers = FieldNormReaders::open(fieldnorm_data)?;
let fieldnorms_data = segment.open_read(SegmentComponent::FIELDNORMS)?;
let fieldnorms_composite = CompositeFile::open(&fieldnorms_data)?;
let delete_bitset_opt = if segment.meta().has_deletes() {
let delete_data = segment.open_read(SegmentComponent::DELETE)?;
let delete_bitset = DeleteBitSet::open(delete_data)?;
Some(delete_bitset)
Some(DeleteBitSet::open(delete_data))
} else {
None
};
Ok(SegmentReader {
inv_idx_reader_cache: Default::default(),
inv_idx_reader_cache: Arc::new(RwLock::new(HashMap::new())),
max_doc: segment.meta().max_doc(),
num_docs: segment.meta().num_docs(),
termdict_composite,
postings_composite,
fast_fields_readers: fast_field_readers,
fieldnorm_readers,
fieldnorms_composite,
segment_id: segment.id(),
store_file,
store_source,
delete_bitset_opt,
positions_composite,
positions_idx_composite,
@@ -221,64 +212,58 @@ impl SegmentReader {
/// The field reader is in charge of iterating through the
/// term dictionary associated to a specific field,
/// and opening the posting list associated to any term.
///
/// If the field is marked as index, a warn is logged and an empty `InvertedIndexReader`
/// is returned.
/// Similarly if the field is marked as indexed but no term has been indexed for the given
/// index. an empty `InvertedIndexReader` is returned (but no warning is logged).
pub fn inverted_index(&self, field: Field) -> crate::Result<Arc<InvertedIndexReader>> {
pub fn inverted_index(&self, field: Field) -> Arc<InvertedIndexReader> {
if let Some(inv_idx_reader) = self
.inv_idx_reader_cache
.read()
.expect("Lock poisoned. This should never happen")
.get(&field)
{
return Ok(Arc::clone(inv_idx_reader));
return Arc::clone(inv_idx_reader);
}
let field_entry = self.schema.get_field_entry(field);
let field_type = field_entry.field_type();
let record_option_opt = field_type.get_index_record_option();
if record_option_opt.is_none() {
warn!("Field {:?} does not seem indexed.", field_entry.name());
panic!("Field {:?} does not seem indexed.", field_entry.name());
}
let postings_file_opt = self.postings_composite.open_read(field);
let record_option = record_option_opt.unwrap();
if postings_file_opt.is_none() || record_option_opt.is_none() {
let postings_source_opt = self.postings_composite.open_read(field);
if postings_source_opt.is_none() {
// no documents in the segment contained this field.
// As a result, no data is associated to the inverted index.
//
// Returns an empty inverted index.
let record_option = record_option_opt.unwrap_or(IndexRecordOption::Basic);
return Ok(Arc::new(InvertedIndexReader::empty(record_option)));
return Arc::new(InvertedIndexReader::empty(field_type));
}
let record_option = record_option_opt.unwrap();
let postings_file = postings_file_opt.unwrap();
let postings_source = postings_source_opt.unwrap();
let termdict_file: FileSlice = self.termdict_composite.open_read(field)
.ok_or_else(||
DataCorruption::comment_only(format!("Failed to open field {:?}'s term dictionary in the composite file. Has the schema been modified?", field_entry.name()))
)?;
let termdict_source = self.termdict_composite.open_read(field).expect(
"Failed to open field term dictionary in composite file. Is the field indexed?",
);
let positions_file = self
let positions_source = self
.positions_composite
.open_read(field)
.expect("Index corrupted. Failed to open field positions in composite file.");
let positions_idx_file = self
let positions_idx_source = self
.positions_idx_composite
.open_read(field)
.expect("Index corrupted. Failed to open field positions in composite file.");
let inv_idx_reader = Arc::new(InvertedIndexReader::new(
TermDictionary::open(termdict_file)?,
postings_file,
positions_file,
positions_idx_file,
TermDictionary::from_source(&termdict_source),
postings_source,
positions_source,
positions_idx_source,
record_option,
)?);
));
// by releasing the lock in between, we may end up opening the inverting index
// twice, but this is fine.
@@ -287,7 +272,7 @@ impl SegmentReader {
.expect("Field reader cache lock poisoned. This should never happen.")
.insert(field, Arc::clone(&inv_idx_reader));
Ok(inv_idx_reader)
inv_idx_reader
}
/// Returns the segment id
@@ -310,26 +295,26 @@ impl SegmentReader {
}
/// Returns an iterator that will iterate over the alive document ids
pub fn doc_ids_alive<'a>(&'a self) -> impl Iterator<Item = DocId> + 'a {
(0u32..self.max_doc).filter(move |doc| !self.is_deleted(*doc))
pub fn doc_ids_alive(&self) -> SegmentReaderAliveDocsIterator<'_> {
SegmentReaderAliveDocsIterator::new(&self)
}
/// Summarize total space usage of this segment.
pub fn space_usage(&self) -> io::Result<SegmentSpaceUsage> {
Ok(SegmentSpaceUsage::new(
pub fn space_usage(&self) -> SegmentSpaceUsage {
SegmentSpaceUsage::new(
self.num_docs(),
self.termdict_composite.space_usage(),
self.postings_composite.space_usage(),
self.positions_composite.space_usage(),
self.positions_idx_composite.space_usage(),
self.fast_fields_readers.space_usage(),
self.fieldnorm_readers.space_usage(),
self.get_store_reader()?.space_usage(),
self.fieldnorms_composite.space_usage(),
self.get_store_reader().space_usage(),
self.delete_bitset_opt
.as_ref()
.map(DeleteBitSet::space_usage)
.unwrap_or(0),
))
)
}
}
@@ -339,6 +324,52 @@ impl fmt::Debug for SegmentReader {
}
}
/// Implements the iterator trait to allow easy iteration
/// over non-deleted ("alive") DocIds in a SegmentReader
pub struct SegmentReaderAliveDocsIterator<'a> {
reader: &'a SegmentReader,
max_doc: DocId,
current: DocId,
}
impl<'a> SegmentReaderAliveDocsIterator<'a> {
pub fn new(reader: &'a SegmentReader) -> SegmentReaderAliveDocsIterator<'a> {
SegmentReaderAliveDocsIterator {
reader,
max_doc: reader.max_doc(),
current: 0,
}
}
}
impl<'a> Iterator for SegmentReaderAliveDocsIterator<'a> {
type Item = DocId;
fn next(&mut self) -> Option<Self::Item> {
// TODO: Use TinySet (like in BitSetDocSet) to speed this process up
if self.current >= self.max_doc {
return None;
}
// find the next alive doc id
while self.reader.is_deleted(self.current) {
self.current += 1;
if self.current >= self.max_doc {
return None;
}
}
// capture the current alive DocId
let result = Some(self.current);
// move down the chain
self.current += 1;
result
}
}
#[cfg(test)]
mod test {
use crate::core::Index;
@@ -346,7 +377,7 @@ mod test {
use crate::DocId;
#[test]
fn test_alive_docs_iterator() -> crate::Result<()> {
fn test_alive_docs_iterator() {
let mut schema_builder = Schema::builder();
schema_builder.add_text_field("name", TEXT | STORED);
let schema = schema_builder.build();
@@ -354,26 +385,26 @@ mod test {
let name = schema.get_field("name").unwrap();
{
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(name => "tantivy"));
index_writer.add_document(doc!(name => "horse"));
index_writer.add_document(doc!(name => "jockey"));
index_writer.add_document(doc!(name => "cap"));
// we should now have one segment with two docs
index_writer.commit()?;
index_writer.commit().unwrap();
}
{
let mut index_writer2 = index.writer(50_000_000)?;
let mut index_writer2 = index.writer(50_000_000).unwrap();
index_writer2.delete_term(Term::from_field_text(name, "horse"));
index_writer2.delete_term(Term::from_field_text(name, "cap"));
// ok, now we should have a deleted doc
index_writer2.commit()?;
index_writer2.commit().unwrap();
}
let searcher = index.reader()?.searcher();
let searcher = index.reader().unwrap().searcher();
let docs: Vec<DocId> = searcher.segment_reader(0).doc_ids_alive().collect();
assert_eq!(vec![0u32, 2u32], docs);
Ok(())
}
}

19
src/directory/directory.rs
View File

@@ -3,7 +3,7 @@ use crate::directory::error::LockError;
use crate::directory::error::{DeleteError, OpenReadError, OpenWriteError};
use crate::directory::WatchCallback;
use crate::directory::WatchHandle;
use crate::directory::{FileSlice, WritePtr};
use crate::directory::{ReadOnlySource, WritePtr};
use std::fmt;
use std::io;
use std::io::Write;
@@ -11,6 +11,7 @@ use std::marker::Send;
use std::marker::Sync;
use std::path::Path;
use std::path::PathBuf;
use std::result;
use std::thread;
use std::time::Duration;
@@ -79,7 +80,7 @@ fn try_acquire_lock(
) -> Result<DirectoryLock, TryAcquireLockError> {
let mut write = directory.open_write(filepath).map_err(|e| match e {
OpenWriteError::FileAlreadyExists(_) => TryAcquireLockError::FileExists,
OpenWriteError::IOError { io_error, .. } => TryAcquireLockError::IOError(io_error),
OpenWriteError::IOError(io_error) => TryAcquireLockError::IOError(io_error.into()),
})?;
write.flush().map_err(TryAcquireLockError::IOError)?;
Ok(DirectoryLock::from(Box::new(DirectoryLockGuard {
@@ -116,19 +117,19 @@ pub trait Directory: DirectoryClone + fmt::Debug + Send + Sync + 'static {
/// change.
///
/// Specifically, subsequent writes or flushes should
/// have no effect on the returned `FileSlice` object.
/// have no effect on the returned `ReadOnlySource` object.
///
/// You should only use this to read files create with [Directory::open_write].
fn open_read(&self, path: &Path) -> Result<FileSlice, OpenReadError>;
fn open_read(&self, path: &Path) -> result::Result<ReadOnlySource, OpenReadError>;
/// Removes a file
///
/// Removing a file will not affect an eventual
/// existing FileSlice pointing to it.
/// existing ReadOnlySource pointing to it.
///
/// Removing a nonexistent file, yields a
/// `DeleteError::DoesNotExist`.
fn delete(&self, path: &Path) -> Result<(), DeleteError>;
fn delete(&self, path: &Path) -> result::Result<(), DeleteError>;
/// Returns true iff the file exists
fn exists(&self, path: &Path) -> bool;
@@ -138,7 +139,7 @@ pub trait Directory: DirectoryClone + fmt::Debug + Send + Sync + 'static {
///
/// Right after this call, the file should be created
/// and any subsequent call to `open_read` for the
/// same path should return a `FileSlice`.
/// same path should return a `ReadOnlySource`.
///
/// Write operations may be aggressively buffered.
/// The client of this trait is responsible for calling flush
@@ -152,7 +153,7 @@ pub trait Directory: DirectoryClone + fmt::Debug + Send + Sync + 'static {
/// was not called.
///
/// The file may not previously exist.
fn open_write(&self, path: &Path) -> Result<WritePtr, OpenWriteError>;
fn open_write(&mut self, path: &Path) -> Result<WritePtr, OpenWriteError>;
/// Reads the full content file that has been written using
/// atomic_write.
@@ -168,7 +169,7 @@ pub trait Directory: DirectoryClone + fmt::Debug + Send + Sync + 'static {
/// a partially written file.
///
/// The file may or may not previously exist.
fn atomic_write(&self, path: &Path, data: &[u8]) -> io::Result<()>;
fn atomic_write(&mut self, path: &Path, data: &[u8]) -> io::Result<()>;
/// Acquire a lock in the given directory.
///

256
src/directory/error.rs
View File

@@ -1,67 +1,162 @@
use crate::Version;
use std::error::Error as StdError;
use std::fmt;
use std::io;
use std::path::PathBuf;
/// Error while trying to acquire a directory lock.
#[derive(Debug, Error)]
#[derive(Debug, Fail)]
pub enum LockError {
/// Failed to acquired a lock as it is already held by another
/// client.
/// - In the context of a blocking lock, this means the lock was not released within some `timeout` period.
/// - In the context of a non-blocking lock, this means the lock was busy at the moment of the call.
#[error("Could not acquire lock as it is already held, possibly by a different process.")]
#[fail(
display = "Could not acquire lock as it is already held, possibly by a different process."
)]
LockBusy,
/// Trying to acquire a lock failed with an `IOError`
#[error("Failed to acquire the lock due to an io:Error.")]
#[fail(display = "Failed to acquire the lock due to an io:Error.")]
IOError(io::Error),
}
/// General IO error with an optional path to the offending file.
#[derive(Debug)]
pub struct IOError {
path: Option<PathBuf>,
err: io::Error,
}
impl Into<io::Error> for IOError {
fn into(self) -> io::Error {
self.err
}
}
impl fmt::Display for IOError {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
match self.path {
Some(ref path) => write!(f, "io error occurred on path '{:?}': '{}'", path, self.err),
None => write!(f, "io error occurred: '{}'", self.err),
}
}
}
impl StdError for IOError {
fn description(&self) -> &str {
"io error occurred"
}
fn cause(&self) -> Option<&dyn StdError> {
Some(&self.err)
}
}
impl IOError {
pub(crate) fn with_path(path: PathBuf, err: io::Error) -> Self {
IOError {
path: Some(path),
err,
}
}
}
impl From<io::Error> for IOError {
fn from(err: io::Error) -> IOError {
IOError { path: None, err }
}
}
/// Error that may occur when opening a directory
#[derive(Debug, Error)]
#[derive(Debug)]
pub enum OpenDirectoryError {
/// The underlying directory does not exists.
#[error("Directory does not exist: '{0}'.")]
DoesNotExist(PathBuf),
/// The path exists but is not a directory.
#[error("Path exists but is not a directory: '{0}'.")]
NotADirectory(PathBuf),
/// Failed to create a temp directory.
#[error("Failed to create a temporary directory: '{0}'.")]
FailedToCreateTempDir(io::Error),
/// IoError
#[error("IOError '{io_error:?}' while create directory in: '{directory_path:?}'.")]
IoError {
/// underlying io Error.
io_error: io::Error,
/// directory we tried to open.
directory_path: PathBuf,
},
IoError(io::Error),
}
impl From<io::Error> for OpenDirectoryError {
fn from(io_err: io::Error) -> Self {
OpenDirectoryError::IoError(io_err)
}
}
impl fmt::Display for OpenDirectoryError {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
match *self {
OpenDirectoryError::DoesNotExist(ref path) => {
write!(f, "the underlying directory '{:?}' does not exist", path)
}
OpenDirectoryError::NotADirectory(ref path) => {
write!(f, "the path '{:?}' exists but is not a directory", path)
}
OpenDirectoryError::IoError(ref err) => write!(
f,
"IOError while trying to open/create the directory. {:?}",
err
),
}
}
}
impl StdError for OpenDirectoryError {
fn description(&self) -> &str {
"error occurred while opening a directory"
}
fn cause(&self) -> Option<&dyn StdError> {
None
}
}
/// Error that may occur when starting to write in a file
#[derive(Debug, Error)]
#[derive(Debug)]
pub enum OpenWriteError {
/// Our directory is WORM, writing an existing file is forbidden.
/// Checkout the `Directory` documentation.
#[error("File already exists: '{0}'")]
FileAlreadyExists(PathBuf),
/// Any kind of IO error that happens when
/// writing in the underlying IO device.
#[error("IOError '{io_error:?}' while opening file for write: '{filepath}'.")]
IOError {
/// The underlying `io::Error`.
io_error: io::Error,
/// File path of the file that tantivy failed to open for write.
filepath: PathBuf,
},
IOError(IOError),
}
impl OpenWriteError {
pub(crate) fn wrap_io_error(io_error: io::Error, filepath: PathBuf) -> Self {
Self::IOError { io_error, filepath }
impl From<IOError> for OpenWriteError {
fn from(err: IOError) -> OpenWriteError {
OpenWriteError::IOError(err)
}
}
impl fmt::Display for OpenWriteError {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
match *self {
OpenWriteError::FileAlreadyExists(ref path) => {
write!(f, "the file '{:?}' already exists", path)
}
OpenWriteError::IOError(ref err) => write!(
f,
"an io error occurred while opening a file for writing: '{}'",
err
),
}
}
}
impl StdError for OpenWriteError {
fn description(&self) -> &str {
"error occurred while opening a file for writing"
}
fn cause(&self) -> Option<&dyn StdError> {
match *self {
OpenWriteError::FileAlreadyExists(_) => None,
OpenWriteError::IOError(ref err) => Some(err),
}
}
}
/// Type of index incompatibility between the library and the index found on disk
/// Used to catch and provide a hint to solve this incompatibility issue
pub enum Incompatibility {
@@ -122,46 +217,55 @@ impl fmt::Debug for Incompatibility {
}
/// Error that may occur when accessing a file read
#[derive(Debug, Error)]
#[derive(Debug)]
pub enum OpenReadError {
/// The file does not exists.
#[error("Files does not exists: {0:?}")]
FileDoesNotExist(PathBuf),
/// Any kind of io::Error.
#[error(
"IOError: '{io_error:?}' happened while opening the following file for Read: {filepath}."
)]
IOError {
/// The underlying `io::Error`.
io_error: io::Error,
/// File path of the file that tantivy failed to open for read.
filepath: PathBuf,
},
/// This library does not support the index version found in file footer.
#[error("Index version unsupported: {0:?}")]
IncompatibleIndex(Incompatibility),
}
impl OpenReadError {
pub(crate) fn wrap_io_error(io_error: io::Error, filepath: PathBuf) -> Self {
Self::IOError { io_error, filepath }
}
}
/// Error that may occur when trying to delete a file
#[derive(Debug, Error)]
pub enum DeleteError {
/// The file does not exists.
#[error("File does not exists: '{0}'.")]
FileDoesNotExist(PathBuf),
/// Any kind of IO error that happens when
/// interacting with the underlying IO device.
#[error("The following IO error happened while deleting file '{filepath}': '{io_error:?}'.")]
IOError {
/// The underlying `io::Error`.
io_error: io::Error,
/// File path of the file that tantivy failed to delete.
filepath: PathBuf,
},
IOError(IOError),
/// This library doesn't support the index version found on disk
IncompatibleIndex(Incompatibility),
}
impl From<IOError> for OpenReadError {
fn from(err: IOError) -> OpenReadError {
OpenReadError::IOError(err)
}
}
impl fmt::Display for OpenReadError {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
match *self {
OpenReadError::FileDoesNotExist(ref path) => {
write!(f, "the file '{:?}' does not exist", path)
}
OpenReadError::IOError(ref err) => write!(
f,
"an io error occurred while opening a file for reading: '{}'",
err
),
OpenReadError::IncompatibleIndex(ref footer) => {
write!(f, "Incompatible index format: {:?}", footer)
}
}
}
}
/// Error that may occur when trying to delete a file
#[derive(Debug)]
pub enum DeleteError {
/// The file does not exists.
FileDoesNotExist(PathBuf),
/// Any kind of IO error that happens when
/// interacting with the underlying IO device.
IOError(IOError),
}
impl From<IOError> for DeleteError {
fn from(err: IOError) -> DeleteError {
DeleteError::IOError(err)
}
}
impl From<Incompatibility> for OpenReadError {
@@ -169,3 +273,29 @@ impl From<Incompatibility> for OpenReadError {
OpenReadError::IncompatibleIndex(incompatibility)
}
}
impl fmt::Display for DeleteError {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
match *self {
DeleteError::FileDoesNotExist(ref path) => {
write!(f, "the file '{:?}' does not exist", path)
}
DeleteError::IOError(ref err) => {
write!(f, "an io error occurred while deleting a file: '{}'", err)
}
}
}
}
impl StdError for DeleteError {
fn description(&self) -> &str {
"error occurred while deleting a file"
}
fn cause(&self) -> Option<&dyn StdError> {
match *self {
DeleteError::FileDoesNotExist(_) => None,
DeleteError::IOError(ref err) => Some(err),
}
}
}

237
src/directory/file_slice.rs
View File

@@ -1,237 +0,0 @@
use stable_deref_trait::StableDeref;
use crate::common::HasLen;
use crate::directory::OwnedBytes;
use std::sync::Arc;
use std::{io, ops::Deref};
pub type BoxedData = Box<dyn Deref<Target = [u8]> + Send + Sync + 'static>;
/// Objects that represents files sections in tantivy.
///
/// By contract, whatever happens to the directory file, as long as a FileHandle
/// is alive, the data associated with it cannot be altered or destroyed.
///
/// The underlying behavior is therefore specific to the `Directory` that created it.
/// Despite its name, a `FileSlice` may or may not directly map to an actual file
/// on the filesystem.
pub trait FileHandle: 'static + Send + Sync + HasLen {
/// Reads a slice of bytes.
///
/// This method may panic if the range requested is invalid.
fn read_bytes(&self, from: usize, to: usize) -> io::Result<OwnedBytes>;
}
impl FileHandle for &'static [u8] {
fn read_bytes(&self, from: usize, to: usize) -> io::Result<OwnedBytes> {
let bytes = &self[from..to];
Ok(OwnedBytes::new(bytes))
}
}
impl<T: Deref<Target = [u8]>> HasLen for T {
fn len(&self) -> usize {
self.as_ref().len()
}
}
impl<B> From<B> for FileSlice
where
B: StableDeref + Deref<Target = [u8]> + 'static + Send + Sync,
{
fn from(bytes: B) -> FileSlice {
FileSlice::new(OwnedBytes::new(bytes))
}
}
/// Logical slice of read only file in tantivy.
//
/// It can be cloned and sliced cheaply.
///
#[derive(Clone)]
pub struct FileSlice {
data: Arc<Box<dyn FileHandle>>,
start: usize,
stop: usize,
}
impl FileSlice {
/// Wraps a FileHandle.
pub fn new<D>(data: D) -> Self
where
D: FileHandle,
{
let len = data.len();
FileSlice {
data: Arc::new(Box::new(data)),
start: 0,
stop: len,
}
}
/// Creates a fileslice that is just a view over a slice of the data.
///
/// # Panics
///
/// Panics if `to < from` or if `to` exceeds the filesize.
pub fn slice(&self, from: usize, to: usize) -> FileSlice {
assert!(to <= self.len());
assert!(to >= from);
FileSlice {
data: self.data.clone(),
start: self.start + from,
stop: self.start + to,
}
}
/// Creates an empty FileSlice
pub fn empty() -> FileSlice {
const EMPTY_SLICE: &[u8] = &[];
FileSlice::from(EMPTY_SLICE)
}
/// Returns a `OwnedBytes` with all of the data in the `FileSlice`.
///
/// The behavior is strongly dependant on the implementation of the underlying
/// `Directory` and the `FileSliceTrait` it creates.
/// In particular, it is up to the `Directory` implementation
/// to handle caching if needed.
pub fn read_bytes(&self) -> io::Result<OwnedBytes> {
self.data.read_bytes(self.start, self.stop)
}
/// Reads a specific slice of data.
///
/// This is equivalent to running `file_slice.slice(from, to).read_bytes()`.
pub fn read_bytes_slice(&self, from: usize, to: usize) -> io::Result<OwnedBytes> {
assert!(from <= to);
assert!(
self.start + to <= self.stop,
"`to` exceeds the fileslice length"
);
self.data.read_bytes(self.start + from, self.start + to)
}
/// Splits the FileSlice at the given offset and return two file slices.
/// `file_slice[..split_offset]` and `file_slice[split_offset..]`.
///
/// This operation is cheap and must not copy any underlying data.
pub fn split(self, left_len: usize) -> (FileSlice, FileSlice) {
let left = self.slice_to(left_len);
let right = self.slice_from(left_len);
(left, right)
}
/// Splits the file slice at the given offset and return two file slices.
/// `file_slice[..split_offset]` and `file_slice[split_offset..]`.
pub fn split_from_end(self, right_len: usize) -> (FileSlice, FileSlice) {
let left_len = self.len() - right_len;
self.split(left_len)
}
/// Like `.slice(...)` but enforcing only the `from`
/// boundary.
///
/// Equivalent to `.slice(from_offset, self.len())`
pub fn slice_from(&self, from_offset: usize) -> FileSlice {
self.slice(from_offset, self.len())
}
/// Like `.slice(...)` but enforcing only the `to`
/// boundary.
///
/// Equivalent to `.slice(0, to_offset)`
pub fn slice_to(&self, to_offset: usize) -> FileSlice {
self.slice(0, to_offset)
}
}
impl HasLen for FileSlice {
fn len(&self) -> usize {
self.stop - self.start
}
}
#[cfg(test)]
mod tests {
use super::{FileHandle, FileSlice};
use crate::common::HasLen;
use std::io;
#[test]
fn test_file_slice() -> io::Result<()> {
let file_slice = FileSlice::new(b"abcdef".as_ref());
assert_eq!(file_slice.len(), 6);
assert_eq!(file_slice.slice_from(2).read_bytes()?.as_slice(), b"cdef");
assert_eq!(file_slice.slice_to(2).read_bytes()?.as_slice(), b"ab");
assert_eq!(
file_slice
.slice_from(1)
.slice_to(2)
.read_bytes()?
.as_slice(),
b"bc"
);
{
let (left, right) = file_slice.clone().split(0);
assert_eq!(left.read_bytes()?.as_slice(), b"");
assert_eq!(right.read_bytes()?.as_slice(), b"abcdef");
}
{
let (left, right) = file_slice.clone().split(2);
assert_eq!(left.read_bytes()?.as_slice(), b"ab");
assert_eq!(right.read_bytes()?.as_slice(), b"cdef");
}
{
let (left, right) = file_slice.clone().split_from_end(0);
assert_eq!(left.read_bytes()?.as_slice(), b"abcdef");
assert_eq!(right.read_bytes()?.as_slice(), b"");
}
{
let (left, right) = file_slice.clone().split_from_end(2);
assert_eq!(left.read_bytes()?.as_slice(), b"abcd");
assert_eq!(right.read_bytes()?.as_slice(), b"ef");
}
Ok(())
}
#[test]
fn test_file_slice_trait_slice_len() {
let blop: &'static [u8] = b"abc";
let owned_bytes: Box<dyn FileHandle> = Box::new(blop);
assert_eq!(owned_bytes.len(), 3);
}
#[test]
fn test_slice_simple_read() -> io::Result<()> {
let slice = FileSlice::new(&b"abcdef"[..]);
assert_eq!(slice.len(), 6);
assert_eq!(slice.read_bytes()?.as_ref(), b"abcdef");
assert_eq!(slice.slice(1, 4).read_bytes()?.as_ref(), b"bcd");
Ok(())
}
#[test]
fn test_slice_read_slice() -> io::Result<()> {
let slice_deref = FileSlice::new(&b"abcdef"[..]);
assert_eq!(slice_deref.read_bytes_slice(1, 4)?.as_ref(), b"bcd");
Ok(())
}
#[test]
#[should_panic(expected = "assertion failed: from <= to")]
fn test_slice_read_slice_invalid_range() {
let slice_deref = FileSlice::new(&b"abcdef"[..]);
assert_eq!(slice_deref.read_bytes_slice(1, 0).unwrap().as_ref(), b"bcd");
}
#[test]
#[should_panic(expected = "`to` exceeds the fileslice length")]
fn test_slice_read_slice_invalid_range_exceeds() {
let slice_deref = FileSlice::new(&b"abcdef"[..]);
assert_eq!(
slice_deref.read_bytes_slice(0, 10).unwrap().as_ref(),
b"bcd"
);
}
}

108
src/directory/footer.rs
View File

@@ -1,8 +1,9 @@
use crate::common::{BinarySerializable, CountingWriter, FixedSize, HasLen, VInt};
use crate::common::{BinarySerializable, CountingWriter, FixedSize, VInt};
use crate::directory::error::Incompatibility;
use crate::directory::FileSlice;
use crate::directory::read_only_source::ReadOnlySource;
use crate::directory::{AntiCallToken, TerminatingWrite};
use crate::Version;
use byteorder::{ByteOrder, LittleEndian, WriteBytesExt};
use crc32fast::Hasher;
use std::io;
use std::io::Write;
@@ -63,26 +64,26 @@ impl Footer {
let mut counting_write = CountingWriter::wrap(&mut write);
self.serialize(&mut counting_write)?;
let written_len = counting_write.written_bytes();
(written_len as u32).serialize(write)?;
write.write_u32::<LittleEndian>(written_len as u32)?;
Ok(())
}
pub fn extract_footer(file: FileSlice) -> io::Result<(Footer, FileSlice)> {
if file.len() < 4 {
pub fn extract_footer(source: ReadOnlySource) -> Result<(Footer, ReadOnlySource), io::Error> {
if source.len() < 4 {
return Err(io::Error::new(
io::ErrorKind::UnexpectedEof,
format!(
"File corrupted. The file is smaller than 4 bytes (len={}).",
file.len()
source.len()
),
));
}
let (body_footer, footer_len_file) = file.split_from_end(u32::SIZE_IN_BYTES);
let mut footer_len_bytes = footer_len_file.read_bytes()?;
let footer_len = u32::deserialize(&mut footer_len_bytes)? as usize;
let (body, footer) = body_footer.split_from_end(footer_len);
let mut footer_bytes = footer.read_bytes()?;
let footer = Footer::deserialize(&mut footer_bytes)?;
let (body_footer, footer_len_bytes) = source.split_from_end(u32::SIZE_IN_BYTES);
let footer_len = LittleEndian::read_u32(footer_len_bytes.as_slice()) as usize;
let body_len = body_footer.len() - footer_len;
let (body, footer_data) = body_footer.split(body_len);
let mut cursor = footer_data.as_slice();
let footer = Footer::deserialize(&mut cursor)?;
Ok((footer, body))
}
@@ -93,24 +94,12 @@ impl Footer {
match &self.versioned_footer {
VersionedFooter::V1 {
crc32: _crc,
store_compression,
store_compression: compression,
} => {
if &library_version.store_compression != store_compression {
if &library_version.store_compression != compression {
return Err(Incompatibility::CompressionMismatch {
library_compression_format: library_version.store_compression.to_string(),
index_compression_format: store_compression.to_string(),
});
}
Ok(())
}
VersionedFooter::V2 {
crc32: _crc,
store_compression,
} => {
if &library_version.store_compression != store_compression {
return Err(Incompatibility::CompressionMismatch {
library_compression_format: library_version.store_compression.to_string(),
index_compression_format: store_compression.to_string(),
index_compression_format: compression.to_string(),
});
}
Ok(())
@@ -131,29 +120,24 @@ pub enum VersionedFooter {
crc32: CrcHashU32,
store_compression: String,
},
// Introduction of the Block WAND information.
V2 {
crc32: CrcHashU32,
store_compression: String,
},
}
impl BinarySerializable for VersionedFooter {
fn serialize<W: io::Write>(&self, writer: &mut W) -> io::Result<()> {
let mut buf = Vec::new();
match self {
VersionedFooter::V2 {
VersionedFooter::V1 {
crc32,
store_compression: compression,
} => {
// Serializes a valid `VersionedFooter` or panics if the version is unknown
// [ version | crc_hash | compression_mode ]
// [ 0..4 | 4..8 | variable ]
BinarySerializable::serialize(&2u32, &mut buf)?;
BinarySerializable::serialize(&1u32, &mut buf)?;
BinarySerializable::serialize(crc32, &mut buf)?;
BinarySerializable::serialize(compression, &mut buf)?;
}
VersionedFooter::V1 { .. } | VersionedFooter::UnknownVersion => {
VersionedFooter::UnknownVersion => {
return Err(io::Error::new(
io::ErrorKind::InvalidInput,
"Cannot serialize an unknown versioned footer ",
@@ -182,30 +166,22 @@ impl BinarySerializable for VersionedFooter {
reader.read_exact(&mut buf[..])?;
let mut cursor = &buf[..];
let version = u32::deserialize(&mut cursor)?;
if version != 1 && version != 2 {
return Ok(VersionedFooter::UnknownVersion);
}
let crc32 = u32::deserialize(&mut cursor)?;
let store_compression = String::deserialize(&mut cursor)?;
Ok(if version == 1 {
VersionedFooter::V1 {
if version == 1 {
let crc32 = u32::deserialize(&mut cursor)?;
let compression = String::deserialize(&mut cursor)?;
Ok(VersionedFooter::V1 {
crc32,
store_compression,
}
store_compression: compression,
})
} else {
assert_eq!(version, 2);
VersionedFooter::V2 {
crc32,
store_compression,
}
})
Ok(VersionedFooter::UnknownVersion)
}
}
}
impl VersionedFooter {
pub fn crc(&self) -> Option<CrcHashU32> {
match self {
VersionedFooter::V2 { crc32, .. } => Some(*crc32),
VersionedFooter::V1 { crc32, .. } => Some(*crc32),
VersionedFooter::UnknownVersion { .. } => None,
}
@@ -243,7 +219,7 @@ impl<W: TerminatingWrite> Write for FooterProxy<W> {
impl<W: TerminatingWrite> TerminatingWrite for FooterProxy<W> {
fn terminate_ref(&mut self, _: AntiCallToken) -> io::Result<()> {
let crc32 = self.hasher.take().unwrap().finalize();
let footer = Footer::new(VersionedFooter::V2 {
let footer = Footer::new(VersionedFooter::V1 {
crc32,
store_compression: crate::store::COMPRESSION.to_string(),
});
@@ -270,17 +246,17 @@ mod tests {
let mut vec = Vec::new();
let footer_proxy = FooterProxy::new(&mut vec);
assert!(footer_proxy.terminate().is_ok());
if crate::store::COMPRESSION == "lz4" {
assert_eq!(vec.len(), 158);
} else {
assert_eq!(vec.len(), 167);
}
assert_eq!(vec.len(), 167);
let footer = Footer::deserialize(&mut &vec[..]).unwrap();
assert!(matches!(
footer.versioned_footer,
VersionedFooter::V2 { store_compression, .. }
if store_compression == crate::store::COMPRESSION
));
if let VersionedFooter::V1 {
crc32: _,
store_compression,
} = footer.versioned_footer
{
assert_eq!(store_compression, crate::store::COMPRESSION);
} else {
panic!("Versioned footer should be V1.");
}
assert_eq!(&footer.version, crate::version());
}
@@ -288,7 +264,7 @@ mod tests {
fn test_serialize_deserialize_footer() {
let mut buffer = Vec::new();
let crc32 = 123456u32;
let footer: Footer = Footer::new(VersionedFooter::V2 {
let footer: Footer = Footer::new(VersionedFooter::V1 {
crc32,
store_compression: "lz4".to_string(),
});
@@ -300,7 +276,7 @@ mod tests {
#[test]
fn footer_length() {
let crc32 = 1111111u32;
let versioned_footer = VersionedFooter::V2 {
let versioned_footer = VersionedFooter::V1 {
crc32,
store_compression: "lz4".to_string(),
};
@@ -321,7 +297,7 @@ mod tests {
// versionned footer length
12 | 128,
// index format version
2,
1,
0,
0,
0,
@@ -340,7 +316,7 @@ mod tests {
let versioned_footer = VersionedFooter::deserialize(&mut cursor).unwrap();
assert!(cursor.is_empty());
let expected_crc: u32 = LittleEndian::read_u32(&v_footer_bytes[5..9]) as CrcHashU32;
let expected_versioned_footer: VersionedFooter = VersionedFooter::V2 {
let expected_versioned_footer: VersionedFooter = VersionedFooter::V1 {
crc32: expected_crc,
store_compression: "lz4".to_string(),
};

92
src/directory/managed_directory.rs
View File

@@ -1,16 +1,17 @@
use crate::core::{MANAGED_FILEPATH, META_FILEPATH};
use crate::directory::error::{DeleteError, LockError, OpenReadError, OpenWriteError};
use crate::core::MANAGED_FILEPATH;
use crate::directory::error::{DeleteError, IOError, LockError, OpenReadError, OpenWriteError};
use crate::directory::footer::{Footer, FooterProxy};
use crate::directory::DirectoryLock;
use crate::directory::GarbageCollectionResult;
use crate::directory::Lock;
use crate::directory::META_LOCK;
use crate::directory::{FileSlice, WritePtr};
use crate::directory::{ReadOnlySource, WritePtr};
use crate::directory::{WatchCallback, WatchHandle};
use crate::error::DataCorruption;
use crate::Directory;
use crc32fast::Hasher;
use serde_json;
use std::collections::HashSet;
use std::io;
use std::io::Write;
@@ -53,7 +54,7 @@ struct MetaInformation {
/// Saves the file containing the list of existing files
/// that were created by tantivy.
fn save_managed_paths(
directory: &dyn Directory,
directory: &mut dyn Directory,
wlock: &RwLockWriteGuard<'_, MetaInformation>,
) -> io::Result<()> {
let mut w = serde_json::to_vec(&wlock.managed_paths)?;
@@ -86,7 +87,7 @@ impl ManagedDirectory {
directory: Box::new(directory),
meta_informations: Arc::default(),
}),
io_err @ Err(OpenReadError::IOError { .. }) => Err(io_err.err().unwrap().into()),
Err(OpenReadError::IOError(e)) => Err(From::from(e)),
Err(OpenReadError::IncompatibleIndex(incompatibility)) => {
// For the moment, this should never happen `meta.json`
// do not have any footer and cannot detect incompatibility.
@@ -168,7 +169,7 @@ impl ManagedDirectory {
DeleteError::FileDoesNotExist(_) => {
deleted_files.push(file_to_delete.clone());
}
DeleteError::IOError { .. } => {
DeleteError::IOError(_) => {
failed_to_delete_files.push(file_to_delete.clone());
if !cfg!(target_os = "windows") {
// On windows, delete is expected to fail if the file
@@ -212,7 +213,7 @@ impl ManagedDirectory {
/// File starting by "." are reserved to locks.
/// They are not managed and cannot be subjected
/// to garbage collection.
fn register_file_as_managed(&self, filepath: &Path) -> io::Result<()> {
fn register_file_as_managed(&mut self, filepath: &Path) -> io::Result<()> {
// Files starting by "." (e.g. lock files) are not managed.
if !is_managed(filepath) {
return Ok(());
@@ -223,7 +224,7 @@ impl ManagedDirectory {
.expect("Managed file lock poisoned");
let has_changed = meta_wlock.managed_paths.insert(filepath.to_owned());
if has_changed {
save_managed_paths(self.directory.as_ref(), &meta_wlock)?;
save_managed_paths(self.directory.as_mut(), &meta_wlock)?;
}
Ok(())
}
@@ -231,19 +232,10 @@ impl ManagedDirectory {
/// Verify checksum of a managed file
pub fn validate_checksum(&self, path: &Path) -> result::Result<bool, OpenReadError> {
let reader = self.directory.open_read(path)?;
let (footer, data) =
Footer::extract_footer(reader).map_err(|io_error| OpenReadError::IOError {
io_error,
filepath: path.to_path_buf(),
})?;
let bytes = data
.read_bytes()
.map_err(|io_error| OpenReadError::IOError {
filepath: path.to_path_buf(),
io_error,
})?;
let (footer, data) = Footer::extract_footer(reader)
.map_err(|err| IOError::with_path(path.to_path_buf(), err))?;
let mut hasher = Hasher::new();
hasher.update(bytes.as_slice());
hasher.update(data.as_slice());
let crc = hasher.finalize();
Ok(footer
.versioned_footer
@@ -254,37 +246,35 @@ impl ManagedDirectory {
/// List files for which checksum does not match content
pub fn list_damaged(&self) -> result::Result<HashSet<PathBuf>, OpenReadError> {
let mut managed_paths = self
let mut hashset = HashSet::new();
let managed_paths = self
.meta_informations
.read()
.expect("Managed directory rlock poisoned in list damaged.")
.managed_paths
.clone();
managed_paths.remove(*META_FILEPATH);
let mut damaged_files = HashSet::new();
for path in managed_paths {
for path in managed_paths.into_iter() {
if !self.validate_checksum(&path)? {
damaged_files.insert(path);
hashset.insert(path);
}
}
Ok(damaged_files)
Ok(hashset)
}
}
impl Directory for ManagedDirectory {
fn open_read(&self, path: &Path) -> result::Result<FileSlice, OpenReadError> {
let file_slice = self.directory.open_read(path)?;
let (footer, reader) = Footer::extract_footer(file_slice)
.map_err(|io_error| OpenReadError::wrap_io_error(io_error, path.to_path_buf()))?;
fn open_read(&self, path: &Path) -> result::Result<ReadOnlySource, OpenReadError> {
let read_only_source = self.directory.open_read(path)?;
let (footer, reader) = Footer::extract_footer(read_only_source)
.map_err(|err| IOError::with_path(path.to_path_buf(), err))?;
footer.is_compatible()?;
Ok(reader)
}
fn open_write(&self, path: &Path) -> result::Result<WritePtr, OpenWriteError> {
fn open_write(&mut self, path: &Path) -> result::Result<WritePtr, OpenWriteError> {
self.register_file_as_managed(path)
.map_err(|io_error| OpenWriteError::wrap_io_error(io_error, path.to_path_buf()))?;
.map_err(|e| IOError::with_path(path.to_owned(), e))?;
Ok(io::BufWriter::new(Box::new(FooterProxy::new(
self.directory
.open_write(path)?
@@ -294,7 +284,7 @@ impl Directory for ManagedDirectory {
))))
}
fn atomic_write(&self, path: &Path, data: &[u8]) -> io::Result<()> {
fn atomic_write(&mut self, path: &Path, data: &[u8]) -> io::Result<()> {
self.register_file_as_managed(path)?;
self.directory.atomic_write(path, data)
}
@@ -408,37 +398,39 @@ mod tests_mmap_specific {
}
#[test]
fn test_checksum() -> crate::Result<()> {
fn test_checksum() {
let test_path1: &'static Path = Path::new("some_path_for_test");
let test_path2: &'static Path = Path::new("other_test_path");
let tempdir = TempDir::new().unwrap();
let tempdir_path = PathBuf::from(tempdir.path());
let mmap_directory = MmapDirectory::open(&tempdir_path)?;
let managed_directory = ManagedDirectory::wrap(mmap_directory)?;
let mut write = managed_directory.open_write(test_path1)?;
write.write_all(&[0u8, 1u8])?;
write.terminate()?;
let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
let mut managed_directory = ManagedDirectory::wrap(mmap_directory).unwrap();
let mut write = managed_directory.open_write(test_path1).unwrap();
write.write_all(&[0u8, 1u8]).unwrap();
write.terminate().unwrap();
let mut write = managed_directory.open_write(test_path2)?;
write.write_all(&[3u8, 4u8, 5u8])?;
write.terminate()?;
let mut write = managed_directory.open_write(test_path2).unwrap();
write.write_all(&[3u8, 4u8, 5u8]).unwrap();
write.terminate().unwrap();
let read_file = managed_directory.open_read(test_path2)?.read_bytes()?;
assert_eq!(read_file.as_slice(), &[3u8, 4u8, 5u8]);
let read_source = managed_directory.open_read(test_path2).unwrap();
assert_eq!(read_source.as_slice(), &[3u8, 4u8, 5u8]);
assert!(managed_directory.list_damaged().unwrap().is_empty());
let mut corrupted_path = tempdir_path.clone();
corrupted_path.push(test_path2);
let mut file = OpenOptions::new().write(true).open(&corrupted_path)?;
file.write_all(&[255u8])?;
file.flush()?;
let mut file = OpenOptions::new()
.write(true)
.open(&corrupted_path)
.unwrap();
file.write_all(&[255u8]).unwrap();
file.flush().unwrap();
drop(file);
let damaged = managed_directory.list_damaged()?;
let damaged = managed_directory.list_damaged().unwrap();
assert_eq!(damaged.len(), 1);
assert!(damaged.contains(test_path2));
Ok(())
}
}

149
src/directory/mmap_directory.rs
View File

@@ -1,23 +1,29 @@
use fs2;
use notify;
use self::fs2::FileExt;
use self::notify::RawEvent;
use self::notify::RecursiveMode;
use self::notify::Watcher;
use crate::core::META_FILEPATH;
use crate::directory::error::LockError;
use crate::directory::error::{DeleteError, OpenDirectoryError, OpenReadError, OpenWriteError};
use crate::directory::error::{
DeleteError, IOError, OpenDirectoryError, OpenReadError, OpenWriteError,
};
use crate::directory::read_only_source::BoxedData;
use crate::directory::AntiCallToken;
use crate::directory::BoxedData;
use crate::directory::Directory;
use crate::directory::DirectoryLock;
use crate::directory::FileSlice;
use crate::directory::Lock;
use crate::directory::ReadOnlySource;
use crate::directory::WatchCallback;
use crate::directory::WatchCallbackList;
use crate::directory::WatchHandle;
use crate::directory::{TerminatingWrite, WritePtr};
use fs2::FileExt;
use atomicwrites;
use memmap::Mmap;
use notify::RawEvent;
use notify::RecursiveMode;
use notify::Watcher;
use serde::{Deserialize, Serialize};
use stable_deref_trait::StableDeref;
use std::collections::HashMap;
use std::convert::From;
use std::fmt;
use std::fs::OpenOptions;
@@ -32,7 +38,6 @@ use std::sync::Mutex;
use std::sync::RwLock;
use std::sync::Weak;
use std::thread;
use std::{collections::HashMap, ops::Deref};
use tempfile::TempDir;
/// Create a default io error given a string.
@@ -43,17 +48,17 @@ pub(crate) fn make_io_err(msg: String) -> io::Error {
/// Returns None iff the file exists, can be read, but is empty (and hence
/// cannot be mmapped)
fn open_mmap(full_path: &Path) -> result::Result<Option<Mmap>, OpenReadError> {
let file = File::open(full_path).map_err(|io_err| {
if io_err.kind() == io::ErrorKind::NotFound {
OpenReadError::FileDoesNotExist(full_path.to_path_buf())
let file = File::open(full_path).map_err(|e| {
if e.kind() == io::ErrorKind::NotFound {
OpenReadError::FileDoesNotExist(full_path.to_owned())
} else {
OpenReadError::wrap_io_error(io_err, full_path.to_path_buf())
OpenReadError::IOError(IOError::with_path(full_path.to_owned(), e))
}
})?;
let meta_data = file
.metadata()
.map_err(|io_err| OpenReadError::wrap_io_error(io_err, full_path.to_owned()))?;
.map_err(|e| IOError::with_path(full_path.to_owned(), e))?;
if meta_data.len() == 0 {
// if the file size is 0, it will not be possible
// to mmap the file, so we return None
@@ -63,7 +68,7 @@ fn open_mmap(full_path: &Path) -> result::Result<Option<Mmap>, OpenReadError> {
unsafe {
memmap::Mmap::map(&file)
.map(Some)
.map_err(|io_err| OpenReadError::wrap_io_error(io_err, full_path.to_path_buf()))
.map_err(|e| From::from(IOError::with_path(full_path.to_owned(), e)))
}
}
@@ -182,10 +187,6 @@ impl WatcherWrapper {
}
}
}
})
.map_err(|io_error| OpenDirectoryError::IoError {
io_error,
directory_path: path.to_path_buf(),
})?;
Ok(WatcherWrapper {
_watcher: Mutex::new(watcher),
@@ -223,13 +224,17 @@ struct MmapDirectoryInner {
}
impl MmapDirectoryInner {
fn new(root_path: PathBuf, temp_directory: Option<TempDir>) -> MmapDirectoryInner {
MmapDirectoryInner {
fn new(
root_path: PathBuf,
temp_directory: Option<TempDir>,
) -> Result<MmapDirectoryInner, OpenDirectoryError> {
let mmap_directory_inner = MmapDirectoryInner {
root_path,
mmap_cache: Default::default(),
_temp_directory: temp_directory,
watcher: RwLock::new(None),
}
};
Ok(mmap_directory_inner)
}
fn watch(&self, watch_callback: WatchCallback) -> crate::Result<WatchHandle> {
@@ -263,11 +268,14 @@ impl fmt::Debug for MmapDirectory {
}
impl MmapDirectory {
fn new(root_path: PathBuf, temp_directory: Option<TempDir>) -> MmapDirectory {
let inner = MmapDirectoryInner::new(root_path, temp_directory);
MmapDirectory {
fn new(
root_path: PathBuf,
temp_directory: Option<TempDir>,
) -> Result<MmapDirectory, OpenDirectoryError> {
let inner = MmapDirectoryInner::new(root_path, temp_directory)?;
Ok(MmapDirectory {
inner: Arc::new(inner),
}
})
}
/// Creates a new MmapDirectory in a temporary directory.
@@ -275,11 +283,9 @@ impl MmapDirectory {
/// This is mostly useful to test the MmapDirectory itself.
/// For your unit tests, prefer the RAMDirectory.
pub fn create_from_tempdir() -> Result<MmapDirectory, OpenDirectoryError> {
let tempdir = TempDir::new().map_err(OpenDirectoryError::FailedToCreateTempDir)?;
Ok(MmapDirectory::new(
tempdir.path().to_path_buf(),
Some(tempdir),
))
let tempdir = TempDir::new().map_err(OpenDirectoryError::IoError)?;
let tempdir_path = PathBuf::from(tempdir.path());
MmapDirectory::new(tempdir_path, Some(tempdir))
}
/// Opens a MmapDirectory in a directory.
@@ -297,7 +303,7 @@ impl MmapDirectory {
directory_path,
)))
} else {
Ok(MmapDirectory::new(PathBuf::from(directory_path), None))
Ok(MmapDirectory::new(PathBuf::from(directory_path), None)?)
}
}
@@ -401,20 +407,8 @@ impl TerminatingWrite for SafeFileWriter {
}
}
#[derive(Clone)]
struct MmapArc(Arc<Box<dyn Deref<Target = [u8]> + Send + Sync>>);
impl Deref for MmapArc {
type Target = [u8];
fn deref(&self) -> &[u8] {
self.0.deref()
}
}
unsafe impl StableDeref for MmapArc {}
impl Directory for MmapDirectory {
fn open_read(&self, path: &Path) -> result::Result<FileSlice, OpenReadError> {
fn open_read(&self, path: &Path) -> result::Result<ReadOnlySource, OpenReadError> {
debug!("Open Read {:?}", path);
let full_path = self.resolve_path(path);
@@ -424,14 +418,12 @@ impl Directory for MmapDirectory {
on mmap cache while reading {:?}",
path
);
let io_err = make_io_err(msg);
OpenReadError::wrap_io_error(io_err, path.to_path_buf())
IOError::with_path(path.to_owned(), make_io_err(msg))
})?;
if let Some(mmap_arc) = mmap_cache.get_mmap(&full_path)? {
Ok(FileSlice::from(MmapArc(mmap_arc)))
} else {
Ok(FileSlice::empty())
}
Ok(mmap_cache
.get_mmap(&full_path)?
.map(ReadOnlySource::from)
.unwrap_or_else(ReadOnlySource::empty))
}
/// Any entry associated to the path in the mmap will be
@@ -439,18 +431,14 @@ impl Directory for MmapDirectory {
fn delete(&self, path: &Path) -> result::Result<(), DeleteError> {
let full_path = self.resolve_path(path);
match fs::remove_file(&full_path) {
Ok(_) => self.sync_directory().map_err(|e| DeleteError::IOError {
io_error: e,
filepath: path.to_path_buf(),
}),
Ok(_) => self
.sync_directory()
.map_err(|e| IOError::with_path(path.to_owned(), e).into()),
Err(e) => {
if e.kind() == io::ErrorKind::NotFound {
Err(DeleteError::FileDoesNotExist(path.to_owned()))
} else {
Err(DeleteError::IOError {
io_error: e,
filepath: path.to_path_buf(),
})
Err(IOError::with_path(path.to_owned(), e).into())
}
}
}
@@ -461,7 +449,7 @@ impl Directory for MmapDirectory {
full_path.exists()
}
fn open_write(&self, path: &Path) -> Result<WritePtr, OpenWriteError> {
fn open_write(&mut self, path: &Path) -> Result<WritePtr, OpenWriteError> {
debug!("Open Write {:?}", path);
let full_path = self.resolve_path(path);
@@ -470,22 +458,22 @@ impl Directory for MmapDirectory {
.create_new(true)
.open(full_path);
let mut file = open_res.map_err(|io_err| {
if io_err.kind() == io::ErrorKind::AlreadyExists {
OpenWriteError::FileAlreadyExists(path.to_path_buf())
let mut file = open_res.map_err(|err| {
if err.kind() == io::ErrorKind::AlreadyExists {
OpenWriteError::FileAlreadyExists(path.to_owned())
} else {
OpenWriteError::wrap_io_error(io_err, path.to_path_buf())
IOError::with_path(path.to_owned(), err).into()
}
})?;
// making sure the file is created.
file.flush()
.map_err(|io_error| OpenWriteError::wrap_io_error(io_error, path.to_path_buf()))?;
.map_err(|e| IOError::with_path(path.to_owned(), e))?;
// Apparetntly, on some filesystem syncing the parent
// directory is required.
self.sync_directory()
.map_err(|io_err| OpenWriteError::wrap_io_error(io_err, path.to_path_buf()))?;
.map_err(|e| IOError::with_path(path.to_owned(), e))?;
let writer = SafeFileWriter::new(file);
Ok(BufWriter::new(Box::new(writer)))
@@ -496,28 +484,25 @@ impl Directory for MmapDirectory {
let mut buffer = Vec::new();
match File::open(&full_path) {
Ok(mut file) => {
file.read_to_end(&mut buffer).map_err(|io_error| {
OpenReadError::wrap_io_error(io_error, path.to_path_buf())
})?;
file.read_to_end(&mut buffer)
.map_err(|e| IOError::with_path(path.to_owned(), e))?;
Ok(buffer)
}
Err(io_error) => {
if io_error.kind() == io::ErrorKind::NotFound {
Err(e) => {
if e.kind() == io::ErrorKind::NotFound {
Err(OpenReadError::FileDoesNotExist(path.to_owned()))
} else {
Err(OpenReadError::wrap_io_error(io_error, path.to_path_buf()))
Err(IOError::with_path(path.to_owned(), e).into())
}
}
}
}
fn atomic_write(&self, path: &Path, content: &[u8]) -> io::Result<()> {
fn atomic_write(&mut self, path: &Path, data: &[u8]) -> io::Result<()> {
debug!("Atomic Write {:?}", path);
let mut tempfile = tempfile::Builder::new().tempfile_in(&self.inner.root_path)?;
tempfile.write_all(content)?;
tempfile.flush()?;
let full_path = self.resolve_path(path);
tempfile.into_temp_path().persist(full_path)?;
let meta_file = atomicwrites::AtomicFile::new(full_path, atomicwrites::AllowOverwrite);
meta_file.write(|f| f.write_all(data))?;
Ok(())
}
@@ -553,10 +538,10 @@ mod tests {
// The following tests are specific to the MmapDirectory
use super::*;
use crate::indexer::LogMergePolicy;
use crate::schema::{Schema, SchemaBuilder, TEXT};
use crate::Index;
use crate::ReloadPolicy;
use crate::{common::HasLen, indexer::LogMergePolicy};
use std::fs;
use std::sync::atomic::{AtomicUsize, Ordering};
@@ -571,7 +556,7 @@ mod tests {
// cannot be mmapped.
//
// In that case the directory returns a SharedVecSlice.
let mmap_directory = MmapDirectory::create_from_tempdir().unwrap();
let mut mmap_directory = MmapDirectory::create_from_tempdir().unwrap();
let path = PathBuf::from("test");
{
let mut w = mmap_directory.open_write(&path).unwrap();
@@ -587,7 +572,7 @@ mod tests {
// here we test if the cache releases
// mmaps correctly.
let mmap_directory = MmapDirectory::create_from_tempdir().unwrap();
let mut mmap_directory = MmapDirectory::create_from_tempdir().unwrap();
let num_paths = 10;
let paths: Vec<PathBuf> = (0..num_paths)
.map(|i| PathBuf::from(&*format!("file_{}", i)))
@@ -678,7 +663,7 @@ mod tests {
{
let index = Index::create(mmap_directory.clone(), schema).unwrap();
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
let mut log_merge_policy = LogMergePolicy::default();
log_merge_policy.set_min_merge_size(3);
index_writer.set_merge_policy(Box::new(log_merge_policy));

20
src/directory/mod.rs
View File

@@ -9,12 +9,14 @@ mod mmap_directory;
mod directory;
mod directory_lock;
mod file_slice;
mod footer;
mod managed_directory;
mod owned_bytes;
mod ram_directory;
mod read_only_source;
mod spilling_writer;
mod watch_event_router;
mod persistor;
/// Errors specific to the directory module.
pub mod error;
@@ -22,14 +24,12 @@ pub mod error;
pub use self::directory::DirectoryLock;
pub use self::directory::{Directory, DirectoryClone};
pub use self::directory_lock::{Lock, INDEX_WRITER_LOCK, META_LOCK};
pub(crate) use self::file_slice::BoxedData;
pub use self::file_slice::{FileHandle, FileSlice};
pub use self::owned_bytes::OwnedBytes;
pub use self::ram_directory::RAMDirectory;
pub use self::read_only_source::ReadOnlySource;
pub(crate) use self::spilling_writer::{SpillingResult, SpillingWriter};
pub use self::watch_event_router::{WatchCallback, WatchCallbackList, WatchHandle};
use std::io::{self, BufWriter, Write};
use std::path::PathBuf;
/// Outcome of the Garbage collection
pub struct GarbageCollectionResult {
/// List of files that were deleted in this cycle
@@ -83,10 +83,16 @@ impl<W: TerminatingWrite> TerminatingWrite for BufWriter<W> {
}
}
impl TerminatingWrite for Vec<u8> {
fn terminate_ref(&mut self, _a: AntiCallToken) -> io::Result<()> {
Ok(())
}
}
#[cfg(test)]
impl<'a> TerminatingWrite for &'a mut Vec<u8> {
fn terminate_ref(&mut self, _a: AntiCallToken) -> io::Result<()> {
self.flush()
Ok(())
}
}

255
src/directory/owned_bytes.rs
View File

@@ -1,255 +0,0 @@
use crate::directory::FileHandle;
use stable_deref_trait::StableDeref;
use std::mem;
use std::ops::Deref;
use std::sync::Arc;
use std::{fmt, io};
/// An OwnedBytes simply wraps an object that owns a slice of data and exposes
/// this data as a static slice.
///
/// The backing object is required to be `StableDeref`.
#[derive(Clone)]
pub struct OwnedBytes {
data: &'static [u8],
box_stable_deref: Arc<dyn Deref<Target = [u8]> + Sync + Send>,
}
impl FileHandle for OwnedBytes {
fn read_bytes(&self, from: usize, to: usize) -> io::Result<OwnedBytes> {
Ok(self.slice(from, to))
}
}
impl OwnedBytes {
/// Creates an empty `OwnedBytes`.
pub fn empty() -> OwnedBytes {
OwnedBytes::new(&[][..])
}
/// Creates an `OwnedBytes` intance given a `StableDeref` object.
pub fn new<T: StableDeref + Deref<Target = [u8]> + 'static + Send + Sync>(
data_holder: T,
) -> OwnedBytes {
let box_stable_deref = Arc::new(data_holder);
let bytes: &[u8] = box_stable_deref.as_ref();
let data = unsafe { mem::transmute::<_, &'static [u8]>(bytes.deref()) };
OwnedBytes {
box_stable_deref,
data,
}
}
/// creates a fileslice that is just a view over a slice of the data.
pub fn slice(&self, from: usize, to: usize) -> Self {
OwnedBytes {
data: &self.data[from..to],
box_stable_deref: self.box_stable_deref.clone(),
}
}
/// Returns the underlying slice of data.
/// `Deref` and `AsRef` are also available.
#[inline(always)]
pub fn as_slice(&self) -> &[u8] {
self.data
}
/// Returns the len of the slice.
#[inline(always)]
pub fn len(&self) -> usize {
self.data.len()
}
/// Splits the OwnedBytes into two OwnedBytes `(left, right)`.
///
/// Left will hold `split_len` bytes.
///
/// This operation is cheap and does not require to copy any memory.
/// On the other hand, both `left` and `right` retain a handle over
/// the entire slice of memory. In other words, the memory will only
/// be released when both left and right are dropped.
pub fn split(self, split_len: usize) -> (OwnedBytes, OwnedBytes) {
let right_box_stable_deref = self.box_stable_deref.clone();
let left = OwnedBytes {
data: &self.data[..split_len],
box_stable_deref: self.box_stable_deref,
};
let right = OwnedBytes {
data: &self.data[split_len..],
box_stable_deref: right_box_stable_deref,
};
(left, right)
}
/// Returns true iff this `OwnedBytes` is empty.
#[inline(always)]
pub fn is_empty(&self) -> bool {
self.as_slice().is_empty()
}
/// Drops the left most `advance_len` bytes.
///
/// See also [.clip(clip_len: usize))](#method.clip).
#[inline(always)]
pub fn advance(&mut self, advance_len: usize) {
self.data = &self.data[advance_len..]
}
}
impl fmt::Debug for OwnedBytes {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
// We truncate the bytes in order to make sure the debug string
// is not too long.
let bytes_truncated: &[u8] = if self.len() > 8 {
&self.as_slice()[..10]
} else {
self.as_slice()
};
write!(f, "OwnedBytes({:?}, len={})", bytes_truncated, self.len())
}
}
impl Deref for OwnedBytes {
type Target = [u8];
fn deref(&self) -> &Self::Target {
self.as_slice()
}
}
impl io::Read for OwnedBytes {
fn read(&mut self, buf: &mut [u8]) -> io::Result<usize> {
let read_len = {
let data = self.as_slice();
if data.len() >= buf.len() {
let buf_len = buf.len();
buf.copy_from_slice(&data[..buf_len]);
buf.len()
} else {
let data_len = data.len();
buf[..data_len].copy_from_slice(data);
data_len
}
};
self.advance(read_len);
Ok(read_len)
}
fn read_to_end(&mut self, buf: &mut Vec<u8>) -> io::Result<usize> {
let read_len = {
let data = self.as_slice();
buf.extend(data);
data.len()
};
self.advance(read_len);
Ok(read_len)
}
fn read_exact(&mut self, buf: &mut [u8]) -> io::Result<()> {
let read_len = self.read(buf)?;
if read_len != buf.len() {
return Err(io::Error::new(
io::ErrorKind::UnexpectedEof,
"failed to fill whole buffer",
));
}
Ok(())
}
}
impl AsRef<[u8]> for OwnedBytes {
fn as_ref(&self) -> &[u8] {
self.as_slice()
}
}
#[cfg(test)]
mod tests {
use std::io::{self, Read};
use super::OwnedBytes;
#[test]
fn test_owned_bytes_debug() {
let short_bytes = OwnedBytes::new(b"abcd".as_ref());
assert_eq!(
format!("{:?}", short_bytes),
"OwnedBytes([97, 98, 99, 100], len=4)"
);
let long_bytes = OwnedBytes::new(b"abcdefghijklmnopq".as_ref());
assert_eq!(
format!("{:?}", long_bytes),
"OwnedBytes([97, 98, 99, 100, 101, 102, 103, 104, 105, 106], len=17)"
);
}
#[test]
fn test_owned_bytes_read() -> io::Result<()> {
let mut bytes = OwnedBytes::new(b"abcdefghiklmnopqrstuvwxyz".as_ref());
{
let mut buf = [0u8; 5];
bytes.read_exact(&mut buf[..]).unwrap();
assert_eq!(&buf, b"abcde");
assert_eq!(bytes.as_slice(), b"fghiklmnopqrstuvwxyz")
}
{
let mut buf = [0u8; 2];
bytes.read_exact(&mut buf[..]).unwrap();
assert_eq!(&buf, b"fg");
assert_eq!(bytes.as_slice(), b"hiklmnopqrstuvwxyz")
}
Ok(())
}
#[test]
fn test_owned_bytes_read_right_at_the_end() -> io::Result<()> {
let mut bytes = OwnedBytes::new(b"abcde".as_ref());
let mut buf = [0u8; 5];
assert_eq!(bytes.read(&mut buf[..]).unwrap(), 5);
assert_eq!(&buf, b"abcde");
assert_eq!(bytes.as_slice(), b"");
assert_eq!(bytes.read(&mut buf[..]).unwrap(), 0);
assert_eq!(&buf, b"abcde");
Ok(())
}
#[test]
fn test_owned_bytes_read_incomplete() -> io::Result<()> {
let mut bytes = OwnedBytes::new(b"abcde".as_ref());
let mut buf = [0u8; 7];
assert_eq!(bytes.read(&mut buf[..]).unwrap(), 5);
assert_eq!(&buf[..5], b"abcde");
assert_eq!(bytes.read(&mut buf[..]).unwrap(), 0);
Ok(())
}
#[test]
fn test_owned_bytes_read_to_end() -> io::Result<()> {
let mut bytes = OwnedBytes::new(b"abcde".as_ref());
let mut buf = Vec::new();
bytes.read_to_end(&mut buf)?;
assert_eq!(buf.as_slice(), b"abcde".as_ref());
Ok(())
}
#[test]
fn test_owned_bytes_split() {
let bytes = OwnedBytes::new(b"abcdefghi".as_ref());
let (left, right) = bytes.split(3);
assert_eq!(left.as_slice(), b"abc");
assert_eq!(right.as_slice(), b"defghi");
}
#[test]
fn test_owned_bytes_split_boundary() {
let bytes = OwnedBytes::new(b"abcdefghi".as_ref());
{
let (left, right) = bytes.clone().split(0);
assert_eq!(left.as_slice(), b"");
assert_eq!(right.as_slice(), b"abcdefghi");
}
{
let (left, right) = bytes.split(9);
assert_eq!(left.as_slice(), b"abcdefghi");
assert_eq!(right.as_slice(), b"");
}
}
}

38
src/directory/persistor.rs Normal file
View File

@@ -0,0 +1,38 @@
use crate::indexer::{SegmentManager, ResourceManager, MergeOperationInventory};
use std::thread::JoinHandle;
use crate::{IndexWriterConfig, SegmentId};
use std::collections::HashSet;
pub(crate) struct Persistor {
memory_manager: ResourceManager,
thread_handle: JoinHandle<()>,
}
impl Persistor {
pub(crate) fn create_and_start(segment_manager: SegmentManager,
memory_manager: ResourceManager,
config: IndexWriterConfig) -> crate::Result<Persistor> {
let memory_manager_clone = memory_manager.clone();
let thread_handle = std::thread::Builder::new()
.name("persistor-thread".to_string())
.spawn(move || {
while let Ok(_) = memory_manager_clone.wait_until_in_range(config.persist_low..) {
segment_manager.largest_segment_not_in_merge();
}
}).map_err(|_err| crate::TantivyError::ErrorInThread("Failed to start persistor thread.".to_string()))?;
Ok(Persistor {
memory_manager,
thread_handle
})
}
/// Stop the persisting thread.
///
/// The memory manager will be terminated, which will unlock the thread from any waiting
/// position.
/// This method blocks for a short amount of tim until the persistor thread has terminated.
pub fn stop(self) {
self.memory_manager.terminate();
let _ = self.thread_handle.join();
}
}

130
src/directory/ram_directory.rs
View File

@@ -1,9 +1,9 @@
use crate::core::META_FILEPATH;
use crate::directory::error::{DeleteError, OpenReadError, OpenWriteError};
use crate::directory::AntiCallToken;
use crate::directory::WatchCallbackList;
use crate::directory::{Directory, FileSlice, WatchCallback, WatchHandle};
use crate::directory::{Directory, ReadOnlySource, WatchCallback, WatchHandle};
use crate::directory::{TerminatingWrite, WritePtr};
use crate::{common::HasLen, core::META_FILEPATH};
use fail::fail_point;
use std::collections::HashMap;
use std::fmt;
@@ -11,6 +11,7 @@ use std::io::{self, BufWriter, Cursor, Seek, SeekFrom, Write};
use std::path::{Path, PathBuf};
use std::result;
use std::sync::{Arc, RwLock};
use crate::indexer::ResourceManager;
/// Writer associated with the `RAMDirectory`
///
@@ -80,17 +81,18 @@ impl TerminatingWrite for VecWriter {
#[derive(Default)]
struct InnerDirectory {
fs: HashMap<PathBuf, FileSlice>,
fs: HashMap<PathBuf, ReadOnlySource>,
watch_router: WatchCallbackList,
memory_manager: ResourceManager,
}
impl InnerDirectory {
fn write(&mut self, path: PathBuf, data: &[u8]) -> bool {
let data = FileSlice::from(data.to_vec());
let data = ReadOnlySource::new_with_allocation(Vec::from(data), &self.memory_manager);
self.fs.insert(path, data).is_some()
}
fn open_read(&self, path: &Path) -> Result<FileSlice, OpenReadError> {
fn open_read(&self, path: &Path) -> Result<ReadOnlySource, OpenReadError> {
self.fs
.get(path)
.ok_or_else(|| OpenReadError::FileDoesNotExist(PathBuf::from(path)))
@@ -112,8 +114,8 @@ impl InnerDirectory {
self.watch_router.subscribe(watch_handle)
}
fn total_mem_usage(&self) -> usize {
self.fs.values().map(|f| f.len()).sum()
fn total_mem_usage(&self) -> u64 {
self.fs.values().map(|source| source.len() as u64).sum()
}
}
@@ -134,14 +136,30 @@ pub struct RAMDirectory {
}
impl RAMDirectory {
/// Constructor
/// Creates a new RAMDirectory.
///
/// Check `.create_with_memory_manager(..)` if you want to associate an external memory
/// manager to your RAMDirectory.
pub fn create() -> RAMDirectory {
Self::default()
RAMDirectory::default()
}
/// Constructor
pub fn create_with_memory_manager(memory_manager: ResourceManager) -> RAMDirectory {
let inner_directory = InnerDirectory {
fs: Default::default(),
watch_router: Default::default(),
memory_manager
};
RAMDirectory {
fs: Arc::new(RwLock::new(inner_directory))
}
}
/// Returns the sum of the size of the different files
/// in the RAMDirectory.
pub fn total_mem_usage(&self) -> usize {
pub fn total_mem_usage(&self) -> u64 {
self.fs.read().unwrap().total_mem_usage()
}
@@ -151,11 +169,11 @@ impl RAMDirectory {
/// written using the `atomic_write` api.
///
/// If an error is encounterred, files may be persisted partially.
pub fn persist(&self, dest: &dyn Directory) -> crate::Result<()> {
pub fn persist(&self, dest: &mut dyn Directory) -> crate::Result<()> {
let wlock = self.fs.write().unwrap();
for (path, file) in wlock.fs.iter() {
for (path, source) in wlock.fs.iter() {
let mut dest_wrt = dest.open_write(path)?;
dest_wrt.write_all(file.read_bytes()?.as_slice())?;
dest_wrt.write_all(source.as_slice())?;
dest_wrt.terminate()?;
}
Ok(())
@@ -163,16 +181,15 @@ impl RAMDirectory {
}
impl Directory for RAMDirectory {
fn open_read(&self, path: &Path) -> result::Result<FileSlice, OpenReadError> {
fn open_read(&self, path: &Path) -> result::Result<ReadOnlySource, OpenReadError> {
self.fs.read().unwrap().open_read(path)
}
fn delete(&self, path: &Path) -> result::Result<(), DeleteError> {
fail_point!("RAMDirectory::delete", |_| {
Err(DeleteError::IOError {
io_error: io::Error::from(io::ErrorKind::Other),
filepath: path.to_path_buf(),
})
use crate::directory::error::IOError;
let io_error = IOError::from(io::Error::from(io::ErrorKind::Other));
Err(DeleteError::from(io_error))
});
self.fs.write().unwrap().delete(path)
}
@@ -181,7 +198,7 @@ impl Directory for RAMDirectory {
self.fs.read().unwrap().exists(path)
}
fn open_write(&self, path: &Path) -> Result<WritePtr, OpenWriteError> {
fn open_write(&mut self, path: &Path) -> Result<WritePtr, OpenWriteError> {
let mut fs = self.fs.write().unwrap();
let path_buf = PathBuf::from(path);
let vec_writer = VecWriter::new(path_buf.clone(), self.clone());
@@ -195,17 +212,10 @@ impl Directory for RAMDirectory {
}
fn atomic_read(&self, path: &Path) -> Result<Vec<u8>, OpenReadError> {
let bytes =
self.open_read(path)?
.read_bytes()
.map_err(|io_error| OpenReadError::IOError {
io_error,
filepath: path.to_path_buf(),
})?;
Ok(bytes.as_slice().to_owned())
Ok(self.open_read(path)?.as_slice().to_owned())
}
fn atomic_write(&self, path: &Path, data: &[u8]) -> io::Result<()> {
fn atomic_write(&mut self, path: &Path, data: &[u8]) -> io::Result<()> {
fail_point!("RAMDirectory::atomic_write", |msg| Err(io::Error::new(
io::ErrorKind::Other,
msg.unwrap_or_else(|| "Undefined".to_string())
@@ -235,6 +245,9 @@ mod tests {
use crate::Directory;
use std::io::Write;
use std::path::Path;
use crate::indexer::ResourceManager;
use crate::directory::TerminatingWrite;
use std::mem;
#[test]
fn test_persist() {
@@ -242,14 +255,69 @@ mod tests {
let msg_seq: &'static [u8] = b"sequential is the way";
let path_atomic: &'static Path = Path::new("atomic");
let path_seq: &'static Path = Path::new("seq");
let directory = RAMDirectory::create();
let mut directory = RAMDirectory::create();
assert!(directory.atomic_write(path_atomic, msg_atomic).is_ok());
let mut wrt = directory.open_write(path_seq).unwrap();
assert!(wrt.write_all(msg_seq).is_ok());
assert!(wrt.flush().is_ok());
let directory_copy = RAMDirectory::create();
assert!(directory.persist(&directory_copy).is_ok());
let mut directory_copy = RAMDirectory::create();
assert!(directory.persist(&mut directory_copy).is_ok());
assert_eq!(directory_copy.atomic_read(path_atomic).unwrap(), msg_atomic);
assert_eq!(directory_copy.atomic_read(path_seq).unwrap(), msg_seq);
}
#[test]
fn test_memory_manager_several_path() {
let memory_manager = ResourceManager::default();
let mut ram_directory = RAMDirectory::create_with_memory_manager(memory_manager.clone());
assert!(ram_directory.atomic_write(Path::new("/titi"), b"abcd").is_ok());
assert_eq!(memory_manager.total_amount(), 4u64);
assert!(ram_directory.atomic_write(Path::new("/toto"), b"abcde").is_ok());
assert_eq!(memory_manager.total_amount(), 9u64);
}
#[test]
fn test_memory_manager_override() {
let memory_manager = ResourceManager::default();
let mut ram_directory = RAMDirectory::create_with_memory_manager(memory_manager.clone());
assert!(ram_directory.atomic_write(Path::new("/titi"), b"abcde").is_ok());
assert_eq!(memory_manager.total_amount(), 5u64);
assert!(ram_directory.atomic_write(Path::new("/titi"), b"abcdef").is_ok());
assert_eq!(memory_manager.total_amount(), 6u64);
}
#[test]
fn test_memory_manager_seq_wrt() {
let memory_manager = ResourceManager::default();
let mut ram_directory = RAMDirectory::create_with_memory_manager(memory_manager.clone());
let mut wrt = ram_directory.open_write(Path::new("/titi")).unwrap();
assert!(wrt.write_all(b"abcde").is_ok());
assert!(wrt.terminate().is_ok());
assert_eq!(memory_manager.total_amount(), 5u64);
assert!(ram_directory.atomic_write(Path::new("/titi"), b"abcdef").is_ok());
assert_eq!(memory_manager.total_amount(), 6u64);
}
#[test]
fn test_release_on_drop() {
let memory_manager = ResourceManager::default();
let mut ram_directory = RAMDirectory::create_with_memory_manager(memory_manager.clone());
let mut wrt = ram_directory.open_write(Path::new("/titi")).unwrap();
assert!(wrt.write_all(b"abcde").is_ok());
assert!(wrt.terminate().is_ok());
assert_eq!(memory_manager.total_amount(), 5u64);
let mut wrt2 = ram_directory.open_write(Path::new("/toto")).unwrap();
assert!(wrt2.write_all(b"abcdefghijkl").is_ok());
assert!(wrt2.terminate().is_ok());
assert_eq!(memory_manager.total_amount(), 17u64);
let source = ram_directory.open_read(Path::new("/titi")).unwrap();
let source_clone = source.clone();
assert_eq!(memory_manager.total_amount(), 17u64);
mem::drop(ram_directory);
assert_eq!(memory_manager.total_amount(), 5u64);
mem::drop(source);
assert_eq!(memory_manager.total_amount(), 5u64);
mem::drop(source_clone);
assert_eq!(memory_manager.total_amount(), 0u64);
}
}

157
src/directory/read_only_source.rs Normal file
View File

@@ -0,0 +1,157 @@
use crate::common::HasLen;
use stable_deref_trait::{CloneStableDeref, StableDeref};
use std::ops::Deref;
use std::sync::Arc;
use crate::indexer::{Allocation, ResourceManager};
pub type BoxedData = Box<dyn Deref<Target = [u8]> + Send + Sync + 'static>;
/// Read object that represents files in tantivy.
///
/// These read objects are only in charge to deliver
/// the data in the form of a constant read-only `&[u8]`.
/// Whatever happens to the directory file, the data
/// hold by this object should never be altered or destroyed.
pub struct ReadOnlySource {
data: Arc<BoxedData>,
start: usize,
stop: usize,
allocation: Option<Arc<Allocation>>
}
unsafe impl StableDeref for ReadOnlySource {}
unsafe impl CloneStableDeref for ReadOnlySource {}
impl Deref for ReadOnlySource {
type Target = [u8];
fn deref(&self) -> &[u8] {
self.as_slice()
}
}
impl From<Arc<BoxedData>> for ReadOnlySource {
fn from(data: Arc<BoxedData>) -> Self {
let len = data.len();
ReadOnlySource {
data,
start: 0,
stop: len,
allocation: None
}
}
}
impl ReadOnlySource {
pub(crate) fn new<D>(data: D) -> ReadOnlySource
where
D: Deref<Target = [u8]> + Send + Sync + 'static,
{
let len = data.len();
ReadOnlySource {
data: Arc::new(Box::new(data)),
start: 0,
stop: len,
allocation: None
}
}
pub(crate) fn new_with_allocation<D>(data: D, memory_manager: &ResourceManager) -> ReadOnlySource
where
D: Deref<Target = [u8]> + Send + Sync + 'static,
{
let len = data.len();
ReadOnlySource {
data: Arc::new(Box::new(data)),
start: 0,
stop: len,
allocation: Some(Arc::new(memory_manager.allocate(len as u64)))
}
}
/// Creates an empty ReadOnlySource
pub fn empty() -> ReadOnlySource {
ReadOnlySource::new(&[][..])
}
/// Returns the data underlying the ReadOnlySource object.
pub fn as_slice(&self) -> &[u8] {
&self.data[self.start..self.stop]
}
/// Splits into 2 `ReadOnlySource`, at the offset given
/// as an argument.
pub fn split(self, addr: usize) -> (ReadOnlySource, ReadOnlySource) {
let left = self.slice(0, addr);
let right = self.slice_from(addr);
(left, right)
}
/// Splits into 2 `ReadOnlySource`, at the offset `end - right_len`.
pub fn split_from_end(self, right_len: usize) -> (ReadOnlySource, ReadOnlySource) {
let left_len = self.len() - right_len;
self.split(left_len)
}
/// Creates a ReadOnlySource that is just a
/// view over a slice of the data.
///
/// Keep in mind that any living slice extends
/// the lifetime of the original ReadOnlySource,
///
/// For instance, if `ReadOnlySource` wraps 500MB
/// worth of data in anonymous memory, and only a
/// 1KB slice is remaining, the whole `500MBs`
/// are retained in memory.
pub fn slice(&self, start: usize, stop: usize) -> ReadOnlySource {
assert!(
start <= stop,
"Requested negative slice [{}..{}]",
start,
stop
);
assert!(stop <= self.len());
ReadOnlySource {
data: self.data.clone(),
start: self.start + start,
stop: self.start + stop,
allocation: self.allocation.clone()
}
}
/// Like `.slice(...)` but enforcing only the `from`
/// boundary.
///
/// Equivalent to `.slice(from_offset, self.len())`
pub fn slice_from(&self, from_offset: usize) -> ReadOnlySource {
self.slice(from_offset, self.len())
}
/// Like `.slice(...)` but enforcing only the `to`
/// boundary.
///
/// Equivalent to `.slice(0, to_offset)`
pub fn slice_to(&self, to_offset: usize) -> ReadOnlySource {
self.slice(0, to_offset)
}
}
impl HasLen for ReadOnlySource {
fn len(&self) -> usize {
self.stop - self.start
}
}
impl Clone for ReadOnlySource {
fn clone(&self) -> Self {
self.slice_from(0)
}
}
impl From<Vec<u8>> for ReadOnlySource {
fn from(data: Vec<u8>) -> ReadOnlySource {
ReadOnlySource::new(data)
}
}

186
src/directory/spilling_writer.rs Normal file
View File

@@ -0,0 +1,186 @@
use crate::common::MutableEnum;
use crate::directory::{TerminatingWrite, WritePtr};
use std::io::{self, Write};
/// Represents the state of the `SpillingWriter`.
enum SpillingState {
Buffer {
buffer: Vec<u8>,
capacity: usize,
write_factory: Box<dyn FnOnce() -> io::Result<WritePtr>>,
},
Spilled(WritePtr),
}
impl SpillingState {
fn new(
limit: usize,
write_factory: Box<dyn FnOnce() -> io::Result<WritePtr>>,
) -> SpillingState {
SpillingState::Buffer {
buffer: Vec::with_capacity(limit),
capacity: limit,
write_factory,
}
}
// Change the state in such a way that it is ready to accept
// `extra_capacity` bytes.
//
fn reserve(self, extra_capacity: usize) -> io::Result<SpillingState> {
match self {
SpillingState::Buffer {
buffer,
capacity,
write_factory,
} => {
if capacity >= extra_capacity {
Ok(SpillingState::Buffer {
buffer,
capacity: capacity - extra_capacity,
write_factory,
})
} else {
let mut wrt = write_factory()?;
wrt.write_all(&buffer[..])?;
Ok(SpillingState::Spilled(wrt))
}
}
SpillingState::Spilled(wrt) => Ok(SpillingState::Spilled(wrt)),
}
}
}
/// The `SpillingWriter` is a writer that start by writing in a
/// buffer.
///
/// Once a memory limit is reached, the spilling writer will
/// call a given `WritePtr` factory and start spilling into it.
///
/// Spilling here includes:
/// - writing all of the data that were written in the in-memory buffer so far
/// - writing subsequent data as well.
///
/// Once entering "spilling" mode, the `SpillingWriter` stays in this mode.
pub struct SpillingWriter {
state: MutableEnum<SpillingState>,
}
impl SpillingWriter {
//// Creates a new `Spilling Writer`.
pub fn new(
limit: usize,
write_factory: Box<dyn FnOnce() -> io::Result<WritePtr>>,
) -> SpillingWriter {
let state = SpillingState::new(limit, write_factory);
SpillingWriter {
state: MutableEnum::wrap(state),
}
}
/// Finalizes the `SpillingWriter`.
///
/// The `SpillingResult` object is an enum specific
/// to whether the `SpillingWriter` reached the spilling limit
/// (In that case, the buffer is returned).
///
/// If the writer reached the spilling mode, the underlying `WritePtr`
/// is terminated and SpillingResult::Spilled is returned.
pub fn finalize(self) -> io::Result<SpillingResult> {
match self.state.into() {
SpillingState::Spilled(wrt) => {
wrt.terminate()?;
Ok(SpillingResult::Spilled)
}
SpillingState::Buffer { buffer, .. } => Ok(SpillingResult::Buffer(buffer)),
}
}
}
/// enum used as the result of `.finalize()`.
pub enum SpillingResult {
Spilled,
Buffer(Vec<u8>),
}
impl io::Write for SpillingWriter {
fn write(&mut self, payload: &[u8]) -> io::Result<usize> {
self.write_all(payload)?;
Ok(payload.len())
}
fn flush(&mut self) -> io::Result<()> {
if let SpillingState::Spilled(wrt) = &mut *self.state {
wrt.flush()?;
}
Ok(())
}
fn write_all(&mut self, payload: &[u8]) -> io::Result<()> {
self.state.map_mutate(|mut state| {
state = state.reserve(payload.len())?;
match &mut state {
SpillingState::Buffer { buffer, .. } => {
buffer.extend_from_slice(payload);
}
SpillingState::Spilled(wrt) => {
wrt.write_all(payload)?;
}
}
Ok(state)
})
}
}
#[cfg(test)]
mod tests {
use super::SpillingWriter;
use crate::directory::spilling_writer::SpillingResult;
use crate::directory::RAMDirectory;
use crate::Directory;
use std::io::{self, Write};
use std::path::Path;
#[test]
fn test_no_spilling() {
let ram_directory = RAMDirectory::create();
let mut ram_directory_clone = ram_directory.clone();
let path = Path::new("test");
let write_factory = Box::new(move || {
ram_directory_clone
.open_write(path)
.map_err(|err| io::Error::new(io::ErrorKind::Other, err))
});
let mut spilling_wrt = SpillingWriter::new(10, write_factory);
assert!(spilling_wrt.write_all(b"abcd").is_ok());
if let SpillingResult::Buffer(buf) = spilling_wrt.finalize().unwrap() {
assert_eq!(buf, b"abcd")
} else {
panic!("spill writer should not have spilled");
}
assert!(!ram_directory.exists(path));
}
#[test]
fn test_spilling() {
let ram_directory = RAMDirectory::create();
let mut ram_directory_clone = ram_directory.clone();
let path = Path::new("test");
let write_factory = Box::new(move || {
ram_directory_clone
.open_write(path)
.map_err(|err| io::Error::new(io::ErrorKind::Other, err))
});
let mut spilling_wrt = SpillingWriter::new(10, write_factory);
assert!(spilling_wrt.write_all(b"abcd").is_ok());
assert!(spilling_wrt.write_all(b"efghijklmnop").is_ok());
if let SpillingResult::Spilled = spilling_wrt.finalize().unwrap() {
} else {
panic!("spill writer should have spilled");
}
assert_eq!(
ram_directory.atomic_read(path).unwrap(),
b"abcdefghijklmnop"
);
}
}

142
src/directory/tests.rs
View File

@@ -20,47 +20,45 @@ mod mmap_directory_tests {
}
#[test]
fn test_simple() -> crate::Result<()> {
let directory = make_directory();
super::test_simple(&directory)
fn test_simple() {
let mut directory = make_directory();
super::test_simple(&mut directory);
}
#[test]
fn test_write_create_the_file() {
let directory = make_directory();
super::test_write_create_the_file(&directory);
let mut directory = make_directory();
super::test_write_create_the_file(&mut directory);
}
#[test]
fn test_rewrite_forbidden() -> crate::Result<()> {
let directory = make_directory();
super::test_rewrite_forbidden(&directory)?;
Ok(())
fn test_rewrite_forbidden() {
let mut directory = make_directory();
super::test_rewrite_forbidden(&mut directory);
}
#[test]
fn test_directory_delete() -> crate::Result<()> {
let directory = make_directory();
super::test_directory_delete(&directory)?;
Ok(())
fn test_directory_delete() {
let mut directory = make_directory();
super::test_directory_delete(&mut directory);
}
#[test]
fn test_lock_non_blocking() {
let directory = make_directory();
super::test_lock_non_blocking(&directory);
let mut directory = make_directory();
super::test_lock_non_blocking(&mut directory);
}
#[test]
fn test_lock_blocking() {
let directory = make_directory();
super::test_lock_blocking(&directory);
let mut directory = make_directory();
super::test_lock_blocking(&mut directory);
}
#[test]
fn test_watch() {
let directory = make_directory();
super::test_watch(&directory);
let mut directory = make_directory();
super::test_watch(&mut directory);
}
}
@@ -74,47 +72,45 @@ mod ram_directory_tests {
}
#[test]
fn test_simple() -> crate::Result<()> {
let directory = make_directory();
super::test_simple(&directory)
fn test_simple() {
let mut directory = make_directory();
super::test_simple(&mut directory);
}
#[test]
fn test_write_create_the_file() {
let directory = make_directory();
super::test_write_create_the_file(&directory);
let mut directory = make_directory();
super::test_write_create_the_file(&mut directory);
}
#[test]
fn test_rewrite_forbidden() -> crate::Result<()> {
let directory = make_directory();
super::test_rewrite_forbidden(&directory)?;
Ok(())
fn test_rewrite_forbidden() {
let mut directory = make_directory();
super::test_rewrite_forbidden(&mut directory);
}
#[test]
fn test_directory_delete() -> crate::Result<()> {
let directory = make_directory();
super::test_directory_delete(&directory)?;
Ok(())
fn test_directory_delete() {
let mut directory = make_directory();
super::test_directory_delete(&mut directory);
}
#[test]
fn test_lock_non_blocking() {
let directory = make_directory();
super::test_lock_non_blocking(&directory);
let mut directory = make_directory();
super::test_lock_non_blocking(&mut directory);
}
#[test]
fn test_lock_blocking() {
let directory = make_directory();
super::test_lock_blocking(&directory);
let mut directory = make_directory();
super::test_lock_blocking(&mut directory);
}
#[test]
fn test_watch() {
let directory = make_directory();
super::test_watch(&directory);
let mut directory = make_directory();
super::test_watch(&mut directory);
}
}
@@ -122,37 +118,43 @@ mod ram_directory_tests {
#[should_panic]
fn ram_directory_panics_if_flush_forgotten() {
let test_path: &'static Path = Path::new("some_path_for_test");
let ram_directory = RAMDirectory::create();
let mut ram_directory = RAMDirectory::create();
let mut write_file = ram_directory.open_write(test_path).unwrap();
assert!(write_file.write_all(&[4]).is_ok());
}
fn test_simple(directory: &dyn Directory) -> crate::Result<()> {
fn test_simple(directory: &mut dyn Directory) {
let test_path: &'static Path = Path::new("some_path_for_test");
let mut write_file = directory.open_write(test_path)?;
assert!(directory.exists(test_path));
write_file.write_all(&[4])?;
write_file.write_all(&[3])?;
write_file.write_all(&[7, 3, 5])?;
write_file.flush()?;
let read_file = directory.open_read(test_path)?.read_bytes()?;
assert_eq!(read_file.as_slice(), &[4u8, 3u8, 7u8, 3u8, 5u8]);
mem::drop(read_file);
{
let mut write_file = directory.open_write(test_path).unwrap();
assert!(directory.exists(test_path));
write_file.write_all(&[4]).unwrap();
write_file.write_all(&[3]).unwrap();
write_file.write_all(&[7, 3, 5]).unwrap();
write_file.flush().unwrap();
}
{
let read_file = directory.open_read(test_path).unwrap();
let data: &[u8] = &*read_file;
assert_eq!(data, &[4u8, 3u8, 7u8, 3u8, 5u8]);
}
assert!(directory.delete(test_path).is_ok());
assert!(!directory.exists(test_path));
Ok(())
}
fn test_rewrite_forbidden(directory: &dyn Directory) -> crate::Result<()> {
fn test_rewrite_forbidden(directory: &mut dyn Directory) {
let test_path: &'static Path = Path::new("some_path_for_test");
directory.open_write(test_path)?;
assert!(directory.exists(test_path));
assert!(directory.open_write(test_path).is_err());
{
directory.open_write(test_path).unwrap();
assert!(directory.exists(test_path));
}
{
assert!(directory.open_write(test_path).is_err());
}
assert!(directory.delete(test_path).is_ok());
Ok(())
}
fn test_write_create_the_file(directory: &dyn Directory) {
fn test_write_create_the_file(directory: &mut dyn Directory) {
let test_path: &'static Path = Path::new("some_path_for_test");
{
assert!(directory.open_read(test_path).is_err());
@@ -163,20 +165,21 @@ fn test_write_create_the_file(directory: &dyn Directory) {
}
}
fn test_directory_delete(directory: &dyn Directory) -> crate::Result<()> {
fn test_directory_delete(directory: &mut dyn Directory) {
let test_path: &'static Path = Path::new("some_path_for_test");
assert!(directory.open_read(test_path).is_err());
let mut write_file = directory.open_write(&test_path)?;
write_file.write_all(&[1, 2, 3, 4])?;
write_file.flush()?;
let mut write_file = directory.open_write(&test_path).unwrap();
write_file.write_all(&[1, 2, 3, 4]).unwrap();
write_file.flush().unwrap();
{
let read_handle = directory.open_read(&test_path)?.read_bytes()?;
assert_eq!(read_handle.as_slice(), &[1u8, 2u8, 3u8, 4u8]);
let read_handle = directory.open_read(&test_path).unwrap();
assert_eq!(&*read_handle, &[1u8, 2u8, 3u8, 4u8]);
// Mapped files can't be deleted on Windows
if !cfg!(windows) {
assert!(directory.delete(&test_path).is_ok());
assert_eq!(read_handle.as_slice(), &[1u8, 2u8, 3u8, 4u8]);
assert_eq!(&*read_handle, &[1u8, 2u8, 3u8, 4u8]);
}
assert!(directory.delete(Path::new("SomeOtherPath")).is_err());
}
@@ -186,10 +189,9 @@ fn test_directory_delete(directory: &dyn Directory) -> crate::Result<()> {
assert!(directory.open_read(&test_path).is_err());
assert!(directory.delete(&test_path).is_err());
Ok(())
}
fn test_watch(directory: &dyn Directory) {
fn test_watch(directory: &mut dyn Directory) {
let num_progress: Arc<AtomicUsize> = Default::default();
let counter: Arc<AtomicUsize> = Default::default();
let counter_clone = counter.clone();
@@ -209,22 +211,22 @@ fn test_watch(directory: &dyn Directory) {
.unwrap();
for i in 0..10 {
assert!(i <= counter.load(SeqCst));
assert_eq!(i, counter.load(SeqCst));
assert!(directory
.atomic_write(Path::new("meta.json"), b"random_test_data_2")
.is_ok());
assert_eq!(receiver.recv_timeout(Duration::from_millis(500)), Ok(i));
assert!(i + 1 <= counter.load(SeqCst)); // notify can trigger more than once.
assert_eq!(i + 1, counter.load(SeqCst));
}
mem::drop(watch_handle);
assert!(directory
.atomic_write(Path::new("meta.json"), b"random_test_data")
.is_ok());
assert!(receiver.recv_timeout(Duration::from_millis(500)).is_ok());
assert!(10 <= counter.load(SeqCst));
assert_eq!(10, counter.load(SeqCst));
}
fn test_lock_non_blocking(directory: &dyn Directory) {
fn test_lock_non_blocking(directory: &mut dyn Directory) {
{
let lock_a_res = directory.acquire_lock(&Lock {
filepath: PathBuf::from("a.lock"),
@@ -249,7 +251,7 @@ fn test_lock_non_blocking(directory: &dyn Directory) {
assert!(lock_a_res.is_ok());
}
fn test_lock_blocking(directory: &dyn Directory) {
fn test_lock_blocking(directory: &mut dyn Directory) {
let lock_a_res = directory.acquire_lock(&Lock {
filepath: PathBuf::from("a.lock"),
is_blocking: true,

11
src/directory/watch_event_router.rs
View File

@@ -5,7 +5,7 @@ use std::sync::RwLock;
use std::sync::Weak;
/// Type alias for callbacks registered when watching files of a `Directory`.
pub type WatchCallback = Box<dyn Fn() + Sync + Send>;
pub type WatchCallback = Box<dyn Fn() -> () + Sync + Send>;
/// Helper struct to implement the watch method in `Directory` implementations.
///
@@ -29,17 +29,10 @@ impl WatchHandle {
pub fn new(watch_callback: Arc<WatchCallback>) -> WatchHandle {
WatchHandle(watch_callback)
}
/// Returns an empty watch handle.
///
/// This function is only useful when implementing a readonly directory.
pub fn empty() -> WatchHandle {
WatchHandle::new(Arc::new(Box::new(|| {})))
}
}
impl WatchCallbackList {
/// Subscribes a new callback and returns a handle that controls the lifetime of the callback.
/// Suscribes a new callback and returns a handle that controls the lifetime of the callback.
pub fn subscribe(&self, watch_callback: WatchCallback) -> WatchHandle {
let watch_callback_arc = Arc::new(watch_callback);
let watch_callback_weak = Arc::downgrade(&watch_callback_arc);

131
src/docset.rs
View File

@@ -1,48 +1,58 @@
use crate::common::BitSet;
use crate::fastfield::DeleteBitSet;
use crate::DocId;
use std::borrow::Borrow;
use std::borrow::BorrowMut;
use std::cmp::Ordering;
/// Sentinel value returned when a DocSet has been entirely consumed.
///
/// This is not u32::MAX as one would have expected, due to the lack of SSE2 instructions
/// to compare [u32; 4].
pub const TERMINATED: DocId = std::i32::MAX as u32;
/// Expresses the outcome of a call to `DocSet`'s `.skip_next(...)`.
#[derive(PartialEq, Eq, Debug)]
pub enum SkipResult {
/// target was in the docset
Reached,
/// target was not in the docset, skipping stopped as a greater element was found
OverStep,
/// the docset was entirely consumed without finding the target, nor any
/// element greater than the target.
End,
}
/// Represents an iterable set of sorted doc ids.
pub trait DocSet {
/// Goes to the next element.
///
/// The DocId of the next element is returned.
/// In other words we should always have :
/// ```ignore
/// let doc = docset.advance();
/// assert_eq!(doc, docset.doc());
/// ```
///
/// If we reached the end of the DocSet, TERMINATED should be returned.
///
/// Calling `.advance()` on a terminated DocSet should be supported, and TERMINATED should
/// be returned.
/// TODO Test existing docsets.
fn advance(&mut self) -> DocId;
/// `.advance(...)` needs to be called a first time to point to the correct
/// element.
fn advance(&mut self) -> bool;
/// Advances the DocSet forward until reaching the target, or going to the
/// lowest DocId greater than the target.
/// After skipping, position the iterator in such a way that `.doc()`
/// will return a value greater than or equal to target.
///
/// If the end of the DocSet is reached, TERMINATED is returned.
/// SkipResult expresses whether the `target value` was reached, overstepped,
/// or if the `DocSet` was entirely consumed without finding any value
/// greater or equal to the `target`.
///
/// Calling `.seek(target)` on a terminated DocSet is legal. Implementation
/// of DocSet should support it.
/// WARNING: Calling skip always advances the docset.
/// More specifically, if the docset is already positionned on the target
/// skipping will advance to the next position and return SkipResult::Overstep.
///
/// Calling `seek(TERMINATED)` is also legal and is the normal way to consume a DocSet.
fn seek(&mut self, target: DocId) -> DocId {
let mut doc = self.doc();
debug_assert!(doc <= target);
while doc < target {
doc = self.advance();
/// If `.skip_next()` oversteps, then the docset must be positionned correctly
/// on an existing document. In other words, `.doc()` should return the first document
/// greater than `DocId`.
fn skip_next(&mut self, target: DocId) -> SkipResult {
if !self.advance() {
return SkipResult::End;
}
loop {
match self.doc().cmp(&target) {
Ordering::Less => {
if !self.advance() {
return SkipResult::End;
}
}
Ordering::Equal => return SkipResult::Reached,
Ordering::Greater => return SkipResult::OverStep,
}
}
doc
}
/// Fills a given mutable buffer with the next doc ids from the
@@ -61,38 +71,38 @@ pub trait DocSet {
/// use case where batching. The normal way to
/// go through the `DocId`'s is to call `.advance()`.
fn fill_buffer(&mut self, buffer: &mut [DocId]) -> usize {
if self.doc() == TERMINATED {
return 0;
}
for (i, buffer_val) in buffer.iter_mut().enumerate() {
*buffer_val = self.doc();
if self.advance() == TERMINATED {
return i + 1;
if self.advance() {
*buffer_val = self.doc();
} else {
return i;
}
}
buffer.len()
}
/// Returns the current document
/// Right after creating a new DocSet, the docset points to the first document.
///
/// If the DocSet is empty, .doc() should return `TERMINATED`.
fn doc(&self) -> DocId;
/// Returns a best-effort hint of the
/// length of the docset.
fn size_hint(&self) -> u32;
/// Appends all docs to a `bitset`.
fn append_to_bitset(&mut self, bitset: &mut BitSet) {
while self.advance() {
bitset.insert(self.doc());
}
}
/// Returns the number documents matching.
/// Calling this method consumes the `DocSet`.
fn count(&mut self, delete_bitset: &DeleteBitSet) -> u32 {
let mut count = 0u32;
let mut doc = self.doc();
while doc != TERMINATED {
if !delete_bitset.is_deleted(doc) {
while self.advance() {
if !delete_bitset.is_deleted(self.doc()) {
count += 1u32;
}
doc = self.advance();
}
count
}
@@ -104,42 +114,22 @@ pub trait DocSet {
/// given by `count()`.
fn count_including_deleted(&mut self) -> u32 {
let mut count = 0u32;
let mut doc = self.doc();
while doc != TERMINATED {
while self.advance() {
count += 1u32;
doc = self.advance();
}
count
}
}
impl<'a> DocSet for &'a mut dyn DocSet {
fn advance(&mut self) -> u32 {
(**self).advance()
}
fn seek(&mut self, target: DocId) -> DocId {
(**self).seek(target)
}
fn doc(&self) -> u32 {
(**self).doc()
}
fn size_hint(&self) -> u32 {
(**self).size_hint()
}
}
impl<TDocSet: DocSet + ?Sized> DocSet for Box<TDocSet> {
fn advance(&mut self) -> DocId {
fn advance(&mut self) -> bool {
let unboxed: &mut TDocSet = self.borrow_mut();
unboxed.advance()
}
fn seek(&mut self, target: DocId) -> DocId {
fn skip_next(&mut self, target: DocId) -> SkipResult {
let unboxed: &mut TDocSet = self.borrow_mut();
unboxed.seek(target)
unboxed.skip_next(target)
}
fn doc(&self) -> DocId {
@@ -161,4 +151,9 @@ impl<TDocSet: DocSet + ?Sized> DocSet for Box<TDocSet> {
let unboxed: &mut TDocSet = self.borrow_mut();
unboxed.count_including_deleted()
}
fn append_to_bitset(&mut self, bitset: &mut BitSet) {
let unboxed: &mut TDocSet = self.borrow_mut();
unboxed.append_to_bitset(bitset);
}
}

111
src/error.rs
View File

@@ -2,27 +2,22 @@
use std::io;
use crate::directory::error::{IOError, OpenDirectoryError, OpenReadError, OpenWriteError};
use crate::directory::error::{Incompatibility, LockError};
use crate::fastfield::FastFieldNotAvailableError;
use crate::query;
use crate::{
directory::error::{OpenDirectoryError, OpenReadError, OpenWriteError},
schema,
};
use crate::schema;
use serde_json;
use std::fmt;
use std::path::PathBuf;
use std::sync::PoisonError;
/// Represents a `DataCorruption` error.
///
/// When facing data corruption, tantivy actually panic or return this error.
pub struct DataCorruption {
filepath: Option<PathBuf>,
comment: String,
}
impl DataCorruption {
/// Creates a `DataCorruption` Error.
pub fn new(filepath: PathBuf, comment: String) -> DataCorruption {
DataCorruption {
filepath: Some(filepath),
@@ -30,11 +25,10 @@ impl DataCorruption {
}
}
/// Creates a `DataCorruption` Error, when the filepath is irrelevant.
pub fn comment_only<TStr: ToString>(comment: TStr) -> DataCorruption {
pub fn comment_only(comment: String) -> DataCorruption {
DataCorruption {
filepath: None,
comment: comment.to_string(),
comment,
}
}
}
@@ -50,47 +44,44 @@ impl fmt::Debug for DataCorruption {
}
}
/// The library's error enum
#[derive(Debug, Error)]
/// The library's failure based error enum
#[derive(Debug, Fail)]
pub enum TantivyError {
/// Failed to open the directory.
#[error("Failed to open the directory: '{0:?}'")]
OpenDirectoryError(#[from] OpenDirectoryError),
/// Failed to open a file for read.
#[error("Failed to open file for read: '{0:?}'")]
OpenReadError(#[from] OpenReadError),
/// Failed to open a file for write.
#[error("Failed to open file for write: '{0:?}'")]
OpenWriteError(#[from] OpenWriteError),
/// Path does not exist.
#[fail(display = "Path does not exist: '{:?}'", _0)]
PathDoesNotExist(PathBuf),
/// File already exists, this is a problem when we try to write into a new file.
#[fail(display = "File already exists: '{:?}'", _0)]
FileAlreadyExists(PathBuf),
/// Index already exists in this directory
#[error("Index already exists")]
#[fail(display = "Index already exists")]
IndexAlreadyExists,
/// Failed to acquire file lock
#[error("Failed to acquire Lockfile: {0:?}. {1:?}")]
#[fail(display = "Failed to acquire Lockfile: {:?}. {:?}", _0, _1)]
LockFailure(LockError, Option<String>),
/// IO Error.
#[error("An IO error occurred: '{0}'")]
IOError(#[from] io::Error),
#[fail(display = "An IO error occurred: '{}'", _0)]
IOError(#[cause] IOError),
/// Data corruption.
#[error("Data corrupted: '{0:?}'")]
#[fail(display = "{:?}", _0)]
DataCorruption(DataCorruption),
/// A thread holding the locked panicked and poisoned the lock.
#[error("A thread holding the locked panicked and poisoned the lock")]
#[fail(display = "A thread holding the locked panicked and poisoned the lock")]
Poisoned,
/// Invalid argument was passed by the user.
#[error("An invalid argument was passed: '{0}'")]
#[fail(display = "An invalid argument was passed: '{}'", _0)]
InvalidArgument(String),
/// An Error happened in one of the thread.
#[error("An error occurred in a thread: '{0}'")]
#[fail(display = "An error occurred in a thread: '{}'", _0)]
ErrorInThread(String),
/// An Error appeared related to the schema.
#[error("Schema error: '{0}'")]
#[fail(display = "Schema error: '{}'", _0)]
SchemaError(String),
/// System error. (e.g.: We failed spawning a new thread)
#[error("System error.'{0}'")]
#[fail(display = "System error.'{}'", _0)]
SystemError(String),
/// Index incompatible with current version of tantivy
#[error("{0:?}")]
#[fail(display = "{:?}", _0)]
IncompatibleIndex(Incompatibility),
}
@@ -99,17 +90,31 @@ impl From<DataCorruption> for TantivyError {
TantivyError::DataCorruption(data_corruption)
}
}
impl From<FastFieldNotAvailableError> for TantivyError {
fn from(fastfield_error: FastFieldNotAvailableError) -> TantivyError {
TantivyError::SchemaError(format!("{}", fastfield_error))
}
}
impl From<LockError> for TantivyError {
fn from(lock_error: LockError) -> TantivyError {
TantivyError::LockFailure(lock_error, None)
}
}
impl From<IOError> for TantivyError {
fn from(io_error: IOError) -> TantivyError {
TantivyError::IOError(io_error)
}
}
impl From<io::Error> for TantivyError {
fn from(io_error: io::Error) -> TantivyError {
TantivyError::IOError(io_error.into())
}
}
impl From<query::QueryParserError> for TantivyError {
fn from(parsing_error: query::QueryParserError) -> TantivyError {
TantivyError::InvalidArgument(format!("Query is invalid. {:?}", parsing_error))
@@ -122,9 +127,15 @@ impl<Guard> From<PoisonError<Guard>> for TantivyError {
}
}
impl From<chrono::ParseError> for TantivyError {
fn from(err: chrono::ParseError) -> TantivyError {
TantivyError::InvalidArgument(err.to_string())
impl From<OpenReadError> for TantivyError {
fn from(error: OpenReadError) -> TantivyError {
match error {
OpenReadError::FileDoesNotExist(filepath) => TantivyError::PathDoesNotExist(filepath),
OpenReadError::IOError(io_error) => TantivyError::IOError(io_error),
OpenReadError::IncompatibleIndex(incompatibility) => {
TantivyError::IncompatibleIndex(incompatibility)
}
}
}
}
@@ -134,9 +145,35 @@ impl From<schema::DocParsingError> for TantivyError {
}
}
impl From<OpenWriteError> for TantivyError {
fn from(error: OpenWriteError) -> TantivyError {
match error {
OpenWriteError::FileAlreadyExists(filepath) => {
TantivyError::FileAlreadyExists(filepath)
}
OpenWriteError::IOError(io_error) => TantivyError::IOError(io_error),
}
}
}
impl From<OpenDirectoryError> for TantivyError {
fn from(error: OpenDirectoryError) -> TantivyError {
match error {
OpenDirectoryError::DoesNotExist(directory_path) => {
TantivyError::PathDoesNotExist(directory_path)
}
OpenDirectoryError::NotADirectory(directory_path) => {
TantivyError::InvalidArgument(format!("{:?} is not a directory", directory_path))
}
OpenDirectoryError::IoError(err) => TantivyError::IOError(IOError::from(err)),
}
}
}
impl From<serde_json::Error> for TantivyError {
fn from(error: serde_json::Error) -> TantivyError {
TantivyError::IOError(error.into())
let io_err = io::Error::from(error);
TantivyError::IOError(io_err.into())
}
}

111
src/fastfield/bytes/mod.rs
View File

@@ -6,114 +6,31 @@ pub use self::writer::BytesFastFieldWriter;
#[cfg(test)]
mod tests {
use crate::schema::{BytesOptions, IndexRecordOption, Schema, Value};
use crate::{query::TermQuery, schema::FAST, schema::INDEXED, schema::STORED};
use crate::{DocAddress, DocSet, Index, Searcher, Term};
use std::ops::Deref;
use crate::schema::Schema;
use crate::Index;
#[test]
fn test_bytes() -> crate::Result<()> {
fn test_bytes() {
let mut schema_builder = Schema::builder();
let bytes_field = schema_builder.add_bytes_field("bytesfield", FAST);
let field = schema_builder.add_bytes_field("bytesfield");
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(bytes_field=>vec![0u8, 1, 2, 3]));
index_writer.add_document(doc!(bytes_field=>vec![]));
index_writer.add_document(doc!(bytes_field=>vec![255u8]));
index_writer.add_document(doc!(bytes_field=>vec![1u8, 3, 5, 7, 9]));
index_writer.add_document(doc!(bytes_field=>vec![0u8; 1000]));
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(field=>vec![0u8, 1, 2, 3]));
index_writer.add_document(doc!(field=>vec![]));
index_writer.add_document(doc!(field=>vec![255u8]));
index_writer.add_document(doc!(field=>vec![1u8, 3, 5, 7, 9]));
index_writer.add_document(doc!(field=>vec![0u8; 1000]));
assert!(index_writer.commit().is_ok());
let searcher = index.reader().unwrap().searcher();
let segment_reader = searcher.segment_reader(0);
let bytes_reader = segment_reader.fast_fields().bytes(bytes_field).unwrap();
let bytes_reader = segment_reader.fast_fields().bytes(field).unwrap();
assert_eq!(bytes_reader.get_bytes(0), &[0u8, 1, 2, 3]);
assert!(bytes_reader.get_bytes(1).is_empty());
assert_eq!(bytes_reader.get_bytes(2), &[255u8]);
assert_eq!(bytes_reader.get_bytes(3), &[1u8, 3, 5, 7, 9]);
let long = vec![0u8; 1000];
assert_eq!(bytes_reader.get_bytes(4), long.as_slice());
Ok(())
}
fn create_index_for_test<T: Into<BytesOptions>>(
byte_options: T,
) -> crate::Result<impl Deref<Target = Searcher>> {
let mut schema_builder = Schema::builder();
let field = schema_builder.add_bytes_field("string_bytes", byte_options.into());
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(
field => b"tantivy".as_ref(),
field => b"lucene".as_ref()
));
index_writer.commit()?;
Ok(index.reader()?.searcher())
}
#[test]
fn test_stored_bytes() -> crate::Result<()> {
let searcher = create_index_for_test(STORED)?;
assert_eq!(searcher.num_docs(), 1);
let retrieved_doc = searcher.doc(DocAddress(0u32, 0u32))?;
let field = searcher.schema().get_field("string_bytes").unwrap();
let values: Vec<&Value> = retrieved_doc.get_all(field).collect();
assert_eq!(values.len(), 2);
let values_bytes: Vec<&[u8]> = values
.into_iter()
.flat_map(|value| value.bytes_value())
.collect();
assert_eq!(values_bytes, &[&b"tantivy"[..], &b"lucene"[..]]);
Ok(())
}
#[test]
fn test_non_stored_bytes() -> crate::Result<()> {
let searcher = create_index_for_test(INDEXED)?;
assert_eq!(searcher.num_docs(), 1);
let retrieved_doc = searcher.doc(DocAddress(0u32, 0u32))?;
let field = searcher.schema().get_field("string_bytes").unwrap();
assert!(retrieved_doc.get_first(field).is_none());
Ok(())
}
#[test]
fn test_index_bytes() -> crate::Result<()> {
let searcher = create_index_for_test(INDEXED)?;
assert_eq!(searcher.num_docs(), 1);
let field = searcher.schema().get_field("string_bytes").unwrap();
let term = Term::from_field_bytes(field, b"lucene".as_ref());
let term_query = TermQuery::new(term, IndexRecordOption::Basic);
let term_weight = term_query.specialized_weight(&searcher, true)?;
let term_scorer = term_weight.specialized_scorer(searcher.segment_reader(0), 1.0f32)?;
assert_eq!(term_scorer.doc(), 0u32);
Ok(())
}
#[test]
fn test_non_index_bytes() -> crate::Result<()> {
let searcher = create_index_for_test(STORED)?;
assert_eq!(searcher.num_docs(), 1);
let field = searcher.schema().get_field("string_bytes").unwrap();
let term = Term::from_field_bytes(field, b"lucene".as_ref());
let term_query = TermQuery::new(term, IndexRecordOption::Basic);
let term_weight_res = term_query.specialized_weight(&searcher, false);
assert!(matches!(
term_weight_res,
Err(crate::TantivyError::SchemaError(_))
));
Ok(())
}
#[test]
fn test_fast_bytes_multivalue_value() -> crate::Result<()> {
let searcher = create_index_for_test(FAST)?;
assert_eq!(searcher.num_docs(), 1);
let fast_fields = searcher.segment_reader(0u32).fast_fields();
let field = searcher.schema().get_field("string_bytes").unwrap();
let fast_field_reader = fast_fields.bytes(field).unwrap();
assert_eq!(fast_field_reader.get_bytes(0u32), b"tantivy");
Ok(())
}
}

17
src/fastfield/bytes/reader.rs
View File

@@ -1,5 +1,6 @@
use crate::directory::FileSlice;
use crate::directory::OwnedBytes;
use owning_ref::OwningRef;
use crate::directory::ReadOnlySource;
use crate::fastfield::FastFieldReader;
use crate::DocId;
@@ -16,16 +17,16 @@ use crate::DocId;
#[derive(Clone)]
pub struct BytesFastFieldReader {
idx_reader: FastFieldReader<u64>,
values: OwnedBytes,
values: OwningRef<ReadOnlySource, [u8]>,
}
impl BytesFastFieldReader {
pub(crate) fn open(
idx_reader: FastFieldReader<u64>,
values_file: FileSlice,
) -> crate::Result<BytesFastFieldReader> {
let values = values_file.read_bytes()?;
Ok(BytesFastFieldReader { idx_reader, values })
values_source: ReadOnlySource,
) -> BytesFastFieldReader {
let values = OwningRef::new(values_source).map(|source| &source[..]);
BytesFastFieldReader { idx_reader, values }
}
fn range(&self, doc: DocId) -> (usize, usize) {
@@ -37,7 +38,7 @@ impl BytesFastFieldReader {
/// Returns the bytes associated to the given `doc`
pub fn get_bytes(&self, doc: DocId) -> &[u8] {
let (start, stop) = self.range(doc);
&self.values.as_slice()[start..stop]
&self.values[start..stop]
}
/// Returns the overall number of bytes in this bytes fast field.

39
src/fastfield/bytes/writer.rs
View File

@@ -49,10 +49,16 @@ impl BytesFastFieldWriter {
/// matching field values present in the document.
pub fn add_document(&mut self, doc: &Document) {
self.next_doc();
for field_value in doc.get_all(self.field) {
if let Value::Bytes(ref bytes) = field_value {
self.vals.extend_from_slice(bytes);
return;
for field_value in doc.field_values() {
if field_value.field() == self.field {
if let Value::Bytes(ref bytes) = *field_value.value() {
self.vals.extend_from_slice(bytes);
} else {
panic!(
"Bytes field contained non-Bytes Value!. Field {:?} = {:?}",
self.field, field_value
);
}
}
}
}
@@ -70,18 +76,21 @@ impl BytesFastFieldWriter {
/// Serializes the fast field values by pushing them to the `FastFieldSerializer`.
pub fn serialize(&self, serializer: &mut FastFieldSerializer) -> io::Result<()> {
// writing the offset index
let mut doc_index_serializer =
serializer.new_u64_fast_field_with_idx(self.field, 0, self.vals.len() as u64, 0)?;
for &offset in &self.doc_index {
doc_index_serializer.add_val(offset)?;
{
// writing the offset index
let mut doc_index_serializer =
serializer.new_u64_fast_field_with_idx(self.field, 0, self.vals.len() as u64, 0)?;
for &offset in &self.doc_index {
doc_index_serializer.add_val(offset)?;
}
doc_index_serializer.add_val(self.vals.len() as u64)?;
doc_index_serializer.close_field()?;
}
{
// writing the values themselves
let mut value_serializer = serializer.new_bytes_fast_field_with_idx(self.field, 1)?;
value_serializer.write_all(&self.vals)?;
}
doc_index_serializer.add_val(self.vals.len() as u64)?;
doc_index_serializer.close_field()?;
// writing the values themselves
serializer
.new_bytes_fast_field_with_idx(self.field, 1)?
.write_all(&self.vals)?;
Ok(())
}
}

93
src/fastfield/delete.rs
View File

@@ -1,6 +1,5 @@
use crate::common::{BitSet, HasLen};
use crate::directory::FileSlice;
use crate::directory::OwnedBytes;
use crate::directory::ReadOnlySource;
use crate::directory::WritePtr;
use crate::space_usage::ByteCount;
use crate::DocId;
@@ -10,8 +9,6 @@ use std::io::Write;
/// Write a delete `BitSet`
///
/// where `delete_bitset` is the set of deleted `DocId`.
/// Warning: this function does not call terminate. The caller is in charge of
/// closing the writer properly.
pub fn write_delete_bitset(
delete_bitset: &BitSet,
max_doc: u32,
@@ -40,41 +37,22 @@ pub fn write_delete_bitset(
/// Set of deleted `DocId`s.
#[derive(Clone)]
pub struct DeleteBitSet {
data: OwnedBytes,
data: ReadOnlySource,
len: usize,
}
impl DeleteBitSet {
#[cfg(test)]
pub(crate) fn for_test(docs: &[DocId], max_doc: u32) -> DeleteBitSet {
use crate::directory::{Directory, RAMDirectory, TerminatingWrite};
use std::path::Path;
assert!(docs.iter().all(|&doc| doc < max_doc));
let mut bitset = BitSet::with_max_value(max_doc);
for &doc in docs {
bitset.insert(doc);
}
let directory = RAMDirectory::create();
let path = Path::new("dummydeletebitset");
let mut wrt = directory.open_write(path).unwrap();
write_delete_bitset(&bitset, max_doc, &mut wrt).unwrap();
wrt.terminate().unwrap();
let file = directory.open_read(path).unwrap();
Self::open(file).unwrap()
}
/// Opens a delete bitset given its file.
pub fn open(file: FileSlice) -> crate::Result<DeleteBitSet> {
let bytes = file.read_bytes()?;
let num_deleted: usize = bytes
/// Opens a delete bitset given its data source.
pub fn open(data: ReadOnlySource) -> DeleteBitSet {
let num_deleted: usize = data
.as_slice()
.iter()
.map(|b| b.count_ones() as usize)
.sum();
Ok(DeleteBitSet {
data: bytes,
DeleteBitSet {
data,
len: num_deleted,
})
}
}
/// Returns true iff the document is still "alive". In other words, if it has not been deleted.
@@ -86,7 +64,7 @@ impl DeleteBitSet {
#[inline(always)]
pub fn is_deleted(&self, doc: DocId) -> bool {
let byte_offset = doc / 8u32;
let b: u8 = self.data.as_slice()[byte_offset as usize];
let b: u8 = (*self.data)[byte_offset as usize];
let shift = (doc & 7u32) as u8;
b & (1u8 << shift) != 0
}
@@ -105,35 +83,42 @@ impl HasLen for DeleteBitSet {
#[cfg(test)]
mod tests {
use super::DeleteBitSet;
use crate::common::HasLen;
use super::*;
use crate::directory::*;
use std::path::PathBuf;
#[test]
fn test_delete_bitset_empty() {
let delete_bitset = DeleteBitSet::for_test(&[], 10);
for doc in 0..10 {
assert_eq!(delete_bitset.is_deleted(doc), !delete_bitset.is_alive(doc));
fn test_delete_bitset_helper(bitset: &BitSet, max_doc: u32) {
let test_path = PathBuf::from("test");
let mut directory = RAMDirectory::default();
{
let mut writer = directory.open_write(&*test_path).unwrap();
write_delete_bitset(bitset, max_doc, &mut writer).unwrap();
writer.terminate().unwrap();
}
assert_eq!(delete_bitset.len(), 0);
let source = directory.open_read(&test_path).unwrap();
let delete_bitset = DeleteBitSet::open(source);
for doc in 0..max_doc {
assert_eq!(bitset.contains(doc), delete_bitset.is_deleted(doc as DocId));
}
assert_eq!(delete_bitset.len(), bitset.len());
}
#[test]
fn test_delete_bitset() {
let delete_bitset = DeleteBitSet::for_test(&[1, 9], 10);
assert!(delete_bitset.is_alive(0));
assert!(delete_bitset.is_deleted(1));
assert!(delete_bitset.is_alive(2));
assert!(delete_bitset.is_alive(3));
assert!(delete_bitset.is_alive(4));
assert!(delete_bitset.is_alive(5));
assert!(delete_bitset.is_alive(6));
assert!(delete_bitset.is_alive(6));
assert!(delete_bitset.is_alive(7));
assert!(delete_bitset.is_alive(8));
assert!(delete_bitset.is_deleted(9));
for doc in 0..10 {
assert_eq!(delete_bitset.is_deleted(doc), !delete_bitset.is_alive(doc));
{
let mut bitset = BitSet::with_max_value(10);
bitset.insert(1);
bitset.insert(9);
test_delete_bitset_helper(&bitset, 10);
}
{
let mut bitset = BitSet::with_max_value(8);
bitset.insert(1);
bitset.insert(2);
bitset.insert(3);
bitset.insert(5);
bitset.insert(7);
test_delete_bitset_helper(&bitset, 8);
}
assert_eq!(delete_bitset.len(), 2);
}
}

4
src/fastfield/error.rs
View File

@@ -4,8 +4,8 @@ use std::result;
/// `FastFieldNotAvailableError` is returned when the
/// user requested for a fast field reader, and the field was not
/// defined in the schema as a fast field.
#[derive(Debug, Error)]
#[error("Fast field not available: '{field_name:?}'")]
#[derive(Debug, Fail)]
#[fail(display = "Fast field not available: '{:?}'", field_name)]
pub struct FastFieldNotAvailableError {
field_name: String,
}

56
src/fastfield/facet_reader.rs
View File

@@ -73,61 +73,7 @@ impl FacetReader {
}
/// Return the list of facet ordinals associated to a document.
pub fn facet_ords(&self, doc: DocId, output: &mut Vec<u64>) {
pub fn facet_ords(&mut self, doc: DocId, output: &mut Vec<u64>) {
self.term_ords.get_vals(doc, output);
}
}
#[cfg(test)]
mod tests {
use crate::Index;
use crate::{
schema::{Facet, SchemaBuilder},
Document,
};
#[test]
fn test_facet_not_populated_for_all_docs() -> crate::Result<()> {
let mut schema_builder = SchemaBuilder::default();
let facet_field = schema_builder.add_facet_field("facet");
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(facet_field=>Facet::from_text("/a/b")));
index_writer.add_document(Document::default());
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let facet_reader = searcher
.segment_reader(0u32)
.facet_reader(facet_field)
.unwrap();
let mut facet_ords = Vec::new();
facet_reader.facet_ords(0u32, &mut facet_ords);
assert_eq!(&facet_ords, &[2u64]);
facet_reader.facet_ords(1u32, &mut facet_ords);
assert!(facet_ords.is_empty());
Ok(())
}
#[test]
fn test_facet_not_populated_for_any_docs() -> crate::Result<()> {
let mut schema_builder = SchemaBuilder::default();
let facet_field = schema_builder.add_facet_field("facet");
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(Document::default());
index_writer.add_document(Document::default());
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let facet_reader = searcher
.segment_reader(0u32)
.facet_reader(facet_field)
.unwrap();
let mut facet_ords = Vec::new();
facet_reader.facet_ords(0u32, &mut facet_ords);
assert!(facet_ords.is_empty());
facet_reader.facet_ords(1u32, &mut facet_ords);
assert!(facet_ords.is_empty());
Ok(())
}
}

168
src/fastfield/mod.rs
View File

@@ -33,14 +33,11 @@ pub use self::reader::FastFieldReader;
pub use self::readers::FastFieldReaders;
pub use self::serializer::FastFieldSerializer;
pub use self::writer::{FastFieldsWriter, IntFastFieldWriter};
use crate::chrono::{NaiveDateTime, Utc};
use crate::common;
use crate::schema::Cardinality;
use crate::schema::FieldType;
use crate::schema::Value;
use crate::{
chrono::{NaiveDateTime, Utc},
schema::Type,
};
mod bytes;
mod delete;
@@ -79,9 +76,6 @@ pub trait FastValue: Clone + Copy + Send + Sync + PartialOrd {
fn make_zero() -> Self {
Self::from_u64(0i64.to_u64())
}
/// Returns the `schema::Type` for this FastValue.
fn to_type() -> Type;
}
impl FastValue for u64 {
@@ -104,10 +98,6 @@ impl FastValue for u64 {
fn as_u64(&self) -> u64 {
*self
}
fn to_type() -> Type {
Type::U64
}
}
impl FastValue for i64 {
@@ -129,10 +119,6 @@ impl FastValue for i64 {
fn as_u64(&self) -> u64 {
*self as u64
}
fn to_type() -> Type {
Type::I64
}
}
impl FastValue for f64 {
@@ -154,10 +140,6 @@ impl FastValue for f64 {
fn as_u64(&self) -> u64 {
self.to_bits()
}
fn to_type() -> Type {
Type::F64
}
}
impl FastValue for crate::DateTime {
@@ -180,10 +162,6 @@ impl FastValue for crate::DateTime {
fn as_u64(&self) -> u64 {
self.timestamp().as_u64()
}
fn to_type() -> Type {
Type::Date
}
}
fn value_to_u64(value: &Value) -> u64 {
@@ -209,7 +187,6 @@ mod tests {
use crate::schema::FAST;
use crate::schema::{Document, IntOptions};
use crate::{Index, SegmentId, SegmentReader};
use common::HasLen;
use once_cell::sync::Lazy;
use rand::prelude::SliceRandom;
use rand::rngs::StdRng;
@@ -240,9 +217,9 @@ mod tests {
}
#[test]
fn test_intfastfield_small() -> crate::Result<()> {
fn test_intfastfield_small() {
let path = Path::new("test");
let directory: RAMDirectory = RAMDirectory::create();
let mut directory: RAMDirectory = RAMDirectory::create();
{
let write: WritePtr = directory.open_write(Path::new("test")).unwrap();
let mut serializer = FastFieldSerializer::from_write(write).unwrap();
@@ -255,24 +232,27 @@ mod tests {
.unwrap();
serializer.close().unwrap();
}
let file = directory.open_read(&path).unwrap();
assert_eq!(file.len(), 36 as usize);
let composite_file = CompositeFile::open(&file)?;
let file = composite_file.open_read(*FIELD).unwrap();
let fast_field_reader = FastFieldReader::<u64>::open(file)?;
assert_eq!(fast_field_reader.get(0), 13u64);
assert_eq!(fast_field_reader.get(1), 14u64);
assert_eq!(fast_field_reader.get(2), 2u64);
Ok(())
let source = directory.open_read(&path).unwrap();
{
assert_eq!(source.len(), 36 as usize);
}
{
let composite_file = CompositeFile::open(&source).unwrap();
let field_source = composite_file.open_read(*FIELD).unwrap();
let fast_field_reader = FastFieldReader::<u64>::open(field_source);
assert_eq!(fast_field_reader.get(0), 13u64);
assert_eq!(fast_field_reader.get(1), 14u64);
assert_eq!(fast_field_reader.get(2), 2u64);
}
}
#[test]
fn test_intfastfield_large() -> crate::Result<()> {
fn test_intfastfield_large() {
let path = Path::new("test");
let directory: RAMDirectory = RAMDirectory::create();
let mut directory: RAMDirectory = RAMDirectory::create();
{
let write: WritePtr = directory.open_write(Path::new("test"))?;
let mut serializer = FastFieldSerializer::from_write(write)?;
let write: WritePtr = directory.open_write(Path::new("test")).unwrap();
let mut serializer = FastFieldSerializer::from_write(write).unwrap();
let mut fast_field_writers = FastFieldsWriter::from_schema(&SCHEMA);
fast_field_writers.add_document(&doc!(*FIELD=>4u64));
fast_field_writers.add_document(&doc!(*FIELD=>14_082_001u64));
@@ -283,15 +263,19 @@ mod tests {
fast_field_writers.add_document(&doc!(*FIELD=>1_002u64));
fast_field_writers.add_document(&doc!(*FIELD=>1_501u64));
fast_field_writers.add_document(&doc!(*FIELD=>215u64));
fast_field_writers.serialize(&mut serializer, &HashMap::new())?;
serializer.close()?;
fast_field_writers
.serialize(&mut serializer, &HashMap::new())
.unwrap();
serializer.close().unwrap();
}
let file = directory.open_read(&path)?;
assert_eq!(file.len(), 61 as usize);
let source = directory.open_read(&path).unwrap();
{
let fast_fields_composite = CompositeFile::open(&file)?;
assert_eq!(source.len(), 61 as usize);
}
{
let fast_fields_composite = CompositeFile::open(&source).unwrap();
let data = fast_fields_composite.open_read(*FIELD).unwrap();
let fast_field_reader = FastFieldReader::<u64>::open(data)?;
let fast_field_reader = FastFieldReader::<u64>::open(data);
assert_eq!(fast_field_reader.get(0), 4u64);
assert_eq!(fast_field_reader.get(1), 14_082_001u64);
assert_eq!(fast_field_reader.get(2), 3_052u64);
@@ -302,13 +286,12 @@ mod tests {
assert_eq!(fast_field_reader.get(7), 1_501u64);
assert_eq!(fast_field_reader.get(8), 215u64);
}
Ok(())
}
#[test]
fn test_intfastfield_null_amplitude() -> crate::Result<()> {
fn test_intfastfield_null_amplitude() {
let path = Path::new("test");
let directory: RAMDirectory = RAMDirectory::create();
let mut directory: RAMDirectory = RAMDirectory::create();
{
let write: WritePtr = directory.open_write(Path::new("test")).unwrap();
@@ -322,23 +305,24 @@ mod tests {
.unwrap();
serializer.close().unwrap();
}
let file = directory.open_read(&path).unwrap();
assert_eq!(file.len(), 34 as usize);
let source = directory.open_read(&path).unwrap();
{
let fast_fields_composite = CompositeFile::open(&file).unwrap();
assert_eq!(source.len(), 34 as usize);
}
{
let fast_fields_composite = CompositeFile::open(&source).unwrap();
let data = fast_fields_composite.open_read(*FIELD).unwrap();
let fast_field_reader = FastFieldReader::<u64>::open(data)?;
let fast_field_reader = FastFieldReader::<u64>::open(data);
for doc in 0..10_000 {
assert_eq!(fast_field_reader.get(doc), 100_000u64);
}
}
Ok(())
}
#[test]
fn test_intfastfield_large_numbers() -> crate::Result<()> {
fn test_intfastfield_large_numbers() {
let path = Path::new("test");
let directory: RAMDirectory = RAMDirectory::create();
let mut directory: RAMDirectory = RAMDirectory::create();
{
let write: WritePtr = directory.open_write(Path::new("test")).unwrap();
@@ -354,12 +338,14 @@ mod tests {
.unwrap();
serializer.close().unwrap();
}
let file = directory.open_read(&path).unwrap();
assert_eq!(file.len(), 80042 as usize);
let source = directory.open_read(&path).unwrap();
{
let fast_fields_composite = CompositeFile::open(&file)?;
assert_eq!(source.len(), 80042 as usize);
}
{
let fast_fields_composite = CompositeFile::open(&source).unwrap();
let data = fast_fields_composite.open_read(*FIELD).unwrap();
let fast_field_reader = FastFieldReader::<u64>::open(data)?;
let fast_field_reader = FastFieldReader::<u64>::open(data);
assert_eq!(fast_field_reader.get(0), 0u64);
for doc in 1..10_001 {
assert_eq!(
@@ -368,13 +354,12 @@ mod tests {
);
}
}
Ok(())
}
#[test]
fn test_signed_intfastfield() -> crate::Result<()> {
fn test_signed_intfastfield() {
let path = Path::new("test");
let directory: RAMDirectory = RAMDirectory::create();
let mut directory: RAMDirectory = RAMDirectory::create();
let mut schema_builder = Schema::builder();
let i64_field = schema_builder.add_i64_field("field", FAST);
@@ -393,12 +378,14 @@ mod tests {
.unwrap();
serializer.close().unwrap();
}
let file = directory.open_read(&path).unwrap();
assert_eq!(file.len(), 17709 as usize);
let source = directory.open_read(&path).unwrap();
{
let fast_fields_composite = CompositeFile::open(&file)?;
assert_eq!(source.len(), 17709 as usize);
}
{
let fast_fields_composite = CompositeFile::open(&source).unwrap();
let data = fast_fields_composite.open_read(i64_field).unwrap();
let fast_field_reader = FastFieldReader::<i64>::open(data)?;
let fast_field_reader = FastFieldReader::<i64>::open(data);
assert_eq!(fast_field_reader.min_value(), -100i64);
assert_eq!(fast_field_reader.max_value(), 9_999i64);
@@ -411,13 +398,12 @@ mod tests {
assert_eq!(buffer[i], -100i64 + 53i64 + i as i64);
}
}
Ok(())
}
#[test]
fn test_signed_intfastfield_default_val() -> crate::Result<()> {
fn test_signed_intfastfield_default_val() {
let path = Path::new("test");
let directory: RAMDirectory = RAMDirectory::create();
let mut directory: RAMDirectory = RAMDirectory::create();
let mut schema_builder = Schema::builder();
let i64_field = schema_builder.add_i64_field("field", FAST);
let schema = schema_builder.build();
@@ -434,14 +420,13 @@ mod tests {
serializer.close().unwrap();
}
let file = directory.open_read(&path).unwrap();
let source = directory.open_read(&path).unwrap();
{
let fast_fields_composite = CompositeFile::open(&file).unwrap();
let fast_fields_composite = CompositeFile::open(&source).unwrap();
let data = fast_fields_composite.open_read(i64_field).unwrap();
let fast_field_reader = FastFieldReader::<i64>::open(data)?;
let fast_field_reader = FastFieldReader::<i64>::open(data);
assert_eq!(fast_field_reader.get(0u32), 0i64);
}
Ok(())
}
// Warning: this generates the same permutation at each call
@@ -452,26 +437,28 @@ mod tests {
}
#[test]
fn test_intfastfield_permutation() -> crate::Result<()> {
fn test_intfastfield_permutation() {
let path = Path::new("test");
let permutation = generate_permutation();
let n = permutation.len();
let directory = RAMDirectory::create();
let mut directory = RAMDirectory::create();
{
let write: WritePtr = directory.open_write(Path::new("test"))?;
let mut serializer = FastFieldSerializer::from_write(write)?;
let write: WritePtr = directory.open_write(Path::new("test")).unwrap();
let mut serializer = FastFieldSerializer::from_write(write).unwrap();
let mut fast_field_writers = FastFieldsWriter::from_schema(&SCHEMA);
for &x in &permutation {
fast_field_writers.add_document(&doc!(*FIELD=>x));
}
fast_field_writers.serialize(&mut serializer, &HashMap::new())?;
serializer.close()?;
fast_field_writers
.serialize(&mut serializer, &HashMap::new())
.unwrap();
serializer.close().unwrap();
}
let file = directory.open_read(&path)?;
let source = directory.open_read(&path).unwrap();
{
let fast_fields_composite = CompositeFile::open(&file)?;
let fast_fields_composite = CompositeFile::open(&source).unwrap();
let data = fast_fields_composite.open_read(*FIELD).unwrap();
let fast_field_reader = FastFieldReader::<u64>::open(data)?;
let fast_field_reader = FastFieldReader::<u64>::open(data);
let mut a = 0u64;
for _ in 0..n {
@@ -479,7 +466,6 @@ mod tests {
a = fast_field_reader.get(a as u32);
}
}
Ok(())
}
#[test]
@@ -488,7 +474,7 @@ mod tests {
let date_field = schema_builder.add_date_field("date", FAST);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.set_merge_policy(Box::new(NoMergePolicy));
index_writer.add_document(doc!(date_field =>crate::chrono::prelude::Utc::now()));
index_writer.commit().unwrap();
@@ -525,7 +511,7 @@ mod tests {
);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.set_merge_policy(Box::new(NoMergePolicy));
index_writer.add_document(doc!(
date_field => crate::DateTime::from_u64(1i64.to_u64()),
@@ -612,7 +598,7 @@ mod bench {
fn bench_intfastfield_linear_fflookup(b: &mut Bencher) {
let path = Path::new("test");
let permutation = generate_permutation();
let directory: RAMDirectory = RAMDirectory::create();
let mut directory: RAMDirectory = RAMDirectory::create();
{
let write: WritePtr = directory.open_write(Path::new("test")).unwrap();
let mut serializer = FastFieldSerializer::from_write(write).unwrap();
@@ -625,9 +611,9 @@ mod bench {
.unwrap();
serializer.close().unwrap();
}
let file = directory.open_read(&path).unwrap();
let source = directory.open_read(&path).unwrap();
{
let fast_fields_composite = CompositeFile::open(&file).unwrap();
let fast_fields_composite = CompositeFile::open(&source).unwrap();
let data = fast_fields_composite.open_read(*FIELD).unwrap();
let fast_field_reader = FastFieldReader::<u64>::open(data);
@@ -646,7 +632,7 @@ mod bench {
fn bench_intfastfield_fflookup(b: &mut Bencher) {
let path = Path::new("test");
let permutation = generate_permutation();
let directory: RAMDirectory = RAMDirectory::create();
let mut directory: RAMDirectory = RAMDirectory::create();
{
let write: WritePtr = directory.open_write(Path::new("test")).unwrap();
let mut serializer = FastFieldSerializer::from_write(write).unwrap();
@@ -659,9 +645,9 @@ mod bench {
.unwrap();
serializer.close().unwrap();
}
let file = directory.open_read(&path).unwrap();
let source = directory.open_read(&path).unwrap();
{
let fast_fields_composite = CompositeFile::open(&file).unwrap();
let fast_fields_composite = CompositeFile::open(&source).unwrap();
let data = fast_fields_composite.open_read(*FIELD).unwrap();
let fast_field_reader = FastFieldReader::<u64>::open(data);

38
src/fastfield/multivalued/mod.rs
View File

@@ -25,7 +25,7 @@ mod tests {
);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(field=>1u64, field=>3u64));
index_writer.add_document(doc!());
index_writer.add_document(doc!(field=>4u64));
@@ -64,7 +64,7 @@ mod tests {
schema_builder.add_i64_field("time_stamp_i", IntOptions::default().set_stored());
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
let first_time_stamp = chrono::Utc::now();
index_writer.add_document(
doc!(date_field=>first_time_stamp, date_field=>first_time_stamp, time_i=>1i64),
@@ -100,7 +100,6 @@ mod tests {
.get_first(date_field)
.expect("cannot find value")
.date_value()
.unwrap()
.timestamp(),
first_time_stamp.timestamp()
);
@@ -109,7 +108,7 @@ mod tests {
.get_first(time_i)
.expect("cannot find value")
.i64_value(),
Some(1i64)
1i64
);
}
}
@@ -132,7 +131,6 @@ mod tests {
.get_first(date_field)
.expect("cannot find value")
.date_value()
.unwrap()
.timestamp(),
two_secs_ahead.timestamp()
);
@@ -141,7 +139,7 @@ mod tests {
.get_first(time_i)
.expect("cannot find value")
.i64_value(),
Some(3i64)
3i64
);
}
}
@@ -188,7 +186,7 @@ mod tests {
);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(field=> 1i64, field => 3i64));
index_writer.add_document(doc!());
index_writer.add_document(doc!(field=> -4i64));
@@ -199,14 +197,22 @@ mod tests {
let segment_reader = searcher.segment_reader(0);
let mut vals = Vec::new();
let multi_value_reader = segment_reader.fast_fields().i64s(field).unwrap();
multi_value_reader.get_vals(2, &mut vals);
assert_eq!(&vals, &[-4i64]);
multi_value_reader.get_vals(0, &mut vals);
assert_eq!(&vals, &[1i64, 3i64]);
multi_value_reader.get_vals(1, &mut vals);
assert!(vals.is_empty());
multi_value_reader.get_vals(3, &mut vals);
assert_eq!(&vals, &[-5i64, -20i64, 1i64]);
{
multi_value_reader.get_vals(2, &mut vals);
assert_eq!(&vals, &[-4i64]);
}
{
multi_value_reader.get_vals(0, &mut vals);
assert_eq!(&vals, &[1i64, 3i64]);
}
{
multi_value_reader.get_vals(1, &mut vals);
assert!(vals.is_empty());
}
{
multi_value_reader.get_vals(3, &mut vals);
assert_eq!(&vals, &[-5i64, -20i64, 1i64]);
}
}
#[test]
#[ignore]
@@ -215,7 +221,7 @@ mod tests {
let field = schema_builder.add_facet_field("facetfield");
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
for i in 0..100_000 {
index_writer.add_document(doc!(field=> Facet::from(format!("/lang/{}", i).as_str())));
}

2
src/fastfield/multivalued/reader.rs
View File

@@ -74,7 +74,7 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index
.writer_for_tests()
.writer_with_num_threads(1, 30_000_000)
.expect("Failed to create index writer.");
index_writer.add_document(doc!(
facet_field => Facet::from("/category/cat2"),

2
src/fastfield/multivalued/writer.rs
View File

@@ -143,7 +143,7 @@ impl MultiValueIntFastFieldWriter {
.iter()
.map(|val| *mapping.get(val).expect("Missing term ordinal"));
doc_vals.extend(remapped_vals);
doc_vals.sort_unstable();
doc_vals.sort();
for &val in &doc_vals {
value_serializer.add_val(val)?;
}

39
src/fastfield/reader.rs
View File

@@ -3,12 +3,13 @@ use crate::common::bitpacker::BitUnpacker;
use crate::common::compute_num_bits;
use crate::common::BinarySerializable;
use crate::common::CompositeFile;
use crate::directory::FileSlice;
use crate::directory::ReadOnlySource;
use crate::directory::{Directory, RAMDirectory, WritePtr};
use crate::fastfield::{FastFieldSerializer, FastFieldsWriter};
use crate::schema::Schema;
use crate::schema::FAST;
use crate::DocId;
use owning_ref::OwningRef;
use std::collections::HashMap;
use std::marker::PhantomData;
use std::path::Path;
@@ -19,27 +20,34 @@ use std::path::Path;
/// fast field is required.
#[derive(Clone)]
pub struct FastFieldReader<Item: FastValue> {
bit_unpacker: BitUnpacker,
bit_unpacker: BitUnpacker<OwningRef<ReadOnlySource, [u8]>>,
min_value_u64: u64,
max_value_u64: u64,
_phantom: PhantomData<Item>,
}
impl<Item: FastValue> FastFieldReader<Item> {
/// Opens a fast field given a file.
pub fn open(file: FileSlice) -> crate::Result<Self> {
let mut bytes = file.read_bytes()?;
let min_value = u64::deserialize(&mut bytes)?;
let amplitude = u64::deserialize(&mut bytes)?;
/// Opens a fast field given a source.
pub fn open(data: ReadOnlySource) -> Self {
let min_value: u64;
let amplitude: u64;
{
let mut cursor = data.as_slice();
min_value =
u64::deserialize(&mut cursor).expect("Failed to read the min_value of fast field.");
amplitude =
u64::deserialize(&mut cursor).expect("Failed to read the amplitude of fast field.");
}
let max_value = min_value + amplitude;
let num_bits = compute_num_bits(amplitude);
let bit_unpacker = BitUnpacker::new(bytes, num_bits);
Ok(FastFieldReader {
let owning_ref = OwningRef::new(data).map(|data| &data[16..]);
let bit_unpacker = BitUnpacker::new(owning_ref, num_bits);
FastFieldReader {
min_value_u64: min_value,
max_value_u64: max_value,
bit_unpacker,
_phantom: PhantomData,
})
}
}
pub(crate) fn into_u64_reader(self) -> FastFieldReader<u64> {
@@ -127,7 +135,7 @@ impl<Item: FastValue> From<Vec<Item>> for FastFieldReader<Item> {
let field = schema_builder.add_u64_field("field", FAST);
let schema = schema_builder.build();
let path = Path::new("__dummy__");
let directory: RAMDirectory = RAMDirectory::create();
let mut directory: RAMDirectory = RAMDirectory::create();
{
let write: WritePtr = directory
.open_write(path)
@@ -149,11 +157,12 @@ impl<Item: FastValue> From<Vec<Item>> for FastFieldReader<Item> {
serializer.close().unwrap();
}
let file = directory.open_read(path).expect("Failed to open the file");
let composite_file = CompositeFile::open(&file).expect("Failed to read the composite file");
let field_file = composite_file
let source = directory.open_read(path).expect("Failed to open the file");
let composite_file =
CompositeFile::open(&source).expect("Failed to read the composite file");
let field_source = composite_file
.open_read(field)
.expect("File component not found");
FastFieldReader::open(field_file).unwrap()
FastFieldReader::open(field_source)
}
}

53
src/fastfield/readers.rs
View File

@@ -68,52 +68,45 @@ impl FastFieldReaders {
};
for (field, field_entry) in schema.fields() {
let field_type = field_entry.field_type();
if let FieldType::Bytes(bytes_option) = field_type {
if !bytes_option.is_fast() {
continue;
}
let fast_field_idx_file = fast_fields_composite
if field_type == &FieldType::Bytes {
let idx_reader = fast_fields_composite
.open_read_with_idx(field, 0)
.ok_or_else(|| FastFieldNotAvailableError::new(field_entry))?;
let idx_reader = FastFieldReader::open(fast_field_idx_file)?;
.ok_or_else(|| FastFieldNotAvailableError::new(field_entry))
.map(FastFieldReader::open)?;
let data = fast_fields_composite
.open_read_with_idx(field, 1)
.ok_or_else(|| FastFieldNotAvailableError::new(field_entry))?;
let bytes_fast_field_reader = BytesFastFieldReader::open(idx_reader, data)?;
fast_field_readers
.fast_bytes
.insert(field, bytes_fast_field_reader);
.insert(field, BytesFastFieldReader::open(idx_reader, data));
} else if let Some((fast_type, cardinality)) = type_and_cardinality(field_type) {
match cardinality {
Cardinality::SingleValue => {
if let Some(fast_field_data) = fast_fields_composite.open_read(field) {
match fast_type {
FastType::U64 => {
let fast_field_reader = FastFieldReader::open(fast_field_data)?;
let fast_field_reader = FastFieldReader::open(fast_field_data);
fast_field_readers
.fast_field_u64
.insert(field, fast_field_reader);
}
FastType::I64 => {
let fast_field_reader =
FastFieldReader::open(fast_field_data.clone())?;
fast_field_readers
.fast_field_i64
.insert(field, fast_field_reader);
fast_field_readers.fast_field_i64.insert(
field,
FastFieldReader::open(fast_field_data.clone()),
);
}
FastType::F64 => {
let fast_field_reader =
FastFieldReader::open(fast_field_data.clone())?;
fast_field_readers
.fast_field_f64
.insert(field, fast_field_reader);
fast_field_readers.fast_field_f64.insert(
field,
FastFieldReader::open(fast_field_data.clone()),
);
}
FastType::Date => {
let fast_field_reader =
FastFieldReader::open(fast_field_data.clone())?;
fast_field_readers
.fast_field_date
.insert(field, fast_field_reader);
fast_field_readers.fast_field_date.insert(
field,
FastFieldReader::open(fast_field_data.clone()),
);
}
}
} else {
@@ -124,10 +117,10 @@ impl FastFieldReaders {
let idx_opt = fast_fields_composite.open_read_with_idx(field, 0);
let data_opt = fast_fields_composite.open_read_with_idx(field, 1);
if let (Some(fast_field_idx), Some(fast_field_data)) = (idx_opt, data_opt) {
let idx_reader = FastFieldReader::open(fast_field_idx)?;
let idx_reader = FastFieldReader::open(fast_field_idx);
match fast_type {
FastType::I64 => {
let vals_reader = FastFieldReader::open(fast_field_data)?;
let vals_reader = FastFieldReader::open(fast_field_data);
let multivalued_int_fast_field =
MultiValueIntFastFieldReader::open(idx_reader, vals_reader);
fast_field_readers
@@ -135,7 +128,7 @@ impl FastFieldReaders {
.insert(field, multivalued_int_fast_field);
}
FastType::U64 => {
let vals_reader = FastFieldReader::open(fast_field_data)?;
let vals_reader = FastFieldReader::open(fast_field_data);
let multivalued_int_fast_field =
MultiValueIntFastFieldReader::open(idx_reader, vals_reader);
fast_field_readers
@@ -143,7 +136,7 @@ impl FastFieldReaders {
.insert(field, multivalued_int_fast_field);
}
FastType::F64 => {
let vals_reader = FastFieldReader::open(fast_field_data)?;
let vals_reader = FastFieldReader::open(fast_field_data);
let multivalued_int_fast_field =
MultiValueIntFastFieldReader::open(idx_reader, vals_reader);
fast_field_readers
@@ -151,7 +144,7 @@ impl FastFieldReaders {
.insert(field, multivalued_int_fast_field);
}
FastType::Date => {
let vals_reader = FastFieldReader::open(fast_field_data)?;
let vals_reader = FastFieldReader::open(fast_field_data);
let multivalued_int_fast_field =
MultiValueIntFastFieldReader::open(idx_reader, vals_reader);
fast_field_readers

11
src/fastfield/writer.rs
View File

@@ -33,7 +33,7 @@ impl FastFieldsWriter {
let mut bytes_value_writers = Vec::new();
for (field, field_entry) in schema.fields() {
match field_entry.field_type() {
match *field_entry.field_type() {
FieldType::I64(ref int_options)
| FieldType::U64(ref int_options)
| FieldType::F64(ref int_options)
@@ -56,11 +56,9 @@ impl FastFieldsWriter {
let fast_field_writer = MultiValueIntFastFieldWriter::new(field, true);
multi_values_writers.push(fast_field_writer);
}
FieldType::Bytes(bytes_option) => {
if bytes_option.is_fast() {
let fast_field_writer = BytesFastFieldWriter::new(field);
bytes_value_writers.push(fast_field_writer);
}
FieldType::Bytes => {
let fast_field_writer = BytesFastFieldWriter::new(field);
bytes_value_writers.push(fast_field_writer);
}
_ => {}
}
@@ -128,7 +126,6 @@ impl FastFieldsWriter {
for field_writer in &self.single_value_writers {
field_writer.serialize(serializer)?;
}
for field_writer in &self.multi_values_writers {
let field = field_writer.field();
field_writer.serialize(serializer, mapping.get(&field))?;

2
src/fieldnorm/mod.rs
View File

@@ -21,7 +21,7 @@ mod reader;
mod serializer;
mod writer;
pub use self::reader::{FieldNormReader, FieldNormReaders};
pub use self::reader::FieldNormReader;
pub use self::serializer::FieldNormsSerializer;
pub use self::writer::FieldNormsWriter;

115
src/fieldnorm/reader.rs
View File

@@ -1,47 +1,6 @@
use super::{fieldnorm_to_id, id_to_fieldnorm};
use crate::common::CompositeFile;
use crate::directory::FileSlice;
use crate::directory::OwnedBytes;
use crate::schema::Field;
use crate::space_usage::PerFieldSpaceUsage;
use crate::directory::ReadOnlySource;
use crate::DocId;
use std::sync::Arc;
/// Reader for the fieldnorm (for each document, the number of tokens indexed in the
/// field) of all indexed fields in the index.
///
/// Each fieldnorm is approximately compressed over one byte. We refer to this byte as
/// `fieldnorm_id`.
/// The mapping from `fieldnorm` to `fieldnorm_id` is given by monotonic.
#[derive(Clone)]
pub struct FieldNormReaders {
data: Arc<CompositeFile>,
}
impl FieldNormReaders {
/// Creates a field norm reader.
pub fn open(file: FileSlice) -> crate::Result<FieldNormReaders> {
let data = CompositeFile::open(&file)?;
Ok(FieldNormReaders {
data: Arc::new(data),
})
}
/// Returns the FieldNormReader for a specific field.
pub fn get_field(&self, field: Field) -> crate::Result<Option<FieldNormReader>> {
if let Some(file) = self.data.open_read(field) {
let fieldnorm_reader = FieldNormReader::open(file)?;
Ok(Some(fieldnorm_reader))
} else {
Ok(None)
}
}
/// Return a break down of the space usage per field.
pub fn space_usage(&self) -> PerFieldSpaceUsage {
self.data.space_usage()
}
}
/// Reads the fieldnorm associated to a document.
/// The fieldnorm represents the length associated to
@@ -49,7 +8,7 @@ impl FieldNormReaders {
///
/// This metric is important to compute the score of a
/// document : a document having a query word in one its short fields
/// (e.g. title)is likely to be more relevant than in one of its longer field
/// (e.g. title) is likely to be more relevant than in one of its longer field
/// (e.g. body).
///
/// tantivy encodes `fieldnorm` on one byte with some precision loss,
@@ -60,32 +19,14 @@ impl FieldNormReaders {
/// Apart from compression, this scale also makes it possible to
/// precompute computationally expensive functions of the fieldnorm
/// in a very short array.
#[derive(Clone)]
pub enum FieldNormReader {
ConstFieldNorm { fieldnorm_id: u8, num_docs: u32 },
OneByte(OwnedBytes),
pub struct FieldNormReader {
data: ReadOnlySource,
}
impl FieldNormReader {
pub fn const_fieldnorm_id(fieldnorm_id: u8, num_docs: u32) -> FieldNormReader {
FieldNormReader::ConstFieldNorm {
fieldnorm_id,
num_docs,
}
}
/// Opens a field norm reader given its file.
pub fn open(fieldnorm_file: FileSlice) -> crate::Result<Self> {
let data = fieldnorm_file.read_bytes()?;
Ok(FieldNormReader::OneByte(data))
}
/// Returns the number of documents in this segment.
pub fn num_docs(&self) -> u32 {
match self {
Self::ConstFieldNorm { num_docs, .. } => *num_docs,
FieldNormReader::OneByte(vals) => vals.len() as u32,
}
/// Opens a field norm reader given its data source.
pub fn open(data: ReadOnlySource) -> Self {
FieldNormReader { data }
}
/// Returns the `fieldnorm` associated to a doc id.
@@ -97,7 +38,6 @@ impl FieldNormReader {
///
/// The fieldnorm is effectively decoded from the
/// `fieldnorm_id` by doing a simple table lookup.
#[inline(always)]
pub fn fieldnorm(&self, doc_id: DocId) -> u32 {
let fieldnorm_id = self.fieldnorm_id(doc_id);
id_to_fieldnorm(fieldnorm_id)
@@ -106,11 +46,8 @@ impl FieldNormReader {
/// Returns the `fieldnorm_id` associated to a document.
#[inline(always)]
pub fn fieldnorm_id(&self, doc_id: DocId) -> u8 {
match self {
FieldNormReader::ConstFieldNorm { fieldnorm_id, .. } => *fieldnorm_id,
FieldNormReader::OneByte(data) => data.as_slice()[doc_id as usize],
}
let fielnorms_data = self.data.as_slice();
fielnorms_data[doc_id as usize]
}
/// Converts a `fieldnorm_id` into a fieldnorm.
@@ -125,32 +62,18 @@ impl FieldNormReader {
pub fn fieldnorm_to_id(fieldnorm: u32) -> u8 {
fieldnorm_to_id(fieldnorm)
}
#[cfg(test)]
pub fn for_test(field_norms: &[u32]) -> FieldNormReader {
let field_norms_id = field_norms
.iter()
.cloned()
.map(FieldNormReader::fieldnorm_to_id)
.collect::<Vec<u8>>();
let field_norms_data = OwnedBytes::new(field_norms_id);
FieldNormReader::OneByte(field_norms_data)
}
}
#[cfg(test)]
mod tests {
use crate::fieldnorm::FieldNormReader;
#[test]
fn test_from_fieldnorms_array() {
let fieldnorms = &[1, 2, 3, 4, 1_000_000];
let fieldnorm_reader = FieldNormReader::for_test(fieldnorms);
assert_eq!(fieldnorm_reader.num_docs(), 5);
assert_eq!(fieldnorm_reader.fieldnorm(0), 1);
assert_eq!(fieldnorm_reader.fieldnorm(1), 2);
assert_eq!(fieldnorm_reader.fieldnorm(2), 3);
assert_eq!(fieldnorm_reader.fieldnorm(3), 4);
assert_eq!(fieldnorm_reader.fieldnorm(4), 983_064);
impl From<Vec<u32>> for FieldNormReader {
fn from(field_norms: Vec<u32>) -> FieldNormReader {
let field_norms_id = field_norms
.into_iter()
.map(FieldNormReader::fieldnorm_to_id)
.collect::<Vec<u8>>();
let field_norms_data = ReadOnlySource::from(field_norms_id);
FieldNormReader {
data: field_norms_data,
}
}
}

47
src/fieldnorm/writer.rs
View File

@@ -13,7 +13,7 @@ use std::io;
/// byte per document per field.
pub struct FieldNormsWriter {
fields: Vec<Field>,
fieldnorms_buffer: Vec<Option<Vec<u8>>>,
fieldnorms_buffer: Vec<Vec<u8>>,
}
impl FieldNormsWriter {
@@ -23,7 +23,7 @@ impl FieldNormsWriter {
schema
.fields()
.filter_map(|(field, field_entry)| {
if field_entry.has_fieldnorms() {
if field_entry.is_indexed() {
Some(field)
} else {
None
@@ -36,14 +36,15 @@ impl FieldNormsWriter {
/// specified in the schema.
pub fn for_schema(schema: &Schema) -> FieldNormsWriter {
let fields = FieldNormsWriter::fields_with_fieldnorm(schema);
let num_fields = schema.num_fields();
let mut fieldnorms_buffer: Vec<Option<Vec<u8>>> = vec![None; num_fields];
for field in &fields {
fieldnorms_buffer[field.field_id() as usize] = Some(Vec::new());
}
let max_field = fields
.iter()
.map(Field::field_id)
.max()
.map(|max_field_id| max_field_id as usize + 1)
.unwrap_or(0);
FieldNormsWriter {
fields,
fieldnorms_buffer,
fieldnorms_buffer: (0..max_field).map(|_| Vec::new()).collect::<Vec<_>>(),
}
}
@@ -52,10 +53,8 @@ impl FieldNormsWriter {
///
/// Will extend with 0-bytes for documents that have not been seen.
pub fn fill_up_to_max_doc(&mut self, max_doc: DocId) {
for buffer_opt in self.fieldnorms_buffer.iter_mut() {
if let Some(buffer) = buffer_opt {
buffer.resize(max_doc as usize, 0u8);
}
for field in self.fields.iter() {
self.fieldnorms_buffer[field.field_id() as usize].resize(max_doc as usize, 0u8);
}
}
@@ -68,24 +67,22 @@ impl FieldNormsWriter {
/// * field - the field being set
/// * fieldnorm - the number of terms present in document `doc` in field `field`
pub fn record(&mut self, doc: DocId, field: Field, fieldnorm: u32) {
if let Some(fieldnorm_buffer) = self.fieldnorms_buffer[field.field_id() as usize].as_mut() {
assert!(
fieldnorm_buffer.len() <= doc as usize,
"Cannot register a given fieldnorm twice" // we fill intermediary `DocId` as having a fieldnorm of 0.
);
fieldnorm_buffer.resize(doc as usize + 1, 0u8);
fieldnorm_buffer[doc as usize] = fieldnorm_to_id(fieldnorm);
}
let fieldnorm_buffer: &mut Vec<u8> = &mut self.fieldnorms_buffer[field.field_id() as usize];
assert!(
fieldnorm_buffer.len() <= doc as usize,
"Cannot register a given fieldnorm twice"
);
// we fill intermediary `DocId` as having a fieldnorm of 0.
fieldnorm_buffer.resize(doc as usize + 1, 0u8);
fieldnorm_buffer[doc as usize] = fieldnorm_to_id(fieldnorm);
}
/// Serialize the seen fieldnorm values to the serializer for all fields.
pub fn serialize(&self, mut fieldnorms_serializer: FieldNormsSerializer) -> io::Result<()> {
pub fn serialize(&self, fieldnorms_serializer: &mut FieldNormsSerializer) -> io::Result<()> {
for &field in self.fields.iter() {
if let Some(buffer) = self.fieldnorms_buffer[field.field_id() as usize].as_ref() {
fieldnorms_serializer.serialize_field(field, &buffer[..])?;
}
let fieldnorm_values: &[u8] = &self.fieldnorms_buffer[field.field_id() as usize][..];
fieldnorms_serializer.serialize_field(field, fieldnorm_values)?;
}
fieldnorms_serializer.close()?;
Ok(())
}
}

293
src/indexer/index_writer.rs
View File

@@ -8,22 +8,26 @@ use crate::core::SegmentComponent;
use crate::core::SegmentId;
use crate::core::SegmentMeta;
use crate::core::SegmentReader;
use crate::directory::TerminatingWrite;
use crate::directory::{DirectoryLock, GarbageCollectionResult};
use crate::docset::{DocSet, TERMINATED};
use crate::directory::{TerminatingWrite, WatchCallbackList};
use crate::docset::DocSet;
use crate::error::TantivyError;
use crate::fastfield::write_delete_bitset;
use crate::indexer::delete_queue::{DeleteCursor, DeleteQueue};
use crate::indexer::doc_opstamp_mapping::DocToOpstampMapping;
use crate::indexer::operation::DeleteOperation;
use crate::indexer::segment_manager::SegmentRegisters;
use crate::indexer::segment_register::SegmentRegister;
use crate::indexer::stamper::Stamper;
use crate::indexer::MergePolicy;
use crate::indexer::SegmentEntry;
use crate::indexer::{SegmentEntry, ResourceManager};
use crate::indexer::SegmentWriter;
use crate::indexer::{IndexWriterConfig, MergePolicy};
use crate::reader::NRTReader;
use crate::schema::Document;
use crate::schema::IndexRecordOption;
use crate::schema::Term;
use crate::Opstamp;
use crate::tokenizer::TokenizerManager;
use crate::{IndexReader, Opstamp};
use crossbeam::channel;
use futures::executor::block_on;
use futures::future::Future;
@@ -31,18 +35,10 @@ use smallvec::smallvec;
use smallvec::SmallVec;
use std::mem;
use std::ops::Range;
use std::sync::Arc;
use std::sync::{Arc, RwLock};
use std::thread;
use std::thread::JoinHandle;
// Size of the margin for the heap. A segment is closed when the remaining memory
// in the heap goes below MARGIN_IN_BYTES.
pub const MARGIN_IN_BYTES: usize = 1_000_000;
// We impose the memory per thread to be at least 3 MB.
pub const HEAP_SIZE_MIN: usize = ((MARGIN_IN_BYTES as u32) * 3u32) as usize;
pub const HEAP_SIZE_MAX: usize = u32::max_value() as usize - MARGIN_IN_BYTES;
// Add document will block if the number of docs waiting in the queue to be indexed
// reaches `PIPELINE_MAX_SIZE_IN_DOCS`
const PIPELINE_MAX_SIZE_IN_DOCS: usize = 10_000;
@@ -69,8 +65,9 @@ pub struct IndexWriter {
_directory_lock: Option<DirectoryLock>,
index: Index,
config: IndexWriterConfig,
heap_size_in_bytes_per_thread: usize,
segment_registers: Arc<RwLock<SegmentRegisters>>,
workers_join_handle: Vec<JoinHandle<crate::Result<()>>>,
@@ -80,13 +77,14 @@ pub struct IndexWriter {
segment_updater: SegmentUpdater,
worker_id: usize,
num_threads: usize,
delete_queue: DeleteQueue,
stamper: Stamper,
committed_opstamp: Opstamp,
on_commit: WatchCallbackList,
memory_manager: ResourceManager,
}
fn compute_deleted_bitset(
@@ -108,19 +106,19 @@ fn compute_deleted_bitset(
// Limit doc helps identify the first document
// that may be affected by the delete operation.
let limit_doc = doc_opstamps.compute_doc_limit(delete_op.opstamp);
let inverted_index = segment_reader.inverted_index(delete_op.term.field())?;
let inverted_index = segment_reader.inverted_index(delete_op.term.field());
if let Some(mut docset) =
inverted_index.read_postings(&delete_op.term, IndexRecordOption::Basic)?
inverted_index.read_postings(&delete_op.term, IndexRecordOption::Basic)
{
let mut deleted_doc = docset.doc();
while deleted_doc != TERMINATED {
while docset.advance() {
let deleted_doc = docset.doc();
if deleted_doc < limit_doc {
delete_bitset.insert(deleted_doc);
might_have_changed = true;
}
deleted_doc = docset.advance();
}
}
delete_cursor.advance();
}
Ok(might_have_changed)
@@ -133,7 +131,6 @@ fn compute_deleted_bitset(
/// For instance, there was no delete operation between the state of the `segment_entry` and
/// the `target_opstamp`, `segment_entry` is not updated.
pub(crate) fn advance_deletes(
mut segment: Segment,
segment_entry: &mut SegmentEntry,
target_opstamp: Opstamp,
) -> crate::Result<()> {
@@ -142,25 +139,33 @@ pub(crate) fn advance_deletes(
return Ok(());
}
if segment_entry.delete_bitset().is_none() && segment_entry.delete_cursor().get().is_none() {
let delete_bitset_opt = segment_entry.take_delete_bitset();
// We avoid directly advancing the `SegmentEntry` delete cursor, because
// we do not want to end up in an invalid state if the delete bitset
// serialization fails.
let mut delete_cursor = segment_entry.delete_cursor();
if delete_bitset_opt.is_none() && segment_entry.delete_cursor().get().is_none() {
// There has been no `DeleteOperation` between the segment status and `target_opstamp`.
return Ok(());
}
// We open our current serialized segment to compute the new deleted bitset.
let segment = segment_entry.segment().clone();
let segment_reader = SegmentReader::open(&segment)?;
let max_doc = segment_reader.max_doc();
let mut delete_bitset: BitSet = match segment_entry.delete_bitset() {
Some(previous_delete_bitset) => (*previous_delete_bitset).clone(),
None => BitSet::with_max_value(max_doc),
};
let mut delete_bitset: BitSet =
delete_bitset_opt.unwrap_or_else(|| BitSet::with_max_value(max_doc));
let num_deleted_docs_before = segment.meta().num_deleted_docs();
compute_deleted_bitset(
&mut delete_bitset,
&segment_reader,
segment_entry.delete_cursor(),
&mut delete_cursor,
&DocToOpstampMapping::None,
target_opstamp,
)?;
@@ -179,32 +184,40 @@ pub(crate) fn advance_deletes(
let num_deleted_docs: u32 = delete_bitset.len() as u32;
if num_deleted_docs > num_deleted_docs_before {
// There are new deletes. We need to write a new delete file.
segment = segment.with_delete_meta(num_deleted_docs as u32, target_opstamp);
let mut delete_file = segment.open_write(SegmentComponent::DELETE)?;
let mut delete_file = segment
.with_delete_meta(num_deleted_docs as u32, target_opstamp)
.open_write(SegmentComponent::DELETE)?;
write_delete_bitset(&delete_bitset, max_doc, &mut delete_file)?;
delete_file.terminate()?;
segment_entry.reset_delete_meta(num_deleted_docs as u32, target_opstamp);
}
segment_entry.set_meta(segment.meta().clone());
// Regardless of whether we did end up having to write a new file or not
// we advance the `delete_cursor`. This is an optimisation. We want to ensure we do not
// check that a given deleted term does not match any of our docs more than once.
segment_entry.set_delete_cursor(delete_cursor);
Ok(())
}
fn index_documents(
memory_budget: usize,
config: IndexWriterConfig,
segment: Segment,
grouped_document_iterator: &mut dyn Iterator<Item = OperationGroup>,
segment_updater: &mut SegmentUpdater,
tokenizers: &TokenizerManager,
mut delete_cursor: DeleteCursor,
memory_manager: ResourceManager
) -> crate::Result<bool> {
let schema = segment.schema();
let mut segment_writer = SegmentWriter::for_segment(memory_budget, segment.clone(), &schema)?;
let mut segment_writer = SegmentWriter::for_segment(
&config, segment, &schema, tokenizers, memory_manager)?;
for document_group in grouped_document_iterator {
for doc in document_group {
segment_writer.add_document(doc, &schema)?;
}
let mem_usage = segment_writer.mem_usage();
if mem_usage >= memory_budget - MARGIN_IN_BYTES {
if mem_usage >= config.heap_size_before_flushing() {
info!(
"Buffer limit reached, flushing segment with maxdoc={}.",
segment_writer.max_doc()
@@ -223,24 +236,14 @@ fn index_documents(
// the worker thread.
assert!(max_doc > 0);
let doc_opstamps: Vec<Opstamp> = segment_writer.finalize()?;
let segment_with_max_doc = segment.with_max_doc(max_doc);
let (segment, doc_opstamps): (Segment, Vec<Opstamp>) = segment_writer.finalize()?;
let last_docstamp: Opstamp = *(doc_opstamps.last().unwrap());
let delete_bitset_opt = apply_deletes(
&segment_with_max_doc,
&mut delete_cursor,
&doc_opstamps,
last_docstamp,
)?;
let delete_bitset_opt =
apply_deletes(&segment, &mut delete_cursor, &doc_opstamps, last_docstamp)?;
let segment_entry = SegmentEntry::new(
segment_with_max_doc.meta().clone(),
delete_cursor,
delete_bitset_opt,
);
let segment_entry = SegmentEntry::new(segment, delete_cursor, delete_bitset_opt);
block_on(segment_updater.schedule_add_segment(segment_entry))?;
Ok(true)
}
@@ -292,21 +295,10 @@ impl IndexWriter {
/// If the heap size per thread is too small, panics.
pub(crate) fn new(
index: &Index,
num_threads: usize,
heap_size_in_bytes_per_thread: usize,
mut config: IndexWriterConfig,
directory_lock: DirectoryLock,
) -> crate::Result<IndexWriter> {
if heap_size_in_bytes_per_thread < HEAP_SIZE_MIN {
let err_msg = format!(
"The heap size per thread needs to be at least {}.",
HEAP_SIZE_MIN
);
return Err(TantivyError::InvalidArgument(err_msg));
}
if heap_size_in_bytes_per_thread >= HEAP_SIZE_MAX {
let err_msg = format!("The heap size per thread cannot exceed {}", HEAP_SIZE_MAX);
return Err(TantivyError::InvalidArgument(err_msg));
}
config.validate()?;
let (document_sender, document_receiver): (OperationSender, OperationReceiver) =
channel::bounded(PIPELINE_MAX_SIZE_IN_DOCS);
@@ -314,16 +306,26 @@ impl IndexWriter {
let current_opstamp = index.load_metas()?.opstamp;
let meta = index.load_metas()?;
let stamper = Stamper::new(current_opstamp);
let commited_segments = SegmentRegister::new(
index.directory(),
&index.schema(),
meta.segments,
&delete_queue.cursor(),
);
let segment_registers = Arc::new(RwLock::new(SegmentRegisters::new(commited_segments)));
let segment_updater =
SegmentUpdater::create(index.clone(), stamper.clone(), &delete_queue.cursor())?;
SegmentUpdater::create(segment_registers.clone(), index.clone(), stamper.clone())?;
let mut index_writer = IndexWriter {
_directory_lock: Some(directory_lock),
heap_size_in_bytes_per_thread,
index: index.clone(),
config,
operation_receiver: document_receiver,
operation_sender: document_sender,
@@ -331,7 +333,6 @@ impl IndexWriter {
segment_updater,
workers_join_handle: vec![],
num_threads,
delete_queue,
@@ -339,6 +340,10 @@ impl IndexWriter {
stamper,
worker_id: 0,
segment_registers,
on_commit: Default::default(),
memory_manager: Default::default()
};
index_writer.start_workers()?;
Ok(index_writer)
@@ -346,7 +351,7 @@ impl IndexWriter {
fn drop_sender(&mut self) {
let (sender, _receiver) = channel::bounded(1);
self.operation_sender = sender;
mem::replace(&mut self.operation_sender, sender);
}
/// If there are some merging threads, blocks until they all finish their work and
@@ -366,23 +371,11 @@ impl IndexWriter {
})?;
}
let result = self
self
.segment_updater
.wait_merging_thread()
.map_err(|_| TantivyError::ErrorInThread("Failed to join merging thread.".into()));
.wait_merging_thread();
if let Err(ref e) = result {
error!("Some merging thread failed {:?}", e);
}
result
}
#[doc(hidden)]
pub fn add_segment(&self, segment_meta: SegmentMeta) -> crate::Result<()> {
let delete_cursor = self.delete_queue.cursor();
let segment_entry = SegmentEntry::new(segment_meta, delete_cursor, None);
block_on(self.segment_updater.schedule_add_segment(segment_entry))
Ok(())
}
/// Creates a new segment.
@@ -405,8 +398,9 @@ impl IndexWriter {
let mut delete_cursor = self.delete_queue.cursor();
let mem_budget = self.heap_size_in_bytes_per_thread;
let index = self.index.clone();
let config = self.config.clone();
let memory_manager = self.memory_manager.clone();
let join_handle: JoinHandle<crate::Result<()>> = thread::Builder::new()
.name(format!("thrd-tantivy-index{}", self.worker_id))
.spawn(move || {
@@ -435,11 +429,13 @@ impl IndexWriter {
}
let segment = index.new_segment();
index_documents(
mem_budget,
config.clone(),
segment,
&mut document_iterator,
&mut segment_updater,
index.tokenizers(),
delete_cursor.clone(),
memory_manager.clone()
)?;
}
})?;
@@ -459,7 +455,7 @@ impl IndexWriter {
}
fn start_workers(&mut self) -> crate::Result<()> {
for _ in 0..self.num_threads {
for _ in 0..self.config.max_indexing_threads {
self.add_indexing_worker()?;
}
Ok(())
@@ -536,7 +532,6 @@ impl IndexWriter {
/// when no documents are remaining.
///
/// Returns the former segment_ready channel.
#[allow(unused_must_use)]
fn recreate_document_channel(&mut self) -> OperationReceiver {
let (document_sender, document_receiver): (OperationSender, OperationReceiver) =
channel::bounded(PIPELINE_MAX_SIZE_IN_DOCS);
@@ -565,18 +560,14 @@ impl IndexWriter {
.take()
.expect("The IndexWriter does not have any lock. This is a bug, please report.");
let new_index_writer: IndexWriter = IndexWriter::new(
&self.index,
self.num_threads,
self.heap_size_in_bytes_per_thread,
directory_lock,
)?;
let new_index_writer: IndexWriter =
IndexWriter::new(&self.index, self.config.clone(), directory_lock)?;
// the current `self` is dropped right away because of this call.
//
// This will drop the document queue, and the thread
// should terminate.
*self = new_index_writer;
mem::replace(self, new_index_writer);
// Drains the document receiver pipeline :
// Workers don't need to index the pending documents.
@@ -609,7 +600,7 @@ impl IndexWriter {
/// It is also possible to add a payload to the `commit`
/// using this API.
/// See [`PreparedCommit::set_payload()`](PreparedCommit.html)
pub fn prepare_commit(&mut self) -> crate::Result<PreparedCommit> {
pub fn prepare_commit(&mut self, soft_commit: bool) -> crate::Result<PreparedCommit> {
// Here, because we join all of the worker threads,
// all of the segment update for this commit have been
// sent.
@@ -637,7 +628,7 @@ impl IndexWriter {
}
let commit_opstamp = self.stamper.stamp();
let prepared_commit = PreparedCommit::new(self, commit_opstamp);
let prepared_commit = PreparedCommit::new(self, commit_opstamp, soft_commit);
info!("Prepared commit {}", commit_opstamp);
Ok(prepared_commit)
}
@@ -657,7 +648,25 @@ impl IndexWriter {
/// that made it in the commit.
///
pub fn commit(&mut self) -> crate::Result<Opstamp> {
self.prepare_commit()?.commit()
self.prepare_commit(false)?.commit()
}
pub fn soft_commit(&mut self) -> crate::Result<Opstamp> {
self.prepare_commit(true)?.commit()
}
pub(crate) fn trigger_commit(&self) -> impl Future<Output = ()> {
self.on_commit.broadcast()
}
pub fn reader(&self, num_searchers: usize) -> crate::Result<IndexReader> {
let nrt_reader = NRTReader::create(
num_searchers,
self.index.clone(),
self.segment_registers.clone(),
&self.on_commit,
)?;
Ok(IndexReader::NRT(nrt_reader))
}
pub(crate) fn segment_updater(&self) -> &SegmentUpdater {
@@ -800,7 +809,7 @@ mod tests {
let mut schema_builder = schema::Schema::builder();
let text_field = schema_builder.add_text_field("text", schema::TEXT);
let index = Index::create_in_ram(schema_builder.build());
let index_writer = index.writer_for_tests().unwrap();
let index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
let operations = vec![
UserOperation::Add(doc!(text_field=>"a")),
UserOperation::Add(doc!(text_field=>"b")),
@@ -815,7 +824,7 @@ mod tests {
let text_field = schema_builder.add_text_field("text", schema::TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(text_field => "hello1"));
index_writer.add_document(doc!(text_field => "hello2"));
assert!(index_writer.commit().is_ok());
@@ -864,7 +873,7 @@ mod tests {
.reload_policy(ReloadPolicy::Manual)
.try_into()
.unwrap();
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
let a_term = Term::from_field_text(text_field, "a");
let b_term = Term::from_field_text(text_field, "b");
let operations = vec![
@@ -926,8 +935,8 @@ mod tests {
fn test_lockfile_already_exists_error_msg() {
let schema_builder = schema::Schema::builder();
let index = Index::create_in_ram(schema_builder.build());
let _index_writer = index.writer_for_tests().unwrap();
match index.writer_for_tests() {
let _index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
match index.writer_with_num_threads(1, 3_000_000) {
Err(err) => {
let err_msg = err.to_string();
assert!(err_msg.contains("already an `IndexWriter`"));
@@ -979,7 +988,7 @@ mod tests {
let num_docs_containing = |s: &str| {
let searcher = reader.searcher();
let term = Term::from_field_text(text_field, s);
searcher.doc_freq(&term).unwrap()
searcher.doc_freq(&term)
};
{
@@ -1015,7 +1024,7 @@ mod tests {
.unwrap();
let num_docs_containing = |s: &str| {
let term_a = Term::from_field_text(text_field, s);
reader.searcher().doc_freq(&term_a).unwrap()
reader.searcher().doc_freq(&term_a)
};
{
// writing the segment
@@ -1055,7 +1064,8 @@ mod tests {
index_writer.add_document(doc!(text_field => "a"));
}
{
let mut prepared_commit = index_writer.prepare_commit().expect("commit failed");
let mut prepared_commit =
index_writer.prepare_commit(false).expect("commit failed");
prepared_commit.set_payload("first commit");
prepared_commit.commit().expect("commit failed");
}
@@ -1088,7 +1098,8 @@ mod tests {
index_writer.add_document(doc!(text_field => "a"));
}
{
let mut prepared_commit = index_writer.prepare_commit().expect("commit failed");
let mut prepared_commit =
index_writer.prepare_commit(false).expect("commit failed");
prepared_commit.set_payload("first commit");
prepared_commit.abort().expect("commit failed");
}
@@ -1110,7 +1121,6 @@ mod tests {
.unwrap()
.searcher()
.doc_freq(&term_a)
.unwrap()
};
assert_eq!(num_docs_containing("a"), 0);
assert_eq!(num_docs_containing("b"), 100);
@@ -1130,7 +1140,7 @@ mod tests {
reader.reload().unwrap();
let searcher = reader.searcher();
let term = Term::from_field_text(text_field, s);
searcher.doc_freq(&term).unwrap()
searcher.doc_freq(&term)
};
let mut index_writer = index.writer_with_num_threads(4, 12_000_000).unwrap();
@@ -1181,15 +1191,7 @@ mod tests {
// working with an empty index == no documents
let term_b = Term::from_field_text(text_field, "b");
assert_eq!(
index
.reader()
.unwrap()
.searcher()
.doc_freq(&term_b)
.unwrap(),
0
);
assert_eq!(index.reader().unwrap().searcher().doc_freq(&term_b), 0);
}
#[test]
@@ -1209,15 +1211,7 @@ mod tests {
let term_a = Term::from_field_text(text_field, "a");
// expect the document with that term to be in the index
assert_eq!(
index
.reader()
.unwrap()
.searcher()
.doc_freq(&term_a)
.unwrap(),
1
);
assert_eq!(index.reader().unwrap().searcher().doc_freq(&term_a), 1);
}
#[test]
@@ -1243,15 +1237,7 @@ mod tests {
// Find original docs in the index
let term_a = Term::from_field_text(text_field, "a");
// expect the document with that term to be in the index
assert_eq!(
index
.reader()
.unwrap()
.searcher()
.doc_freq(&term_a)
.unwrap(),
1
);
assert_eq!(index.reader().unwrap().searcher().doc_freq(&term_a), 1);
}
#[test]
@@ -1286,9 +1272,46 @@ mod tests {
let idfield = schema_builder.add_text_field("id", STRING);
schema_builder.add_text_field("optfield", STRING);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(idfield=>"myid"));
let commit = index_writer.commit();
assert!(commit.is_ok());
}
#[test]
fn test_index_writer_reader() {
let mut schema_builder = schema::Schema::builder();
let idfield = schema_builder.add_text_field("id", STRING);
schema_builder.add_text_field("optfield", STRING);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(idfield=>"myid"));
assert!(index_writer.commit().is_ok());
let reader = index_writer.reader(2).unwrap();
let searcher = reader.searcher();
assert_eq!(searcher.num_docs(), 1u64);
index_writer.add_document(doc!(idfield=>"myid"));
assert!(index_writer.commit().is_ok());
assert_eq!(reader.searcher().num_docs(), 2u64);
assert_eq!(searcher.num_docs(), 1u64);
}
#[test]
fn test_index_writer_reader_soft_commit() {
let mut schema_builder = schema::Schema::builder();
let idfield = schema_builder.add_text_field("id", STRING);
schema_builder.add_text_field("optfield", STRING);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(idfield=>"myid"));
assert!(index_writer.soft_commit().is_ok());
let nrt_reader = index_writer.reader(2).unwrap();
let normal_reader = index.reader_builder().try_into().unwrap();
assert_eq!(nrt_reader.searcher().num_docs(), 1u64);
assert_eq!(normal_reader.searcher().num_docs(), 0u64);
assert!(index_writer.commit().is_ok());
assert!(normal_reader.reload().is_ok());
assert_eq!(nrt_reader.searcher().num_docs(), 1u64);
assert_eq!(normal_reader.searcher().num_docs(), 1u64);
}
}

133
src/indexer/index_writer_config.rs Normal file
View File

@@ -0,0 +1,133 @@
use serde::{Deserialize, Serialize};
// Size of the margin for the heap. A segment is closed when the remaining memory
// in the heap goes below MARGIN_IN_BYTES.
const MARGIN_IN_BYTES: u64 = 1_000_000;
// We impose the memory per thread to be at least 3 MB.
const HEAP_SIZE_MIN: u64 = MARGIN_IN_BYTES * 3u64;
const HEAP_SIZE_MAX: u64 = u32::max_value() as u64 - MARGIN_IN_BYTES;
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct IndexWriterConfig {
pub max_indexing_threads: usize,
pub max_merging_threads: usize,
pub memory_budget: u64,
pub store_flush_num_bytes: u64,
pub persist_low: u64,
pub persist_high: u64,
}
impl Default for IndexWriterConfig {
fn default() -> Self {
IndexWriterConfig {
max_indexing_threads: 1,
max_merging_threads: 3,
memory_budget: 50_000_000u64,
store_flush_num_bytes: 10_000_000u64,
persist_low: 10_000_000u64,
persist_high: 50_000_000u64
}
}
}
impl IndexWriterConfig {
#[cfg(test)]
pub fn for_test() -> IndexWriterConfig {
IndexWriterConfig {
max_indexing_threads: 1,
max_merging_threads: 5,
memory_budget: 4_000_000u64,
store_flush_num_bytes: 500_000u64,
persist_low: 2_000_000u64,
persist_high: 3_000_000u64,
}
}
// Ensures the `IndexWriterConfig` is correct.
//
// This method checks that the values in the `IndexWriterConfig`
// are valid. If it is not, it may mutate some of the values (like `max_num_threads`) to
// fit the contracts or return an error with an explicit error message.
//
// If called twice, the config is guaranteed to not be updated the second time.
pub fn validate(&mut self) -> crate::Result<()> {
if self.memory_budget < HEAP_SIZE_MIN {
let err_msg = format!(
"The heap size per thread needs to be at least {}.",
HEAP_SIZE_MIN
);
return Err(crate::TantivyError::InvalidArgument(err_msg));
}
let heap_size_in_bytes_per_thread = self.heap_size_in_byte_per_thread();
if heap_size_in_bytes_per_thread >= HEAP_SIZE_MAX {
let err_msg = format!("The heap size per thread cannot exceed {}", HEAP_SIZE_MAX);
return Err(crate::TantivyError::InvalidArgument(err_msg));
}
if heap_size_in_bytes_per_thread < HEAP_SIZE_MIN {
self.max_indexing_threads = (self.memory_budget / HEAP_SIZE_MIN) as usize;
}
Ok(())
}
pub fn heap_size_in_byte_per_thread(&self) -> u64 {
self.memory_budget / self.max_indexing_threads as u64
}
pub fn heap_size_before_flushing(&self) -> u64 {
self.heap_size_in_byte_per_thread() - MARGIN_IN_BYTES
}
}
#[cfg(test)]
mod tests {
use crate::IndexWriterConfig;
#[test]
fn test_index_writer_config_simple() {
let mut index = IndexWriterConfig {
max_indexing_threads: 3,
memory_budget: super::HEAP_SIZE_MIN * 3,
..Default::default()
};
assert!(index.validate().is_ok());
assert_eq!(index.max_indexing_threads, 3);
assert_eq!(index.heap_size_in_byte_per_thread(), super::HEAP_SIZE_MIN);
}
#[test]
fn test_index_writer_config_reduce_num_threads() {
let mut index = IndexWriterConfig {
max_indexing_threads: 3,
memory_budget: super::HEAP_SIZE_MIN,
..Default::default()
};
assert!(index.validate().is_ok());
assert_eq!(index.max_indexing_threads, 1);
assert_eq!(index.heap_size_in_byte_per_thread(), super::HEAP_SIZE_MIN);
}
#[test]
fn test_index_writer_config_not_enough_memory() {
let mut index = IndexWriterConfig {
max_indexing_threads: 1,
memory_budget: super::HEAP_SIZE_MIN - 1,
..Default::default()
};
assert!(
matches!(index.validate(), Err(crate::TantivyError::InvalidArgument(msg) ) if msg.contains("The heap size per thread needs to be at least"))
);
}
#[test]
fn test_index_writer_config_too_much_memory() {
let mut index = IndexWriterConfig {
max_indexing_threads: 1,
memory_budget: (u32::max_value() as u64) + 1,
..Default::default()
};
assert!(
matches!(index.validate(), Err(crate::TantivyError::InvalidArgument(msg) ) if msg.contains("The heap size per thread cannot exceed"))
);
}
}

52
src/indexer/log_merge_policy.rs
View File

@@ -54,6 +54,10 @@ impl LogMergePolicy {
impl MergePolicy for LogMergePolicy {
fn compute_merge_candidates(&self, segments: &[SegmentMeta]) -> Vec<MergeCandidate> {
if segments.is_empty() {
return Vec::new();
}
let mut size_sorted_tuples = segments
.iter()
.map(SegmentMeta::num_docs)
@@ -63,35 +67,27 @@ impl MergePolicy for LogMergePolicy {
size_sorted_tuples.sort_by(|x, y| y.1.cmp(&(x.1)));
if size_sorted_tuples.len() <= 1 {
return Vec::new();
}
let size_sorted_log_tuples: Vec<_> = size_sorted_tuples
.into_iter()
.map(|(ind, num_docs)| (ind, f64::from(self.clip_min_size(num_docs)).log2()))
.collect();
if let Some(&(first_ind, first_score)) = size_sorted_log_tuples.first() {
let mut current_max_log_size = first_score;
let mut levels = vec![vec![first_ind]];
for &(ind, score) in (&size_sorted_log_tuples).iter().skip(1) {
if score < (current_max_log_size - self.level_log_size) {
current_max_log_size = score;
levels.push(Vec::new());
}
levels.last_mut().unwrap().push(ind);
let (first_ind, first_score) = size_sorted_log_tuples[0];
let mut current_max_log_size = first_score;
let mut levels = vec![vec![first_ind]];
for &(ind, score) in (&size_sorted_log_tuples).iter().skip(1) {
if score < (current_max_log_size - self.level_log_size) {
current_max_log_size = score;
levels.push(Vec::new());
}
levels
.iter()
.filter(|level| level.len() >= self.min_merge_size)
.map(|ind_vec| {
MergeCandidate(ind_vec.iter().map(|&ind| segments[ind].id()).collect())
})
.collect()
} else {
return vec![];
levels.last_mut().unwrap().push(ind);
}
levels
.iter()
.filter(|level| level.len() >= self.min_merge_size)
.map(|ind_vec| MergeCandidate(ind_vec.iter().map(|&ind| segments[ind].id()).collect()))
.collect()
}
}
@@ -183,7 +179,6 @@ mod tests {
let result_list = test_merge_policy().compute_merge_candidates(&test_input);
assert_eq!(result_list.len(), 2);
}
#[test]
fn test_log_merge_policy_small_segments() {
// segments under min_layer_size are merged together
@@ -199,17 +194,6 @@ mod tests {
assert_eq!(result_list.len(), 1);
}
#[test]
fn test_log_merge_policy_all_segments_too_large_to_merge() {
let eight_large_segments: Vec<SegmentMeta> =
std::iter::repeat_with(|| create_random_segment_meta(100_001))
.take(8)
.collect();
assert!(test_merge_policy()
.compute_merge_candidates(&eight_large_segments)
.is_empty());
}
#[test]
fn test_large_merge_segments() {
let test_input = vec![

28
src/indexer/merge_operation.rs
View File

@@ -1,17 +1,22 @@
use crate::indexer::resource_manager::{Allocation, ResourceManager};
use crate::Opstamp;
use crate::SegmentId;
use census::{Inventory, TrackedObject};
use std::collections::HashSet;
use std::fmt;
use std::ops::Deref;
#[derive(Default)]
pub(crate) struct MergeOperationInventory(Inventory<InnerMergeOperation>);
#[derive(Default, Clone)]
pub(crate) struct MergeOperationInventory {
inventory: Inventory<InnerMergeOperation>,
num_merge_watcher: ResourceManager,
}
impl Deref for MergeOperationInventory {
type Target = Inventory<InnerMergeOperation>;
fn deref(&self) -> &Self::Target {
&self.0
&self.inventory
}
}
@@ -25,6 +30,10 @@ impl MergeOperationInventory {
}
segment_in_merge
}
pub fn wait_until_empty(&self) {
let _ = self.num_merge_watcher.wait_until_in_range(0..1);
}
}
/// A `MergeOperation` has two roles.
@@ -47,6 +56,17 @@ pub struct MergeOperation {
pub(crate) struct InnerMergeOperation {
target_opstamp: Opstamp,
segment_ids: Vec<SegmentId>,
_allocation: Allocation,
}
impl fmt::Debug for InnerMergeOperation {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
write!(
f,
"MergeOp(target_opstamp={:?}, segment_ids={:?})",
self.target_opstamp, self.segment_ids
)
}
}
impl MergeOperation {
@@ -55,9 +75,11 @@ impl MergeOperation {
target_opstamp: Opstamp,
segment_ids: Vec<SegmentId>,
) -> MergeOperation {
let allocation = inventory.num_merge_watcher.allocate(1);
let inner_merge_operation = InnerMergeOperation {
target_opstamp,
segment_ids,
_allocation: allocation,
};
MergeOperation {
inner: inventory.track(inner_merge_operation),

578
src/indexer/merger.rs
View File

File diff suppressed because it is too large Load Diff

11
src/indexer/mod.rs
View File

@@ -2,12 +2,14 @@ pub mod delete_queue;
mod doc_opstamp_mapping;
pub mod index_writer;
mod index_writer_config;
mod log_merge_policy;
mod merge_operation;
pub mod merge_policy;
pub mod merger;
pub mod operation;
mod prepared_commit;
mod resource_manager;
mod segment_entry;
mod segment_manager;
mod segment_register;
@@ -16,22 +18,25 @@ pub mod segment_updater;
mod segment_writer;
mod stamper;
pub(crate) use self::resource_manager::{Allocation, ResourceManager};
pub(crate) use self::merge_operation::MergeOperationInventory;
pub use self::index_writer::IndexWriter;
pub use self::index_writer_config::IndexWriterConfig;
pub use self::log_merge_policy::LogMergePolicy;
pub use self::merge_operation::MergeOperation;
pub use self::merge_policy::{MergeCandidate, MergePolicy, NoMergePolicy};
pub use self::prepared_commit::PreparedCommit;
pub use self::segment_entry::SegmentEntry;
pub use self::segment_manager::SegmentManager;
pub(crate) use self::segment_manager::SegmentRegisters;
pub use self::segment_serializer::SegmentSerializer;
pub use self::segment_writer::SegmentWriter;
/// Alias for the default merge policy, which is the `LogMergePolicy`.
pub type DefaultMergePolicy = LogMergePolicy;
#[cfg(feature = "mmap")]
#[cfg(test)]
mod tests_mmap {
mod tests {
use crate::schema::{self, Schema};
use crate::{Index, Term};
@@ -40,7 +45,7 @@ mod tests_mmap {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", schema::TEXT);
let index = Index::create_from_tempdir(schema_builder.build()).unwrap();
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
// there must be one deleted document in the segment
index_writer.add_document(doc!(text_field=>"b"));
index_writer.delete_term(Term::from_field_text(text_field, "b"));

19
src/indexer/prepared_commit.rs
View File

@@ -7,14 +7,20 @@ pub struct PreparedCommit<'a> {
index_writer: &'a mut IndexWriter,
payload: Option<String>,
opstamp: Opstamp,
soft_commit: bool,
}
impl<'a> PreparedCommit<'a> {
pub(crate) fn new(index_writer: &'a mut IndexWriter, opstamp: Opstamp) -> PreparedCommit<'_> {
pub(crate) fn new(
index_writer: &'a mut IndexWriter,
opstamp: Opstamp,
soft_commit: bool,
) -> PreparedCommit<'_> {
PreparedCommit {
index_writer,
payload: None,
opstamp,
soft_commit,
}
}
@@ -32,11 +38,12 @@ impl<'a> PreparedCommit<'a> {
pub fn commit(self) -> crate::Result<Opstamp> {
info!("committing {}", self.opstamp);
let _ = block_on(
self.index_writer
.segment_updater()
.schedule_commit(self.opstamp, self.payload),
);
block_on(self.index_writer.segment_updater().schedule_commit(
self.opstamp,
self.payload,
self.soft_commit,
))?;
block_on(self.index_writer.trigger_commit());
Ok(self.opstamp)
}
}

213
src/indexer/resource_manager.rs Normal file
View File

@@ -0,0 +1,213 @@
use std::ops::RangeBounds;
use std::sync::{Arc, Condvar, Mutex, MutexGuard, RwLock};
struct LockedData {
count: u64,
enabled: bool
}
impl Default for LockedData {
fn default() -> Self {
LockedData {
count: 0u64,
enabled: true
}
}
}
#[derive(Default)]
struct Inner {
resource_level: Mutex<LockedData>,
convdvar: Condvar,
}
/// The resource manager makes it possible to track the amount of level of a given resource.
/// There is no magic here : it is to the description of the user to declare how much
/// of the resource is being held.
///
/// Allocation of a resource is bound to the lifetime of a `Allocation` instance.
///
/// ```rust
/// let resource_manager = ResourceManager::default();
///
/// ```
///
/// In tantivy, this is used to check the number of merging thread and the number of memory
/// used by the volatile segments.
///
#[derive(Clone, Default)]
pub struct ResourceManager {
inner: Arc<Inner>,
}
impl ResourceManager {
/// Return the total amount of reousrce allocated
pub fn total_amount(&self) -> u64 {
self.lock().count
}
fn lock(&self) -> MutexGuard<LockedData> {
self.inner
.resource_level
.lock()
.expect("Failed to obtain lock for ReservedMemory. This should never happen.")
}
fn record_delta(&self, delta: i64) {
if delta == 0i64 {
return;
}
let mut lock = self.lock();
let new_val = lock.count as i64 + delta;
lock.count = new_val as u64;
self.inner.convdvar.notify_all();
}
/// Records a new allocation.
///
/// The returned allocate object is used to automatically release the allocated resource
/// on drop.
pub fn allocate(&self, amount: u64) -> Allocation {
self.record_delta(amount as i64);
Allocation {
resource_manager: self.clone(),
amount: RwLock::new(amount),
}
}
/// Stops the resource manager.
///
/// If any thread is waiting via `.wait_until_in_range(...)`, the method will stop
/// being blocking and will return an error.
pub fn terminate(&self) {
self.lock().enabled = false;
self.inner.convdvar.notify_all();
}
/// Blocks the current thread until the resource level reaches the given range,
/// in a cpu-efficient way.
///
/// This method does not necessarily wakes up the current thread at every transition
/// into the targetted range, but any durable entry in the range will be detected.
pub fn wait_until_in_range<R: RangeBounds<u64>>(&self, range: R) -> Result<u64, u64> {
let mut levels = self.lock();
if !levels.enabled {
return Err(levels.count)
}
while !range.contains(&levels.count) {
levels = self.inner.convdvar.wait(levels).unwrap();
if !levels.enabled {
return Err(levels.count)
}
}
Ok(levels.count)
}
}
pub struct Allocation {
resource_manager: ResourceManager,
amount: RwLock<u64>,
}
impl Allocation {
pub fn amount(&self) -> u64 {
*self.amount.read().unwrap()
}
pub fn modify(&self, new_amount: u64) {
let mut wlock = self.amount.write().unwrap();
let delta = new_amount as i64 - *wlock as i64;
*wlock = new_amount;
self.resource_manager.record_delta(delta);
}
}
impl Drop for Allocation {
fn drop(&mut self) {
let amount = self.amount();
self.resource_manager.record_delta(-(amount as i64))
}
}
#[cfg(test)]
mod tests {
use super::ResourceManager;
use futures::channel::oneshot;
use futures::executor::block_on;
use std::{mem, thread};
#[test]
fn test_simple_allocation() {
let memory = ResourceManager::default();
assert_eq!(memory.total_amount(), 0u64);
let _allocation = memory.allocate(10u64);
assert_eq!(memory.total_amount(), 10u64);
}
#[test]
fn test_multiple_allocation() {
let memory = ResourceManager::default();
assert_eq!(memory.total_amount(), 0u64);
let _allocation = memory.allocate(10u64);
let _allocation_2 = memory.allocate(11u64);
assert_eq!(memory.total_amount(), 21u64);
}
#[test]
fn test_release_on_drop() {
let memory = ResourceManager::default();
assert_eq!(memory.total_amount(), 0u64);
let allocation = memory.allocate(10u64);
let allocation_2 = memory.allocate(11u64);
assert_eq!(memory.total_amount(), 21u64);
mem::drop(allocation);
assert_eq!(memory.total_amount(), 11u64);
mem::drop(allocation_2);
assert_eq!(memory.total_amount(), 0u64);
}
#[test]
fn test_wait_until() {
let memory = ResourceManager::default();
let (send, recv) = oneshot::channel::<()>();
let memory_clone = memory.clone();
thread::spawn(move || {
let _allocation1 = memory_clone.allocate(2u64);
let _allocation2 = memory_clone.allocate(3u64);
let _allocation3 = memory_clone.allocate(4u64);
std::mem::drop(_allocation3);
assert!(block_on(recv).is_ok());
});
assert_eq!(memory.wait_until_in_range(5u64..8u64), Ok(5u64));
assert!(send.send(()).is_ok());
}
#[test]
fn test_modify_amount() {
let memory = ResourceManager::default();
let alloc = memory.allocate(2u64);
assert_eq!(memory.total_amount(), 2u64);
assert_eq!(alloc.amount(), 2u64);
let alloc2 = memory.allocate(3u64);
assert_eq!(memory.total_amount(), 2u64 + 3u64);
assert_eq!(alloc2.amount(), 3u64);
alloc.modify(14u64);
assert_eq!(alloc.amount(), 14u64);
assert_eq!(memory.total_amount(), 14u64 + 3u64)
}
#[test]
fn test_stop_resource_manager() {
let resource_manager = ResourceManager::default();
let resource_manager_clone = resource_manager.clone();
let (sender, recv) = oneshot::channel();
let join_handle = thread::spawn(move || {
assert!(sender.send(()).is_ok());
resource_manager_clone.wait_until_in_range(10..20)
});
let _ = block_on(recv);
resource_manager.terminate();
assert_eq!(join_handle.join().unwrap(), Err(0u64));
}
}

54
src/indexer/segment_entry.rs
View File

@@ -1,7 +1,9 @@
use crate::common::BitSet;
use crate::core::SegmentId;
use crate::core::SegmentMeta;
use crate::directory::ManagedDirectory;
use crate::indexer::delete_queue::DeleteCursor;
use crate::{Opstamp, Segment};
use std::fmt;
/// A segment entry describes the state of
@@ -19,7 +21,7 @@ use std::fmt;
/// in the .del file or in the `delete_bitset`.
#[derive(Clone)]
pub struct SegmentEntry {
meta: SegmentMeta,
segment: Segment,
delete_bitset: Option<BitSet>,
delete_cursor: DeleteCursor,
}
@@ -27,47 +29,67 @@ pub struct SegmentEntry {
impl SegmentEntry {
/// Create a new `SegmentEntry`
pub fn new(
segment_meta: SegmentMeta,
segment: Segment,
delete_cursor: DeleteCursor,
delete_bitset: Option<BitSet>,
) -> SegmentEntry {
SegmentEntry {
meta: segment_meta,
segment,
delete_bitset,
delete_cursor,
}
}
/// Return a reference to the segment entry deleted bitset.
///
/// `DocId` in this bitset are flagged as deleted.
pub fn delete_bitset(&self) -> Option<&BitSet> {
self.delete_bitset.as_ref()
pub fn persist(&mut self, dest_directory: ManagedDirectory) -> crate::Result<()> {
// TODO take in account delete bitset?
self.segment.persist(dest_directory)?;
Ok(())
}
/// Set the `SegmentMeta` for this segment.
pub fn set_meta(&mut self, segment_meta: SegmentMeta) {
self.meta = segment_meta;
pub fn set_delete_cursor(&mut self, delete_cursor: DeleteCursor) {
self.delete_cursor = delete_cursor;
}
/// `Takes` (as in Option::take) the delete bitset of a segment entry.
/// `DocId` in this bitset are flagged as deleted.
pub fn take_delete_bitset(&mut self) -> Option<BitSet> {
self.delete_bitset.take()
}
/// Reset the delete information in this segment.
///
/// The `SegmentEntry` segment's `SegmentMeta` gets updated, and
/// any delete bitset is drop and set to None.
pub fn reset_delete_meta(&mut self, num_deleted_docs: u32, target_opstamp: Opstamp) {
self.segment = self
.segment
.clone()
.with_delete_meta(num_deleted_docs, target_opstamp);
self.delete_bitset = None;
}
/// Return a reference to the segment_entry's delete cursor
pub fn delete_cursor(&mut self) -> &mut DeleteCursor {
&mut self.delete_cursor
pub fn delete_cursor(&mut self) -> DeleteCursor {
self.delete_cursor.clone()
}
/// Returns the segment id.
pub fn segment_id(&self) -> SegmentId {
self.meta.id()
self.meta().id()
}
/// Returns the `segment` associated to the `SegmentEntry`.
pub fn segment(&self) -> &Segment {
&self.segment
}
/// Accessor to the `SegmentMeta`
pub fn meta(&self) -> &SegmentMeta {
&self.meta
self.segment.meta()
}
}
impl fmt::Debug for SegmentEntry {
fn fmt(&self, formatter: &mut fmt::Formatter<'_>) -> fmt::Result {
write!(formatter, "SegmentEntry({:?})", self.meta)
write!(formatter, "SegmentEntry({:?})", self.meta())
}
}

88
src/indexer/segment_manager.rs
View File

@@ -2,15 +2,15 @@ use super::segment_register::SegmentRegister;
use crate::core::SegmentId;
use crate::core::SegmentMeta;
use crate::error::TantivyError;
use crate::indexer::delete_queue::DeleteCursor;
use crate::indexer::SegmentEntry;
use crate::indexer::{SegmentEntry, MergeOperationInventory, MergeCandidate, MergeOperation};
use crate::{Segment, Opstamp};
use std::collections::hash_set::HashSet;
use std::fmt::{self, Debug, Formatter};
use std::sync::RwLock;
use std::sync::{Arc, RwLock};
use std::sync::{RwLockReadGuard, RwLockWriteGuard};
#[derive(Default)]
struct SegmentRegisters {
pub(crate) struct SegmentRegisters {
uncommitted: SegmentRegister,
committed: SegmentRegister,
}
@@ -22,6 +22,17 @@ pub(crate) enum SegmentsStatus {
}
impl SegmentRegisters {
pub fn new(committed: SegmentRegister) -> SegmentRegisters {
SegmentRegisters {
uncommitted: Default::default(),
committed,
}
}
pub fn committed_segment(&self) -> Vec<Segment> {
self.committed.segments()
}
/// Check if all the segments are committed or uncommited.
///
/// If some segment is missing or segments are in a different state (this should not happen
@@ -44,7 +55,8 @@ impl SegmentRegisters {
/// changes (merges especially)
#[derive(Default)]
pub struct SegmentManager {
registers: RwLock<SegmentRegisters>,
registers: Arc<RwLock<SegmentRegisters>>,
merge_operations: MergeOperationInventory,
}
impl Debug for SegmentManager {
@@ -58,34 +70,28 @@ impl Debug for SegmentManager {
}
}
pub fn get_mergeable_segments(
in_merge_segment_ids: &HashSet<SegmentId>,
segment_manager: &SegmentManager,
) -> (Vec<SegmentMeta>, Vec<SegmentMeta>) {
let registers_lock = segment_manager.read();
(
registers_lock
.committed
.get_mergeable_segments(in_merge_segment_ids),
registers_lock
.uncommitted
.get_mergeable_segments(in_merge_segment_ids),
)
}
impl SegmentManager {
pub fn from_segments(
segment_metas: Vec<SegmentMeta>,
delete_cursor: &DeleteCursor,
) -> SegmentManager {
pub(crate) fn new(registers: Arc<RwLock<SegmentRegisters>>) -> SegmentManager {
SegmentManager {
registers: RwLock::new(SegmentRegisters {
uncommitted: SegmentRegister::default(),
committed: SegmentRegister::new(segment_metas, delete_cursor),
}),
registers,
merge_operations: Default::default()
}
}
pub fn new_merge_operation(&self, opstamp: Opstamp, merge_candidate: MergeCandidate) -> MergeOperation {
MergeOperation::new(
&self.merge_operations,
opstamp,
merge_candidate.0
)
}
pub fn wait_merging_thread(&self) {
self.merge_operations.wait_until_empty()
}
/// Returns all of the segment entries (committed or uncommitted)
pub fn segment_entries(&self) -> Vec<SegmentEntry> {
let registers_lock = self.read();
@@ -94,6 +100,34 @@ impl SegmentManager {
segment_entries
}
/// Returns the segments that are currently not in merge.
///
/// They are split over two `Vec`. The committed segments on one hand
/// and the uncommitted ones on the other hand.
///
/// This method is useful when searching for merge candidate or segments
/// to persists.
pub fn segments_not_in_merge(&self) -> (Vec<SegmentMeta>, Vec<SegmentMeta>) {
let in_merge_segment_ids: HashSet<SegmentId> = self.merge_operations.segment_in_merge();
let registers_lock = self.read();
(
registers_lock
.committed
.get_mergeable_segments(&in_merge_segment_ids),
registers_lock
.uncommitted
.get_mergeable_segments(&in_merge_segment_ids),
)
}
pub fn largest_segment_not_in_merge(&self) -> Option<SegmentMeta> {
let (committed, uncommitted) = self.segments_not_in_merge();
let mut segments = vec![];
segments.extend(committed);
segments.extend(uncommitted);
None
}
// Lock poisoning should never happen :
// The lock is acquired and released within this class,
// and the operations cannot panic.

37
src/indexer/segment_register.rs
View File

@@ -1,7 +1,10 @@
use crate::core::SegmentId;
use crate::core::SegmentMeta;
use crate::directory::ManagedDirectory;
use crate::indexer::delete_queue::DeleteCursor;
use crate::indexer::segment_entry::SegmentEntry;
use crate::schema::Schema;
use crate::Segment;
use std::collections::HashMap;
use std::collections::HashSet;
use std::fmt::{self, Debug, Formatter};
@@ -46,6 +49,13 @@ impl SegmentRegister {
.collect()
}
pub fn segments(&self) -> Vec<Segment> {
self.segment_states
.values()
.map(|segment_entry| segment_entry.segment().clone())
.collect()
}
pub fn segment_entries(&self) -> Vec<SegmentEntry> {
self.segment_states.values().cloned().collect()
}
@@ -79,11 +89,17 @@ impl SegmentRegister {
self.segment_states.get(segment_id).cloned()
}
pub fn new(segment_metas: Vec<SegmentMeta>, delete_cursor: &DeleteCursor) -> SegmentRegister {
pub fn new(
directory: &ManagedDirectory,
schema: &Schema,
segment_metas: Vec<SegmentMeta>,
delete_cursor: &DeleteCursor,
) -> SegmentRegister {
let mut segment_states = HashMap::new();
for segment_meta in segment_metas {
let segment_id = segment_meta.id();
let segment_entry = SegmentEntry::new(segment_meta, delete_cursor.clone(), None);
let segment = Segment::new_persisted(segment_meta, directory.clone(), schema.clone());
let segment_entry = SegmentEntry::new(segment, delete_cursor.clone(), None);
segment_states.insert(segment_id, segment_entry);
}
SegmentRegister { segment_states }
@@ -95,6 +111,7 @@ mod tests {
use super::*;
use crate::core::{SegmentId, SegmentMetaInventory};
use crate::indexer::delete_queue::*;
use crate::indexer::ResourceManager;
fn segment_ids(segment_register: &SegmentRegister) -> Vec<SegmentId> {
segment_register
@@ -108,28 +125,34 @@ mod tests {
fn test_segment_register() {
let inventory = SegmentMetaInventory::default();
let delete_queue = DeleteQueue::new();
let schema = Schema::builder().build();
let mut segment_register = SegmentRegister::default();
let segment_id_a = SegmentId::generate_random();
let segment_id_b = SegmentId::generate_random();
let segment_id_merged = SegmentId::generate_random();
let memory_manager = ResourceManager::default();
{
let segment_meta = inventory.new_segment_meta(segment_id_a, 0u32);
let segment_entry = SegmentEntry::new(segment_meta, delete_queue.cursor(), None);
let meta = inventory.new_segment_meta(segment_id_a, 0u32);
let segment = Segment::new_volatile(meta, schema.clone(), memory_manager.clone());
let segment_entry = SegmentEntry::new(segment, delete_queue.cursor(), None);
segment_register.add_segment_entry(segment_entry);
}
assert_eq!(segment_ids(&segment_register), vec![segment_id_a]);
{
let segment_meta = inventory.new_segment_meta(segment_id_b, 0u32);
let segment_entry = SegmentEntry::new(segment_meta, delete_queue.cursor(), None);
let meta = inventory.new_segment_meta(segment_id_b, 0u32);
let segment = Segment::new_volatile(meta, schema.clone(), memory_manager.clone());
let segment_entry = SegmentEntry::new(segment, delete_queue.cursor(), None);
segment_register.add_segment_entry(segment_entry);
}
segment_register.remove_segment(&segment_id_a);
segment_register.remove_segment(&segment_id_b);
{
let segment_meta_merged = inventory.new_segment_meta(segment_id_merged, 0u32);
let segment_entry = SegmentEntry::new(segment_meta_merged, delete_queue.cursor(), None);
let segment_merged = Segment::new_volatile(segment_meta_merged, schema.clone(), memory_manager.clone());
let segment_entry = SegmentEntry::new(segment_merged, delete_queue.cursor(), None);
segment_register.add_segment_entry(segment_entry);
}
assert_eq!(segment_ids(&segment_register), vec![segment_id_merged]);

39
src/indexer/segment_serializer.rs
View File

@@ -3,43 +3,32 @@ use crate::core::SegmentComponent;
use crate::fastfield::FastFieldSerializer;
use crate::fieldnorm::FieldNormsSerializer;
use crate::postings::InvertedIndexSerializer;
use crate::store::StoreWriter;
/// Segment serializer is in charge of laying out on disk
/// the data accumulated and sorted by the `SegmentWriter`.
pub struct SegmentSerializer {
segment: Segment,
store_writer: StoreWriter,
fast_field_serializer: FastFieldSerializer,
fieldnorms_serializer: Option<FieldNormsSerializer>,
fieldnorms_serializer: FieldNormsSerializer,
postings_serializer: InvertedIndexSerializer,
}
impl SegmentSerializer {
/// Creates a new `SegmentSerializer`.
pub fn for_segment(mut segment: Segment) -> crate::Result<SegmentSerializer> {
let store_write = segment.open_write(SegmentComponent::STORE)?;
pub fn for_segment(segment: &mut Segment) -> crate::Result<SegmentSerializer> {
let fast_field_write = segment.open_write(SegmentComponent::FASTFIELDS)?;
let fast_field_serializer = FastFieldSerializer::from_write(fast_field_write)?;
let fieldnorms_write = segment.open_write(SegmentComponent::FIELDNORMS)?;
let fieldnorms_serializer = FieldNormsSerializer::from_write(fieldnorms_write)?;
let postings_serializer = InvertedIndexSerializer::open(&mut segment)?;
let postings_serializer = InvertedIndexSerializer::open(segment)?;
Ok(SegmentSerializer {
segment,
store_writer: StoreWriter::new(store_write),
fast_field_serializer,
fieldnorms_serializer: Some(fieldnorms_serializer),
fieldnorms_serializer,
postings_serializer,
})
}
pub fn segment(&self) -> &Segment {
&self.segment
}
/// Accessor to the `PostingsSerializer`.
pub fn get_postings_serializer(&mut self) -> &mut InvertedIndexSerializer {
&mut self.postings_serializer
@@ -50,26 +39,16 @@ impl SegmentSerializer {
&mut self.fast_field_serializer
}
/// Extract the field norm serializer.
///
/// Note the fieldnorms serializer can only be extracted once.
pub fn extract_fieldnorms_serializer(&mut self) -> Option<FieldNormsSerializer> {
self.fieldnorms_serializer.take()
}
/// Accessor to the `StoreWriter`.
pub fn get_store_writer(&mut self) -> &mut StoreWriter {
&mut self.store_writer
/// Accessor to the field norm serializer.
pub fn get_fieldnorms_serializer(&mut self) -> &mut FieldNormsSerializer {
&mut self.fieldnorms_serializer
}
/// Finalize the segment serialization.
pub fn close(mut self) -> crate::Result<()> {
if let Some(fieldnorms_serializer) = self.extract_fieldnorms_serializer() {
fieldnorms_serializer.close()?;
}
pub fn close(self) -> crate::Result<()> {
self.fast_field_serializer.close()?;
self.postings_serializer.close()?;
self.store_writer.close()?;
self.fieldnorms_serializer.close()?;
Ok(())
}
}

124
src/indexer/segment_updater.rs
View File

@@ -1,4 +1,4 @@
use super::segment_manager::{get_mergeable_segments, SegmentManager};
use super::segment_manager::SegmentManager;
use crate::core::Index;
use crate::core::IndexMeta;
use crate::core::Segment;
@@ -7,22 +7,21 @@ use crate::core::SegmentMeta;
use crate::core::SerializableSegment;
use crate::core::META_FILEPATH;
use crate::directory::{Directory, DirectoryClone, GarbageCollectionResult};
use crate::indexer::delete_queue::DeleteCursor;
use crate::indexer::index_writer::advance_deletes;
use crate::indexer::merge_operation::MergeOperationInventory;
use crate::indexer::merger::IndexMerger;
use crate::indexer::segment_manager::SegmentsStatus;
use crate::indexer::segment_manager::{SegmentRegisters, SegmentsStatus};
use crate::indexer::stamper::Stamper;
use crate::indexer::SegmentEntry;
use crate::indexer::SegmentSerializer;
use crate::indexer::{DefaultMergePolicy, MergePolicy};
use crate::indexer::{MergeCandidate, MergeOperation};
use crate::schema::Schema;
use crate::Opstamp;
use crate::{Opstamp, SegmentComponent};
use futures::channel::oneshot;
use futures::executor::{ThreadPool, ThreadPoolBuilder};
use futures::future::Future;
use futures::future::TryFutureExt;
use serde_json;
use std::borrow::BorrowMut;
use std::collections::HashSet;
use std::io::Write;
@@ -43,7 +42,7 @@ const NUM_MERGE_THREADS: usize = 4;
/// and flushed.
///
/// This method is not part of tantivy's public API
pub fn save_new_metas(schema: Schema, directory: &dyn Directory) -> crate::Result<()> {
pub fn save_new_metas(schema: Schema, directory: &mut dyn Directory) -> crate::Result<()> {
save_metas(
&IndexMeta {
segments: Vec::new(),
@@ -64,7 +63,7 @@ pub fn save_new_metas(schema: Schema, directory: &dyn Directory) -> crate::Resul
/// and flushed.
///
/// This method is not part of tantivy's public API
fn save_metas(metas: &IndexMeta, directory: &dyn Directory) -> crate::Result<()> {
fn save_metas(metas: &IndexMeta, directory: &mut dyn Directory) -> crate::Result<()> {
info!("save metas");
let mut buffer = serde_json::to_vec_pretty(metas)?;
// Just adding a new line at the end of the buffer.
@@ -112,33 +111,36 @@ fn merge(
target_opstamp: Opstamp,
) -> crate::Result<SegmentEntry> {
// first we need to apply deletes to our segment.
let merged_segment = index.new_segment();
let mut merged_segment = index.new_segment();
// First we apply all of the delet to the merged segment, up to the target opstamp.
for segment_entry in &mut segment_entries {
let segment = index.segment(segment_entry.meta().clone());
advance_deletes(segment, segment_entry, target_opstamp)?;
advance_deletes(segment_entry, target_opstamp)?;
}
let delete_cursor = segment_entries[0].delete_cursor().clone();
let delete_cursor = segment_entries[0].delete_cursor();
let segments: Vec<Segment> = segment_entries
.iter()
.map(|segment_entry| index.segment(segment_entry.meta().clone()))
.map(|segment_entry| segment_entry.segment().clone())
.collect();
// An IndexMerger is like a "view" of our merged segments.
let merger: IndexMerger = IndexMerger::open(index.schema(), &segments[..])?;
// ... we just serialize this index merger in our new segment to merge the two segments.
let segment_serializer = SegmentSerializer::for_segment(merged_segment.clone())?;
let segment_serializer = SegmentSerializer::for_segment(&mut merged_segment)?;
let num_docs = merger.write(segment_serializer)?;
let store_wrt = merged_segment.open_write(SegmentComponent::STORE)?;
merger.write_storable_fields(store_wrt)?;
let merged_segment_id = merged_segment.id();
let max_doc = merger.write(segment_serializer)?;
let segment_meta = index.new_segment_meta(merged_segment_id, num_docs);
Ok(SegmentEntry::new(segment_meta, delete_cursor, None))
Ok(SegmentEntry::new(
merged_segment.with_max_doc(max_doc),
delete_cursor,
None,
))
}
pub(crate) struct InnerSegmentUpdater {
@@ -157,17 +159,15 @@ pub(crate) struct InnerSegmentUpdater {
merge_policy: RwLock<Arc<Box<dyn MergePolicy>>>,
killed: AtomicBool,
stamper: Stamper,
merge_operations: MergeOperationInventory,
}
impl SegmentUpdater {
pub fn create(
segment_registers: Arc<RwLock<SegmentRegisters>>,
index: Index,
stamper: Stamper,
delete_cursor: &DeleteCursor,
) -> crate::Result<SegmentUpdater> {
let segments = index.searchable_segment_metas()?;
let segment_manager = SegmentManager::from_segments(segments, delete_cursor);
let segment_manager = SegmentManager::new(segment_registers);
let pool = ThreadPoolBuilder::new()
.name_prefix("segment_updater")
.pool_size(1)
@@ -196,7 +196,6 @@ impl SegmentUpdater {
merge_policy: RwLock::new(Arc::new(Box::new(DefaultMergePolicy::default()))),
killed: AtomicBool::new(false),
stamper,
merge_operations: Default::default(),
})))
}
@@ -262,8 +261,7 @@ impl SegmentUpdater {
fn purge_deletes(&self, target_opstamp: Opstamp) -> crate::Result<Vec<SegmentEntry>> {
let mut segment_entries = self.segment_manager.segment_entries();
for segment_entry in &mut segment_entries {
let segment = self.index.segment(segment_entry.meta().clone());
advance_deletes(segment, segment_entry, target_opstamp)?;
advance_deletes(segment_entry, target_opstamp)?;
}
Ok(segment_entries)
}
@@ -331,12 +329,21 @@ impl SegmentUpdater {
&self,
opstamp: Opstamp,
payload: Option<String>,
soft_commit: bool,
) -> impl Future<Output = crate::Result<()>> {
let segment_updater: SegmentUpdater = self.clone();
let directory = self.index.directory().clone();
self.schedule_future(async move {
let segment_entries = segment_updater.purge_deletes(opstamp)?;
let mut segment_entries = segment_updater.purge_deletes(opstamp)?;
if !soft_commit {
for segment_entry in &mut segment_entries {
segment_entry.persist(directory.clone())?;
}
}
segment_updater.segment_manager.commit(segment_entries);
segment_updater.save_metas(opstamp, payload)?;
if !soft_commit {
segment_updater.save_metas(opstamp, payload)?;
}
let _ = garbage_collect_files(segment_updater.clone()).await;
segment_updater.consider_merge_options().await;
Ok(())
@@ -353,7 +360,7 @@ impl SegmentUpdater {
pub(crate) fn make_merge_operation(&self, segment_ids: &[SegmentId]) -> MergeOperation {
let commit_opstamp = self.load_metas().opstamp;
MergeOperation::new(&self.merge_operations, commit_opstamp, segment_ids.to_vec())
self.segment_manager.new_merge_operation(commit_opstamp, MergeCandidate(segment_ids.to_vec()))
}
// Starts a merge operation. This function will block until the merge operation is effectively
@@ -427,9 +434,8 @@ impl SegmentUpdater {
}
async fn consider_merge_options(&self) {
let merge_segment_ids: HashSet<SegmentId> = self.merge_operations.segment_in_merge();
let (committed_segments, uncommitted_segments) =
get_mergeable_segments(&merge_segment_ids, &self.segment_manager);
self.segment_manager.segments_not_in_merge();
// Committed segments cannot be merged with uncommitted_segments.
// We therefore consider merges using these two sets of segments independently.
@@ -440,7 +446,7 @@ impl SegmentUpdater {
.compute_merge_candidates(&uncommitted_segments)
.into_iter()
.map(|merge_candidate| {
MergeOperation::new(&self.merge_operations, current_opstamp, merge_candidate.0)
self.segment_manager.new_merge_operation(current_opstamp, merge_candidate)
})
.collect();
@@ -449,9 +455,10 @@ impl SegmentUpdater {
.compute_merge_candidates(&committed_segments)
.into_iter()
.map(|merge_candidate: MergeCandidate| {
MergeOperation::new(&self.merge_operations, commit_opstamp, merge_candidate.0)
});
merge_candidates.extend(committed_merge_candidates);
self.segment_manager.new_merge_operation(commit_opstamp, merge_candidate)
})
.collect::<Vec<_>>();
merge_candidates.extend(committed_merge_candidates.into_iter());
for merge_operation in merge_candidates {
if let Err(err) = self.start_merge(merge_operation) {
@@ -473,17 +480,13 @@ impl SegmentUpdater {
let end_merge_future = self.schedule_future(async move {
info!("End merge {:?}", after_merge_segment_entry.meta());
{
let mut delete_cursor = after_merge_segment_entry.delete_cursor().clone();
let mut delete_cursor = after_merge_segment_entry.delete_cursor();
if let Some(delete_operation) = delete_cursor.get() {
let committed_opstamp = segment_updater.load_metas().opstamp;
if delete_operation.opstamp < committed_opstamp {
let index = &segment_updater.index;
let segment = index.segment(after_merge_segment_entry.meta().clone());
if let Err(e) = advance_deletes(
segment,
&mut after_merge_segment_entry,
committed_opstamp,
) {
if let Err(e) =
advance_deletes(&mut after_merge_segment_entry, committed_opstamp)
{
error!(
"Merge of {:?} was cancelled (advancing deletes failed): {:?}",
merge_operation.segment_ids(),
@@ -521,7 +524,7 @@ impl SegmentUpdater {
///
/// Upon termination of the current merging threads,
/// merge opportunity may appear.
///
//
/// We keep waiting until the merge policy judges that
/// no opportunity is available.
///
@@ -532,9 +535,8 @@ impl SegmentUpdater {
///
/// Obsolete files will eventually be cleaned up
/// by the directory garbage collector.
pub fn wait_merging_thread(&self) -> crate::Result<()> {
self.merge_operations.wait_until_empty();
Ok(())
pub fn wait_merging_thread(&self) {
self.segment_manager.wait_merging_thread()
}
}
@@ -543,7 +545,8 @@ mod tests {
use crate::indexer::merge_policy::tests::MergeWheneverPossible;
use crate::schema::*;
use crate::Index;
use crate::{Index, SegmentId};
use futures::executor::block_on;
#[test]
fn test_delete_during_merge() {
@@ -554,7 +557,7 @@ mod tests {
let index = Index::create_in_ram(schema);
// writing the segment
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.set_merge_policy(Box::new(MergeWheneverPossible));
{
@@ -607,7 +610,7 @@ mod tests {
let index = Index::create_in_ram(schema);
// writing the segment
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
{
for _ in 0..100 {
@@ -678,7 +681,7 @@ mod tests {
let index = Index::create_in_ram(schema);
// writing the segment
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
{
for _ in 0..100 {
@@ -694,4 +697,27 @@ mod tests {
.segment_entries();
assert!(seg_vec.is_empty());
}
#[test]
fn test_merge_over_soft_commit() {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
// writing the segment
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(text_field=>"a"));
assert!(index_writer.soft_commit().is_ok());
index_writer.add_document(doc!(text_field=>"a"));
assert!(index_writer.soft_commit().is_ok());
let reader = index_writer.reader(1).unwrap();
let segment_ids: Vec<SegmentId> = reader
.searcher()
.segment_readers()
.iter()
.map(|reader| reader.segment_id())
.collect();
assert!(block_on(index_writer.merge(&segment_ids)).is_ok());
}
}

213
src/indexer/segment_writer.rs
View File

@@ -1,9 +1,11 @@
use super::operation::AddOperation;
use crate::core::Segment;
use crate::core::SerializableSegment;
use crate::directory::{SpillingResult, SpillingWriter, TerminatingWrite};
use crate::fastfield::FastFieldsWriter;
use crate::fieldnorm::{FieldNormReaders, FieldNormsWriter};
use crate::fieldnorm::FieldNormsWriter;
use crate::indexer::segment_serializer::SegmentSerializer;
use crate::indexer::IndexWriterConfig;
use crate::postings::compute_table_size;
use crate::postings::MultiFieldPostingsWriter;
use crate::schema::FieldType;
@@ -11,19 +13,24 @@ use crate::schema::Schema;
use crate::schema::Term;
use crate::schema::Value;
use crate::schema::{Field, FieldEntry};
use crate::store::StoreWriter;
use crate::tokenizer::{BoxTokenStream, PreTokenizedStream};
use crate::tokenizer::{FacetTokenizer, TextAnalyzer};
use crate::tokenizer::{FacetTokenizer, TextAnalyzer, TokenizerManager};
use crate::tokenizer::{TokenStreamChain, Tokenizer};
use crate::Opstamp;
use crate::{DocId, SegmentComponent};
use std::io;
use std::io::Write;
use std::str;
use crate::indexer::resource_manager::ResourceManager;
/// Computes the initial size of the hash table.
///
/// Returns a number of bit `b`, such that the recommended initial table size is 2^b.
fn initial_table_size(per_thread_memory_budget: usize) -> crate::Result<usize> {
let table_memory_upper_bound = per_thread_memory_budget / 3;
fn initial_table_size(per_thread_memory_budget: u64) -> crate::Result<usize> {
let table_memory_upper_bound = per_thread_memory_budget / 3u64;
if let Some(limit) = (10..)
.take_while(|num_bits: &usize| compute_table_size(*num_bits) < table_memory_upper_bound)
.take_while(|num_bits| compute_table_size(*num_bits) < table_memory_upper_bound)
.last()
{
Ok(limit.min(19)) // we cap it at 2^19 = 512K.
@@ -41,12 +48,13 @@ fn initial_table_size(per_thread_memory_budget: usize) -> crate::Result<usize> {
pub struct SegmentWriter {
max_doc: DocId,
multifield_postings: MultiFieldPostingsWriter,
segment_serializer: SegmentSerializer,
segment: Segment,
fast_field_writers: FastFieldsWriter,
fieldnorms_writer: FieldNormsWriter,
doc_opstamps: Vec<Opstamp>,
tokenizers: Vec<Option<TextAnalyzer>>,
term_buffer: Term,
store_writer: StoreWriter<SpillingWriter>,
memory_manager: ResourceManager,
}
impl SegmentWriter {
@@ -60,13 +68,13 @@ impl SegmentWriter {
/// - segment: The segment being written
/// - schema
pub fn for_segment(
memory_budget: usize,
config: &IndexWriterConfig,
segment: Segment,
schema: &Schema,
tokenizer_manager: &TokenizerManager,
memory_manager: ResourceManager
) -> crate::Result<SegmentWriter> {
let tokenizer_manager = segment.index().tokenizers().clone();
let table_num_bits = initial_table_size(memory_budget)?;
let segment_serializer = SegmentSerializer::for_segment(segment)?;
let table_num_bits = initial_table_size(config.heap_size_in_byte_per_thread())?;
let multifield_postings = MultiFieldPostingsWriter::new(schema, table_num_bits);
let tokenizers = schema
.fields()
@@ -82,15 +90,26 @@ impl SegmentWriter {
},
)
.collect();
let mut segment_clone = segment.clone();
let spilling_wrt = SpillingWriter::new(
50_000_000,
Box::new(move || {
segment_clone
.open_write(SegmentComponent::STORE)
.map_err(|err| io::Error::new(io::ErrorKind::Other, err))
}),
);
let store_writer = StoreWriter::new(spilling_wrt);
Ok(SegmentWriter {
max_doc: 0,
multifield_postings,
fieldnorms_writer: FieldNormsWriter::for_schema(schema),
segment_serializer,
segment,
fast_field_writers: FastFieldsWriter::from_schema(schema),
doc_opstamps: Vec::with_capacity(1_000),
tokenizers,
term_buffer: Term::new(),
store_writer,
memory_manager
})
}
@@ -98,29 +117,40 @@ impl SegmentWriter {
///
/// Finalize consumes the `SegmentWriter`, so that it cannot
/// be used afterwards.
pub fn finalize(mut self) -> crate::Result<Vec<u64>> {
pub fn finalize(mut self) -> crate::Result<(Segment, Vec<u64>)> {
self.fieldnorms_writer.fill_up_to_max_doc(self.max_doc);
let spilling_wrt = self.store_writer.close()?;
let mut segment: Segment;
match spilling_wrt.finalize()? {
SpillingResult::Spilled => {
segment = self.segment.clone();
}
SpillingResult::Buffer(buf) => {
segment = self.segment.into_volatile(self.memory_manager.clone());
let mut store_wrt = segment.open_write(SegmentComponent::STORE)?;
store_wrt.write_all(&buf[..])?;
store_wrt.terminate()?;
}
}
let segment_serializer = SegmentSerializer::for_segment(&mut segment)?;
segment = segment.with_max_doc(self.max_doc);
write(
&self.multifield_postings,
&self.fast_field_writers,
&self.fieldnorms_writer,
self.segment_serializer,
segment_serializer,
)?;
Ok(self.doc_opstamps)
Ok((segment, self.doc_opstamps))
}
pub fn mem_usage(&self) -> usize {
pub fn mem_usage(&self) -> u64 {
self.multifield_postings.mem_usage()
}
/// Indexes a new document
///
/// As a user, you should rather use `IndexWriter`'s add_document.
pub fn add_document(
&mut self,
add_operation: AddOperation,
schema: &Schema,
) -> crate::Result<()> {
pub fn add_document(&mut self, add_operation: AddOperation, schema: &Schema) -> io::Result<()> {
let doc_id = self.max_doc;
let mut doc = add_operation.document;
self.doc_opstamps.push(add_operation.opstamp);
@@ -128,45 +158,34 @@ impl SegmentWriter {
self.fast_field_writers.add_document(&doc);
for (field, field_values) in doc.get_sorted_field_values() {
let field_entry = schema.get_field_entry(field);
let make_schema_error = || {
crate::TantivyError::SchemaError(format!(
"Expected a {:?} for field {:?}",
field_entry.field_type().value_type(),
field_entry.name()
))
};
if !field_entry.is_indexed() {
let field_options = schema.get_field_entry(field);
if !field_options.is_indexed() {
continue;
}
let (term_buffer, multifield_postings) =
(&mut self.term_buffer, &mut self.multifield_postings);
match *field_entry.field_type() {
match *field_options.field_type() {
FieldType::HierarchicalFacet => {
term_buffer.set_field(field);
let facets =
field_values
.iter()
.flat_map(|field_value| match *field_value.value() {
Value::Facet(ref facet) => Some(facet.encoded_str()),
_ => {
panic!("Expected hierarchical facet");
}
});
for facet_str in facets {
let facets: Vec<&str> = field_values
.iter()
.flat_map(|field_value| match *field_value.value() {
Value::Facet(ref facet) => Some(facet.encoded_str()),
_ => {
panic!("Expected hierarchical facet");
}
})
.collect();
let mut term = Term::for_field(field); // we set the Term
for fake_str in facets {
let mut unordered_term_id_opt = None;
FacetTokenizer
.token_stream(facet_str)
.process(&mut |token| {
term_buffer.set_text(&token.text);
let unordered_term_id =
multifield_postings.subscribe(doc_id, &term_buffer);
unordered_term_id_opt = Some(unordered_term_id);
});
FacetTokenizer.token_stream(fake_str).process(&mut |token| {
term.set_text(&token.text);
let unordered_term_id =
self.multifield_postings.subscribe(doc_id, &term);
unordered_term_id_opt = Some(unordered_term_id);
});
if let Some(unordered_term_id) = unordered_term_id_opt {
self.fast_field_writers
.get_multivalue_writer(field)
.expect("writer for facet missing")
.expect("multified writer for facet missing")
.add_val(unordered_term_id);
}
}
@@ -183,6 +202,7 @@ impl SegmentWriter {
if let Some(last_token) = tok_str.tokens.last() {
total_offset += last_token.offset_to;
}
token_streams
.push(PreTokenizedStream::from(tok_str.clone()).into());
}
@@ -192,6 +212,7 @@ impl SegmentWriter {
{
offsets.push(total_offset);
total_offset += text.len();
token_streams.push(tokenizer.token_stream(text));
}
}
@@ -203,12 +224,8 @@ impl SegmentWriter {
0
} else {
let mut token_stream = TokenStreamChain::new(offsets, token_streams);
multifield_postings.index_text(
doc_id,
field,
&mut token_stream,
term_buffer,
)
self.multifield_postings
.index_text(doc_id, field, &mut token_stream)
};
self.fieldnorms_writer.record(doc_id, field, num_tokens);
@@ -216,74 +233,55 @@ impl SegmentWriter {
FieldType::U64(ref int_option) => {
if int_option.is_indexed() {
for field_value in field_values {
term_buffer.set_field(field_value.field());
let u64_val = field_value
.value()
.u64_value()
.ok_or_else(make_schema_error)?;
term_buffer.set_u64(u64_val);
multifield_postings.subscribe(doc_id, &term_buffer);
let term = Term::from_field_u64(
field_value.field(),
field_value.value().u64_value(),
);
self.multifield_postings.subscribe(doc_id, &term);
}
}
}
FieldType::Date(ref int_option) => {
if int_option.is_indexed() {
for field_value in field_values {
term_buffer.set_field(field_value.field());
let date_val = field_value
.value()
.date_value()
.ok_or_else(make_schema_error)?;
term_buffer.set_i64(date_val.timestamp());
multifield_postings.subscribe(doc_id, &term_buffer);
let term = Term::from_field_i64(
field_value.field(),
field_value.value().date_value().timestamp(),
);
self.multifield_postings.subscribe(doc_id, &term);
}
}
}
FieldType::I64(ref int_option) => {
if int_option.is_indexed() {
for field_value in field_values {
term_buffer.set_field(field_value.field());
let i64_val = field_value
.value()
.i64_value()
.ok_or_else(make_schema_error)?;
term_buffer.set_i64(i64_val);
multifield_postings.subscribe(doc_id, &term_buffer);
let term = Term::from_field_i64(
field_value.field(),
field_value.value().i64_value(),
);
self.multifield_postings.subscribe(doc_id, &term);
}
}
}
FieldType::F64(ref int_option) => {
if int_option.is_indexed() {
for field_value in field_values {
term_buffer.set_field(field_value.field());
let f64_val = field_value
.value()
.f64_value()
.ok_or_else(make_schema_error)?;
term_buffer.set_f64(f64_val);
multifield_postings.subscribe(doc_id, &term_buffer);
let term = Term::from_field_f64(
field_value.field(),
field_value.value().f64_value(),
);
self.multifield_postings.subscribe(doc_id, &term);
}
}
}
FieldType::Bytes(ref option) => {
if option.is_indexed() {
for field_value in field_values {
term_buffer.set_field(field_value.field());
let bytes = field_value
.value()
.bytes_value()
.ok_or_else(make_schema_error)?;
term_buffer.set_bytes(bytes);
self.multifield_postings.subscribe(doc_id, &term_buffer);
}
}
FieldType::Bytes => {
// Do nothing. Bytes only supports fast fields.
}
}
}
doc.filter_fields(|field| schema.get_field_entry(field).is_stored());
doc.prepare_for_store();
let doc_writer = self.segment_serializer.get_store_writer();
doc_writer.store(&doc)?;
self.store_writer.store(&doc)?;
self.max_doc += 1;
Ok(())
}
@@ -316,16 +314,9 @@ fn write(
fieldnorms_writer: &FieldNormsWriter,
mut serializer: SegmentSerializer,
) -> crate::Result<()> {
if let Some(fieldnorms_serializer) = serializer.extract_fieldnorms_serializer() {
fieldnorms_writer.serialize(fieldnorms_serializer)?;
}
let fieldnorm_data = serializer
.segment()
.open_read(SegmentComponent::FIELDNORMS)?;
let fieldnorm_readers = FieldNormReaders::open(fieldnorm_data)?;
let term_ord_map =
multifield_postings.serialize(serializer.get_postings_serializer(), fieldnorm_readers)?;
let term_ord_map = multifield_postings.serialize(serializer.get_postings_serializer())?;
fast_field_writers.serialize(serializer.get_fast_field_serializer(), &term_ord_map)?;
fieldnorms_writer.serialize(serializer.get_fieldnorms_serializer())?;
serializer.close()?;
Ok(())
}

531
src/lib.rs
View File

@@ -105,7 +105,7 @@ extern crate serde_json;
extern crate log;
#[macro_use]
extern crate thiserror;
extern crate failure;
#[cfg(all(test, feature = "unstable"))]
extern crate test;
@@ -134,7 +134,7 @@ mod core;
mod indexer;
#[allow(unused_doc_comments)]
pub mod error;
mod error;
pub mod tokenizer;
pub mod collector;
@@ -156,15 +156,14 @@ mod snippet;
pub use self::snippet::{Snippet, SnippetGenerator};
mod docset;
pub use self::docset::{DocSet, TERMINATED};
pub use crate::common::HasLen;
pub use self::docset::{DocSet, SkipResult};
pub use crate::common::{f64_to_u64, i64_to_u64, u64_to_f64, u64_to_i64};
pub use crate::core::{Executor, SegmentComponent};
pub use crate::core::{Index, IndexMeta, Searcher, Segment, SegmentId, SegmentMeta};
pub use crate::core::{InvertedIndexReader, SegmentReader};
pub use crate::directory::Directory;
pub use crate::indexer::operation::UserOperation;
pub use crate::indexer::IndexWriter;
pub use crate::indexer::{IndexWriter, IndexWriterConfig};
pub use crate::postings::Postings;
pub use crate::reader::LeasedItem;
pub use crate::schema::{Document, Term};
@@ -174,7 +173,7 @@ use once_cell::sync::Lazy;
use serde::{Deserialize, Serialize};
/// Index format version.
const INDEX_FORMAT_VERSION: u32 = 2;
const INDEX_FORMAT_VERSION: u32 = 1;
/// Structure version for the index.
#[derive(Clone, PartialEq, Eq, Serialize, Deserialize)]
@@ -246,10 +245,11 @@ pub type DocId = u32;
/// with opstamp `n+1`.
pub type Opstamp = u64;
/// A Score that represents the relevance of the document to the query
/// A f32 that represents the relevance of the document to the query
///
/// This is modelled internally as a `f32`. The larger the number, the more relevant
/// the document to the search query.
/// This is modelled internally as a `f32`. The
/// larger the number, the more relevant the document
/// to the search
pub type Score = f32;
/// A `SegmentLocalId` identifies a segment.
@@ -277,18 +277,20 @@ impl DocAddress {
///
/// The id used for the segment is actually an ordinal
/// in the list of `Segment`s held by a `Searcher`.
#[derive(Debug, Clone, Copy, PartialEq, Eq, PartialOrd, Ord, Hash)]
#[derive(Debug, Clone, Copy, PartialEq, Eq, PartialOrd, Ord)]
pub struct DocAddress(pub SegmentLocalId, pub DocId);
#[cfg(test)]
mod tests {
use crate::collector::tests::TEST_COLLECTOR_WITH_SCORE;
use crate::core::SegmentReader;
use crate::docset::{DocSet, TERMINATED};
use crate::docset::DocSet;
use crate::query::BooleanQuery;
use crate::schema::*;
use crate::DocAddress;
use crate::Index;
use crate::IndexWriter;
use crate::Postings;
use crate::ReloadPolicy;
use rand::distributions::Bernoulli;
@@ -296,26 +298,17 @@ mod tests {
use rand::rngs::StdRng;
use rand::{Rng, SeedableRng};
/// Checks if left and right are close one to each other.
/// Panics if the two values are more than 0.5% apart.
#[macro_export]
macro_rules! assert_nearly_equals {
($left:expr, $right:expr) => {{
match (&$left, &$right) {
(left_val, right_val) => {
let diff = (left_val - right_val).abs();
let add = left_val.abs() + right_val.abs();
if diff > 0.0005 * add {
panic!(
r#"assertion failed: `(left ~= right)`
left: `{:?}`,
right: `{:?}`"#,
&*left_val, &*right_val
)
}
}
}
}};
pub fn assert_nearly_equals(expected: f32, val: f32) {
assert!(
nearly_equals(val, expected),
"Got {}, expected {}.",
val,
expected
);
}
pub fn nearly_equals(a: f32, b: f32) -> bool {
(a - b).abs() < 0.0005 * (a + b).abs()
}
pub fn generate_nonunique_unsorted(max_value: u32, n_elems: usize) -> Vec<u32> {
@@ -353,14 +346,14 @@ mod tests {
#[test]
#[cfg(feature = "mmap")]
fn test_indexing() -> crate::Result<()> {
fn test_indexing() {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_from_tempdir(schema).unwrap();
{
// writing the segment
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
{
let doc = doc!(text_field=>"af b");
index_writer.add_document(doc);
@@ -375,91 +368,120 @@ mod tests {
}
assert!(index_writer.commit().is_ok());
}
Ok(())
}
#[test]
fn test_docfreq1() -> crate::Result<()> {
fn test_docfreq1() {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"a b c"));
index_writer.commit()?;
index_writer.add_document(doc!(text_field=>"a"));
index_writer.add_document(doc!(text_field=>"a a"));
index_writer.commit()?;
index_writer.add_document(doc!(text_field=>"c"));
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let term_a = Term::from_field_text(text_field, "a");
assert_eq!(searcher.doc_freq(&term_a)?, 3);
let term_b = Term::from_field_text(text_field, "b");
assert_eq!(searcher.doc_freq(&term_b)?, 1);
let term_c = Term::from_field_text(text_field, "c");
assert_eq!(searcher.doc_freq(&term_c)?, 2);
let term_d = Term::from_field_text(text_field, "d");
assert_eq!(searcher.doc_freq(&term_d)?, 0);
Ok(())
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
{
index_writer.add_document(doc!(text_field=>"a b c"));
index_writer.commit().unwrap();
}
{
{
let doc = doc!(text_field=>"a");
index_writer.add_document(doc);
}
{
let doc = doc!(text_field=>"a a");
index_writer.add_document(doc);
}
index_writer.commit().unwrap();
}
{
let doc = doc!(text_field=>"c");
index_writer.add_document(doc);
index_writer.commit().unwrap();
}
{
let reader = index.reader().unwrap();
let searcher = reader.searcher();
let term_a = Term::from_field_text(text_field, "a");
assert_eq!(searcher.doc_freq(&term_a), 3);
let term_b = Term::from_field_text(text_field, "b");
assert_eq!(searcher.doc_freq(&term_b), 1);
let term_c = Term::from_field_text(text_field, "c");
assert_eq!(searcher.doc_freq(&term_c), 2);
let term_d = Term::from_field_text(text_field, "d");
assert_eq!(searcher.doc_freq(&term_d), 0);
}
}
#[test]
fn test_fieldnorm_no_docs_with_field() -> crate::Result<()> {
fn test_fieldnorm_no_docs_with_field() {
let mut schema_builder = Schema::builder();
let title_field = schema_builder.add_text_field("title", TEXT);
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"a b c"));
index_writer.commit()?;
let index_reader = index.reader()?;
let searcher = index_reader.searcher();
let reader = searcher.segment_reader(0);
{
let fieldnorm_reader = reader.get_fieldnorms_reader(text_field)?;
assert_eq!(fieldnorm_reader.fieldnorm(0), 3);
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
{
let doc = doc!(text_field=>"a b c");
index_writer.add_document(doc);
}
index_writer.commit().unwrap();
}
{
let fieldnorm_reader = reader.get_fieldnorms_reader(title_field)?;
assert_eq!(fieldnorm_reader.fieldnorm_id(0), 0);
let index_reader = index.reader().unwrap();
let searcher = index_reader.searcher();
let reader = searcher.segment_reader(0);
{
let fieldnorm_reader = reader.get_fieldnorms_reader(text_field);
assert_eq!(fieldnorm_reader.fieldnorm(0), 3);
}
{
let fieldnorm_reader = reader.get_fieldnorms_reader(title_field);
assert_eq!(fieldnorm_reader.fieldnorm_id(0), 0);
}
}
Ok(())
}
#[test]
fn test_fieldnorm() -> crate::Result<()> {
fn test_fieldnorm() {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"a b c"));
index_writer.add_document(doc!());
index_writer.add_document(doc!(text_field=>"a b"));
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let segment_reader: &SegmentReader = searcher.segment_reader(0);
let fieldnorms_reader = segment_reader.get_fieldnorms_reader(text_field)?;
assert_eq!(fieldnorms_reader.fieldnorm(0), 3);
assert_eq!(fieldnorms_reader.fieldnorm(1), 0);
assert_eq!(fieldnorms_reader.fieldnorm(2), 2);
Ok(())
{
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
{
let doc = doc!(text_field=>"a b c");
index_writer.add_document(doc);
}
{
let doc = doc!();
index_writer.add_document(doc);
}
{
let doc = doc!(text_field=>"a b");
index_writer.add_document(doc);
}
index_writer.commit().unwrap();
}
{
let reader = index.reader().unwrap();
let searcher = reader.searcher();
let segment_reader: &SegmentReader = searcher.segment_reader(0);
let fieldnorms_reader = segment_reader.get_fieldnorms_reader(text_field);
assert_eq!(fieldnorms_reader.fieldnorm(0), 3);
assert_eq!(fieldnorms_reader.fieldnorm(1), 0);
assert_eq!(fieldnorms_reader.fieldnorm(2), 2);
}
}
fn advance_undeleted(docset: &mut dyn DocSet, reader: &SegmentReader) -> bool {
let mut doc = docset.advance();
while doc != TERMINATED {
if !reader.is_deleted(doc) {
while docset.advance() {
if !reader.is_deleted(docset.doc()) {
return true;
}
doc = docset.advance();
}
false
}
#[test]
fn test_delete_postings1() -> crate::Result<()> {
fn test_delete_postings1() {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let term_abcd = Term::from_field_text(text_field, "abcd");
@@ -475,7 +497,7 @@ mod tests {
.unwrap();
{
// writing the segment
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
// 0
index_writer.add_document(doc!(text_field=>"a b"));
// 1
@@ -491,19 +513,19 @@ mod tests {
index_writer.add_document(doc!(text_field=>" b c"));
// 5
index_writer.add_document(doc!(text_field=>" a"));
index_writer.commit()?;
index_writer.commit().unwrap();
}
{
reader.reload()?;
reader.reload().unwrap();
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(0);
let inverted_index = segment_reader.inverted_index(text_field)?;
let inverted_index = segment_reader.inverted_index(text_field);
assert!(inverted_index
.read_postings(&term_abcd, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term_abcd, IndexRecordOption::WithFreqsAndPositions)
.is_none());
{
let mut postings = inverted_index
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert!(advance_undeleted(&mut postings, segment_reader));
assert_eq!(postings.doc(), 5);
@@ -511,7 +533,7 @@ mod tests {
}
{
let mut postings = inverted_index
.read_postings(&term_b, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term_b, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert!(advance_undeleted(&mut postings, segment_reader));
assert_eq!(postings.doc(), 3);
@@ -522,25 +544,25 @@ mod tests {
}
{
// writing the segment
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
// 0
index_writer.add_document(doc!(text_field=>"a b"));
// 1
index_writer.delete_term(Term::from_field_text(text_field, "c"));
index_writer.rollback()?;
index_writer.rollback().unwrap();
}
{
reader.reload()?;
reader.reload().unwrap();
let searcher = reader.searcher();
let seg_reader = searcher.segment_reader(0);
let inverted_index = seg_reader.inverted_index(term_abcd.field())?;
let inverted_index = seg_reader.inverted_index(term_abcd.field());
assert!(inverted_index
.read_postings(&term_abcd, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term_abcd, IndexRecordOption::WithFreqsAndPositions)
.is_none());
{
let mut postings = inverted_index
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert!(advance_undeleted(&mut postings, seg_reader));
assert_eq!(postings.doc(), 5);
@@ -548,7 +570,7 @@ mod tests {
}
{
let mut postings = inverted_index
.read_postings(&term_b, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term_b, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert!(advance_undeleted(&mut postings, seg_reader));
assert_eq!(postings.doc(), 3);
@@ -559,30 +581,30 @@ mod tests {
}
{
// writing the segment
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(text_field=>"a b"));
index_writer.delete_term(Term::from_field_text(text_field, "c"));
index_writer.rollback()?;
index_writer.rollback().unwrap();
index_writer.delete_term(Term::from_field_text(text_field, "a"));
index_writer.commit()?;
index_writer.commit().unwrap();
}
{
reader.reload()?;
reader.reload().unwrap();
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(0);
let inverted_index = segment_reader.inverted_index(term_abcd.field())?;
let inverted_index = segment_reader.inverted_index(term_abcd.field());
assert!(inverted_index
.read_postings(&term_abcd, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term_abcd, IndexRecordOption::WithFreqsAndPositions)
.is_none());
{
let mut postings = inverted_index
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert!(!advance_undeleted(&mut postings, segment_reader));
}
{
let mut postings = inverted_index
.read_postings(&term_b, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term_b, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert!(advance_undeleted(&mut postings, segment_reader));
assert_eq!(postings.doc(), 3);
@@ -592,107 +614,104 @@ mod tests {
}
{
let mut postings = inverted_index
.read_postings(&term_c, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term_c, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert!(advance_undeleted(&mut postings, segment_reader));
assert_eq!(postings.doc(), 4);
assert!(!advance_undeleted(&mut postings, segment_reader));
}
}
Ok(())
}
#[test]
fn test_indexed_u64() -> crate::Result<()> {
fn test_indexed_u64() {
let mut schema_builder = Schema::builder();
let field = schema_builder.add_u64_field("value", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(field=>1u64));
index_writer.commit()?;
let reader = index.reader()?;
index_writer.commit().unwrap();
let reader = index.reader().unwrap();
let searcher = reader.searcher();
let term = Term::from_field_u64(field, 1u64);
let mut postings = searcher
.segment_reader(0)
.inverted_index(term.field())?
.read_postings(&term, IndexRecordOption::Basic)?
.inverted_index(term.field())
.read_postings(&term, IndexRecordOption::Basic)
.unwrap();
assert!(postings.advance());
assert_eq!(postings.doc(), 0);
assert_eq!(postings.advance(), TERMINATED);
Ok(())
assert!(!postings.advance());
}
#[test]
fn test_indexed_i64() -> crate::Result<()> {
fn test_indexed_i64() {
let mut schema_builder = Schema::builder();
let value_field = schema_builder.add_i64_field("value", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
let negative_val = -1i64;
index_writer.add_document(doc!(value_field => negative_val));
index_writer.commit()?;
let reader = index.reader()?;
index_writer.commit().unwrap();
let reader = index.reader().unwrap();
let searcher = reader.searcher();
let term = Term::from_field_i64(value_field, negative_val);
let mut postings = searcher
.segment_reader(0)
.inverted_index(term.field())?
.read_postings(&term, IndexRecordOption::Basic)?
.inverted_index(term.field())
.read_postings(&term, IndexRecordOption::Basic)
.unwrap();
assert!(postings.advance());
assert_eq!(postings.doc(), 0);
assert_eq!(postings.advance(), TERMINATED);
Ok(())
assert!(!postings.advance());
}
#[test]
fn test_indexed_f64() -> crate::Result<()> {
fn test_indexed_f64() {
let mut schema_builder = Schema::builder();
let value_field = schema_builder.add_f64_field("value", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
let val = std::f64::consts::PI;
index_writer.add_document(doc!(value_field => val));
index_writer.commit()?;
let reader = index.reader()?;
index_writer.commit().unwrap();
let reader = index.reader().unwrap();
let searcher = reader.searcher();
let term = Term::from_field_f64(value_field, val);
let mut postings = searcher
.segment_reader(0)
.inverted_index(term.field())?
.read_postings(&term, IndexRecordOption::Basic)?
.inverted_index(term.field())
.read_postings(&term, IndexRecordOption::Basic)
.unwrap();
assert!(postings.advance());
assert_eq!(postings.doc(), 0);
assert_eq!(postings.advance(), TERMINATED);
Ok(())
assert!(!postings.advance());
}
#[test]
fn test_indexedfield_not_in_documents() -> crate::Result<()> {
fn test_indexedfield_not_in_documents() {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let absent_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(2, 6_000_000).unwrap();
index_writer.add_document(doc!(text_field=>"a"));
assert!(index_writer.commit().is_ok());
let reader = index.reader()?;
let reader = index.reader().unwrap();
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(0);
let inverted_index = segment_reader.inverted_index(absent_field)?;
assert_eq!(inverted_index.terms().num_terms(), 0);
Ok(())
segment_reader.inverted_index(absent_field); //< should not panic
}
#[test]
fn test_delete_postings2() -> crate::Result<()> {
fn test_delete_postings2() {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
@@ -700,112 +719,128 @@ mod tests {
let reader = index
.reader_builder()
.reload_policy(ReloadPolicy::Manual)
.try_into()?;
.try_into()
.unwrap();
// writing the segment
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"63"));
index_writer.add_document(doc!(text_field=>"70"));
index_writer.add_document(doc!(text_field=>"34"));
index_writer.add_document(doc!(text_field=>"1"));
index_writer.add_document(doc!(text_field=>"38"));
index_writer.add_document(doc!(text_field=>"33"));
index_writer.add_document(doc!(text_field=>"40"));
index_writer.add_document(doc!(text_field=>"17"));
index_writer.delete_term(Term::from_field_text(text_field, "38"));
index_writer.delete_term(Term::from_field_text(text_field, "34"));
index_writer.commit()?;
reader.reload()?;
assert_eq!(reader.searcher().num_docs(), 6);
Ok(())
let mut index_writer = index.writer_with_num_threads(2, 6_000_000).unwrap();
let add_document = |index_writer: &mut IndexWriter, val: &'static str| {
let doc = doc!(text_field=>val);
index_writer.add_document(doc);
};
let remove_document = |index_writer: &mut IndexWriter, val: &'static str| {
let delterm = Term::from_field_text(text_field, val);
index_writer.delete_term(delterm);
};
add_document(&mut index_writer, "63");
add_document(&mut index_writer, "70");
add_document(&mut index_writer, "34");
add_document(&mut index_writer, "1");
add_document(&mut index_writer, "38");
add_document(&mut index_writer, "33");
add_document(&mut index_writer, "40");
add_document(&mut index_writer, "17");
remove_document(&mut index_writer, "38");
remove_document(&mut index_writer, "34");
index_writer.commit().unwrap();
reader.reload().unwrap();
let searcher = reader.searcher();
assert_eq!(searcher.num_docs(), 6);
}
#[test]
fn test_termfreq() -> crate::Result<()> {
fn test_termfreq() {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
// writing the segment
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"af af af bc bc"));
index_writer.commit()?;
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
{
let doc = doc!(text_field=>"af af af bc bc");
index_writer.add_document(doc);
}
index_writer.commit().unwrap();
}
{
let index_reader = index.reader()?;
let index_reader = index.reader().unwrap();
let searcher = index_reader.searcher();
let reader = searcher.segment_reader(0);
let inverted_index = reader.inverted_index(text_field)?;
let inverted_index = reader.inverted_index(text_field);
let term_abcd = Term::from_field_text(text_field, "abcd");
assert!(inverted_index
.read_postings(&term_abcd, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term_abcd, IndexRecordOption::WithFreqsAndPositions)
.is_none());
let term_af = Term::from_field_text(text_field, "af");
let mut postings = inverted_index
.read_postings(&term_af, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term_af, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert!(postings.advance());
assert_eq!(postings.doc(), 0);
assert_eq!(postings.term_freq(), 3);
assert_eq!(postings.advance(), TERMINATED);
assert!(!postings.advance());
}
Ok(())
}
#[test]
fn test_searcher_1() -> crate::Result<()> {
fn test_searcher_1() {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let reader = index.reader()?;
// writing the segment
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"af af af b"));
index_writer.add_document(doc!(text_field=>"a b c"));
index_writer.add_document(doc!(text_field=>"a b c d"));
index_writer.commit()?;
reader.reload()?;
let searcher = reader.searcher();
let get_doc_ids = |terms: Vec<Term>| {
let query = BooleanQuery::new_multiterms_query(terms);
searcher
.search(&query, &TEST_COLLECTOR_WITH_SCORE)
.map(|topdocs| topdocs.docs().to_vec())
};
assert_eq!(
get_doc_ids(vec![Term::from_field_text(text_field, "a")])?,
vec![DocAddress(0, 1), DocAddress(0, 2)]
);
assert_eq!(
get_doc_ids(vec![Term::from_field_text(text_field, "af")])?,
vec![DocAddress(0, 0)]
);
assert_eq!(
get_doc_ids(vec![Term::from_field_text(text_field, "b")])?,
vec![DocAddress(0, 0), DocAddress(0, 1), DocAddress(0, 2)]
);
assert_eq!(
get_doc_ids(vec![Term::from_field_text(text_field, "c")])?,
vec![DocAddress(0, 1), DocAddress(0, 2)]
);
assert_eq!(
get_doc_ids(vec![Term::from_field_text(text_field, "d")])?,
vec![DocAddress(0, 2)]
);
assert_eq!(
get_doc_ids(vec![
Term::from_field_text(text_field, "b"),
Term::from_field_text(text_field, "a"),
])?,
vec![DocAddress(0, 0), DocAddress(0, 1), DocAddress(0, 2)]
);
Ok(())
let reader = index.reader().unwrap();
{
// writing the segment
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(text_field=>"af af af b"));
index_writer.add_document(doc!(text_field=>"a b c"));
index_writer.add_document(doc!(text_field=>"a b c d"));
index_writer.commit().unwrap();
}
{
reader.reload().unwrap();
let searcher = reader.searcher();
let get_doc_ids = |terms: Vec<Term>| {
let query = BooleanQuery::new_multiterms_query(terms);
let topdocs = searcher.search(&query, &TEST_COLLECTOR_WITH_SCORE).unwrap();
topdocs.docs().to_vec()
};
assert_eq!(
get_doc_ids(vec![Term::from_field_text(text_field, "a")]),
vec![DocAddress(0, 1), DocAddress(0, 2)]
);
assert_eq!(
get_doc_ids(vec![Term::from_field_text(text_field, "af")]),
vec![DocAddress(0, 0)]
);
assert_eq!(
get_doc_ids(vec![Term::from_field_text(text_field, "b")]),
vec![DocAddress(0, 0), DocAddress(0, 1), DocAddress(0, 2)]
);
assert_eq!(
get_doc_ids(vec![Term::from_field_text(text_field, "c")]),
vec![DocAddress(0, 1), DocAddress(0, 2)]
);
assert_eq!(
get_doc_ids(vec![Term::from_field_text(text_field, "d")]),
vec![DocAddress(0, 2)]
);
assert_eq!(
get_doc_ids(vec![
Term::from_field_text(text_field, "b"),
Term::from_field_text(text_field, "a"),
]),
vec![DocAddress(0, 0), DocAddress(0, 1), DocAddress(0, 2)]
);
}
}
#[test]
fn test_searcher_2() -> crate::Result<()> {
fn test_searcher_2() {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
@@ -813,17 +848,19 @@ mod tests {
let reader = index
.reader_builder()
.reload_policy(ReloadPolicy::Manual)
.try_into()?;
.try_into()
.unwrap();
assert_eq!(reader.searcher().num_docs(), 0u64);
// writing the segment
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"af b"));
index_writer.add_document(doc!(text_field=>"a b c"));
index_writer.add_document(doc!(text_field=>"a b c d"));
index_writer.commit()?;
reader.reload()?;
{
// writing the segment
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(text_field=>"af b"));
index_writer.add_document(doc!(text_field=>"a b c"));
index_writer.add_document(doc!(text_field=>"a b c d"));
index_writer.commit().unwrap();
}
reader.reload().unwrap();
assert_eq!(reader.searcher().num_docs(), 3u64);
Ok(())
}
#[test]
@@ -835,17 +872,17 @@ mod tests {
text_field => "some other value",
other_text_field => "short");
assert_eq!(document.len(), 3);
let values: Vec<&Value> = document.get_all(text_field).collect();
let values = document.get_all(text_field);
assert_eq!(values.len(), 2);
assert_eq!(values[0].text(), Some("tantivy"));
assert_eq!(values[1].text(), Some("some other value"));
let values: Vec<&Value> = document.get_all(other_text_field).collect();
let values = document.get_all(other_text_field);
assert_eq!(values.len(), 1);
assert_eq!(values[0].text(), Some("short"));
}
#[test]
fn test_wrong_fast_field_type() -> crate::Result<()> {
fn test_wrong_fast_field_type() {
let mut schema_builder = Schema::builder();
let fast_field_unsigned = schema_builder.add_u64_field("unsigned", FAST);
let fast_field_signed = schema_builder.add_i64_field("signed", FAST);
@@ -855,14 +892,14 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(1, 50_000_000).unwrap();
{
let document =
doc!(fast_field_unsigned => 4u64, fast_field_signed=>4i64, fast_field_float=>4f64);
index_writer.add_document(document);
index_writer.commit()?;
index_writer.commit().unwrap();
}
let reader = index.reader()?;
let reader = index.reader().unwrap();
let searcher = reader.searcher();
let segment_reader: &SegmentReader = searcher.segment_reader(0);
{
@@ -901,12 +938,11 @@ mod tests {
let fast_field_reader = fast_field_reader_opt.unwrap();
assert_eq!(fast_field_reader.get(0), 4f64)
}
Ok(())
}
// motivated by #729
#[test]
fn test_update_via_delete_insert() -> crate::Result<()> {
fn test_update_via_delete_insert() {
use crate::collector::Count;
use crate::indexer::NoMergePolicy;
use crate::query::AllQuery;
@@ -920,17 +956,17 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema.clone());
let index_reader = index.reader()?;
let index_reader = index.reader().unwrap();
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer(3_000_000).unwrap();
index_writer.set_merge_policy(Box::new(NoMergePolicy));
for doc_id in 0u64..DOC_COUNT {
index_writer.add_document(doc!(id => doc_id));
}
index_writer.commit()?;
index_writer.commit().unwrap();
index_reader.reload()?;
index_reader.reload().unwrap();
let searcher = index_reader.searcher();
assert_eq!(
@@ -941,11 +977,12 @@ mod tests {
// update the 10 elements by deleting and re-adding
for doc_id in 0u64..DOC_COUNT {
index_writer.delete_term(Term::from_field_u64(id, doc_id));
index_writer.commit()?;
index_reader.reload()?;
index_writer.add_document(doc!(id => doc_id));
index_writer.commit()?;
index_reader.reload()?;
index_writer.commit().unwrap();
index_reader.reload().unwrap();
let doc = doc!(id => doc_id);
index_writer.add_document(doc);
index_writer.commit().unwrap();
index_reader.reload().unwrap();
let searcher = index_reader.searcher();
// The number of document should be stable.
assert_eq!(
@@ -954,7 +991,7 @@ mod tests {
);
}
index_reader.reload()?;
index_reader.reload().unwrap();
let searcher = index_reader.searcher();
let segment_ids: Vec<SegmentId> = searcher
.segment_readers()
@@ -963,18 +1000,12 @@ mod tests {
.collect();
block_on(index_writer.merge(&segment_ids)).unwrap();
index_reader.reload()?;
index_reader.reload().unwrap();
let searcher = index_reader.searcher();
assert_eq!(searcher.search(&AllQuery, &Count)?, DOC_COUNT as usize);
Ok(())
}
#[test]
fn test_validate_checksum() -> crate::Result<()> {
let index_path = tempfile::tempdir().expect("dir");
let schema = Schema::builder().build();
let index = Index::create_in_dir(&index_path, schema)?;
assert!(index.validate_checksum()?.is_empty());
Ok(())
assert_eq!(
searcher.search(&AllQuery, &Count).unwrap(),
DOC_COUNT as usize
);
}
}

110
src/positions/mod.rs
View File

@@ -37,12 +37,12 @@ const LONG_SKIP_INTERVAL: u64 = (LONG_SKIP_IN_BLOCKS * COMPRESSION_BLOCK_SIZE) a
#[cfg(test)]
pub mod tests {
use super::PositionSerializer;
use crate::positions::reader::PositionReader;
use crate::{common::HasLen, directory::FileSlice};
use super::{PositionReader, PositionSerializer};
use crate::directory::ReadOnlySource;
use crate::positions::COMPRESSION_BLOCK_SIZE;
use std::iter;
fn create_stream_buffer(vals: &[u32]) -> (FileSlice, FileSlice) {
fn create_stream_buffer(vals: &[u32]) -> (ReadOnlySource, ReadOnlySource) {
let mut skip_buffer = vec![];
let mut stream_buffer = vec![];
{
@@ -53,7 +53,10 @@ pub mod tests {
}
serializer.close().unwrap();
}
(FileSlice::from(stream_buffer), FileSlice::from(skip_buffer))
(
ReadOnlySource::from(stream_buffer),
ReadOnlySource::from(skip_buffer),
)
}
#[test]
@@ -62,10 +65,10 @@ pub mod tests {
let (stream, skip) = create_stream_buffer(&v[..]);
assert_eq!(skip.len(), 12);
assert_eq!(stream.len(), 1168);
let mut position_reader = PositionReader::new(stream, skip, 0u64).unwrap();
let mut position_reader = PositionReader::new(stream, skip, 0u64);
for &n in &[1, 10, 127, 128, 130, 312] {
let mut v = vec![0u32; n];
position_reader.read(0, &mut v[..]);
position_reader.read(&mut v[..n]);
for i in 0..n {
assert_eq!(v[i], i as u32);
}
@@ -73,19 +76,19 @@ pub mod tests {
}
#[test]
fn test_position_read_with_offset() {
let v: Vec<u32> = (0..1000).collect();
fn test_position_skip() {
let v: Vec<u32> = (0..1_000).collect();
let (stream, skip) = create_stream_buffer(&v[..]);
assert_eq!(skip.len(), 12);
assert_eq!(stream.len(), 1168);
let mut position_reader = PositionReader::new(stream, skip, 0u64).unwrap();
for &offset in &[1u64, 10u64, 127u64, 128u64, 130u64, 312u64] {
for &len in &[1, 10, 130, 500] {
let mut v = vec![0u32; len];
position_reader.read(offset, &mut v[..]);
for i in 0..len {
assert_eq!(v[i], i as u32 + offset as u32);
}
let mut position_reader = PositionReader::new(stream, skip, 0u64);
position_reader.skip(10);
for &n in &[10, 127, COMPRESSION_BLOCK_SIZE, 130, 312] {
let mut v = vec![0u32; n];
position_reader.read(&mut v[..n]);
for i in 0..n {
assert_eq!(v[i], 10u32 + i as u32);
}
}
}
@@ -97,15 +100,14 @@ pub mod tests {
assert_eq!(skip.len(), 12);
assert_eq!(stream.len(), 1168);
let mut position_reader = PositionReader::new(stream, skip, 0u64).unwrap();
let mut position_reader = PositionReader::new(stream, skip, 0u64);
let mut buf = [0u32; 7];
let mut c = 0;
let mut offset = 0;
for _ in 0..100 {
position_reader.read(offset, &mut buf);
position_reader.read(offset, &mut buf);
offset += 7;
position_reader.read(&mut buf);
position_reader.read(&mut buf);
position_reader.skip(4);
position_reader.skip(3);
for &el in &buf {
assert_eq!(c, el);
c += 1;
@@ -113,59 +115,6 @@ pub mod tests {
}
}
#[test]
fn test_position_reread_anchor_different_than_block() {
let v: Vec<u32> = (0..2_000_000).collect();
let (stream, skip) = create_stream_buffer(&v[..]);
assert_eq!(skip.len(), 15_749);
assert_eq!(stream.len(), 4_987_872);
let mut position_reader = PositionReader::new(stream.clone(), skip.clone(), 0).unwrap();
let mut buf = [0u32; 256];
position_reader.read(128, &mut buf);
for i in 0..256 {
assert_eq!(buf[i], (128 + i) as u32);
}
position_reader.read(128, &mut buf);
for i in 0..256 {
assert_eq!(buf[i], (128 + i) as u32);
}
}
#[test]
#[should_panic(expected = "offset arguments should be increasing.")]
fn test_position_panic_if_called_previous_anchor() {
let v: Vec<u32> = (0..2_000_000).collect();
let (stream, skip) = create_stream_buffer(&v[..]);
assert_eq!(skip.len(), 15_749);
assert_eq!(stream.len(), 4_987_872);
let mut buf = [0u32; 1];
let mut position_reader =
PositionReader::new(stream.clone(), skip.clone(), 200_000).unwrap();
position_reader.read(230, &mut buf);
position_reader.read(9, &mut buf);
}
#[test]
fn test_positions_bug() {
let mut v: Vec<u32> = vec![];
for i in 1..200 {
for j in 0..i {
v.push(j);
}
}
let (stream, skip) = create_stream_buffer(&v[..]);
let mut buf = Vec::new();
let mut position_reader = PositionReader::new(stream.clone(), skip.clone(), 0).unwrap();
let mut offset = 0;
for i in 1..24 {
buf.resize(i, 0);
position_reader.read(offset, &mut buf[..]);
offset += i as u64;
let r: Vec<u32> = (0..i).map(|el| el as u32).collect();
assert_eq!(buf, &r[..]);
}
}
#[test]
fn test_position_long_skip_const() {
const CONST_VAL: u32 = 9u32;
@@ -173,9 +122,9 @@ pub mod tests {
let (stream, skip) = create_stream_buffer(&v[..]);
assert_eq!(skip.len(), 15_749);
assert_eq!(stream.len(), 1_000_000);
let mut position_reader = PositionReader::new(stream, skip, 128 * 1024).unwrap();
let mut position_reader = PositionReader::new(stream, skip, 128 * 1024);
let mut buf = [0u32; 1];
position_reader.read(0, &mut buf);
position_reader.read(&mut buf);
assert_eq!(buf[0], CONST_VAL);
}
@@ -192,10 +141,9 @@ pub mod tests {
128 * 1024 + 7,
128 * 10 * 1024 + 10,
] {
let mut position_reader =
PositionReader::new(stream.clone(), skip.clone(), offset).unwrap();
let mut position_reader = PositionReader::new(stream.clone(), skip.clone(), offset);
let mut buf = [0u32; 1];
position_reader.read(0, &mut buf);
position_reader.read(&mut buf);
assert_eq!(buf[0], offset as u32);
}
}

233
src/positions/reader.rs
View File

@@ -1,13 +1,9 @@
use std::io;
use crate::common::{BinarySerializable, FixedSize};
use crate::directory::FileSlice;
use crate::directory::OwnedBytes;
use crate::directory::ReadOnlySource;
use crate::positions::COMPRESSION_BLOCK_SIZE;
use crate::positions::LONG_SKIP_INTERVAL;
use crate::positions::LONG_SKIP_IN_BLOCKS;
use bitpacking::{BitPacker, BitPacker4x};
use crate::postings::compression::compressed_block_size;
/// Positions works as a long sequence of compressed block.
/// All terms are chained one after the other.
///
@@ -28,28 +24,28 @@ use bitpacking::{BitPacker, BitPacker4x};
/// A given block obviously takes `(128 x num_bit_for_the_block / num_bits_in_a_byte)`,
/// so skipping a block without decompressing it is just a matter of advancing that many
/// bytes.
use bitpacking::{BitPacker, BitPacker4x};
use owned_read::OwnedRead;
struct Positions {
bit_packer: BitPacker4x,
skip_file: FileSlice,
position_file: FileSlice,
long_skip_data: OwnedBytes,
skip_source: ReadOnlySource,
position_source: ReadOnlySource,
long_skip_source: ReadOnlySource,
}
impl Positions {
pub fn new(position_file: FileSlice, skip_file: FileSlice) -> io::Result<Positions> {
let (body, footer) = skip_file.split_from_end(u32::SIZE_IN_BYTES);
let footer_data = footer.read_bytes()?;
let num_long_skips = u32::deserialize(&mut footer_data.as_slice())?;
let (skip_file, long_skip_file) =
pub fn new(position_source: ReadOnlySource, skip_source: ReadOnlySource) -> Positions {
let (body, footer) = skip_source.split_from_end(u32::SIZE_IN_BYTES);
let num_long_skips = u32::deserialize(&mut footer.as_slice()).expect("Index corrupted");
let (skip_source, long_skip_source) =
body.split_from_end(u64::SIZE_IN_BYTES * (num_long_skips as usize));
let long_skip_data = long_skip_file.read_bytes()?;
Ok(Positions {
Positions {
bit_packer: BitPacker4x::new(),
skip_file,
long_skip_data,
position_file,
})
skip_source,
long_skip_source,
position_source,
}
}
/// Returns the offset of the block associated to the given `long_skip_id`.
@@ -59,116 +55,143 @@ impl Positions {
if long_skip_id == 0 {
return 0;
}
let long_skip_slice = self.long_skip_data.as_slice();
let long_skip_slice = self.long_skip_source.as_slice();
let mut long_skip_blocks: &[u8] = &long_skip_slice[(long_skip_id - 1) * 8..][..8];
u64::deserialize(&mut long_skip_blocks).expect("Index corrupted")
}
fn reader(&self, offset: u64) -> io::Result<PositionReader> {
fn reader(&self, offset: u64) -> PositionReader {
let long_skip_id = (offset / LONG_SKIP_INTERVAL) as usize;
let small_skip = (offset % LONG_SKIP_INTERVAL) as usize;
let offset_num_bytes: u64 = self.long_skip(long_skip_id);
let position_read = self
.position_file
.slice_from(offset_num_bytes as usize)
.read_bytes()?;
let skip_read = self
.skip_file
.slice_from(long_skip_id * LONG_SKIP_IN_BLOCKS)
.read_bytes()?;
Ok(PositionReader {
let mut position_read = OwnedRead::new(self.position_source.clone());
position_read.advance(offset_num_bytes as usize);
let mut skip_read = OwnedRead::new(self.skip_source.clone());
skip_read.advance(long_skip_id * LONG_SKIP_IN_BLOCKS);
let mut position_reader = PositionReader {
bit_packer: self.bit_packer,
skip_read,
position_read,
inner_offset: 0,
buffer: Box::new([0u32; 128]),
block_offset: std::i64::MAX as u64,
anchor_offset: (long_skip_id as u64) * LONG_SKIP_INTERVAL,
abs_offset: offset,
})
ahead: None,
};
position_reader.skip(small_skip);
position_reader
}
}
#[derive(Clone)]
pub struct PositionReader {
skip_read: OwnedBytes,
position_read: OwnedBytes,
skip_read: OwnedRead,
position_read: OwnedRead,
bit_packer: BitPacker4x,
buffer: Box<[u32; COMPRESSION_BLOCK_SIZE]>,
inner_offset: usize,
buffer: Box<[u32; 128]>,
ahead: Option<usize>, // if None, no block is loaded.
// if Some(num_blocks), the block currently loaded is num_blocks ahead
// of the block of the next int to read.
}
block_offset: u64,
anchor_offset: u64,
abs_offset: u64,
// `ahead` represents the offset of the block currently loaded
// compared to the cursor of the actual stream.
//
// By contract, when this function is called, the current block has to be
// decompressed.
//
// If the requested number of els ends exactly at a given block, the next
// block is not decompressed.
fn read_impl(
bit_packer: BitPacker4x,
mut position: &[u8],
buffer: &mut [u32; 128],
mut inner_offset: usize,
num_bits: &[u8],
output: &mut [u32],
) -> usize {
let mut output_start = 0;
let mut output_len = output.len();
let mut ahead = 0;
loop {
let available_len = COMPRESSION_BLOCK_SIZE - inner_offset;
// We have enough elements in the current block.
// Let's copy the requested elements in the output buffer,
// and return.
if output_len <= available_len {
output[output_start..].copy_from_slice(&buffer[inner_offset..][..output_len]);
return ahead;
}
output[output_start..][..available_len].copy_from_slice(&buffer[inner_offset..]);
output_len -= available_len;
output_start += available_len;
inner_offset = 0;
let num_bits = num_bits[ahead];
bit_packer.decompress(position, &mut buffer[..], num_bits);
let block_len = compressed_block_size(num_bits);
position = &position[block_len..];
ahead += 1;
}
}
impl PositionReader {
pub fn new(
position_file: FileSlice,
skip_file: FileSlice,
position_source: ReadOnlySource,
skip_source: ReadOnlySource,
offset: u64,
) -> io::Result<PositionReader> {
let positions = Positions::new(position_file, skip_file)?;
positions.reader(offset)
) -> PositionReader {
Positions::new(position_source, skip_source).reader(offset)
}
fn advance_num_blocks(&mut self, num_blocks: usize) {
let num_bits: usize = self.skip_read.as_ref()[..num_blocks]
.iter()
.cloned()
.map(|num_bits| num_bits as usize)
.sum();
let num_bytes_to_skip = num_bits * COMPRESSION_BLOCK_SIZE / 8;
self.skip_read.advance(num_blocks as usize);
self.position_read.advance(num_bytes_to_skip);
}
/// Fills a buffer with the positions `[offset..offset+output.len())` integers.
///
/// `offset` is required to have a value >= to the offsets given in previous calls
/// for the given `PositionReaderAbsolute` instance.
pub fn read(&mut self, mut offset: u64, mut output: &mut [u32]) {
offset += self.abs_offset;
assert!(
offset >= self.anchor_offset,
"offset arguments should be increasing."
);
let delta_to_block_offset = offset as i64 - self.block_offset as i64;
if delta_to_block_offset < 0 || delta_to_block_offset >= 128 {
// The first position is not within the first block.
// We need to decompress the first block.
let delta_to_anchor_offset = offset - self.anchor_offset;
let num_blocks_to_skip =
(delta_to_anchor_offset / (COMPRESSION_BLOCK_SIZE as u64)) as usize;
self.advance_num_blocks(num_blocks_to_skip);
self.anchor_offset = offset - (offset % COMPRESSION_BLOCK_SIZE as u64);
self.block_offset = self.anchor_offset;
let num_bits = self.skip_read.as_slice()[0];
self.bit_packer
.decompress(self.position_read.as_ref(), self.buffer.as_mut(), num_bits);
} else {
let num_blocks_to_skip =
((self.block_offset - self.anchor_offset) / COMPRESSION_BLOCK_SIZE as u64) as usize;
self.advance_num_blocks(num_blocks_to_skip);
self.anchor_offset = self.block_offset;
}
let mut num_bits = self.skip_read.as_slice()[0];
let mut position_data = self.position_read.as_ref();
for i in 1.. {
let offset_in_block = (offset as usize) % COMPRESSION_BLOCK_SIZE;
let remaining_in_block = COMPRESSION_BLOCK_SIZE - offset_in_block;
if remaining_in_block >= output.len() {
output.copy_from_slice(&self.buffer[offset_in_block..][..output.len()]);
break;
}
output[..remaining_in_block].copy_from_slice(&self.buffer[offset_in_block..]);
output = &mut output[remaining_in_block..];
offset += remaining_in_block as u64;
position_data = &position_data[(num_bits as usize * COMPRESSION_BLOCK_SIZE / 8)..];
num_bits = self.skip_read.as_slice()[i];
/// Fills a buffer with the next `output.len()` integers.
/// This does not consume / advance the stream.
pub fn read(&mut self, output: &mut [u32]) {
let skip_data = self.skip_read.as_ref();
let position_data = self.position_read.as_ref();
let num_bits = self.skip_read.get(0);
if self.ahead != Some(0) {
// the block currently available is not the block
// for the current position
self.bit_packer
.decompress(position_data, self.buffer.as_mut(), num_bits);
self.block_offset += COMPRESSION_BLOCK_SIZE as u64;
self.ahead = Some(0);
}
let block_len = compressed_block_size(num_bits);
self.ahead = Some(read_impl(
self.bit_packer,
&position_data[block_len..],
self.buffer.as_mut(),
self.inner_offset,
&skip_data[1..],
output,
));
}
/// Skip the next `skip_len` integer.
///
/// If a full block is skipped, calling
/// `.skip(...)` will avoid decompressing it.
///
/// May panic if the end of the stream is reached.
pub fn skip(&mut self, skip_len: usize) {
let skip_len_plus_inner_offset = skip_len + self.inner_offset;
let num_blocks_to_advance = skip_len_plus_inner_offset / COMPRESSION_BLOCK_SIZE;
self.inner_offset = skip_len_plus_inner_offset % COMPRESSION_BLOCK_SIZE;
self.ahead = self.ahead.and_then(|num_blocks| {
if num_blocks >= num_blocks_to_advance {
Some(num_blocks - num_blocks_to_advance)
} else {
None
}
});
let skip_len_in_bits = self.skip_read.as_ref()[..num_blocks_to_advance]
.iter()
.map(|num_bits| *num_bits as usize)
.sum::<usize>()
* COMPRESSION_BLOCK_SIZE;
let skip_len_in_bytes = skip_len_in_bits / 8;
self.skip_read.advance(num_blocks_to_advance);
self.position_read.advance(skip_len_in_bytes);
}
}

38
src/postings/block_search.rs
View File

@@ -87,7 +87,6 @@ fn exponential_search(arr: &[u32], target: u32) -> (usize, usize) {
(begin, end)
}
#[inline(never)]
fn galloping(block_docs: &[u32], target: u32) -> usize {
let (start, end) = exponential_search(&block_docs, target);
start + linear_search(&block_docs[start..end], target)
@@ -130,18 +129,23 @@ impl BlockSearcher {
///
/// If SSE2 instructions are available in the `(platform, running CPU)`,
/// then we use a different implementation that does an exhaustive linear search over
/// the block regardless of whether the block is full or not.
///
/// Indeed, if the block is not full, the remaining items are TERMINATED.
/// It is surprisingly faster, most likely because of the lack of branch misprediction.
pub(crate) fn search_in_block(self, block_docs: &AlignedBuffer, target: u32) -> usize {
/// the full block whenever the block is full (`len == 128`). It is surprisingly faster, most likely because of the lack
/// of branch.
pub(crate) fn search_in_block(
self,
block_docs: &AlignedBuffer,
len: usize,
start: usize,
target: u32,
) -> usize {
#[cfg(target_arch = "x86_64")]
{
if self == BlockSearcher::SSE2 {
use crate::postings::compression::COMPRESSION_BLOCK_SIZE;
if self == BlockSearcher::SSE2 && len == COMPRESSION_BLOCK_SIZE {
return sse2::linear_search_sse2_128(block_docs, target);
}
}
galloping(&block_docs.0[..], target)
start + galloping(&block_docs.0[start..len], target)
}
}
@@ -162,7 +166,6 @@ mod tests {
use super::exponential_search;
use super::linear_search;
use super::BlockSearcher;
use crate::docset::TERMINATED;
use crate::postings::compression::{AlignedBuffer, COMPRESSION_BLOCK_SIZE};
#[test]
@@ -193,12 +196,19 @@ mod tests {
fn util_test_search_in_block(block_searcher: BlockSearcher, block: &[u32], target: u32) {
let cursor = search_in_block_trivial_but_slow(block, target);
assert!(block.len() < COMPRESSION_BLOCK_SIZE);
let mut output_buffer = [TERMINATED; COMPRESSION_BLOCK_SIZE];
let mut output_buffer = [u32::max_value(); COMPRESSION_BLOCK_SIZE];
output_buffer[..block.len()].copy_from_slice(block);
assert_eq!(
block_searcher.search_in_block(&AlignedBuffer(output_buffer), target),
cursor
);
for i in 0..cursor {
assert_eq!(
block_searcher.search_in_block(
&AlignedBuffer(output_buffer),
block.len(),
i,
target
),
cursor
);
}
}
fn util_test_search_in_block_all(block_searcher: BlockSearcher, block: &[u32]) {

530
src/postings/block_segment_postings.rs
View File

@@ -1,530 +0,0 @@
use std::io;
use crate::common::{BinarySerializable, VInt};
use crate::directory::FileSlice;
use crate::directory::OwnedBytes;
use crate::fieldnorm::FieldNormReader;
use crate::postings::compression::{
AlignedBuffer, BlockDecoder, VIntDecoder, COMPRESSION_BLOCK_SIZE,
};
use crate::postings::{BlockInfo, FreqReadingOption, SkipReader};
use crate::query::BM25Weight;
use crate::schema::IndexRecordOption;
use crate::{DocId, Score, TERMINATED};
fn max_score<I: Iterator<Item = Score>>(mut it: I) -> Option<Score> {
if let Some(first) = it.next() {
Some(it.fold(first, Score::max))
} else {
None
}
}
/// `BlockSegmentPostings` is a cursor iterating over blocks
/// of documents.
///
/// # Warning
///
/// While it is useful for some very specific high-performance
/// use cases, you should prefer using `SegmentPostings` for most usage.
#[derive(Clone)]
pub struct BlockSegmentPostings {
pub(crate) doc_decoder: BlockDecoder,
loaded_offset: usize,
freq_decoder: BlockDecoder,
freq_reading_option: FreqReadingOption,
block_max_score_cache: Option<Score>,
doc_freq: u32,
data: OwnedBytes,
pub(crate) skip_reader: SkipReader,
}
fn decode_bitpacked_block(
doc_decoder: &mut BlockDecoder,
freq_decoder_opt: Option<&mut BlockDecoder>,
data: &[u8],
doc_offset: DocId,
doc_num_bits: u8,
tf_num_bits: u8,
) {
let num_consumed_bytes = doc_decoder.uncompress_block_sorted(data, doc_offset, doc_num_bits);
if let Some(freq_decoder) = freq_decoder_opt {
freq_decoder.uncompress_block_unsorted(&data[num_consumed_bytes..], tf_num_bits);
}
}
fn decode_vint_block(
doc_decoder: &mut BlockDecoder,
freq_decoder_opt: Option<&mut BlockDecoder>,
data: &[u8],
doc_offset: DocId,
num_vint_docs: usize,
) {
let num_consumed_bytes =
doc_decoder.uncompress_vint_sorted(data, doc_offset, num_vint_docs, TERMINATED);
if let Some(freq_decoder) = freq_decoder_opt {
freq_decoder.uncompress_vint_unsorted(
&data[num_consumed_bytes..],
num_vint_docs,
TERMINATED,
);
}
}
fn split_into_skips_and_postings(
doc_freq: u32,
mut bytes: OwnedBytes,
) -> (Option<OwnedBytes>, OwnedBytes) {
if doc_freq < COMPRESSION_BLOCK_SIZE as u32 {
return (None, bytes);
}
let skip_len = VInt::deserialize(&mut bytes).expect("Data corrupted").0 as usize;
let (skip_data, postings_data) = bytes.split(skip_len);
(Some(skip_data), postings_data)
}
impl BlockSegmentPostings {
pub(crate) fn open(
doc_freq: u32,
data: FileSlice,
record_option: IndexRecordOption,
requested_option: IndexRecordOption,
) -> io::Result<BlockSegmentPostings> {
let freq_reading_option = match (record_option, requested_option) {
(IndexRecordOption::Basic, _) => FreqReadingOption::NoFreq,
(_, IndexRecordOption::Basic) => FreqReadingOption::SkipFreq,
(_, _) => FreqReadingOption::ReadFreq,
};
let (skip_data_opt, postings_data) =
split_into_skips_and_postings(doc_freq, data.read_bytes()?);
let skip_reader = match skip_data_opt {
Some(skip_data) => SkipReader::new(skip_data, doc_freq, record_option),
None => SkipReader::new(OwnedBytes::empty(), doc_freq, record_option),
};
let mut block_segment_postings = BlockSegmentPostings {
doc_decoder: BlockDecoder::with_val(TERMINATED),
loaded_offset: std::usize::MAX,
freq_decoder: BlockDecoder::with_val(1),
freq_reading_option,
block_max_score_cache: None,
doc_freq,
data: postings_data,
skip_reader,
};
block_segment_postings.load_block();
Ok(block_segment_postings)
}
/// Returns the block_max_score for the current block.
/// It does not require the block to be loaded. For instance, it is ok to call this method
/// after having called `.shallow_advance(..)`.
///
/// See `TermScorer::block_max_score(..)` for more information.
pub fn block_max_score(
&mut self,
fieldnorm_reader: &FieldNormReader,
bm25_weight: &BM25Weight,
) -> Score {
if let Some(score) = self.block_max_score_cache {
return score;
}
if let Some(skip_reader_max_score) = self.skip_reader.block_max_score(bm25_weight) {
// if we are on a full block, the skip reader should have the block max information
// for us
self.block_max_score_cache = Some(skip_reader_max_score);
return skip_reader_max_score;
}
// this is the last block of the segment posting list.
// If it is actually loaded, we can compute block max manually.
if self.block_is_loaded() {
let docs = self.doc_decoder.output_array().iter().cloned();
let freqs = self.freq_decoder.output_array().iter().cloned();
let bm25_scores = docs.zip(freqs).map(|(doc, term_freq)| {
let fieldnorm_id = fieldnorm_reader.fieldnorm_id(doc);
bm25_weight.score(fieldnorm_id, term_freq)
});
let block_max_score = max_score(bm25_scores).unwrap_or(0.0);
self.block_max_score_cache = Some(block_max_score);
return block_max_score;
}
// We do not have access to any good block max value. We return bm25_weight.max_score()
// as it is a valid upperbound.
//
// We do not cache it however, so that it gets computed when once block is loaded.
bm25_weight.max_score()
}
pub(crate) fn freq_reading_option(&self) -> FreqReadingOption {
self.freq_reading_option
}
// Resets the block segment postings on another position
// in the postings file.
//
// This is useful for enumerating through a list of terms,
// and consuming the associated posting lists while avoiding
// reallocating a `BlockSegmentPostings`.
//
// # Warning
//
// This does not reset the positions list.
pub(crate) fn reset(&mut self, doc_freq: u32, postings_data: OwnedBytes) {
let (skip_data_opt, postings_data) = split_into_skips_and_postings(doc_freq, postings_data);
self.data = postings_data;
self.block_max_score_cache = None;
self.loaded_offset = std::usize::MAX;
if let Some(skip_data) = skip_data_opt {
self.skip_reader.reset(skip_data, doc_freq);
} else {
self.skip_reader.reset(OwnedBytes::empty(), doc_freq);
}
self.doc_freq = doc_freq;
self.load_block();
}
/// Returns the overall number of documents in the block postings.
/// It does not take in account whether documents are deleted or not.
///
/// This `doc_freq` is simply the sum of the length of all of the blocks
/// length, and it does not take in account deleted documents.
pub fn doc_freq(&self) -> u32 {
self.doc_freq
}
/// Returns the array of docs in the current block.
///
/// Before the first call to `.advance()`, the block
/// returned by `.docs()` is empty.
#[inline]
pub fn docs(&self) -> &[DocId] {
debug_assert!(self.block_is_loaded());
self.doc_decoder.output_array()
}
/// Returns a full block, regardless of whetehr the block is complete or incomplete (
/// as it happens for the last block of the posting list).
///
/// In the latter case, the block is guaranteed to be padded with the sentinel value:
/// `TERMINATED`. The array is also guaranteed to be aligned on 16 bytes = 128 bits.
///
/// This method is useful to run SSE2 linear search.
#[inline(always)]
pub(crate) fn docs_aligned(&self) -> &AlignedBuffer {
debug_assert!(self.block_is_loaded());
self.doc_decoder.output_aligned()
}
/// Return the document at index `idx` of the block.
#[inline(always)]
pub fn doc(&self, idx: usize) -> u32 {
self.doc_decoder.output(idx)
}
/// Return the array of `term freq` in the block.
#[inline]
pub fn freqs(&self) -> &[u32] {
debug_assert!(self.block_is_loaded());
self.freq_decoder.output_array()
}
/// Return the frequency at index `idx` of the block.
#[inline]
pub fn freq(&self, idx: usize) -> u32 {
debug_assert!(self.block_is_loaded());
self.freq_decoder.output(idx)
}
/// Returns the length of the current block.
///
/// All blocks have a length of `NUM_DOCS_PER_BLOCK`,
/// except the last block that may have a length
/// of any number between 1 and `NUM_DOCS_PER_BLOCK - 1`
#[inline]
pub fn block_len(&self) -> usize {
debug_assert!(self.block_is_loaded());
self.doc_decoder.output_len
}
/// Position on a block that may contains `target_doc`.
///
/// If all docs are smaller than target, the block loaded may be empty,
/// or be the last an incomplete VInt block.
pub fn seek(&mut self, target_doc: DocId) {
self.shallow_seek(target_doc);
self.load_block();
}
pub(crate) fn position_offset(&self) -> u64 {
self.skip_reader.position_offset()
}
/// Dangerous API! This calls seek on the skip list,
/// but does not `.load_block()` afterwards.
///
/// `.load_block()` needs to be called manually afterwards.
/// If all docs are smaller than target, the block loaded may be empty,
/// or be the last an incomplete VInt block.
pub(crate) fn shallow_seek(&mut self, target_doc: DocId) {
if self.skip_reader.seek(target_doc) {
self.block_max_score_cache = None;
}
}
pub(crate) fn block_is_loaded(&self) -> bool {
self.loaded_offset == self.skip_reader.byte_offset()
}
pub(crate) fn load_block(&mut self) {
let offset = self.skip_reader.byte_offset();
if self.loaded_offset == offset {
return;
}
self.loaded_offset = offset;
match self.skip_reader.block_info() {
BlockInfo::BitPacked {
doc_num_bits,
tf_num_bits,
..
} => {
decode_bitpacked_block(
&mut self.doc_decoder,
if let FreqReadingOption::ReadFreq = self.freq_reading_option {
Some(&mut self.freq_decoder)
} else {
None
},
&self.data.as_slice()[offset..],
self.skip_reader.last_doc_in_previous_block,
doc_num_bits,
tf_num_bits,
);
}
BlockInfo::VInt { num_docs } => {
let data = {
if num_docs == 0 {
&[]
} else {
&self.data.as_slice()[offset..]
}
};
decode_vint_block(
&mut self.doc_decoder,
if let FreqReadingOption::ReadFreq = self.freq_reading_option {
Some(&mut self.freq_decoder)
} else {
None
},
data,
self.skip_reader.last_doc_in_previous_block,
num_docs as usize,
);
}
}
}
/// Advance to the next block.
///
/// Returns false iff there was no remaining blocks.
pub fn advance(&mut self) {
self.skip_reader.advance();
self.block_max_score_cache = None;
self.load_block();
}
/// Returns an empty segment postings object
pub fn empty() -> BlockSegmentPostings {
BlockSegmentPostings {
doc_decoder: BlockDecoder::with_val(TERMINATED),
loaded_offset: 0,
freq_decoder: BlockDecoder::with_val(1),
freq_reading_option: FreqReadingOption::NoFreq,
block_max_score_cache: None,
doc_freq: 0,
data: OwnedBytes::empty(),
skip_reader: SkipReader::new(OwnedBytes::empty(), 0, IndexRecordOption::Basic),
}
}
}
#[cfg(test)]
mod tests {
use super::BlockSegmentPostings;
use crate::common::HasLen;
use crate::core::Index;
use crate::docset::{DocSet, TERMINATED};
use crate::postings::compression::COMPRESSION_BLOCK_SIZE;
use crate::postings::postings::Postings;
use crate::postings::SegmentPostings;
use crate::schema::IndexRecordOption;
use crate::schema::Schema;
use crate::schema::Term;
use crate::schema::INDEXED;
use crate::DocId;
#[test]
fn test_empty_segment_postings() {
let mut postings = SegmentPostings::empty();
assert_eq!(postings.doc(), TERMINATED);
assert_eq!(postings.advance(), TERMINATED);
assert_eq!(postings.advance(), TERMINATED);
assert_eq!(postings.doc_freq(), 0);
assert_eq!(postings.len(), 0);
}
#[test]
fn test_empty_postings_doc_returns_terminated() {
let mut postings = SegmentPostings::empty();
assert_eq!(postings.doc(), TERMINATED);
assert_eq!(postings.advance(), TERMINATED);
}
#[test]
fn test_empty_postings_doc_term_freq_returns_0() {
let postings = SegmentPostings::empty();
assert_eq!(postings.term_freq(), 1);
}
#[test]
fn test_empty_block_segment_postings() {
let mut postings = BlockSegmentPostings::empty();
assert!(postings.docs().is_empty());
assert_eq!(postings.doc_freq(), 0);
postings.advance();
assert!(postings.docs().is_empty());
assert_eq!(postings.doc_freq(), 0);
}
#[test]
fn test_block_segment_postings() {
let mut block_segments = build_block_postings(&(0..100_000).collect::<Vec<u32>>());
let mut offset: u32 = 0u32;
// checking that the `doc_freq` is correct
assert_eq!(block_segments.doc_freq(), 100_000);
loop {
let block = block_segments.docs();
if block.is_empty() {
break;
}
for (i, doc) in block.iter().cloned().enumerate() {
assert_eq!(offset + (i as u32), doc);
}
offset += block.len() as u32;
block_segments.advance();
}
}
#[test]
fn test_skip_right_at_new_block() {
let mut doc_ids = (0..128).collect::<Vec<u32>>();
// 128 is missing
doc_ids.push(129);
doc_ids.push(130);
{
let block_segments = build_block_postings(&doc_ids);
let mut docset = SegmentPostings::from_block_postings(block_segments, None);
assert_eq!(docset.seek(128), 129);
assert_eq!(docset.doc(), 129);
assert_eq!(docset.advance(), 130);
assert_eq!(docset.doc(), 130);
assert_eq!(docset.advance(), TERMINATED);
}
{
let block_segments = build_block_postings(&doc_ids);
let mut docset = SegmentPostings::from_block_postings(block_segments, None);
assert_eq!(docset.seek(129), 129);
assert_eq!(docset.doc(), 129);
assert_eq!(docset.advance(), 130);
assert_eq!(docset.doc(), 130);
assert_eq!(docset.advance(), TERMINATED);
}
{
let block_segments = build_block_postings(&doc_ids);
let mut docset = SegmentPostings::from_block_postings(block_segments, None);
assert_eq!(docset.doc(), 0);
assert_eq!(docset.seek(131), TERMINATED);
assert_eq!(docset.doc(), TERMINATED);
}
}
fn build_block_postings(docs: &[DocId]) -> BlockSegmentPostings {
let mut schema_builder = Schema::builder();
let int_field = schema_builder.add_u64_field("id", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut last_doc = 0u32;
for &doc in docs {
for _ in last_doc..doc {
index_writer.add_document(doc!(int_field=>1u64));
}
index_writer.add_document(doc!(int_field=>0u64));
last_doc = doc + 1;
}
index_writer.commit().unwrap();
let searcher = index.reader().unwrap().searcher();
let segment_reader = searcher.segment_reader(0);
let inverted_index = segment_reader.inverted_index(int_field).unwrap();
let term = Term::from_field_u64(int_field, 0u64);
let term_info = inverted_index.get_term_info(&term).unwrap();
inverted_index
.read_block_postings_from_terminfo(&term_info, IndexRecordOption::Basic)
.unwrap()
}
#[test]
fn test_block_segment_postings_seek() {
let mut docs = vec![0];
for i in 0..1300 {
docs.push((i * i / 100) + i);
}
let mut block_postings = build_block_postings(&docs[..]);
for i in vec![0, 424, 10000] {
block_postings.seek(i);
let docs = block_postings.docs();
assert!(docs[0] <= i);
assert!(docs.last().cloned().unwrap_or(0u32) >= i);
}
block_postings.seek(100_000);
assert_eq!(block_postings.doc(COMPRESSION_BLOCK_SIZE - 1), TERMINATED);
}
#[test]
fn test_reset_block_segment_postings() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let int_field = schema_builder.add_u64_field("id", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
// create two postings list, one containg even number,
// the other containing odd numbers.
for i in 0..6 {
let doc = doc!(int_field=> (i % 2) as u64);
index_writer.add_document(doc);
}
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let segment_reader = searcher.segment_reader(0);
let mut block_segments;
{
let term = Term::from_field_u64(int_field, 0u64);
let inverted_index = segment_reader.inverted_index(int_field)?;
let term_info = inverted_index.get_term_info(&term).unwrap();
block_segments = inverted_index
.read_block_postings_from_terminfo(&term_info, IndexRecordOption::Basic)?;
}
assert_eq!(block_segments.docs(), &[0, 2, 4]);
{
let term = Term::from_field_u64(int_field, 1u64);
let inverted_index = segment_reader.inverted_index(int_field)?;
let term_info = inverted_index.get_term_info(&term).unwrap();
inverted_index.reset_block_postings_from_terminfo(&term_info, &mut block_segments)?;
}
assert_eq!(block_segments.docs(), &[1, 3, 5]);
Ok(())
}
}

101
src/postings/compression/mod.rs
View File

@@ -17,12 +17,6 @@ pub struct BlockEncoder {
pub output_len: usize,
}
impl Default for BlockEncoder {
fn default() -> Self {
BlockEncoder::new()
}
}
impl BlockEncoder {
pub fn new() -> BlockEncoder {
BlockEncoder {
@@ -52,23 +46,19 @@ impl BlockEncoder {
/// We ensure that the OutputBuffer is align on 128 bits
/// in order to run SSE2 linear search on it.
#[repr(align(128))]
#[derive(Clone)]
pub(crate) struct AlignedBuffer(pub [u32; COMPRESSION_BLOCK_SIZE]);
#[derive(Clone)]
pub struct BlockDecoder {
bitpacker: BitPacker4x,
output: AlignedBuffer,
pub output_len: usize,
}
impl Default for BlockDecoder {
fn default() -> Self {
impl BlockDecoder {
pub fn new() -> BlockDecoder {
BlockDecoder::with_val(0u32)
}
}
impl BlockDecoder {
pub fn with_val(val: u32) -> BlockDecoder {
BlockDecoder {
bitpacker: BitPacker4x::new(),
@@ -100,8 +90,8 @@ impl BlockDecoder {
}
#[inline]
pub(crate) fn output_aligned(&self) -> &AlignedBuffer {
&self.output
pub(crate) fn output_aligned(&self) -> (&AlignedBuffer, usize) {
(&self.output, self.output_len)
}
#[inline]
@@ -144,14 +134,11 @@ pub trait VIntDecoder {
/// For instance, if delta encoded are `1, 3, 9`, and the
/// `offset` is 5, then the output will be:
/// `5 + 1 = 6, 6 + 3= 9, 9 + 9 = 18`
///
/// The value given in `padding` will be used to fill the remaining `128 - num_els` values.
fn uncompress_vint_sorted(
fn uncompress_vint_sorted<'a>(
&mut self,
compressed_data: &[u8],
compressed_data: &'a [u8],
offset: u32,
num_els: usize,
padding: u32,
) -> usize;
/// Uncompress an array of `u32s`, compressed using variable
@@ -159,14 +146,7 @@ pub trait VIntDecoder {
///
/// The method takes a number of int to decompress, and returns
/// the amount of bytes that were read to decompress them.
///
/// The value given in `padding` will be used to fill the remaining `128 - num_els` values.
fn uncompress_vint_unsorted(
&mut self,
compressed_data: &[u8],
num_els: usize,
padding: u32,
) -> usize;
fn uncompress_vint_unsorted<'a>(&mut self, compressed_data: &'a [u8], num_els: usize) -> usize;
}
impl VIntEncoder for BlockEncoder {
@@ -180,26 +160,18 @@ impl VIntEncoder for BlockEncoder {
}
impl VIntDecoder for BlockDecoder {
fn uncompress_vint_sorted(
fn uncompress_vint_sorted<'a>(
&mut self,
compressed_data: &[u8],
compressed_data: &'a [u8],
offset: u32,
num_els: usize,
padding: u32,
) -> usize {
self.output_len = num_els;
self.output.0.iter_mut().for_each(|el| *el = padding);
vint::uncompress_sorted(compressed_data, &mut self.output.0[..num_els], offset)
}
fn uncompress_vint_unsorted(
&mut self,
compressed_data: &[u8],
num_els: usize,
padding: u32,
) -> usize {
fn uncompress_vint_unsorted<'a>(&mut self, compressed_data: &'a [u8], num_els: usize) -> usize {
self.output_len = num_els;
self.output.0.iter_mut().for_each(|el| *el = padding);
vint::uncompress_unsorted(compressed_data, &mut self.output.0[..num_els])
}
}
@@ -208,14 +180,13 @@ impl VIntDecoder for BlockDecoder {
pub mod tests {
use super::*;
use crate::TERMINATED;
#[test]
fn test_encode_sorted_block() {
let vals: Vec<u32> = (0u32..128u32).map(|i| i * 7).collect();
let mut encoder = BlockEncoder::new();
let (num_bits, compressed_data) = encoder.compress_block_sorted(&vals, 0);
let mut decoder = BlockDecoder::default();
let mut decoder = BlockDecoder::new();
{
let consumed_num_bytes = decoder.uncompress_block_sorted(compressed_data, 0, num_bits);
assert_eq!(consumed_num_bytes, compressed_data.len());
@@ -228,9 +199,9 @@ pub mod tests {
#[test]
fn test_encode_sorted_block_with_offset() {
let vals: Vec<u32> = (0u32..128u32).map(|i| 11 + i * 7).collect();
let mut encoder = BlockEncoder::default();
let mut encoder = BlockEncoder::new();
let (num_bits, compressed_data) = encoder.compress_block_sorted(&vals, 10);
let mut decoder = BlockDecoder::default();
let mut decoder = BlockDecoder::new();
{
let consumed_num_bytes = decoder.uncompress_block_sorted(compressed_data, 10, num_bits);
assert_eq!(consumed_num_bytes, compressed_data.len());
@@ -245,11 +216,11 @@ pub mod tests {
let mut compressed: Vec<u8> = Vec::new();
let n = 128;
let vals: Vec<u32> = (0..n).map(|i| 11u32 + (i as u32) * 7u32).collect();
let mut encoder = BlockEncoder::default();
let mut encoder = BlockEncoder::new();
let (num_bits, compressed_data) = encoder.compress_block_sorted(&vals, 10);
compressed.extend_from_slice(compressed_data);
compressed.push(173u8);
let mut decoder = BlockDecoder::default();
let mut decoder = BlockDecoder::new();
{
let consumed_num_bytes = decoder.uncompress_block_sorted(&compressed, 10, num_bits);
assert_eq!(consumed_num_bytes, compressed.len() - 1);
@@ -265,11 +236,11 @@ pub mod tests {
let mut compressed: Vec<u8> = Vec::new();
let n = 128;
let vals: Vec<u32> = (0..n).map(|i| 11u32 + (i as u32) * 7u32 % 12).collect();
let mut encoder = BlockEncoder::default();
let mut encoder = BlockEncoder::new();
let (num_bits, compressed_data) = encoder.compress_block_unsorted(&vals);
compressed.extend_from_slice(compressed_data);
compressed.push(173u8);
let mut decoder = BlockDecoder::default();
let mut decoder = BlockDecoder::new();
{
let consumed_num_bytes = decoder.uncompress_block_unsorted(&compressed, num_bits);
assert_eq!(consumed_num_bytes + 1, compressed.len());
@@ -280,27 +251,20 @@ pub mod tests {
}
}
#[test]
fn test_block_decoder_initialization() {
let block = BlockDecoder::with_val(TERMINATED);
assert_eq!(block.output(0), TERMINATED);
}
#[test]
fn test_encode_vint() {
const PADDING_VALUE: u32 = 234_234_345u32;
let expected_length = 154;
let mut encoder = BlockEncoder::new();
let input: Vec<u32> = (0u32..123u32).map(|i| 4 + i * 7 / 2).into_iter().collect();
for offset in &[0u32, 1u32, 2u32] {
let encoded_data = encoder.compress_vint_sorted(&input, *offset);
assert!(encoded_data.len() <= expected_length);
let mut decoder = BlockDecoder::default();
let consumed_num_bytes =
decoder.uncompress_vint_sorted(&encoded_data, *offset, input.len(), PADDING_VALUE);
assert_eq!(consumed_num_bytes, encoded_data.len());
assert_eq!(input, decoder.output_array());
for i in input.len()..COMPRESSION_BLOCK_SIZE {
assert_eq!(decoder.output(i), PADDING_VALUE);
{
let expected_length = 154;
let mut encoder = BlockEncoder::new();
let input: Vec<u32> = (0u32..123u32).map(|i| 4 + i * 7 / 2).into_iter().collect();
for offset in &[0u32, 1u32, 2u32] {
let encoded_data = encoder.compress_vint_sorted(&input, *offset);
assert!(encoded_data.len() <= expected_length);
let mut decoder = BlockDecoder::new();
let consumed_num_bytes =
decoder.uncompress_vint_sorted(&encoded_data, *offset, input.len());
assert_eq!(consumed_num_bytes, encoded_data.len());
assert_eq!(input, decoder.output_array());
}
}
}
@@ -310,7 +274,6 @@ pub mod tests {
mod bench {
use super::*;
use crate::TERMINATED;
use rand::rngs::StdRng;
use rand::Rng;
use rand::SeedableRng;
@@ -341,7 +304,7 @@ mod bench {
let mut encoder = BlockEncoder::new();
let data = generate_array(COMPRESSION_BLOCK_SIZE, 0.1);
let (num_bits, compressed) = encoder.compress_block_sorted(&data, 0u32);
let mut decoder = BlockDecoder::default();
let mut decoder = BlockDecoder::new();
b.iter(|| {
decoder.uncompress_block_sorted(compressed, 0u32, num_bits);
});
@@ -376,9 +339,9 @@ mod bench {
let mut encoder = BlockEncoder::new();
let data = generate_array(NUM_INTS_BENCH_VINT, 0.001);
let compressed = encoder.compress_vint_sorted(&data, 0u32);
let mut decoder = BlockDecoder::default();
let mut decoder = BlockDecoder::new();
b.iter(|| {
decoder.uncompress_vint_sorted(compressed, 0u32, NUM_INTS_BENCH_VINT, TERMINATED);
decoder.uncompress_vint_sorted(compressed, 0u32, NUM_INTS_BENCH_VINT);
});
}
}

2
src/postings/compression/vint.rs
View File

@@ -42,7 +42,7 @@ pub(crate) fn compress_unsorted<'a>(input: &[u32], output: &'a mut [u8]) -> &'a
}
#[inline(always)]
pub fn uncompress_sorted(compressed_data: &[u8], output: &mut [u32], offset: u32) -> usize {
pub fn uncompress_sorted<'a>(compressed_data: &'a [u8], output: &mut [u32], offset: u32) -> usize {
let mut read_byte = 0;
let mut result = offset;
for output_mut in output.iter_mut() {

352
src/postings/mod.rs
View File

@@ -3,8 +3,11 @@ Postings module (also called inverted index)
*/
mod block_search;
mod block_segment_postings;
pub(crate) mod compression;
/// Postings module
///
/// Postings, also called inverted lists, is the key datastructure
/// to full-text search.
mod postings;
mod postings_writer;
mod recorder;
@@ -19,15 +22,18 @@ pub(crate) use self::block_search::BlockSearcher;
pub(crate) use self::postings_writer::MultiFieldPostingsWriter;
pub use self::serializer::{FieldSerializer, InvertedIndexSerializer};
use self::compression::COMPRESSION_BLOCK_SIZE;
pub use self::postings::Postings;
pub(crate) use self::skip::{BlockInfo, SkipReader};
pub(crate) use self::skip::SkipReader;
pub use self::term_info::TermInfo;
pub use self::block_segment_postings::BlockSegmentPostings;
pub use self::segment_postings::SegmentPostings;
pub use self::segment_postings::{BlockSegmentPostings, SegmentPostings};
pub(crate) use self::stacker::compute_table_size;
pub use crate::common::HasLen;
pub(crate) const USE_SKIP_INFO_LIMIT: u32 = COMPRESSION_BLOCK_SIZE as u32;
pub(crate) type UnorderedTermId = u64;
#[cfg_attr(feature = "cargo-clippy", allow(clippy::enum_variant_names))]
@@ -40,15 +46,15 @@ pub(crate) enum FreqReadingOption {
#[cfg(test)]
pub mod tests {
use super::InvertedIndexSerializer;
use super::Postings;
use super::*;
use crate::core::Index;
use crate::core::SegmentComponent;
use crate::core::SegmentReader;
use crate::docset::{DocSet, TERMINATED};
use crate::docset::{DocSet, SkipResult};
use crate::fieldnorm::FieldNormReader;
use crate::indexer::operation::AddOperation;
use crate::indexer::SegmentWriter;
use crate::indexer::{IndexWriterConfig, SegmentWriter, ResourceManager};
use crate::merge_policy::NoMergePolicy;
use crate::query::Scorer;
use crate::schema::{Document, Schema, Term, INDEXED, STRING, TEXT};
@@ -56,7 +62,6 @@ pub mod tests {
use crate::schema::{IndexRecordOption, TextFieldIndexing};
use crate::tokenizer::{SimpleTokenizer, MAX_TOKEN_LEN};
use crate::DocId;
use crate::HasLen;
use crate::Score;
use once_cell::sync::Lazy;
use rand::rngs::StdRng;
@@ -64,101 +69,102 @@ pub mod tests {
use std::iter;
#[test]
pub fn test_position_write() -> crate::Result<()> {
pub fn test_position_write() {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut segment = index.new_segment();
let mut posting_serializer = InvertedIndexSerializer::open(&mut segment)?;
let mut field_serializer = posting_serializer.new_field(text_field, 120 * 4, None)?;
field_serializer.new_term("abc".as_bytes(), 12u32)?;
for doc_id in 0u32..120u32 {
let delta_positions = vec![1, 2, 3, 2];
field_serializer.write_doc(doc_id, 4, &delta_positions)?;
let mut posting_serializer = InvertedIndexSerializer::open(&mut segment).unwrap();
{
let mut field_serializer = posting_serializer.new_field(text_field, 120 * 4).unwrap();
field_serializer.new_term("abc".as_bytes()).unwrap();
for doc_id in 0u32..120u32 {
let delta_positions = vec![1, 2, 3, 2];
field_serializer
.write_doc(doc_id, 4, &delta_positions)
.unwrap();
}
field_serializer.close_term().unwrap();
}
field_serializer.close_term()?;
posting_serializer.close()?;
let read = segment.open_read(SegmentComponent::POSITIONS)?;
posting_serializer.close().unwrap();
let read = segment.open_read(SegmentComponent::POSITIONS).unwrap();
assert!(read.len() <= 140);
Ok(())
}
#[test]
pub fn test_skip_positions() -> crate::Result<()> {
pub fn test_skip_positions() {
let mut schema_builder = Schema::builder();
let title = schema_builder.add_text_field("title", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(1, 30_000_000).unwrap();
index_writer.add_document(doc!(title => r#"abc abc abc"#));
index_writer.add_document(doc!(title => r#"abc be be be be abc"#));
for _ in 0..1_000 {
index_writer.add_document(doc!(title => r#"abc abc abc"#));
}
index_writer.add_document(doc!(title => r#"abc be be be be abc"#));
index_writer.commit()?;
index_writer.commit().unwrap();
let searcher = index.reader()?.searcher();
let inverted_index = searcher.segment_reader(0u32).inverted_index(title)?;
let searcher = index.reader().unwrap().searcher();
let inverted_index = searcher.segment_reader(0u32).inverted_index(title);
let term = Term::from_field_text(title, "abc");
let mut positions = Vec::new();
{
let mut postings = inverted_index
.read_postings(&term, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert_eq!(postings.doc(), 0);
postings.advance();
postings.positions(&mut positions);
assert_eq!(&[0, 1, 2], &positions[..]);
postings.positions(&mut positions);
assert_eq!(&[0, 1, 2], &positions[..]);
assert_eq!(postings.advance(), 1);
postings.advance();
postings.positions(&mut positions);
assert_eq!(&[0, 5], &positions[..]);
}
{
let mut postings = inverted_index
.read_postings(&term, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
postings.advance();
postings.advance();
postings.positions(&mut positions);
assert_eq!(&[0, 5], &positions[..]);
}
{
let mut postings = inverted_index
.read_postings(&term, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert_eq!(postings.skip_next(1), SkipResult::Reached);
assert_eq!(postings.doc(), 1);
postings.positions(&mut positions);
assert_eq!(&[0, 5], &positions[..]);
}
{
let mut postings = inverted_index
.read_postings(&term, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert_eq!(postings.doc(), 0);
assert_eq!(postings.advance(), 1);
postings.positions(&mut positions);
assert_eq!(&[0, 5], &positions[..]);
}
{
let mut postings = inverted_index
.read_postings(&term, IndexRecordOption::WithFreqsAndPositions)?
.unwrap();
assert_eq!(postings.seek(1), 1);
assert_eq!(postings.doc(), 1);
postings.positions(&mut positions);
assert_eq!(&[0, 5], &positions[..]);
}
{
let mut postings = inverted_index
.read_postings(&term, IndexRecordOption::WithFreqsAndPositions)?
.unwrap();
assert_eq!(postings.seek(1002), 1002);
assert_eq!(postings.skip_next(1002), SkipResult::Reached);
assert_eq!(postings.doc(), 1002);
postings.positions(&mut positions);
assert_eq!(&[0, 5], &positions[..]);
}
{
let mut postings = inverted_index
.read_postings(&term, IndexRecordOption::WithFreqsAndPositions)?
.read_postings(&term, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert_eq!(postings.seek(100), 100);
assert_eq!(postings.seek(1002), 1002);
assert_eq!(postings.skip_next(100), SkipResult::Reached);
assert_eq!(postings.skip_next(1002), SkipResult::Reached);
assert_eq!(postings.doc(), 1002);
postings.positions(&mut positions);
assert_eq!(&[0, 5], &positions[..]);
}
Ok(())
}
#[test]
pub fn test_drop_token_that_are_too_long() -> crate::Result<()> {
pub fn test_drop_token_that_are_too_long() {
let ok_token_text: String = iter::repeat('A').take(MAX_TOKEN_LEN).collect();
let mut exceeding_token_text: String = iter::repeat('A').take(MAX_TOKEN_LEN + 1).collect();
exceeding_token_text.push_str(" hello");
@@ -175,7 +181,7 @@ pub mod tests {
.tokenizers()
.register("simple_no_truncation", SimpleTokenizer);
let reader = index.reader().unwrap();
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.set_merge_policy(Box::new(NoMergePolicy));
{
index_writer.add_document(doc!(text_field=>exceeding_token_text));
@@ -183,7 +189,7 @@ pub mod tests {
reader.reload().unwrap();
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(0u32);
let inverted_index = segment_reader.inverted_index(text_field)?;
let inverted_index = segment_reader.inverted_index(text_field);
assert_eq!(inverted_index.terms().num_terms(), 1);
let mut bytes = vec![];
assert!(inverted_index.terms().ord_to_term(0, &mut bytes));
@@ -195,60 +201,70 @@ pub mod tests {
reader.reload().unwrap();
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(1u32);
let inverted_index = segment_reader.inverted_index(text_field)?;
let inverted_index = segment_reader.inverted_index(text_field);
assert_eq!(inverted_index.terms().num_terms(), 1);
let mut bytes = vec![];
assert!(inverted_index.terms().ord_to_term(0, &mut bytes));
assert_eq!(&bytes[..], ok_token_text.as_bytes());
}
Ok(())
}
#[test]
pub fn test_position_and_fieldnorm1() -> crate::Result<()> {
pub fn test_position_and_fieldnorm1() {
let mut positions = Vec::new();
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema.clone());
let segment = index.new_segment();
{
let mut segment_writer =
SegmentWriter::for_segment(3_000_000, segment.clone(), &schema).unwrap();
let resource_manager = ResourceManager::default();
let segment = {
let mut segment_writer = SegmentWriter::for_segment(
&IndexWriterConfig::for_test(),
index.new_segment(),
&schema,
index.tokenizers(),
resource_manager
)
.unwrap();
{
let mut doc = Document::default();
// checking that position works if the field has two values
doc.add_text(text_field, "a b a c a d a a.");
doc.add_text(text_field, "d d d d a");
let op = AddOperation {
opstamp: 0u64,
document: doc!(
text_field => "a b a c a d a a.",
text_field => "d d d d a"
),
document: doc,
};
segment_writer.add_document(op, &schema)?;
segment_writer.add_document(op, &schema).unwrap();
}
{
let mut doc = Document::default();
doc.add_text(text_field, "b a");
let op = AddOperation {
opstamp: 1u64,
document: doc!(text_field => "b a"),
document: doc,
};
segment_writer.add_document(op, &schema).unwrap();
}
for i in 2..1000 {
let mut text: String = iter::repeat("e ").take(i).collect();
let mut doc = Document::default();
let mut text = iter::repeat("e ").take(i).collect::<String>();
text.push_str(" a");
doc.add_text(text_field, &text);
let op = AddOperation {
opstamp: 2u64,
document: doc!(text_field => text),
document: doc,
};
segment_writer.add_document(op, &schema).unwrap();
}
segment_writer.finalize()?;
}
let (segment, _) = segment_writer.finalize().unwrap();
segment
};
{
let segment_reader = SegmentReader::open(&segment)?;
let segment_reader = SegmentReader::open(&segment).unwrap();
{
let fieldnorm_reader = segment_reader.get_fieldnorms_reader(text_field)?;
let fieldnorm_reader = segment_reader.get_fieldnorms_reader(text_field);
assert_eq!(fieldnorm_reader.fieldnorm(0), 8 + 5);
assert_eq!(fieldnorm_reader.fieldnorm(1), 2);
for i in 2..1000 {
@@ -261,41 +277,43 @@ pub mod tests {
{
let term_a = Term::from_field_text(text_field, "abcdef");
assert!(segment_reader
.inverted_index(term_a.field())?
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)?
.inverted_index(term_a.field())
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)
.is_none());
}
{
let term_a = Term::from_field_text(text_field, "a");
let mut postings_a = segment_reader
.inverted_index(term_a.field())?
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)?
.inverted_index(term_a.field())
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert_eq!(postings_a.len(), 1000);
assert!(postings_a.advance());
assert_eq!(postings_a.doc(), 0);
assert_eq!(postings_a.term_freq(), 6);
postings_a.positions(&mut positions);
assert_eq!(&positions[..], [0, 2, 4, 6, 7, 13]);
assert_eq!(postings_a.advance(), 1u32);
assert!(postings_a.advance());
assert_eq!(postings_a.doc(), 1u32);
assert_eq!(postings_a.term_freq(), 1);
for i in 2u32..1000u32 {
assert_eq!(postings_a.advance(), i);
assert!(postings_a.advance());
assert_eq!(postings_a.term_freq(), 1);
postings_a.positions(&mut positions);
assert_eq!(&positions[..], [i]);
assert_eq!(postings_a.doc(), i);
}
assert_eq!(postings_a.advance(), TERMINATED);
assert!(!postings_a.advance());
}
{
let term_e = Term::from_field_text(text_field, "e");
let mut postings_e = segment_reader
.inverted_index(term_e.field())?
.read_postings(&term_e, IndexRecordOption::WithFreqsAndPositions)?
.inverted_index(term_e.field())
.read_postings(&term_e, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert_eq!(postings_e.len(), 1000 - 2);
for i in 2u32..1000u32 {
assert!(postings_e.advance());
assert_eq!(postings_e.term_freq(), i);
postings_e.positions(&mut positions);
assert_eq!(positions.len(), i as usize);
@@ -303,42 +321,48 @@ pub mod tests {
assert_eq!(positions[j], (j as u32));
}
assert_eq!(postings_e.doc(), i);
postings_e.advance();
}
assert_eq!(postings_e.doc(), TERMINATED);
assert!(!postings_e.advance());
}
}
Ok(())
}
#[test]
pub fn test_position_and_fieldnorm2() -> crate::Result<()> {
pub fn test_position_and_fieldnorm2() {
let mut positions: Vec<u32> = Vec::new();
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(text_field => "g b b d c g c"));
index_writer.add_document(doc!(text_field => "g a b b a d c g c"));
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
{
let mut doc = Document::default();
doc.add_text(text_field, "g b b d c g c");
index_writer.add_document(doc);
}
{
let mut doc = Document::default();
doc.add_text(text_field, "g a b b a d c g c");
index_writer.add_document(doc);
}
assert!(index_writer.commit().is_ok());
}
let term_a = Term::from_field_text(text_field, "a");
let searcher = index.reader().unwrap().searcher();
let segment_reader = searcher.segment_reader(0);
let mut postings = segment_reader
.inverted_index(text_field)?
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)?
.inverted_index(text_field)
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert!(postings.advance());
assert_eq!(postings.doc(), 1u32);
postings.positions(&mut positions);
assert_eq!(&positions[..], &[1u32, 4]);
Ok(())
}
#[test]
fn test_skip_next() -> crate::Result<()> {
fn test_skip_next() {
let term_0 = Term::from_field_u64(Field::from_field_id(0), 0);
let term_1 = Term::from_field_u64(Field::from_field_id(0), 1);
let term_2 = Term::from_field_u64(Field::from_field_id(0), 2);
@@ -349,100 +373,105 @@ pub mod tests {
let mut schema_builder = Schema::builder();
let value_field = schema_builder.add_u64_field("value", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_for_tests()?;
for i in 0u64..num_docs as u64 {
let doc = doc!(value_field => 2u64, value_field => i % 2u64);
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
for i in 0..num_docs {
let mut doc = Document::default();
doc.add_u64(value_field, 2);
doc.add_u64(value_field, (i % 2) as u64);
index_writer.add_document(doc);
}
assert!(index_writer.commit().is_ok());
}
index
};
let searcher = index.reader()?.searcher();
let searcher = index.reader().unwrap().searcher();
let segment_reader = searcher.segment_reader(0);
// check that the basic usage works
for i in 0..num_docs - 1 {
for j in i + 1..num_docs {
let mut segment_postings = segment_reader
.inverted_index(term_2.field())?
.read_postings(&term_2, IndexRecordOption::Basic)?
.inverted_index(term_2.field())
.read_postings(&term_2, IndexRecordOption::Basic)
.unwrap();
assert_eq!(segment_postings.seek(i), i);
assert_eq!(segment_postings.skip_next(i), SkipResult::Reached);
assert_eq!(segment_postings.doc(), i);
assert_eq!(segment_postings.seek(j), j);
assert_eq!(segment_postings.skip_next(j), SkipResult::Reached);
assert_eq!(segment_postings.doc(), j);
}
}
{
let mut segment_postings = segment_reader
.inverted_index(term_2.field())?
.read_postings(&term_2, IndexRecordOption::Basic)?
.inverted_index(term_2.field())
.read_postings(&term_2, IndexRecordOption::Basic)
.unwrap();
// check that `skip_next` advances the iterator
assert!(segment_postings.advance());
assert_eq!(segment_postings.doc(), 0);
assert_eq!(segment_postings.seek(1), 1);
assert_eq!(segment_postings.skip_next(1), SkipResult::Reached);
assert_eq!(segment_postings.doc(), 1);
assert_eq!(segment_postings.seek(1), 1);
assert_eq!(segment_postings.doc(), 1);
assert_eq!(segment_postings.skip_next(1), SkipResult::OverStep);
assert_eq!(segment_postings.doc(), 2);
// check that going beyond the end is handled
assert_eq!(segment_postings.seek(num_docs), TERMINATED);
assert_eq!(segment_postings.skip_next(num_docs), SkipResult::End);
}
// check that filtering works
{
let mut segment_postings = segment_reader
.inverted_index(term_0.field())?
.read_postings(&term_0, IndexRecordOption::Basic)?
.inverted_index(term_0.field())
.read_postings(&term_0, IndexRecordOption::Basic)
.unwrap();
for i in 0..num_docs / 2 {
assert_eq!(segment_postings.seek(i * 2), i * 2);
assert_eq!(segment_postings.skip_next(i * 2), SkipResult::Reached);
assert_eq!(segment_postings.doc(), i * 2);
}
let mut segment_postings = segment_reader
.inverted_index(term_0.field())?
.read_postings(&term_0, IndexRecordOption::Basic)?
.inverted_index(term_0.field())
.read_postings(&term_0, IndexRecordOption::Basic)
.unwrap();
for i in 0..num_docs / 2 - 1 {
assert!(segment_postings.seek(i * 2 + 1) > (i * 1) * 2);
assert_eq!(segment_postings.skip_next(i * 2 + 1), SkipResult::OverStep);
assert_eq!(segment_postings.doc(), (i + 1) * 2);
}
}
// delete some of the documents
{
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.delete_term(term_0);
assert!(index_writer.commit().is_ok());
}
let searcher = index.reader()?.searcher();
assert_eq!(searcher.segment_readers().len(), 1);
let searcher = index.reader().unwrap().searcher();
let segment_reader = searcher.segment_reader(0);
// make sure seeking still works
for i in 0..num_docs {
let mut segment_postings = segment_reader
.inverted_index(term_2.field())?
.read_postings(&term_2, IndexRecordOption::Basic)?
.inverted_index(term_2.field())
.read_postings(&term_2, IndexRecordOption::Basic)
.unwrap();
if i % 2 == 0 {
assert_eq!(segment_postings.seek(i), i);
assert_eq!(segment_postings.skip_next(i), SkipResult::Reached);
assert_eq!(segment_postings.doc(), i);
assert!(segment_reader.is_deleted(i));
} else {
assert_eq!(segment_postings.seek(i), i);
assert_eq!(segment_postings.skip_next(i), SkipResult::Reached);
assert_eq!(segment_postings.doc(), i);
}
}
@@ -450,23 +479,19 @@ pub mod tests {
// now try with a longer sequence
{
let mut segment_postings = segment_reader
.inverted_index(term_2.field())?
.read_postings(&term_2, IndexRecordOption::Basic)?
.inverted_index(term_2.field())
.read_postings(&term_2, IndexRecordOption::Basic)
.unwrap();
let mut last = 2; // start from 5 to avoid seeking to 3 twice
let mut cur = 3;
loop {
let seek = segment_postings.seek(cur);
if seek == TERMINATED {
break;
}
assert_eq!(seek, segment_postings.doc());
if seek == cur {
assert_eq!(segment_postings.doc(), cur);
} else {
assert_eq!(segment_postings.doc(), cur + 1);
match segment_postings.skip_next(cur) {
SkipResult::End => break,
SkipResult::Reached => assert_eq!(segment_postings.doc(), cur),
SkipResult::OverStep => assert_eq!(segment_postings.doc(), cur + 1),
}
let next = cur + last;
last = cur;
cur = next;
@@ -476,19 +501,20 @@ pub mod tests {
// delete everything else
{
let mut index_writer = index.writer_for_tests()?;
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.delete_term(term_1);
assert!(index_writer.commit().is_ok());
}
let searcher = index.reader()?.searcher();
let searcher = index.reader().unwrap().searcher();
// finally, check that it's empty
{
let searchable_segment_ids = index.searchable_segment_ids()?;
let searchable_segment_ids = index
.searchable_segment_ids()
.expect("could not get index segment ids");
assert!(searchable_segment_ids.is_empty());
assert_eq!(searcher.num_docs(), 0);
}
Ok(())
}
pub static TERM_A: Lazy<Term> = Lazy::new(|| {
@@ -518,7 +544,7 @@ pub mod tests {
let index = Index::create_in_ram(schema);
let posting_list_size = 1_000_000;
{
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
for _ in 0..posting_list_size {
let mut doc = Document::default();
if rng.gen_bool(1f64 / 15f64) {
@@ -551,7 +577,7 @@ pub mod tests {
}
impl<TDocSet: DocSet> DocSet for UnoptimizedDocSet<TDocSet> {
fn advance(&mut self) -> DocId {
fn advance(&mut self) -> bool {
self.0.advance()
}
@@ -576,26 +602,31 @@ pub mod tests {
) {
for target in targets {
let mut postings_opt = postings_factory();
if target < postings_opt.doc() {
continue;
}
let mut postings_unopt = UnoptimizedDocSet::wrap(postings_factory());
let skip_result_opt = postings_opt.seek(target);
let skip_result_unopt = postings_unopt.seek(target);
let skip_result_opt = postings_opt.skip_next(target);
let skip_result_unopt = postings_unopt.skip_next(target);
assert_eq!(
skip_result_unopt, skip_result_opt,
"Failed while skipping to {}",
target
);
assert!(skip_result_opt >= target);
assert_eq!(skip_result_opt, postings_opt.doc());
if skip_result_opt == TERMINATED {
return;
match skip_result_opt {
SkipResult::Reached => assert_eq!(postings_opt.doc(), target),
SkipResult::OverStep => assert!(postings_opt.doc() > target),
SkipResult::End => {
return;
}
}
while postings_opt.doc() != TERMINATED {
assert_eq!(postings_opt.doc(), postings_unopt.doc());
assert_eq!(postings_opt.advance(), postings_unopt.advance());
while postings_opt.advance() {
assert!(postings_unopt.advance());
assert_eq!(
postings_opt.doc(),
postings_unopt.doc(),
"Failed while skipping to {}",
target
);
}
assert!(!postings_unopt.advance());
}
}
}
@@ -604,7 +635,7 @@ pub mod tests {
mod bench {
use super::tests::*;
use crate::docset::TERMINATED;
use crate::docset::SkipResult;
use crate::query::Intersection;
use crate::schema::IndexRecordOption;
use crate::tests;
@@ -620,9 +651,9 @@ mod bench {
b.iter(|| {
let mut segment_postings = segment_reader
.inverted_index(TERM_A.field())
.read_postings(&*TERM_A, IndexRecordOption::Basic)?
.read_postings(&*TERM_A, IndexRecordOption::Basic)
.unwrap();
while segment_postings.advance() != TERMINATED {}
while segment_postings.advance() {}
});
}
@@ -635,22 +666,18 @@ mod bench {
let segment_postings_a = segment_reader
.inverted_index(TERM_A.field())
.read_postings(&*TERM_A, IndexRecordOption::Basic)
.unwrap()
.unwrap();
let segment_postings_b = segment_reader
.inverted_index(TERM_B.field())
.read_postings(&*TERM_B, IndexRecordOption::Basic)
.unwrap()
.unwrap();
let segment_postings_c = segment_reader
.inverted_index(TERM_C.field())
.read_postings(&*TERM_C, IndexRecordOption::Basic)
.unwrap()
.unwrap();
let segment_postings_d = segment_reader
.inverted_index(TERM_D.field())
.read_postings(&*TERM_D, IndexRecordOption::Basic)
.unwrap()
.unwrap();
let mut intersection = Intersection::new(vec![
segment_postings_a,
@@ -658,7 +685,7 @@ mod bench {
segment_postings_c,
segment_postings_d,
]);
while intersection.advance() != TERMINATED {}
while intersection.advance() {}
});
}
@@ -671,14 +698,14 @@ mod bench {
let mut segment_postings = segment_reader
.inverted_index(TERM_A.field())
.read_postings(&*TERM_A, IndexRecordOption::Basic)
.unwrap()
.unwrap();
let mut existing_docs = Vec::new();
segment_postings.advance();
for doc in &docs {
if *doc >= segment_postings.doc() {
existing_docs.push(*doc);
if segment_postings.seek(*doc) == TERMINATED {
if segment_postings.skip_next(*doc) == SkipResult::End {
break;
}
}
@@ -690,7 +717,7 @@ mod bench {
.read_postings(&*TERM_A, IndexRecordOption::Basic)
.unwrap();
for doc in &existing_docs {
if segment_postings.seek(*doc) == TERMINATED {
if segment_postings.skip_next(*doc) == SkipResult::End {
break;
}
}
@@ -729,9 +756,8 @@ mod bench {
.read_postings(&*TERM_A, IndexRecordOption::Basic)
.unwrap();
let mut s = 0u32;
while segment_postings.doc() != TERMINATED {
while segment_postings.advance() {
s += (segment_postings.doc() & n) % 1024;
segment_postings.advance();
}
s
});

41
src/postings/postings_writer.rs
View File

@@ -1,6 +1,5 @@
use super::stacker::{Addr, MemoryArena, TermHashMap};
use crate::fieldnorm::FieldNormReaders;
use crate::postings::recorder::{
BufferLender, NothingRecorder, Recorder, TFAndPositionRecorder, TermFrequencyRecorder,
};
@@ -38,8 +37,12 @@ fn posting_from_field_entry(field_entry: &FieldEntry) -> Box<dyn PostingsWriter>
| FieldType::I64(_)
| FieldType::F64(_)
| FieldType::Date(_)
| FieldType::Bytes(_)
| FieldType::HierarchicalFacet => SpecializedPostingsWriter::<NothingRecorder>::new_boxed(),
FieldType::Bytes => {
// FieldType::Bytes cannot actually be indexed.
// TODO fix during the indexer refactoring described in #276
SpecializedPostingsWriter::<NothingRecorder>::new_boxed()
}
}
}
@@ -92,7 +95,7 @@ impl MultiFieldPostingsWriter {
}
}
pub fn mem_usage(&self) -> usize {
pub fn mem_usage(&self) -> u64 {
self.term_index.mem_usage() + self.heap.mem_usage()
}
@@ -101,7 +104,6 @@ impl MultiFieldPostingsWriter {
doc: DocId,
field: Field,
token_stream: &mut dyn TokenStream,
term_buffer: &mut Term,
) -> u32 {
let postings_writer =
self.per_field_postings_writers[field.field_id() as usize].deref_mut();
@@ -111,7 +113,6 @@ impl MultiFieldPostingsWriter {
field,
token_stream,
&mut self.heap,
term_buffer,
)
}
@@ -127,7 +128,6 @@ impl MultiFieldPostingsWriter {
pub fn serialize(
&self,
serializer: &mut InvertedIndexSerializer,
fieldnorm_readers: FieldNormReaders,
) -> crate::Result<HashMap<Field, FnvHashMap<UnorderedTermId, TermOrdinal>>> {
let mut term_offsets: Vec<(&[u8], Addr, UnorderedTermId)> =
self.term_index.iter().collect();
@@ -157,17 +157,12 @@ impl MultiFieldPostingsWriter {
unordered_term_mappings.insert(field, mapping);
}
FieldType::U64(_) | FieldType::I64(_) | FieldType::F64(_) | FieldType::Date(_) => {}
FieldType::Bytes(_) => {}
FieldType::Bytes => {}
}
let postings_writer =
self.per_field_postings_writers[field.field_id() as usize].as_ref();
let fieldnorm_reader = fieldnorm_readers.get_field(field)?;
let mut field_serializer = serializer.new_field(
field,
postings_writer.total_num_tokens(),
fieldnorm_reader,
)?;
let postings_writer = &self.per_field_postings_writers[field.field_id() as usize];
let mut field_serializer =
serializer.new_field(field, postings_writer.total_num_tokens())?;
postings_writer.serialize(
&term_offsets[start..stop],
&mut field_serializer,
@@ -219,20 +214,13 @@ pub trait PostingsWriter {
field: Field,
token_stream: &mut dyn TokenStream,
heap: &mut MemoryArena,
term_buffer: &mut Term,
) -> u32 {
term_buffer.set_field(field);
let mut term = Term::for_field(field);
let mut sink = |token: &Token| {
// We skip all tokens with a len greater than u16.
if token.text.len() <= MAX_TOKEN_LEN {
term_buffer.set_text(token.text.as_str());
self.subscribe(
term_index,
doc_id,
token.position as u32,
&term_buffer,
heap,
);
term.set_text(token.text.as_str());
self.subscribe(term_index, doc_id, token.position as u32, &term, heap);
} else {
info!(
"A token exceeding MAX_TOKEN_LEN ({}>{}) was dropped. Search for \
@@ -309,8 +297,7 @@ impl<Rec: Recorder + 'static> PostingsWriter for SpecializedPostingsWriter<Rec>
let mut buffer_lender = BufferLender::default();
for &(term_bytes, addr, _) in term_addrs {
let recorder: Rec = termdict_heap.read(addr);
let term_doc_freq = recorder.term_doc_freq().unwrap_or(0u32);
serializer.new_term(&term_bytes[4..], term_doc_freq)?;
serializer.new_term(&term_bytes[4..])?;
recorder.serialize(&mut buffer_lender, serializer, heap)?;
serializer.close_term()?;
}

23
src/postings/recorder.rs
View File

@@ -75,10 +75,6 @@ pub(crate) trait Recorder: Copy + 'static {
serializer: &mut FieldSerializer<'_>,
heap: &MemoryArena,
) -> io::Result<()>;
/// Returns the number of document containing this term.
///
/// Returns `None` if not available.
fn term_doc_freq(&self) -> Option<u32>;
}
/// Only records the doc ids
@@ -117,16 +113,11 @@ impl Recorder for NothingRecorder {
) -> io::Result<()> {
let buffer = buffer_lender.lend_u8();
self.stack.read_to_end(heap, buffer);
// TODO avoid reading twice.
for doc in VInt32Reader::new(&buffer[..]) {
serializer.write_doc(doc as u32, 0u32, &[][..])?;
}
Ok(())
}
fn term_doc_freq(&self) -> Option<u32> {
None
}
}
/// Recorder encoding document ids, and term frequencies
@@ -135,7 +126,6 @@ pub struct TermFrequencyRecorder {
stack: ExpUnrolledLinkedList,
current_doc: DocId,
current_tf: u32,
term_doc_freq: u32,
}
impl Recorder for TermFrequencyRecorder {
@@ -144,7 +134,6 @@ impl Recorder for TermFrequencyRecorder {
stack: ExpUnrolledLinkedList::new(),
current_doc: u32::max_value(),
current_tf: 0u32,
term_doc_freq: 0u32,
}
}
@@ -153,7 +142,6 @@ impl Recorder for TermFrequencyRecorder {
}
fn new_doc(&mut self, doc: DocId, heap: &mut MemoryArena) {
self.term_doc_freq += 1;
self.current_doc = doc;
let _ = write_u32_vint(doc, &mut self.stack.writer(heap));
}
@@ -184,10 +172,6 @@ impl Recorder for TermFrequencyRecorder {
Ok(())
}
fn term_doc_freq(&self) -> Option<u32> {
Some(self.term_doc_freq)
}
}
/// Recorder encoding term frequencies as well as positions.
@@ -195,14 +179,12 @@ impl Recorder for TermFrequencyRecorder {
pub struct TFAndPositionRecorder {
stack: ExpUnrolledLinkedList,
current_doc: DocId,
term_doc_freq: u32,
}
impl Recorder for TFAndPositionRecorder {
fn new() -> Self {
TFAndPositionRecorder {
stack: ExpUnrolledLinkedList::new(),
current_doc: u32::max_value(),
term_doc_freq: 0u32,
}
}
@@ -212,7 +194,6 @@ impl Recorder for TFAndPositionRecorder {
fn new_doc(&mut self, doc: DocId, heap: &mut MemoryArena) {
self.current_doc = doc;
self.term_doc_freq += 1u32;
let _ = write_u32_vint(doc, &mut self.stack.writer(heap));
}
@@ -252,10 +233,6 @@ impl Recorder for TFAndPositionRecorder {
}
Ok(())
}
fn term_doc_freq(&self) -> Option<u32> {
Some(self.term_doc_freq)
}
}
#[cfg(test)]

809
src/postings/segment_postings.rs
View File

@@ -1,65 +1,81 @@
use crate::common::BitSet;
use crate::common::HasLen;
use crate::directory::FileSlice;
use crate::docset::DocSet;
use crate::fastfield::DeleteBitSet;
use crate::common::{BinarySerializable, VInt};
use crate::docset::{DocSet, SkipResult};
use crate::positions::PositionReader;
use crate::postings::compression::COMPRESSION_BLOCK_SIZE;
use crate::postings::compression::{compressed_block_size, AlignedBuffer};
use crate::postings::compression::{BlockDecoder, VIntDecoder, COMPRESSION_BLOCK_SIZE};
use crate::postings::serializer::PostingsSerializer;
use crate::postings::BlockSearcher;
use crate::postings::BlockSegmentPostings;
use crate::postings::FreqReadingOption;
use crate::postings::Postings;
use crate::postings::SkipReader;
use crate::postings::USE_SKIP_INFO_LIMIT;
use crate::schema::IndexRecordOption;
use crate::{DocId, TERMINATED};
use crate::DocId;
use owned_read::OwnedRead;
use std::cmp::Ordering;
use tantivy_fst::Streamer;
struct PositionComputer {
// store the amount of position int
// before reading positions.
//
// if none, position are already loaded in
// the positions vec.
position_to_skip: usize,
position_reader: PositionReader,
}
impl PositionComputer {
pub fn new(position_reader: PositionReader) -> PositionComputer {
PositionComputer {
position_to_skip: 0,
position_reader,
}
}
pub fn add_skip(&mut self, num_skip: usize) {
self.position_to_skip += num_skip;
}
// Positions can only be read once.
pub fn positions_with_offset(&mut self, offset: u32, output: &mut [u32]) {
self.position_reader.skip(self.position_to_skip);
self.position_to_skip = 0;
self.position_reader.read(output);
let mut cum = offset;
for output_mut in output.iter_mut() {
cum += *output_mut;
*output_mut = cum;
}
}
}
/// `SegmentPostings` represents the inverted list or postings associated to
/// a term in a `Segment`.
///
/// As we iterate through the `SegmentPostings`, the frequencies are optionally decoded.
/// Positions on the other hand, are optionally entirely decoded upfront.
#[derive(Clone)]
pub struct SegmentPostings {
pub(crate) block_cursor: BlockSegmentPostings,
block_cursor: BlockSegmentPostings,
cur: usize,
position_reader: Option<PositionReader>,
position_computer: Option<PositionComputer>,
block_searcher: BlockSearcher,
}
impl SegmentPostings {
/// Returns an empty segment postings object
pub fn empty() -> Self {
let empty_block_cursor = BlockSegmentPostings::empty();
SegmentPostings {
block_cursor: BlockSegmentPostings::empty(),
cur: 0,
position_reader: None,
block_cursor: empty_block_cursor,
cur: COMPRESSION_BLOCK_SIZE,
position_computer: None,
block_searcher: BlockSearcher::default(),
}
}
/// Compute the number of non-deleted documents.
///
/// This method will clone and scan through the posting lists.
/// (this is a rather expensive operation).
pub fn doc_freq_given_deletes(&self, delete_bitset: &DeleteBitSet) -> u32 {
let mut docset = self.clone();
let mut doc_freq = 0;
loop {
let doc = docset.doc();
if doc == TERMINATED {
return doc_freq;
}
if delete_bitset.is_alive(doc) {
doc_freq += 1u32;
}
docset.advance();
}
}
/// Returns the overall number of documents in the block postings.
/// It does not take in account whether documents are deleted or not.
pub fn doc_freq(&self) -> u32 {
self.block_cursor.doc_freq()
}
/// Creates a segment postings object with the given documents
/// and no frequency encoded.
///
@@ -71,9 +87,7 @@ impl SegmentPostings {
pub fn create_from_docs(docs: &[u32]) -> SegmentPostings {
let mut buffer = Vec::new();
{
let mut postings_serializer =
PostingsSerializer::new(&mut buffer, 0.0, IndexRecordOption::Basic, None);
postings_serializer.new_term(docs.len() as u32);
let mut postings_serializer = PostingsSerializer::new(&mut buffer, false, false);
for &doc in docs {
postings_serializer.write_doc(doc, 1u32);
}
@@ -81,61 +95,17 @@ impl SegmentPostings {
.close_term(docs.len() as u32)
.expect("In memory Serialization should never fail.");
}
let block_segment_postings = BlockSegmentPostings::open(
let block_segment_postings = BlockSegmentPostings::from_data(
docs.len() as u32,
FileSlice::from(buffer),
OwnedRead::new(buffer),
IndexRecordOption::Basic,
IndexRecordOption::Basic,
)
.unwrap();
SegmentPostings::from_block_postings(block_segment_postings, None)
}
/// Helper functions to create `SegmentPostings` for tests.
#[cfg(test)]
pub fn create_from_docs_and_tfs(
doc_and_tfs: &[(u32, u32)],
fieldnorms: Option<&[u32]>,
) -> SegmentPostings {
use crate::fieldnorm::FieldNormReader;
use crate::Score;
let mut buffer: Vec<u8> = Vec::new();
let fieldnorm_reader = fieldnorms.map(FieldNormReader::for_test);
let average_field_norm = fieldnorms
.map(|fieldnorms| {
if fieldnorms.len() == 0 {
return 0.0;
}
let total_num_tokens: u64 = fieldnorms
.iter()
.map(|&fieldnorm| fieldnorm as u64)
.sum::<u64>();
total_num_tokens as Score / fieldnorms.len() as Score
})
.unwrap_or(0.0);
let mut postings_serializer = PostingsSerializer::new(
&mut buffer,
average_field_norm,
IndexRecordOption::WithFreqs,
fieldnorm_reader,
);
postings_serializer.new_term(doc_and_tfs.len() as u32);
for &(doc, tf) in doc_and_tfs {
postings_serializer.write_doc(doc, tf);
}
postings_serializer
.close_term(doc_and_tfs.len() as u32)
.unwrap();
let block_segment_postings = BlockSegmentPostings::open(
doc_and_tfs.len() as u32,
FileSlice::from(buffer),
IndexRecordOption::WithFreqs,
IndexRecordOption::WithFreqs,
)
.unwrap();
SegmentPostings::from_block_postings(block_segment_postings, None)
}
}
impl SegmentPostings {
/// Reads a Segment postings from an &[u8]
///
/// * `len` - number of document in the posting lists.
@@ -144,12 +114,12 @@ impl SegmentPostings {
/// frequencies and/or positions
pub(crate) fn from_block_postings(
segment_block_postings: BlockSegmentPostings,
position_reader: Option<PositionReader>,
positions_stream_opt: Option<PositionReader>,
) -> SegmentPostings {
SegmentPostings {
block_cursor: segment_block_postings,
cur: 0, // cursor within the block
position_reader,
cur: COMPRESSION_BLOCK_SIZE, // cursor within the block
position_computer: positions_stream_opt.map(PositionComputer::new),
block_searcher: BlockSearcher::default(),
}
}
@@ -159,60 +129,139 @@ impl DocSet for SegmentPostings {
// goes to the next element.
// next needs to be called a first time to point to the correct element.
#[inline]
fn advance(&mut self) -> DocId {
debug_assert!(self.block_cursor.block_is_loaded());
if self.cur == COMPRESSION_BLOCK_SIZE - 1 {
self.cur = 0;
self.block_cursor.advance();
} else {
self.cur += 1;
fn advance(&mut self) -> bool {
if self.position_computer.is_some() && self.cur < COMPRESSION_BLOCK_SIZE {
let term_freq = self.term_freq() as usize;
if let Some(position_computer) = self.position_computer.as_mut() {
position_computer.add_skip(term_freq);
}
}
self.doc()
self.cur += 1;
if self.cur >= self.block_cursor.block_len() {
self.cur = 0;
if !self.block_cursor.advance() {
self.cur = COMPRESSION_BLOCK_SIZE;
return false;
}
}
true
}
fn seek(&mut self, target: DocId) -> DocId {
debug_assert!(self.doc() <= target);
if self.doc() >= target {
return self.doc();
fn skip_next(&mut self, target: DocId) -> SkipResult {
if !self.advance() {
return SkipResult::End;
}
match self.doc().cmp(&target) {
Ordering::Equal => {
return SkipResult::Reached;
}
Ordering::Greater => {
return SkipResult::OverStep;
}
_ => {
// ...
}
}
self.block_cursor.seek(target);
// In the following, thanks to the call to advance above,
// we know that the position is not loaded and we need
// to skip every doc_freq we cross.
// At this point we are on the block, that might contain our document.
let output = self.block_cursor.docs_aligned();
self.cur = self.block_searcher.search_in_block(&output, target);
// skip blocks until one that might contain the target
// check if we need to go to the next block
let mut sum_freqs_skipped: u32 = 0;
if !self
.block_cursor
.docs()
.last()
.map(|doc| *doc >= target)
.unwrap_or(false)
// there should always be at least a document in the block
// since advance returned.
{
// we are not in the right block.
//
// First compute all of the freqs skipped from the current block.
if self.position_computer.is_some() {
sum_freqs_skipped = self.block_cursor.freqs()[self.cur..].iter().sum();
match self.block_cursor.skip_to(target) {
BlockSegmentPostingsSkipResult::Success(block_skip_freqs) => {
sum_freqs_skipped += block_skip_freqs;
}
BlockSegmentPostingsSkipResult::Terminated => {
return SkipResult::End;
}
}
} else if self.block_cursor.skip_to(target)
== BlockSegmentPostingsSkipResult::Terminated
{
// no positions needed. no need to sum freqs.
return SkipResult::End;
}
self.cur = 0;
}
// The last block is not full and padded with the value TERMINATED,
// so that we are guaranteed to have at least doc in the block (a real one or the padding)
// that is greater or equal to the target.
debug_assert!(self.cur < COMPRESSION_BLOCK_SIZE);
let cur = self.cur;
// we're in the right block now, start with an exponential search
let (output, len) = self.block_cursor.docs_aligned();
let new_cur = self
.block_searcher
.search_in_block(&output, len, cur, target);
if let Some(position_computer) = self.position_computer.as_mut() {
sum_freqs_skipped += self.block_cursor.freqs()[cur..new_cur].iter().sum::<u32>();
position_computer.add_skip(sum_freqs_skipped as usize);
}
self.cur = new_cur;
// `doc` is now the first element >= `target`
// If all docs are smaller than target the current block should be incomplemented and padded
// with the value `TERMINATED`.
//
// After the search, the cursor should point to the first value of TERMINATED.
let doc = output.0[self.cur];
let doc = output.0[new_cur];
debug_assert!(doc >= target);
debug_assert_eq!(doc, self.doc());
doc
if doc == target {
SkipResult::Reached
} else {
SkipResult::OverStep
}
}
/// Return the current document's `DocId`.
#[inline(always)]
///
/// # Panics
///
/// Will panics if called without having called advance before.
#[inline]
fn doc(&self) -> DocId {
self.block_cursor.doc(self.cur)
let docs = self.block_cursor.docs();
debug_assert!(
self.cur < docs.len(),
"Have you forgotten to call `.advance()` at least once before calling `.doc()` ."
);
docs[self.cur]
}
fn size_hint(&self) -> u32 {
self.len() as u32
}
fn append_to_bitset(&mut self, bitset: &mut BitSet) {
// finish the current block
if self.advance() {
for &doc in &self.block_cursor.docs()[self.cur..] {
bitset.insert(doc);
}
// ... iterate through the remaining blocks.
while self.block_cursor.advance() {
for &doc in self.block_cursor.docs() {
bitset.insert(doc);
}
}
}
}
}
impl HasLen for SegmentPostings {
fn len(&self) -> usize {
self.block_cursor.doc_freq() as usize
self.block_cursor.doc_freq()
}
}
@@ -241,63 +290,515 @@ impl Postings for SegmentPostings {
fn positions_with_offset(&mut self, offset: u32, output: &mut Vec<u32>) {
let term_freq = self.term_freq() as usize;
if let Some(position_reader) = self.position_reader.as_mut() {
let read_offset = self.block_cursor.position_offset()
+ (self.block_cursor.freqs()[..self.cur]
.iter()
.cloned()
.sum::<u32>() as u64);
if let Some(position_comp) = self.position_computer.as_mut() {
output.resize(term_freq, 0u32);
position_reader.read(read_offset, &mut output[..]);
let mut cum = offset;
for output_mut in output.iter_mut() {
cum += *output_mut;
*output_mut = cum;
}
position_comp.positions_with_offset(offset, &mut output[..]);
} else {
output.clear();
}
}
}
/// `BlockSegmentPostings` is a cursor iterating over blocks
/// of documents.
///
/// # Warning
///
/// While it is useful for some very specific high-performance
/// use cases, you should prefer using `SegmentPostings` for most usage.
pub struct BlockSegmentPostings {
doc_decoder: BlockDecoder,
freq_decoder: BlockDecoder,
freq_reading_option: FreqReadingOption,
doc_freq: usize,
doc_offset: DocId,
num_vint_docs: usize,
remaining_data: OwnedRead,
skip_reader: SkipReader,
}
fn split_into_skips_and_postings(
doc_freq: u32,
mut data: OwnedRead,
) -> (Option<OwnedRead>, OwnedRead) {
if doc_freq >= USE_SKIP_INFO_LIMIT {
let skip_len = VInt::deserialize(&mut data).expect("Data corrupted").0 as usize;
let mut postings_data = data.clone();
postings_data.advance(skip_len);
data.clip(skip_len);
(Some(data), postings_data)
} else {
(None, data)
}
}
#[derive(Debug, Eq, PartialEq)]
pub enum BlockSegmentPostingsSkipResult {
Terminated,
Success(u32), //< number of term freqs to skip
}
impl BlockSegmentPostings {
pub(crate) fn from_data(
doc_freq: u32,
data: OwnedRead,
record_option: IndexRecordOption,
requested_option: IndexRecordOption,
) -> BlockSegmentPostings {
let freq_reading_option = match (record_option, requested_option) {
(IndexRecordOption::Basic, _) => FreqReadingOption::NoFreq,
(_, IndexRecordOption::Basic) => FreqReadingOption::SkipFreq,
(_, _) => FreqReadingOption::ReadFreq,
};
let (skip_data_opt, postings_data) = split_into_skips_and_postings(doc_freq, data);
let skip_reader = match skip_data_opt {
Some(skip_data) => SkipReader::new(skip_data, record_option),
None => SkipReader::new(OwnedRead::new(&[][..]), record_option),
};
let doc_freq = doc_freq as usize;
let num_vint_docs = doc_freq % COMPRESSION_BLOCK_SIZE;
BlockSegmentPostings {
num_vint_docs,
doc_decoder: BlockDecoder::new(),
freq_decoder: BlockDecoder::with_val(1),
freq_reading_option,
doc_offset: 0,
doc_freq,
remaining_data: postings_data,
skip_reader,
}
}
// Resets the block segment postings on another position
// in the postings file.
//
// This is useful for enumerating through a list of terms,
// and consuming the associated posting lists while avoiding
// reallocating a `BlockSegmentPostings`.
//
// # Warning
//
// This does not reset the positions list.
pub(crate) fn reset(&mut self, doc_freq: u32, postings_data: OwnedRead) {
let (skip_data_opt, postings_data) = split_into_skips_and_postings(doc_freq, postings_data);
let num_vint_docs = (doc_freq as usize) & (COMPRESSION_BLOCK_SIZE - 1);
self.num_vint_docs = num_vint_docs;
self.remaining_data = postings_data;
if let Some(skip_data) = skip_data_opt {
self.skip_reader.reset(skip_data);
} else {
self.skip_reader.reset(OwnedRead::new(&[][..]))
}
self.doc_offset = 0;
self.doc_freq = doc_freq as usize;
}
/// Returns the document frequency associated to this block postings.
///
/// This `doc_freq` is simply the sum of the length of all of the blocks
/// length, and it does not take in account deleted documents.
pub fn doc_freq(&self) -> usize {
self.doc_freq
}
/// Returns the array of docs in the current block.
///
/// Before the first call to `.advance()`, the block
/// returned by `.docs()` is empty.
#[inline]
pub fn docs(&self) -> &[DocId] {
self.doc_decoder.output_array()
}
pub(crate) fn docs_aligned(&self) -> (&AlignedBuffer, usize) {
self.doc_decoder.output_aligned()
}
/// Return the document at index `idx` of the block.
#[inline]
pub fn doc(&self, idx: usize) -> u32 {
self.doc_decoder.output(idx)
}
/// Return the array of `term freq` in the block.
#[inline]
pub fn freqs(&self) -> &[u32] {
self.freq_decoder.output_array()
}
/// Return the frequency at index `idx` of the block.
#[inline]
pub fn freq(&self, idx: usize) -> u32 {
self.freq_decoder.output(idx)
}
/// Returns the length of the current block.
///
/// All blocks have a length of `NUM_DOCS_PER_BLOCK`,
/// except the last block that may have a length
/// of any number between 1 and `NUM_DOCS_PER_BLOCK - 1`
#[inline]
fn block_len(&self) -> usize {
self.doc_decoder.output_len
}
/// position on a block that may contains `doc_id`.
/// Always advance the current block.
///
/// Returns true if a block that has an element greater or equal to the target is found.
/// Returning true does not guarantee that the smallest element of the block is smaller
/// than the target. It only guarantees that the last element is greater or equal.
///
/// Returns false iff all of the document remaining are smaller than
/// `doc_id`. In that case, all of these document are consumed.
///
pub fn skip_to(&mut self, target_doc: DocId) -> BlockSegmentPostingsSkipResult {
let mut skip_freqs = 0u32;
while self.skip_reader.advance() {
if self.skip_reader.doc() >= target_doc {
// the last document of the current block is larger
// than the target.
//
// We found our block!
let num_bits = self.skip_reader.doc_num_bits();
let num_consumed_bytes = self.doc_decoder.uncompress_block_sorted(
self.remaining_data.as_ref(),
self.doc_offset,
num_bits,
);
self.remaining_data.advance(num_consumed_bytes);
let tf_num_bits = self.skip_reader.tf_num_bits();
match self.freq_reading_option {
FreqReadingOption::NoFreq => {}
FreqReadingOption::SkipFreq => {
let num_bytes_to_skip = compressed_block_size(tf_num_bits);
self.remaining_data.advance(num_bytes_to_skip);
}
FreqReadingOption::ReadFreq => {
let num_consumed_bytes = self
.freq_decoder
.uncompress_block_unsorted(self.remaining_data.as_ref(), tf_num_bits);
self.remaining_data.advance(num_consumed_bytes);
}
}
self.doc_offset = self.skip_reader.doc();
return BlockSegmentPostingsSkipResult::Success(skip_freqs);
} else {
skip_freqs += self.skip_reader.tf_sum();
let advance_len = self.skip_reader.total_block_len();
self.doc_offset = self.skip_reader.doc();
self.remaining_data.advance(advance_len);
}
}
// we are now on the last, incomplete, variable encoded block.
if self.num_vint_docs > 0 {
let num_compressed_bytes = self.doc_decoder.uncompress_vint_sorted(
self.remaining_data.as_ref(),
self.doc_offset,
self.num_vint_docs,
);
self.remaining_data.advance(num_compressed_bytes);
match self.freq_reading_option {
FreqReadingOption::NoFreq | FreqReadingOption::SkipFreq => {}
FreqReadingOption::ReadFreq => {
self.freq_decoder
.uncompress_vint_unsorted(self.remaining_data.as_ref(), self.num_vint_docs);
}
}
self.num_vint_docs = 0;
return self
.docs()
.last()
.map(|last_doc| {
if *last_doc >= target_doc {
BlockSegmentPostingsSkipResult::Success(skip_freqs)
} else {
BlockSegmentPostingsSkipResult::Terminated
}
})
.unwrap_or(BlockSegmentPostingsSkipResult::Terminated);
}
BlockSegmentPostingsSkipResult::Terminated
}
/// Advance to the next block.
///
/// Returns false iff there was no remaining blocks.
pub fn advance(&mut self) -> bool {
if self.skip_reader.advance() {
let num_bits = self.skip_reader.doc_num_bits();
let num_consumed_bytes = self.doc_decoder.uncompress_block_sorted(
self.remaining_data.as_ref(),
self.doc_offset,
num_bits,
);
self.remaining_data.advance(num_consumed_bytes);
let tf_num_bits = self.skip_reader.tf_num_bits();
match self.freq_reading_option {
FreqReadingOption::NoFreq => {}
FreqReadingOption::SkipFreq => {
let num_bytes_to_skip = compressed_block_size(tf_num_bits);
self.remaining_data.advance(num_bytes_to_skip);
}
FreqReadingOption::ReadFreq => {
let num_consumed_bytes = self
.freq_decoder
.uncompress_block_unsorted(self.remaining_data.as_ref(), tf_num_bits);
self.remaining_data.advance(num_consumed_bytes);
}
}
// it will be used as the next offset.
self.doc_offset = self.doc_decoder.output(COMPRESSION_BLOCK_SIZE - 1);
true
} else if self.num_vint_docs > 0 {
let num_compressed_bytes = self.doc_decoder.uncompress_vint_sorted(
self.remaining_data.as_ref(),
self.doc_offset,
self.num_vint_docs,
);
self.remaining_data.advance(num_compressed_bytes);
match self.freq_reading_option {
FreqReadingOption::NoFreq | FreqReadingOption::SkipFreq => {}
FreqReadingOption::ReadFreq => {
self.freq_decoder
.uncompress_vint_unsorted(self.remaining_data.as_ref(), self.num_vint_docs);
}
}
self.num_vint_docs = 0;
true
} else {
false
}
}
/// Returns an empty segment postings object
pub fn empty() -> BlockSegmentPostings {
BlockSegmentPostings {
num_vint_docs: 0,
doc_decoder: BlockDecoder::new(),
freq_decoder: BlockDecoder::with_val(1),
freq_reading_option: FreqReadingOption::NoFreq,
doc_offset: 0,
doc_freq: 0,
remaining_data: OwnedRead::new(vec![]),
skip_reader: SkipReader::new(OwnedRead::new(vec![]), IndexRecordOption::Basic),
}
}
}
impl<'b> Streamer<'b> for BlockSegmentPostings {
type Item = &'b [DocId];
fn next(&'b mut self) -> Option<&'b [DocId]> {
if self.advance() {
Some(self.docs())
} else {
None
}
}
}
#[cfg(test)]
mod tests {
use super::BlockSegmentPostings;
use super::BlockSegmentPostingsSkipResult;
use super::SegmentPostings;
use crate::common::HasLen;
use crate::docset::{DocSet, TERMINATED};
use crate::fastfield::DeleteBitSet;
use crate::core::Index;
use crate::docset::DocSet;
use crate::postings::postings::Postings;
use crate::schema::IndexRecordOption;
use crate::schema::Schema;
use crate::schema::Term;
use crate::schema::INDEXED;
use crate::DocId;
use crate::SkipResult;
use tantivy_fst::Streamer;
#[test]
fn test_empty_segment_postings() {
let mut postings = SegmentPostings::empty();
assert_eq!(postings.advance(), TERMINATED);
assert_eq!(postings.advance(), TERMINATED);
assert!(!postings.advance());
assert!(!postings.advance());
assert_eq!(postings.len(), 0);
}
#[test]
fn test_empty_postings_doc_returns_terminated() {
let mut postings = SegmentPostings::empty();
assert_eq!(postings.doc(), TERMINATED);
assert_eq!(postings.advance(), TERMINATED);
#[should_panic(expected = "Have you forgotten to call `.advance()`")]
fn test_panic_if_doc_called_before_advance() {
SegmentPostings::empty().doc();
}
#[test]
fn test_empty_postings_doc_term_freq_returns_0() {
let postings = SegmentPostings::empty();
assert_eq!(postings.term_freq(), 1);
#[should_panic(expected = "Have you forgotten to call `.advance()`")]
fn test_panic_if_freq_called_before_advance() {
SegmentPostings::empty().term_freq();
}
#[test]
fn test_doc_freq() {
let docs = SegmentPostings::create_from_docs(&[0, 2, 10]);
assert_eq!(docs.doc_freq(), 3);
let delete_bitset = DeleteBitSet::for_test(&[2], 12);
assert_eq!(docs.doc_freq_given_deletes(&delete_bitset), 2);
let all_deleted = DeleteBitSet::for_test(&[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], 12);
assert_eq!(docs.doc_freq_given_deletes(&all_deleted), 0);
fn test_empty_block_segment_postings() {
let mut postings = BlockSegmentPostings::empty();
assert!(!postings.advance());
assert_eq!(postings.doc_freq(), 0);
}
#[test]
fn test_block_segment_postings() {
let mut block_segments = build_block_postings(&(0..100_000).collect::<Vec<u32>>());
let mut offset: u32 = 0u32;
// checking that the block before calling advance is empty
assert!(block_segments.docs().is_empty());
// checking that the `doc_freq` is correct
assert_eq!(block_segments.doc_freq(), 100_000);
while let Some(block) = block_segments.next() {
for (i, doc) in block.iter().cloned().enumerate() {
assert_eq!(offset + (i as u32), doc);
}
offset += block.len() as u32;
}
}
#[test]
fn test_skip_right_at_new_block() {
let mut doc_ids = (0..128).collect::<Vec<u32>>();
doc_ids.push(129);
doc_ids.push(130);
{
let block_segments = build_block_postings(&doc_ids);
let mut docset = SegmentPostings::from_block_postings(block_segments, None);
assert_eq!(docset.skip_next(128), SkipResult::OverStep);
assert_eq!(docset.doc(), 129);
assert!(docset.advance());
assert_eq!(docset.doc(), 130);
assert!(!docset.advance());
}
{
let block_segments = build_block_postings(&doc_ids);
let mut docset = SegmentPostings::from_block_postings(block_segments, None);
assert_eq!(docset.skip_next(129), SkipResult::Reached);
assert_eq!(docset.doc(), 129);
assert!(docset.advance());
assert_eq!(docset.doc(), 130);
assert!(!docset.advance());
}
{
let block_segments = build_block_postings(&doc_ids);
let mut docset = SegmentPostings::from_block_postings(block_segments, None);
assert_eq!(docset.skip_next(131), SkipResult::End);
}
}
fn build_block_postings(docs: &[DocId]) -> BlockSegmentPostings {
let mut schema_builder = Schema::builder();
let int_field = schema_builder.add_u64_field("id", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
let mut last_doc = 0u32;
for &doc in docs {
for _ in last_doc..doc {
index_writer.add_document(doc!(int_field=>1u64));
}
index_writer.add_document(doc!(int_field=>0u64));
last_doc = doc + 1;
}
index_writer.commit().unwrap();
let searcher = index.reader().unwrap().searcher();
let segment_reader = searcher.segment_reader(0);
let inverted_index = segment_reader.inverted_index(int_field);
let term = Term::from_field_u64(int_field, 0u64);
let term_info = inverted_index.get_term_info(&term).unwrap();
inverted_index.read_block_postings_from_terminfo(&term_info, IndexRecordOption::Basic)
}
#[test]
fn test_block_segment_postings_skip() {
for i in 0..4 {
let mut block_postings = build_block_postings(&[3]);
assert_eq!(
block_postings.skip_to(i),
BlockSegmentPostingsSkipResult::Success(0u32)
);
assert_eq!(
block_postings.skip_to(i),
BlockSegmentPostingsSkipResult::Terminated
);
}
let mut block_postings = build_block_postings(&[3]);
assert_eq!(
block_postings.skip_to(4u32),
BlockSegmentPostingsSkipResult::Terminated
);
}
#[test]
fn test_block_segment_postings_skip2() {
let mut docs = vec![0];
for i in 0..1300 {
docs.push((i * i / 100) + i);
}
let mut block_postings = build_block_postings(&docs[..]);
for i in vec![0, 424, 10000] {
assert_eq!(
block_postings.skip_to(i),
BlockSegmentPostingsSkipResult::Success(0u32)
);
let docs = block_postings.docs();
assert!(docs[0] <= i);
assert!(docs.last().cloned().unwrap_or(0u32) >= i);
}
assert_eq!(
block_postings.skip_to(100_000),
BlockSegmentPostingsSkipResult::Terminated
);
assert_eq!(
block_postings.skip_to(101_000),
BlockSegmentPostingsSkipResult::Terminated
);
}
#[test]
fn test_reset_block_segment_postings() {
let mut schema_builder = Schema::builder();
let int_field = schema_builder.add_u64_field("id", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
// create two postings list, one containg even number,
// the other containing odd numbers.
for i in 0..6 {
let doc = doc!(int_field=> (i % 2) as u64);
index_writer.add_document(doc);
}
index_writer.commit().unwrap();
let searcher = index.reader().unwrap().searcher();
let segment_reader = searcher.segment_reader(0);
let mut block_segments;
{
let term = Term::from_field_u64(int_field, 0u64);
let inverted_index = segment_reader.inverted_index(int_field);
let term_info = inverted_index.get_term_info(&term).unwrap();
block_segments = inverted_index
.read_block_postings_from_terminfo(&term_info, IndexRecordOption::Basic);
}
assert!(block_segments.advance());
assert_eq!(block_segments.docs(), &[0, 2, 4]);
{
let term = Term::from_field_u64(int_field, 1u64);
let inverted_index = segment_reader.inverted_index(int_field);
let term_info = inverted_index.get_term_info(&term).unwrap();
inverted_index.reset_block_postings_from_terminfo(&term_info, &mut block_segments);
}
assert!(block_segments.advance());
assert_eq!(block_segments.docs(), &[1, 3, 5]);
}
}

114
src/postings/serializer.rs
View File

@@ -3,16 +3,14 @@ use crate::common::{BinarySerializable, VInt};
use crate::common::{CompositeWrite, CountingWriter};
use crate::core::Segment;
use crate::directory::WritePtr;
use crate::fieldnorm::FieldNormReader;
use crate::positions::PositionSerializer;
use crate::postings::compression::{BlockEncoder, VIntEncoder, COMPRESSION_BLOCK_SIZE};
use crate::postings::skip::SkipSerializer;
use crate::query::BM25Weight;
use crate::postings::USE_SKIP_INFO_LIMIT;
use crate::schema::Schema;
use crate::schema::{Field, FieldEntry, FieldType};
use crate::schema::{IndexRecordOption, Schema};
use crate::termdict::{TermDictionaryBuilder, TermOrdinal};
use crate::{DocId, Score};
use std::cmp::Ordering;
use crate::DocId;
use std::io::{self, Write};
/// `InvertedIndexSerializer` is in charge of serializing
@@ -92,22 +90,20 @@ impl InvertedIndexSerializer {
&mut self,
field: Field,
total_num_tokens: u64,
fieldnorm_reader: Option<FieldNormReader>,
) -> io::Result<FieldSerializer<'_>> {
let field_entry: &FieldEntry = self.schema.get_field_entry(field);
let term_dictionary_write = self.terms_write.for_field(field);
let postings_write = self.postings_write.for_field(field);
total_num_tokens.serialize(postings_write)?;
let positions_write = self.positions_write.for_field(field);
let positionsidx_write = self.positionsidx_write.for_field(field);
let field_type: FieldType = (*field_entry.field_type()).clone();
FieldSerializer::create(
&field_type,
total_num_tokens,
term_dictionary_write,
postings_write,
positions_write,
positionsidx_write,
fieldnorm_reader,
)
}
@@ -135,32 +131,26 @@ pub struct FieldSerializer<'a> {
impl<'a> FieldSerializer<'a> {
fn create(
field_type: &FieldType,
total_num_tokens: u64,
term_dictionary_write: &'a mut CountingWriter<WritePtr>,
postings_write: &'a mut CountingWriter<WritePtr>,
positions_write: &'a mut CountingWriter<WritePtr>,
positionsidx_write: &'a mut CountingWriter<WritePtr>,
fieldnorm_reader: Option<FieldNormReader>,
) -> io::Result<FieldSerializer<'a>> {
total_num_tokens.serialize(postings_write)?;
let mode = match field_type {
let (term_freq_enabled, position_enabled): (bool, bool) = match field_type {
FieldType::Str(ref text_options) => {
if let Some(text_indexing_options) = text_options.get_indexing_options() {
text_indexing_options.index_option()
let index_option = text_indexing_options.index_option();
(index_option.has_freq(), index_option.has_positions())
} else {
IndexRecordOption::Basic
(false, false)
}
}
_ => IndexRecordOption::Basic,
_ => (false, false),
};
let term_dictionary_builder = TermDictionaryBuilder::create(term_dictionary_write)?;
let average_fieldnorm = fieldnorm_reader
.as_ref()
.map(|ff_reader| (total_num_tokens as Score / ff_reader.num_docs() as Score))
.unwrap_or(0.0);
let postings_serializer =
PostingsSerializer::new(postings_write, average_fieldnorm, mode, fieldnorm_reader);
let positions_serializer_opt = if mode.has_positions() {
PostingsSerializer::new(postings_write, term_freq_enabled, position_enabled);
let positions_serializer_opt = if position_enabled {
Some(PositionSerializer::new(positions_write, positionsidx_write))
} else {
None
@@ -192,20 +182,18 @@ impl<'a> FieldSerializer<'a> {
/// Starts the postings for a new term.
/// * term - the term. It needs to come after the previous term according
/// to the lexicographical order.
/// * term_doc_freq - return the number of document containing the term.
pub fn new_term(&mut self, term: &[u8], term_doc_freq: u32) -> io::Result<TermOrdinal> {
/// * doc_freq - return the number of document containing the term.
pub fn new_term(&mut self, term: &[u8]) -> io::Result<TermOrdinal> {
assert!(
!self.term_open,
"Called new_term, while the previous term was not closed."
);
self.term_open = true;
self.postings_serializer.clear();
self.current_term_info = self.current_term_info();
self.term_dictionary_builder.insert_key(term)?;
let term_ordinal = self.num_terms;
self.num_terms += 1;
self.postings_serializer.new_term(term_doc_freq);
Ok(term_ordinal)
}
@@ -317,21 +305,15 @@ pub struct PostingsSerializer<W: Write> {
postings_write: Vec<u8>,
skip_write: SkipSerializer,
mode: IndexRecordOption,
fieldnorm_reader: Option<FieldNormReader>,
bm25_weight: Option<BM25Weight>,
avg_fieldnorm: Score, // Average number of term in the field for that segment.
// this value is used to compute the block wand information.
termfreq_enabled: bool,
termfreq_sum_enabled: bool,
}
impl<W: Write> PostingsSerializer<W> {
pub fn new(
write: W,
avg_fieldnorm: Score,
mode: IndexRecordOption,
fieldnorm_reader: Option<FieldNormReader>,
termfreq_enabled: bool,
termfreq_sum_enabled: bool,
) -> PostingsSerializer<W> {
PostingsSerializer {
output_write: CountingWriter::wrap(write),
@@ -343,32 +325,11 @@ impl<W: Write> PostingsSerializer<W> {
skip_write: SkipSerializer::new(),
last_doc_id_encoded: 0u32,
mode,
fieldnorm_reader,
bm25_weight: None,
avg_fieldnorm,
termfreq_enabled,
termfreq_sum_enabled,
}
}
/// Returns the number of documents in the segment currently being serialized.
/// This function may return `None` if there are no fieldnorm for that field.
fn num_docs_in_segment(&self) -> Option<u32> {
self.fieldnorm_reader
.as_ref()
.map(|reader| reader.num_docs())
}
pub fn new_term(&mut self, term_doc_freq: u32) {
if self.mode.has_freq() {
return;
}
self.bm25_weight = self.num_docs_in_segment().map(|num_docs| {
BM25Weight::for_one_term(term_doc_freq as u64, num_docs as u64, self.avg_fieldnorm)
});
}
fn write_block(&mut self) {
{
// encode the doc ids
@@ -381,43 +342,17 @@ impl<W: Write> PostingsSerializer<W> {
// last el block 0, offset block 1,
self.postings_write.extend(block_encoded);
}
if self.mode.has_freq() {
if self.termfreq_enabled {
// encode the term_freqs
let (num_bits, block_encoded): (u8, &[u8]) = self
.block_encoder
.compress_block_unsorted(&self.block.term_freqs());
self.postings_write.extend(block_encoded);
self.skip_write.write_term_freq(num_bits);
if self.mode.has_positions() {
// We serialize the sum of term freqs within the skip information
// in order to navigate through positions.
if self.termfreq_sum_enabled {
let sum_freq = self.block.term_freqs().iter().cloned().sum();
self.skip_write.write_total_term_freq(sum_freq);
}
let mut blockwand_params = (0u8, 0u32);
if let Some(bm25_weight) = self.bm25_weight.as_ref() {
if let Some(fieldnorm_reader) = self.fieldnorm_reader.as_ref() {
let docs = self.block.doc_ids().iter().cloned();
let term_freqs = self.block.term_freqs().iter().cloned();
let fieldnorms = docs.map(|doc| fieldnorm_reader.fieldnorm_id(doc));
blockwand_params = fieldnorms
.zip(term_freqs)
.max_by(
|(left_fieldnorm_id, left_term_freq),
(right_fieldnorm_id, right_term_freq)| {
let left_score =
bm25_weight.tf_factor(*left_fieldnorm_id, *left_term_freq);
let right_score =
bm25_weight.tf_factor(*right_fieldnorm_id, *right_term_freq);
left_score
.partial_cmp(&right_score)
.unwrap_or(Ordering::Equal)
},
)
.unwrap();
}
}
let (fieldnorm_id, term_freq) = blockwand_params;
self.skip_write.write_blockwand_max(fieldnorm_id, term_freq);
}
self.block.clear();
}
@@ -448,7 +383,7 @@ impl<W: Write> PostingsSerializer<W> {
self.postings_write.write_all(block_encoded)?;
}
// ... Idem for term frequencies
if self.mode.has_freq() {
if self.termfreq_enabled {
let block_encoded = self
.block_encoder
.compress_vint_unsorted(self.block.term_freqs());
@@ -456,7 +391,7 @@ impl<W: Write> PostingsSerializer<W> {
}
self.block.clear();
}
if doc_freq >= COMPRESSION_BLOCK_SIZE as u32 {
if doc_freq >= USE_SKIP_INFO_LIMIT {
let skip_data = self.skip_write.data();
VInt(skip_data.len() as u64).serialize(&mut self.output_write)?;
self.output_write.write_all(skip_data)?;
@@ -466,7 +401,6 @@ impl<W: Write> PostingsSerializer<W> {
}
self.skip_write.clear();
self.postings_write.clear();
self.bm25_weight = None;
Ok(())
}

362
src/postings/skip.rs
View File

@@ -1,9 +1,8 @@
use crate::common::{read_u32_vint_no_advance, serialize_vint_u32, BinarySerializable};
use crate::directory::OwnedBytes;
use crate::postings::compression::{compressed_block_size, COMPRESSION_BLOCK_SIZE};
use crate::query::BM25Weight;
use crate::common::BinarySerializable;
use crate::postings::compression::COMPRESSION_BLOCK_SIZE;
use crate::schema::IndexRecordOption;
use crate::{DocId, Score, TERMINATED};
use crate::DocId;
use owned_read::OwnedRead;
pub struct SkipSerializer {
buffer: Vec<u8>,
@@ -40,13 +39,6 @@ impl SkipSerializer {
.expect("Should never fail");
}
pub fn write_blockwand_max(&mut self, fieldnorm_id: u8, term_freq: u32) {
self.buffer.push(fieldnorm_id);
let mut buf = [0u8; 8];
let bytes = serialize_vint_u32(term_freq, &mut buf);
self.buffer.extend_from_slice(bytes);
}
pub fn data(&self) -> &[u8] {
&self.buffer[..]
}
@@ -57,210 +49,81 @@ impl SkipSerializer {
}
}
#[derive(Clone)]
pub(crate) struct SkipReader {
last_doc_in_block: DocId,
pub(crate) last_doc_in_previous_block: DocId,
owned_read: OwnedBytes,
doc: DocId,
owned_read: OwnedRead,
doc_num_bits: u8,
tf_num_bits: u8,
tf_sum: u32,
skip_info: IndexRecordOption,
byte_offset: usize,
remaining_docs: u32, // number of docs remaining, including the
// documents in the current block.
block_info: BlockInfo,
position_offset: u64,
}
#[derive(Clone, Eq, PartialEq, Copy, Debug)]
pub(crate) enum BlockInfo {
BitPacked {
doc_num_bits: u8,
tf_num_bits: u8,
tf_sum: u32,
block_wand_fieldnorm_id: u8,
block_wand_term_freq: u32,
},
VInt {
num_docs: u32,
},
}
impl Default for BlockInfo {
fn default() -> Self {
BlockInfo::VInt { num_docs: 0u32 }
}
}
impl SkipReader {
pub fn new(data: OwnedBytes, doc_freq: u32, skip_info: IndexRecordOption) -> SkipReader {
let mut skip_reader = SkipReader {
last_doc_in_block: if doc_freq >= COMPRESSION_BLOCK_SIZE as u32 {
0
} else {
TERMINATED
},
last_doc_in_previous_block: 0u32,
pub fn new(data: OwnedRead, skip_info: IndexRecordOption) -> SkipReader {
SkipReader {
doc: 0u32,
owned_read: data,
skip_info,
block_info: BlockInfo::VInt { num_docs: doc_freq },
byte_offset: 0,
remaining_docs: doc_freq,
position_offset: 0u64,
};
if doc_freq >= COMPRESSION_BLOCK_SIZE as u32 {
skip_reader.read_block_info();
doc_num_bits: 0u8,
tf_num_bits: 0u8,
tf_sum: 0u32,
}
skip_reader
}
pub fn reset(&mut self, data: OwnedBytes, doc_freq: u32) {
self.last_doc_in_block = if doc_freq >= COMPRESSION_BLOCK_SIZE as u32 {
0
} else {
TERMINATED
};
self.last_doc_in_previous_block = 0u32;
pub fn reset(&mut self, data: OwnedRead) {
self.doc = 0u32;
self.owned_read = data;
self.block_info = BlockInfo::VInt { num_docs: doc_freq };
self.byte_offset = 0;
self.remaining_docs = doc_freq;
self.position_offset = 0u64;
if doc_freq >= COMPRESSION_BLOCK_SIZE as u32 {
self.read_block_info();
}
self.doc_num_bits = 0u8;
self.tf_num_bits = 0u8;
self.tf_sum = 0u32;
}
// Returns the block max score for this block if available.
//
// The block max score is available for all full bitpacked block,
// but no available for the last VInt encoded incomplete block.
pub fn block_max_score(&self, bm25_weight: &BM25Weight) -> Option<Score> {
match self.block_info {
BlockInfo::BitPacked {
block_wand_fieldnorm_id,
block_wand_term_freq,
..
} => Some(bm25_weight.score(block_wand_fieldnorm_id, block_wand_term_freq)),
BlockInfo::VInt { .. } => None,
}
pub fn total_block_len(&self) -> usize {
(self.doc_num_bits + self.tf_num_bits) as usize * COMPRESSION_BLOCK_SIZE / 8
}
pub(crate) fn last_doc_in_block(&self) -> DocId {
self.last_doc_in_block
pub fn doc(&self) -> DocId {
self.doc
}
pub fn position_offset(&self) -> u64 {
self.position_offset
pub fn doc_num_bits(&self) -> u8 {
self.doc_num_bits
}
#[inline(always)]
pub fn byte_offset(&self) -> usize {
self.byte_offset
}
fn read_block_info(&mut self) {
let doc_delta = {
let bytes = self.owned_read.as_slice();
let mut buf = [0; 4];
buf.copy_from_slice(&bytes[..4]);
u32::from_le_bytes(buf)
};
self.last_doc_in_block += doc_delta as DocId;
let doc_num_bits = self.owned_read.as_slice()[4];
match self.skip_info {
IndexRecordOption::Basic => {
self.owned_read.advance(5);
self.block_info = BlockInfo::BitPacked {
doc_num_bits,
tf_num_bits: 0,
tf_sum: 0,
block_wand_fieldnorm_id: 0,
block_wand_term_freq: 0,
};
}
IndexRecordOption::WithFreqs => {
let bytes = self.owned_read.as_slice();
let tf_num_bits = bytes[5];
let block_wand_fieldnorm_id = bytes[6];
let (block_wand_term_freq, num_bytes) = read_u32_vint_no_advance(&bytes[7..]);
self.owned_read.advance(7 + num_bytes);
self.block_info = BlockInfo::BitPacked {
doc_num_bits,
tf_num_bits,
tf_sum: 0,
block_wand_fieldnorm_id,
block_wand_term_freq,
};
}
IndexRecordOption::WithFreqsAndPositions => {
let bytes = self.owned_read.as_slice();
let tf_num_bits = bytes[5];
let tf_sum = {
let mut buf = [0; 4];
buf.copy_from_slice(&bytes[6..10]);
u32::from_le_bytes(buf)
};
let block_wand_fieldnorm_id = bytes[10];
let (block_wand_term_freq, num_bytes) = read_u32_vint_no_advance(&bytes[11..]);
self.owned_read.advance(11 + num_bytes);
self.block_info = BlockInfo::BitPacked {
doc_num_bits,
tf_num_bits,
tf_sum,
block_wand_fieldnorm_id,
block_wand_term_freq,
};
}
}
}
pub fn block_info(&self) -> BlockInfo {
self.block_info
}
/// Advance the skip reader to the block that may contain the target.
/// Number of bits used to encode term frequencies
///
/// If the target is larger than all documents, the skip_reader
/// then advance to the last Variable In block.
pub fn seek(&mut self, target: DocId) -> bool {
if self.last_doc_in_block() >= target {
return false;
}
loop {
self.advance();
if self.last_doc_in_block() >= target {
return true;
}
}
/// 0 if term frequencies are not enabled.
pub fn tf_num_bits(&self) -> u8 {
self.tf_num_bits
}
pub fn advance(&mut self) {
match self.block_info {
BlockInfo::BitPacked {
doc_num_bits,
tf_num_bits,
tf_sum,
..
} => {
self.remaining_docs -= COMPRESSION_BLOCK_SIZE as u32;
self.byte_offset += compressed_block_size(doc_num_bits + tf_num_bits);
self.position_offset += tf_sum as u64;
}
BlockInfo::VInt { num_docs } => {
debug_assert_eq!(num_docs, self.remaining_docs);
self.remaining_docs = 0;
self.byte_offset = std::usize::MAX;
}
}
self.last_doc_in_previous_block = self.last_doc_in_block;
if self.remaining_docs >= COMPRESSION_BLOCK_SIZE as u32 {
self.read_block_info();
pub fn tf_sum(&self) -> u32 {
self.tf_sum
}
pub fn advance(&mut self) -> bool {
if self.owned_read.as_ref().is_empty() {
false
} else {
self.last_doc_in_block = TERMINATED;
self.block_info = BlockInfo::VInt {
num_docs: self.remaining_docs,
};
let doc_delta = u32::deserialize(&mut self.owned_read).expect("Skip data corrupted");
self.doc += doc_delta as DocId;
self.doc_num_bits = self.owned_read.get(0);
match self.skip_info {
IndexRecordOption::Basic => {
self.owned_read.advance(1);
}
IndexRecordOption::WithFreqs => {
self.tf_num_bits = self.owned_read.get(1);
self.owned_read.advance(2);
}
IndexRecordOption::WithFreqsAndPositions => {
self.tf_num_bits = self.owned_read.get(1);
self.owned_read.advance(2);
self.tf_sum =
u32::deserialize(&mut self.owned_read).expect("Failed reading tf_sum");
}
}
true
}
}
}
@@ -268,11 +131,9 @@ impl SkipReader {
#[cfg(test)]
mod tests {
use super::BlockInfo;
use super::IndexRecordOption;
use super::{SkipReader, SkipSerializer};
use crate::directory::OwnedBytes;
use crate::postings::compression::COMPRESSION_BLOCK_SIZE;
use owned_read::OwnedRead;
#[test]
fn test_skip_with_freq() {
@@ -280,44 +141,20 @@ mod tests {
let mut skip_serializer = SkipSerializer::new();
skip_serializer.write_doc(1u32, 2u8);
skip_serializer.write_term_freq(3u8);
skip_serializer.write_blockwand_max(13u8, 3u32);
skip_serializer.write_doc(5u32, 5u8);
skip_serializer.write_term_freq(2u8);
skip_serializer.write_blockwand_max(8u8, 2u32);
skip_serializer.data().to_owned()
};
let doc_freq = 3u32 + (COMPRESSION_BLOCK_SIZE * 2) as u32;
let mut skip_reader =
SkipReader::new(OwnedBytes::new(buf), doc_freq, IndexRecordOption::WithFreqs);
assert_eq!(skip_reader.last_doc_in_block(), 1u32);
assert_eq!(
skip_reader.block_info,
BlockInfo::BitPacked {
doc_num_bits: 2u8,
tf_num_bits: 3u8,
tf_sum: 0,
block_wand_fieldnorm_id: 13,
block_wand_term_freq: 3
}
);
skip_reader.advance();
assert_eq!(skip_reader.last_doc_in_block(), 5u32);
assert_eq!(
skip_reader.block_info(),
BlockInfo::BitPacked {
doc_num_bits: 5u8,
tf_num_bits: 2u8,
tf_sum: 0,
block_wand_fieldnorm_id: 8,
block_wand_term_freq: 2
}
);
skip_reader.advance();
assert_eq!(skip_reader.block_info(), BlockInfo::VInt { num_docs: 3u32 });
skip_reader.advance();
assert_eq!(skip_reader.block_info(), BlockInfo::VInt { num_docs: 0u32 });
skip_reader.advance();
assert_eq!(skip_reader.block_info(), BlockInfo::VInt { num_docs: 0u32 });
let mut skip_reader = SkipReader::new(OwnedRead::new(buf), IndexRecordOption::WithFreqs);
assert!(skip_reader.advance());
assert_eq!(skip_reader.doc(), 1u32);
assert_eq!(skip_reader.doc_num_bits(), 2u8);
assert_eq!(skip_reader.tf_num_bits(), 3u8);
assert!(skip_reader.advance());
assert_eq!(skip_reader.doc(), 5u32);
assert_eq!(skip_reader.doc_num_bits(), 5u8);
assert_eq!(skip_reader.tf_num_bits(), 2u8);
assert!(!skip_reader.advance());
}
#[test]
@@ -328,62 +165,13 @@ mod tests {
skip_serializer.write_doc(5u32, 5u8);
skip_serializer.data().to_owned()
};
let doc_freq = 3u32 + (COMPRESSION_BLOCK_SIZE * 2) as u32;
let mut skip_reader =
SkipReader::new(OwnedBytes::new(buf), doc_freq, IndexRecordOption::Basic);
assert_eq!(skip_reader.last_doc_in_block(), 1u32);
assert_eq!(
skip_reader.block_info(),
BlockInfo::BitPacked {
doc_num_bits: 2u8,
tf_num_bits: 0,
tf_sum: 0u32,
block_wand_fieldnorm_id: 0,
block_wand_term_freq: 0
}
);
skip_reader.advance();
assert_eq!(skip_reader.last_doc_in_block(), 5u32);
assert_eq!(
skip_reader.block_info(),
BlockInfo::BitPacked {
doc_num_bits: 5u8,
tf_num_bits: 0,
tf_sum: 0u32,
block_wand_fieldnorm_id: 0,
block_wand_term_freq: 0
}
);
skip_reader.advance();
assert_eq!(skip_reader.block_info(), BlockInfo::VInt { num_docs: 3u32 });
skip_reader.advance();
assert_eq!(skip_reader.block_info(), BlockInfo::VInt { num_docs: 0u32 });
skip_reader.advance();
assert_eq!(skip_reader.block_info(), BlockInfo::VInt { num_docs: 0u32 });
}
#[test]
fn test_skip_multiple_of_block_size() {
let buf = {
let mut skip_serializer = SkipSerializer::new();
skip_serializer.write_doc(1u32, 2u8);
skip_serializer.data().to_owned()
};
let doc_freq = COMPRESSION_BLOCK_SIZE as u32;
let mut skip_reader =
SkipReader::new(OwnedBytes::new(buf), doc_freq, IndexRecordOption::Basic);
assert_eq!(skip_reader.last_doc_in_block(), 1u32);
assert_eq!(
skip_reader.block_info(),
BlockInfo::BitPacked {
doc_num_bits: 2u8,
tf_num_bits: 0,
tf_sum: 0u32,
block_wand_fieldnorm_id: 0,
block_wand_term_freq: 0
}
);
skip_reader.advance();
assert_eq!(skip_reader.block_info(), BlockInfo::VInt { num_docs: 0u32 });
let mut skip_reader = SkipReader::new(OwnedRead::new(buf), IndexRecordOption::Basic);
assert!(skip_reader.advance());
assert_eq!(skip_reader.doc(), 1u32);
assert_eq!(skip_reader.doc_num_bits(), 2u8);
assert!(skip_reader.advance());
assert_eq!(skip_reader.doc(), 5u32);
assert_eq!(skip_reader.doc_num_bits(), 5u8);
assert!(!skip_reader.advance());
}
}

6
src/postings/stacker/expull.rs
View File

@@ -206,8 +206,8 @@ mod tests {
fn test_stack_long() {
let mut heap = MemoryArena::new();
let mut stack = ExpUnrolledLinkedList::new();
let data: Vec<u32> = (0..100).collect();
for &el in &data {
let source: Vec<u32> = (0..100).collect();
for &el in &source {
assert!(stack
.writer(&mut heap)
.write_u32::<LittleEndian>(el)
@@ -221,7 +221,7 @@ mod tests {
result.push(LittleEndian::read_u32(&remaining[..4]));
remaining = &remaining[4..];
}
assert_eq!(&result[..], &data[..]);
assert_eq!(&result[..], &source[..]);
}
#[test]

4
src/postings/stacker/memory_arena.rs
View File

@@ -106,8 +106,8 @@ impl MemoryArena {
///
/// Internally, it counts a number of `1MB` pages
/// and therefore delivers an upperbound.
pub fn mem_usage(&self) -> usize {
self.pages.len() * PAGE_SIZE
pub fn mem_usage(&self) -> u64 {
(self.pages.len() as u64) * (PAGE_SIZE as u64)
}
pub fn write_at<Item: Copy + 'static>(&mut self, addr: Addr, val: Item) {

Some files were not shown because too many files have changed in this diff Show More