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:wasm
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:issue/554
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

305 Commits
wasm ... issue/554

Author SHA1 Message Date
Paul Masurel
94c73363e4 added unit test 2019-06-02 14:53:03 +09:00
Paul Masurel
96f194635f Trying to address #546 2019-05-29 09:17:41 +09:00
Paul Masurel
444662485f Remove mut in add_document and delete_term. Made stamper ordering rel… (#551) 
* Remove mut in add_document and delete_term. Made stamper ordering relaxed.

* Made batch operations &mut self -> &self

* Added example
 2019-05-28 10:26:00 +09:00
Stephen Carman
943c25d0f8 Make IndexMeta public (#553) 2019-05-28 09:27:49 +09:00
Paul Masurel
5c0b2a4579 Merge branch 'stamper_refactor' 2019-05-08 10:02:02 +09:00
Paul Masurel
9870a9258d Removed the mutex implementation of AtomicU64. 
Fixed comment
 2019-05-08 09:59:28 +09:00
Paul Masurel
7102b363f5 Fix build 2019-05-05 14:19:54 +09:00
Paul Masurel
66b4615e4e Issue/542 (#543) 
* Closes 542.

Fast fields are all loaded when the segment reader is created.
 2019-05-05 13:52:43 +09:00
petr-tik
da46913839 Merge branch 'master' into stamper_refactor 2019-04-30 22:28:48 +01:00
Paul Masurel
3df037961f Added more info to fast fields. 2019-04-30 13:14:01 +09:00
petr-tik
8ffae47854 Addressed code review 
moved Opstamp to top-level namespace, added a docstring

Corrected minor typos/whitespace
 2019-04-29 21:23:28 +01:00
petr-tik
1a90a1f3b0 Merge branch 'master' of github.com:tantivy-search/tantivy into stamper_refactor 2019-04-26 08:47:12 +01:00
Paul Masurel
dac50c6aeb Dds merged (#539) 
* add ascii folding support

* Minor change and added Changelog.

* add additional tests

* Add tests for ascii folding (#533)

* first tests for ascii folding

* use a `RawTokenizer` for tokens using punctuation

* add test for all (?) folding, inspired by Lucene

* Simplification of the unit test code
 2019-04-26 10:25:08 +09:00
Paul Masurel
31b22c5acc Added logging when token is dropped. (#538) 2019-04-26 09:23:28 +09:00
petr-tik
8e50921363 Tidied up the Stamper module and upgraded to a 1.34 dependency 
Added stamper.revert method to be used for rollback - rolling back to a previous
commit in case of deleting all documents or rolling operations back should reset
the stamper as well

Added type alias for Opstamp - helps code readibility instead of seeing u64
returned by functions.

Moved to AtomicU64 on stable rust (since 1.34) - where possible use standard
library interfaces.
 2019-04-24 20:46:28 +01:00
Paul Masurel
96a4f503ec Closes #526 (#535) 2019-04-24 20:59:48 +09:00
Paul Masurel
9df288b0c9 Merge branch 'master' of github.com:tantivy-search/tantivy 2019-04-24 12:31:47 +09:00
Paul Masurel
b7c2d0de97 Clippy2 (#534) 
* Clippy comments

Clippy complaints that about the cast of &[u32] to a *const __m128i,
because of the lack of alignment constraints.

This commit passes the OutputBuffer object (which enforces proper
    alignment) instead of `&[u32]`.

* Clippy. Block alignment

* Code simplification

* Added comment. Code simplification

* Removed the extraneous freq block len hack.
 2019-04-24 12:31:32 +09:00
Paul Masurel
62445e0ec8 Merge branch 'master' of github.com:tantivy-search/tantivy 2019-04-23 09:55:55 +09:00
Paul Masurel
a228825462 Clippy comments (#532) 
Clippy complaints that about the cast of &[u32] to a *const __m128i,
because of the lack of alignment constraints.

This commit passes the OutputBuffer object (which enforces proper
    alignment) instead of `&[u32]`.
 2019-04-23 09:54:02 +09:00
Paul Masurel
d3eabd14bc Clippy comments 
Clippy complaints that about the cast of &[u32] to a *const __m128i,
because of the lack of alignment constraints.

This commit passes the OutputBuffer object (which enforces proper
    alignment) instead of `&[u32]`.
 2019-04-22 11:16:21 +09:00
petr-tik
c967031d21 Delete files from target/ dir to avoid caching them on CI (#531) 
* Delete files from target/ dir to avoid caching them on CI

idea from here https://github.com/rust-lang/cargo/issues/5885#issuecomment-432723546

* Delete examples
 2019-04-21 08:02:27 +09:00
Paul Masurel
d823163d52 Closes #527. (#529) 
Fixing the bug that affects the result of `query.count()` in presence of
deletes.
 2019-04-19 09:19:50 +09:00
Paul Masurel
c4f59f202d Bumped combine version 2019-04-11 08:33:56 +09:00
Paul Masurel
acd29b535d Fix comment 2019-04-02 10:05:14 +09:00
Panagiotis Ktistakis
2cd31bcda2 Fix non english stemmers (#521) 2019-03-27 08:54:16 +09:00
Paul Masurel
99870de55c 0.10.0-dev 2019-03-25 08:58:26 +09:00
Paul Masurel
cad2d91845 Disabled tests for android 2019-03-24 22:58:46 +09:00
Paul Masurel
79f3cd6cf4 Added instructions to update 2019-03-24 09:10:31 +09:00
Paul Masurel
e3abb4481b broken link 2019-03-22 09:58:28 +09:00
Paul Masurel
bfa61d2f2f Added patreon button 2019-03-22 09:51:00 +09:00
Paul Masurel
6c0e621fdb Added bench info in README 2019-03-21 09:35:04 +09:00
Paul Masurel
a8cc5208f1 Linear simd (#519) 
* linear simd search within block
 2019-03-20 22:10:05 +09:00
Paul Masurel
83eb0d0cb7 Disabling tests on Android 2019-03-20 10:24:17 +09:00
Paul Masurel
ee6e273365 cleanup for nodefaultfeatures 2019-03-20 10:04:42 +09:00
Paul Masurel
6ea34b3d53 Fix version 2019-03-20 09:39:24 +09:00
Paul Masurel
22cf1004bd Reenabled test on android 2019-03-20 08:54:52 +09:00
Paul Masurel
5768d93171 Rename try to attempt as try is becoming a keyword in rust 2019-03-20 08:54:19 +09:00
Paul Masurel
663dd89c05 Feature/reader (#517) 
Adding IndexReader to the API. Making it possible to watch for changes.

* Closes #500
 2019-03-20 08:39:22 +09:00
barrotsteindev
a934577168 WIP: date field (#487) 
* initial version, still a work in progress

* remove redudant or

* add chrono::DateTime and index i64

* add more tests

* fix tests

* pass DateTime by ptr

* remove println!

* document query_parser rfc 3339 date support

* added some more docs about implementation to schema.rs

* enforce DateTime is UTC, and re-export chrono

* added DateField to changelog

* fixed conflict

* use INDEXED instead of INT_INDEXED for date fields
 2019-03-15 22:10:37 +09:00
Paul Masurel
94f1885334 Issue/513 (#514) 
* Closes #513

* Clean up and doc

* Updated changelog
 2019-03-07 09:39:30 +09:00
Jonathan Fok kan
2ccfdb97b5 WIP: compiling to wasm (#512) 
* First work to enable compile to wasm

* Added back fst-regex/mmap to mmap feature

* Removed fst-regex. Forced uuid version 0.7.2.
 2019-03-06 10:40:54 +09:00
Paul Masurel
e67883138d Cargo fmt 2019-03-06 10:31:00 +09:00
Paul Masurel
f5c65f1f60 Added comment on the constructor fo TopDocSByField 2019-03-06 10:30:37 +09:00
Mauri de Souza Nunes
ec73a9a284 Remove note about panicking in get_field docs (#503) 
Since get_field rely on calling get on the underlying InnerSchema HashMap
it shouldn't fail if the field was not found, it simply returns None.
 2019-02-28 09:23:00 +09:00
Thomas Schaller
a814a31f1e Remove semicolon from doc! expansion (#509) 2019-02-28 09:20:43 +09:00
Paul Masurel
9acadb3756 Code cleaning 2019-02-26 10:50:36 +09:00
Paul Masurel
774fcecf23 cargo fmt 2019-02-26 10:44:59 +09:00
Paul Masurel
27c9fa6028 Jannickj prove bug with facets (#508) 
* prove bug with facets

* Closing #505

Introduce a term id in the TermHashMap
 2019-02-25 22:33:17 +09:00
Paul Masurel
fdefea9e26 Removed path reference to tantivy-fst 2019-02-23 10:42:44 +09:00
Paul Masurel
b422f9c389 Partially addresses #500 (#502) 
Using `tantivy_fst`. Storing `Weak<Mmap>` in the Mmap cache.
 2019-02-23 10:33:59 +09:00
petr-tik
9451fd5b09 MsQueue to channel (#495) 
* Format

Made the docstring consistent
remove empty line

* Move matches to dev deps

* Replace MsQueue with an unbounded crossbeam-channel

Questions:
queue.push ignores Result return

How to test pop() calls, if they block

* Format

Made the docstring consistent
remove empty line

* Unwrap the Result of queue.pop

* Addressed Paul's review

wrap the Result-returning send call with expect()

implemented the test not to fail after popping from empty queue

removed references to the Michael-Scott Queue

formatted
 2019-02-23 09:06:50 +09:00
Jason Goldberger
788b3803d9 updated changelog (#501) 
* updated changelog

* Update CHANGELOG.md

* Update CHANGELOG.md
 2019-02-19 00:25:18 +09:00
Paul Masurel
5b11228083 Merge branch 'master' of github.com:tantivy-search/tantivy 2019-02-15 08:30:55 +09:00
Paul Masurel
515adff644 Merge branch 'hotfix/0.8.2' 2019-02-15 08:30:27 +09:00
Paul Masurel
e70a45426a 0.8.2 release 
Backporting a fix for non x86_64 platforms
 2019-02-14 09:16:27 +09:00
Jason Goldberger
e14701e9cd Add grouped operations (#493) 
* [WIP] added UserOperation enum, added IndexWriter.run, and added MultiStamp

* removed MultiStamp in favor of std::ops::Range

* changed IndexWriter::run to return u64, Stamper::stamps to return a Range, added tests, and added docs

* changed delete_cursor skipping to use first operation's opstamp vice last. change index_writer test to use 1 thread

* added test for order batch of operations

* added a test comment
 2019-02-14 08:56:01 +09:00
Paul Masurel
45e62d4329 Code simplification and adding comments 2019-02-06 10:05:15 +09:00
petr-tik
76d2b4dab6 Add integer range search example (#490) 
Copied and simplified the example in the range_query mod
 2019-02-05 23:34:06 +01:00
Paul Masurel
04e9606638 simplification of positions 2019-02-05 15:36:13 +01:00
Paul Masurel
a5c57ebbd9 Positions simplification 2019-02-05 14:50:51 +01:00
Paul Masurel
96eaa5bc63 Positions 2019-02-05 14:50:16 +01:00
Paul Masurel
f1d30ab196 fastfield reader fix 2019-02-05 14:10:16 +01:00
Paul Masurel
4507df9255 Closes #461 (#489) 
Multivalued fast field uses `u64` indexes.
 2019-02-04 13:24:00 +01:00
Paul Masurel
e8625548b7 Closes #461 (#488) 
Multivalued fast field uses `u64` indexes.
 2019-02-04 13:20:20 +01:00
Paul Masurel
50ed6fb534 Code cleanup 
Fixed compilation without the mmap directory
 2019-02-05 12:39:30 +01:00
Panagiotis Ktistakis
76609deadf Add Greek stemmer (#486) 2019-02-01 06:30:49 +01:00
Paul Masurel
749e62c40b renamed 2019-01-30 16:29:17 +01:00
Paul Masurel
259ce567d1 Using linear search 2019-01-29 15:59:24 +01:00
Paul Masurel
4c93b096eb Rustfmt 2019-01-29 11:45:30 +01:00
Paul Masurel
6a547b0b5f Issue/483 (#484) 
* Downcast_ref

* fixing unit test
 2019-01-28 11:43:42 +01:00
Paul Masurel
e99d1a2355 Better exponential search 2019-01-29 11:29:17 +01:00
Paul Masurel
c7bddc5fe3 Inlined exponential search 2019-01-28 17:28:07 +01:00
Paul Masurel
7b97dde335 Clippy + cargo fmt 2019-01-28 12:37:55 +01:00
Paul Masurel
644b4bd0a1 Issue/468b (#482) 
* Moving lock to directory/

* added fs2

* doc

* Using fs2 for locking

* Added unit test

* Fixed error message related unit test

* Fixing location of import
 2019-01-27 12:32:21 +01:00
Paul Masurel
bf94fd77db Issue/471 (#481) 
* Closes 471

Removing writing_segments in the segment manager as it is now useless.
Removing the target merged segment id as it is useless as well.

* RAII for tracking which segment is in merge.

Closes #471

* fmt

* Using Inventory::default().
 2019-01-27 12:18:59 +09:00
Paul Masurel
097eaf4aa6 impl Future as a result of merges 2019-01-28 03:56:43 +01:00
Paul Masurel
1fd46c1e9b Clippy 2019-01-28 03:46:23 +01:00
Paul Masurel
2fb219d017 Changelog 2019-01-24 09:12:07 +09:00
Paul Masurel
63b593bd0a Lower RAM usage in tests. 2019-01-24 09:10:38 +09:00
Paul Masurel
286bb75a0c Updated changelog 2019-01-24 09:03:58 +09:00
barrotsteindev
222b7f2580 Tantivy-288 (#472) 
* add unit test

* improved test

* added SegmentManager#remove_empty_segments

* update old tests for new behaviour

* cleaner filter for empty segments

* PR adjustments

* rename x in closures

* simplify assert_eq!(vec.len(), 0)

* wait_merging_threads

* acquire searchers

* add comments to test

* rebased on latest master

* harden test

* fix merger#test_merge_multivalued_int_fields_all_deleted test
 2019-01-24 08:58:56 +09:00
pentlander
5292e78860 Allow stemmers in languages other than English (#473) 
Allow users to create stemmers for languages other than English. Add a
default stemmer for English.
 2019-01-23 22:24:32 +09:00
Paul Masurel
c0cc6aac83 Updated changelog 2019-01-23 22:22:34 +09:00
Paul Masurel
0b0bf59a32 Allow stemmers in languages other than English (#478) 
Allow users to create stemmers for languages other than English. Add a
default stemmer for English.

Closes #478
 2019-01-23 22:21:00 +09:00
Paul Masurel
74f70a5c2c 32bits platforms 2019-01-23 13:21:31 +09:00
Paul Masurel
1acfb2ebb5 cargo fmt 2019-01-23 10:21:39 +09:00
Paul Masurel
4dfd091e67 Bumped version to 0.8.2-dev 2019-01-23 10:20:59 +09:00
Paul Masurel
8eba4ab807 Merge branch 'hotfix-476' 2019-01-23 10:20:33 +09:00
Paul Masurel
5e8e03882b Merge branch 'bug/476' 2019-01-23 10:18:27 +09:00
Paul Masurel
7df3260a15 Version bump 2019-01-23 10:13:18 +09:00
Paul Masurel
176f67a266 Refactoring 2019-01-23 10:06:40 +09:00
Paul Masurel
19babff849 Closes #476 2019-01-23 10:06:39 +09:00
Paul Masurel
bf2576adf9 Added a broken unit test 2019-01-23 10:04:27 +09:00
Paul Masurel
0e8fcd5727 Plastic surgery 2019-01-19 23:13:27 +09:00
Paul Masurel
f745c83bb7 Closes 466. Removing mentions of the chain collector. (#467) 2019-01-16 10:28:19 +09:00
Paul Masurel
ffb16d9103 More efficient indexing (#463) 
* Using unrolled u32 VInt and caching Vec s

* cargo fmt

* Exposing a io::Write in the Expull thing

* expull as a writer. clippy + format

* inline the first block

* simplified -if let Some-

* vint reader iterator

* blop
 2019-01-13 14:51:18 +09:00
Paul Masurel
98ca703daa More efficient indexing (#462) 
* Using unrolled u32 VInt and caching Vec s

* cargo fmt

* Exposing a io::Write in the Expull thing

* expull as a writer. clippy + format

* inline the first block

* simplified -if let Some-

* vint reader iterator
 2019-01-13 14:41:56 +09:00
Paul Masurel
b9d25cda5d Using LittleEndian explicitely 2019-01-08 12:41:58 +09:00
Paul Masurel
beb4289ec2 Less unsafe 2019-01-08 00:48:14 +09:00
Andrew Banchich
bdd72e4683 Update README.md (#459) 
Fix Elasticsearch spelling
 2018-12-27 07:26:49 +09:00
Paul Masurel
45c3cd19be Fixing README: git clone https... 2018-12-26 21:13:33 +09:00
Paul Masurel
b8241c5603 0.8.0 2018-12-26 10:18:34 +09:00
Paul Masurel
a4745151c0 Version to 0.8 2018-12-26 10:11:06 +09:00
Paul Masurel
e2ce326a8c Merge branch 'issue/457' 2018-12-18 10:35:01 +09:00
Paul Masurel
bb21d12a70 Bumping version 2018-12-18 10:14:12 +09:00
Paul Masurel
4565aba62a Added unit test for exponential search 2018-12-18 09:24:31 +09:00
Paul Masurel
545a7ec8dd Closes #457 2018-12-18 09:18:46 +09:00
Paul Masurel
e68775d71c Format and update murmurhash32 version 2018-12-17 19:12:38 +09:00
Paul Masurel
dcc92d287e Facet remove unsafe (#456) 
* Removing some unsafe

* Removing some unsafe (2)

* Remove murmurhash
 2018-12-17 19:08:48 +09:00
Paul Masurel
b48f81c051 Removing unsafe from bitpacking code (#455) 2018-12-17 19:06:37 +09:00
Paul Masurel
a3042e956b Facet remove unsafe (#454) 
* Removing some unsafe

* Removing some unsafe (2)
 2018-12-17 09:31:09 +09:00
dependabot[bot]
1fa10f0a0b Update itertools requirement from 0.7 to 0.8 (#453) 
Updates the requirements on [itertools](https://github.com/bluss/rust-itertools) to permit the latest version.
- [Release notes](https://github.com/bluss/rust-itertools/releases)
- [Commits](https://github.com/bluss/rust-itertools/commits/0.8.0)

Signed-off-by: dependabot[bot] <support@dependabot.com>
 2018-12-17 09:28:36 +09:00
Paul Masurel
279a9eb5e3 Closes #449 (#450) 
Clippy working on stable.
Clippy warnings addressed
 2018-12-10 12:20:59 +09:00
fdb-hiroshima
21a24672d8 Add accessors for Snippet and HighlightSection (#448) 
* Add accessors for Snippet and HighlightSection

And add an example of custom highlighter

* Remove inline(always) and unnecessary empty lines
 2018-12-02 18:00:16 +09:00
dependabot[bot]
a3f1fbaae6 Update scoped-pool requirement from 0.1 to 1.0 (#447) 
Updates the requirements on [scoped-pool](https://github.com/reem/rust-scoped-pool) to permit the latest version.
- [Release notes](https://github.com/reem/rust-scoped-pool/releases)
- [Commits](https://github.com/reem/rust-scoped-pool/commits/1.0.0)

Signed-off-by: dependabot[bot] <support@dependabot.com>
 2018-12-01 13:54:59 +09:00
Paul Masurel
a6e767c877 Cargo fmt 2018-11-30 22:52:45 +09:00
Paul Masurel
6af0488dbe Executor made sorted 2018-11-30 22:52:26 +09:00
Paul Masurel
07d87e154b Collector refactoring and multithreaded search (#437) 
* Split Collector into an overall Collector and a per-segment SegmentCollector. Precursor to cross-segment parallelism, and as a side benefit cleans up any per-segment fields from being Option<T> to just T.

* Attempt to add MultiCollector back

* working. Chained collector is broken though

* Fix chained collector

* Fix test

* Make Weight Send+Sync for parallelization purposes

* Expose parameters of RangeQuery for external usage

* Removed &mut self

* fixing tests

* Restored TestCollectors

* blop

* multicollector working

* chained collector working

* test broken

* fixing unit test

* blop

* blop

* Blop

* simplifying APi

* blop

* better syntax

* Simplifying top_collector

* refactoring

* blop

* Sync with master

* Added multithread search

* Collector refactoring

* Schema::builder

* CR and rustdoc

* CR comments

* blop

* Added an executor

* Sorted the segment readers in the searcher

* Update searcher.rs

* Fixed unit testst

* changed the place where we have the sort-segment-by-count heuristic

* using crossbeam::channel

* inlining

* Comments about panics propagating

* Added unit test for executor panicking

* Readded default

* Removed Default impl

* Added unit test for executor
 2018-11-30 22:46:59 +09:00
Paul Masurel
8b0b0133dd Importing crossbeam_channel from crossbeam reexport. 2018-11-19 09:19:28 +09:00
dependabot[bot]
7b9752f897 Update crossbeam-channel requirement from 0.2 to 0.3 (#436) 
* Update crossbeam-channel requirement from 0.2 to 0.3

Updates the requirements on [crossbeam-channel](https://github.com/crossbeam-rs/crossbeam-channel) to permit the latest version.
- [Release notes](https://github.com/crossbeam-rs/crossbeam-channel/releases)
- [Changelog](https://github.com/crossbeam-rs/crossbeam-channel/blob/master/CHANGELOG.md)
- [Commits](https://github.com/crossbeam-rs/crossbeam-channel/commits/v0.3.0)

Signed-off-by: dependabot[bot] <support@dependabot.com>

* fixing build
 2018-11-16 14:26:59 +09:00
dependabot[bot]
c92f41aea8 Update rand requirement from 0.5 to 0.6 (#440) 
* Update rand requirement from 0.5 to 0.6

Updates the requirements on [rand](https://github.com/rust-random/rand) to permit the latest version.
- [Release notes](https://github.com/rust-random/rand/releases)
- [Changelog](https://github.com/rust-random/rand/blob/master/CHANGELOG.md)
- [Commits](https://github.com/rust-random/rand/commits)

Signed-off-by: dependabot[bot] <support@dependabot.com>

* Updating rand.
 2018-11-16 12:38:01 +09:00
Do Duy
dea16f1d9d Derive Clone for QueryParser (#442) 2018-11-15 18:45:40 +09:00
dependabot[bot]
236cfbec08 Update crossbeam requirement from 0.4 to 0.5 (#438) 
Updates the requirements on [crossbeam](https://github.com/crossbeam-rs/crossbeam) to permit the latest version.
- [Release notes](https://github.com/crossbeam-rs/crossbeam/releases)
- [Changelog](https://github.com/crossbeam-rs/crossbeam/blob/master/CHANGELOG.md)
- [Commits](https://github.com/crossbeam-rs/crossbeam/commits/crossbeam-0.5.0)

Signed-off-by: dependabot[bot] <support@dependabot.com>
 2018-11-15 06:16:22 +09:00
Paul Masurel
edcafb69bb Fixed benches 2018-11-10 17:04:29 -08:00
Paul Masurel
14908479d5 Release 0.7.1 2018-11-02 17:56:25 +09:00
Dru Sellers
ab4593eeb7 Adds open_or_create method (#428) 
* Change the semantic of Index::create_in_dir.

It should return an error if the directory already contains an Index.

* Index::open_or_create is working

* additional test

* Checking that schema matches on open_or_create.

Simplifying unit tests.

* simplifying Eq
 2018-10-31 08:36:39 +09:00
Dru Sellers
e75bb1d6a1 Fix NGram processing of non-ascii characters (#430) 
* A working version

* optimize the ngram parsing

* Decoding codepoint only once.

* Closes #429

* using leading_zeros to make code less cryptic

* lookup in a table
 2018-10-31 08:35:27 +09:00
dependabot[bot]
63b9d62237 Update base64 requirement from 0.9.1 to 0.10.0 (#433) 
Updates the requirements on [base64](https://github.com/alicemaz/rust-base64) to permit the latest version.
- [Release notes](https://github.com/alicemaz/rust-base64/releases)
- [Changelog](https://github.com/alicemaz/rust-base64/blob/master/RELEASE-NOTES.md)
- [Commits](https://github.com/alicemaz/rust-base64/commits/v0.10.0)

Signed-off-by: dependabot[bot] <support@dependabot.com>
 2018-10-31 08:34:44 +09:00
Jason Wolfe
0098e3d428 Compute space usage of a Searcher / SegmentReader / CompositeFile (#282) 
* Compute space usage of a Searcher / SegmentReader / CompositeFile

* Fix typo

* Add serde Serialize/Deserialize for all the SpaceUsage structs

* Fix indexing

* Public methods for consuming space usage information

* #281: Add a space usage method that takes a SegmentComponent to support code that is unaware of particular segment components, and to make it more likely to update methods when a new component type is added.

* Add support for space usage computation of positions skip index file (#281)

* Add some tests for space usage computation (#281)
 2018-10-15 09:04:36 +09:00
Konstantin Gribov
69d5e4b9b1 Added proper references for Apache Lucene & Solr (#432) 
Also, added links to websites for Lucene, Solr & ElasticSearch
 2018-10-12 08:46:07 +09:00
Paul Masurel
e0cdd3114d Fixing README (#427) 
Closes #424.
 2018-09-17 08:52:29 +09:00
Paul Masurel
f32b4a2ebe Removing release build from ci, disabling lto (#425) 2018-09-17 06:41:40 +09:00
Paul Masurel
6ff60b8ed8 Fixing README (#426) 2018-09-17 06:20:44 +09:00
Paul Masurel
8da28fb6cf Added iml filewq 2018-09-16 13:26:54 +09:00
Paul Masurel
0df2a221da Bump version pre-release 2018-09-16 13:24:14 +09:00
Paul Masurel
5449ec3c11 Snippet term score (#423) 2018-09-16 10:21:02 +09:00
Paul Masurel
10f6c07c53 Clippy (#422) 
* Cargo Format
* Clippy
 2018-09-15 20:20:22 +09:00
Paul Masurel
06e7bd18e7 Clippy (#421) 
* Cargo Format

* Clippy

* bugfix

* still clippy stuff

* clippy step 2
 2018-09-15 14:56:14 +09:00
Paul Masurel
37e4280c0a Cargo Format (#420) 2018-09-15 07:44:22 +09:00
Paul Masurel
0ba1cf93f7 Remove Searcher dereference (#419) 2018-09-14 09:54:26 +09:00
Paul Masurel
21a9940726 Update Changelog with #388 (#418) 2018-09-14 09:31:11 +09:00
pentlander
8600b8ea25 Top collector (#413) 
* Make TopCollector generic

Make TopCollector take a generic type instead of only being tied to
score. This will allow for sharing code between a TopCollector that
sorts results by Score and a TopCollector that sorts documents by a fast
field. This commit makes no functional changes to TopCollector.

* Add TopFieldCollector and TopScoreCollector

Create two new collectors that use the refactored TopCollector.
TopFieldCollector has the same functionality that TopCollector
originally had. TopFieldCollector allows for sorting results by a given
fast field. Closes tantivy-search/tantivy#388

* Make TopCollector private

Make TopCollector package private and export TopFieldCollector as
TopCollector to maintain backwards compatibility. Mark TopCollector
as deprecated to encourage use of the non-aliased TopFieldCollector.
Remove Collector implementation for TopCollector since it is not longer
used.
 2018-09-14 09:22:17 +09:00
Paul Masurel
30f4f85d48 Closes #414. (#417) 
Updating documentation for load_searchers.
 2018-09-14 09:11:07 +09:00
Paul Masurel
82d25b8397 Fixing snippet example 2018-09-13 12:39:42 +09:00
Paul Masurel
2104c0277c Updating uuid 2018-09-13 09:13:37 +09:00
Paul Masurel
dd37e109f2 Merge branch 'issue/368b' 2018-09-11 20:16:14 +09:00
Paul Masurel
cc23194c58 Editing document 2018-09-11 20:15:38 +09:00
Paul Masurel
63868733a3 Added SnippetGenerator 2018-09-11 09:45:27 +09:00
Paul Masurel
644d8a3a10 Added snippet generator 2018-09-10 16:39:45 +09:00
Paul Masurel
e32dba1a97 Phrase weight 2018-09-10 09:26:33 +09:00
Paul Masurel
a78aa4c259 updating doc 2018-09-09 17:23:30 +09:00
Paul Masurel
7e5f697d00 Closes #387 2018-09-09 16:23:56 +09:00
Paul Masurel
a78f4cca37 Merge branch 'issue/368' into issue/368b 2018-09-09 16:04:20 +09:00
Paul Masurel
2e44f0f099 blop 2018-09-09 14:23:24 +09:00
Vignesh Sarma K
9ccba9f864 Merge branch 'master' into issue/368 2018-09-07 20:27:38 +05:30
Paul Masurel
9101bf5753 Fragments 2018-09-07 09:57:12 +09:00
Paul Masurel
23e97da9f6 Merge branch 'master' of github.com:tantivy-search/tantivy 2018-09-07 08:44:14 +09:00
Paul Masurel
1d439e96f5 Using sort unstable by key. 2018-09-07 08:43:44 +09:00
Paul Masurel
934933582e Closes #402 (#403) 2018-09-06 10:12:26 +09:00
Paul Masurel
98c7fbdc6f Issue/378 (#392) 
* Added failing unit test

* Closes #378. Handling queries that end up empty after going through the analyzer.

* Fixed stop word example
 2018-09-06 10:11:54 +09:00
Paul Masurel
cec9956a01 Issue/389 (#405) 
* Setting up the dependency.

* Completed README
 2018-09-06 10:10:40 +09:00
Paul Masurel
c64972e039 Apply unicode lowercasing. (#408) 
Checks if the str is ASCII, and uses a fast track if it is the case.
If not, the std's definition of a lowercase character.

Closes #406
 2018-09-05 09:43:56 +09:00
Paul Masurel
b3b2421e8a Issue/367 (#404) 
* First stab

* Closes #367
 2018-09-04 09:17:00 +09:00
Paul Masurel
f570fe37d4 small changes 2018-08-31 09:03:44 +09:00
Paul Masurel
6704ab6987 Added methods to extract the matching terms. First stab 2018-08-30 09:47:19 +09:00
Paul Masurel
a12d211330 Extracting terms matching query in the document 2018-08-30 09:23:34 +09:00
Paul Masurel
ee681a4dd1 Added say thanks badge 2018-08-29 11:06:04 +09:00
petr-tik
d15efd6635 Closes #235 - adds a new error type (#398) 
error message suggests possible causes

Addressed code review 1 thread + smaller heap size
 2018-08-29 08:26:59 +09:00
Vignesh Sarma K (വിഘ്നേഷ് ശ൪മ കെ)
18814ba0c1 add a test for second fragment having higher score 2018-08-28 22:27:56 +05:30
Vignesh Sarma K (വിഘ്നേഷ് ശ൪മ കെ)
f247935bb9 Use HighlightSection::new rather than just directly creating the object 2018-08-28 22:16:22 +05:30
Vignesh Sarma K (വിഘ്നേഷ് ശ൪മ കെ)
6a197e023e ran rustfmt 2018-08-28 20:41:58 +05:30
Vignesh Sarma K (വിഘ്നേഷ് ശ൪മ കെ)
96a313c6dd add more tests 2018-08-28 20:41:58 +05:30
Vignesh Sarma K (വിഘ്നേഷ് ശ൪മ കെ)
fb9b1c1f41 add a test and fix the bug of not calculating first token 2018-08-28 20:41:58 +05:30
Vignesh Sarma K (വിഘ്നേഷ് ശ൪മ കെ)
e1bca6db9d update calculate_score to try_add_token 
`try_add_token` will now update the stop_offset as well.
`FragmentCandidate::new` now just takes `start_offset`,
it expects `try_add_token` to be called to add a token.
 2018-08-28 20:41:58 +05:30
Vignesh Sarma K (വിഘ്നേഷ് ശ൪മ കെ)
8438eda01a use while let instead of loop and if. 
as per CR comment
 2018-08-28 20:41:57 +05:30
Vignesh Sarma K (വിഘ്നേഷ് ശ൪മ കെ)
b373f00840 add htmlescape and update to_html fn to use it. 
tests and imports also updated.
 2018-08-28 20:41:57 +05:30
Vignesh Sarma K (വിഘ്നേഷ് ശ൪മ കെ)
46decdb0ea compare against accumulator rather than init value 2018-08-28 20:41:41 +05:30
Vignesh Sarma K (വിഘ്നേഷ് ശ൪മ കെ)
835cdc2fe8 Initial version of snippet 
refer #368
 2018-08-28 20:41:41 +05:30
Paul Masurel
19756bb7d6 Getting started on #368 2018-08-28 20:41:41 +05:30
CJP10
57e1f8ed28 Missed a closing bracket (#397) 2018-08-28 23:17:59 +09:00
Paul Masurel
2649c8a715 Issue/246 (#393) 
* Moving Range and All to Leaves

* Parsing OR/AND

* Simplify user input ast

* AND and OR supported. Returning an error when mixing syntax

Closes #246

* Added support for NOT

* Updated changelog
 2018-08-28 11:03:54 +09:00
Paul Masurel
ede97eded6 Removed use 2018-08-28 09:54:04 +09:00
Paul Masurel
4b7ff78c5a Added fundamentalss 2018-08-28 08:09:27 +09:00
Paul Masurel
948758ad78 First commit for the documentation 2018-08-27 09:49:49 +09:00
Paul Masurel
d71fa43ca3 Moving emoticon on the right side of the parenthesis 2018-08-23 08:59:11 +09:00
Paul Masurel
1e5266d4c9 Merge branch 'master' of github.com:tantivy-search/tantivy 2018-08-23 08:55:30 +09:00
Paul Masurel
537fc27231 Added bench line in features 2018-08-23 08:55:13 +09:00
Dru Sellers
af593b1116 Add default EN stopwords to the default analyzer (#381) 
* Add a default list of en stopwords

* Add the default en stopword filter to the standard tokenizers

* code review feedback
 2018-08-22 10:49:39 +09:00
Paul Masurel
3d73c0c240 Update issue templates 2018-08-21 10:59:08 +09:00
Paul Masurel
3a8e524f77 Added example to show how to access the inverted list directly 2018-08-21 09:36:13 +09:00
Paul Masurel
c0641c2b47 Remove generate html script. It moved to tantivy-search.github.io 2018-08-21 08:26:46 +09:00
Dru Sellers
ef3a16a129 Switch from error-chain to failure crate (#376) 
* Switch from error-chain to failure crate

* Added deprecated alias for

* Started editing the changeld
 2018-08-20 09:40:45 +09:00
Paul Masurel
a0a284fe91 Added a full fledge empty query and relyign on it in QueryParser, instead of using an empty clause. 2018-08-20 09:21:32 +09:00
dependabot[bot]
0feeef2684 Update owning_ref requirement from 0.3 to 0.4 (#379) 
Updates the requirements on [owning_ref](https://github.com/Kimundi/owning-ref-rs) to permit the latest version.
- [Release notes](https://github.com/Kimundi/owning-ref-rs/releases)
- [Commits](https://github.com/Kimundi/owning-ref-rs/commits)

Signed-off-by: dependabot[bot] <support@dependabot.com>
 2018-08-20 09:08:11 +09:00
Dru Sellers
cc50bdb06a Add a basic faceted search example (#383) 
* Add a basic faceted search example

* quieting the compiler
 2018-08-19 08:07:54 +09:00
Paul Masurel
23c2c3ae7c Building all examples on appveyor + running them on travis 2018-08-17 13:24:37 +09:00
Dru Sellers
674524ba91 Add an example of using the stopwords filter (#377) 2018-08-17 12:52:21 +09:00
Paul Masurel
60a9a7f837 Added example showing how to delete/update documents 2018-08-17 09:43:55 +09:00
Paul Masurel
5b5c706581 Simplified examples 2018-08-16 22:38:39 +09:00
Paul Masurel
3e14a76623 Update regex_query.rs 2018-08-15 16:38:32 +09:00
Paul Masurel
8cde1c81e5 Update README.md 2018-08-13 18:03:30 +09:00
Paul Masurel
8d0a29b137 Added sourcerer wall of fame 2018-08-13 18:02:49 +09:00
Paul Masurel
cbfb2fe19d Avoid building twice when doing code coverage 2018-08-13 10:38:01 +09:00
Vignesh Sarma K
09e00f1d42 add position_length to Token (#337) 
* add position_length to Token

refer #291

* Add term offset to `PhraseQuery`

ref #291

* Add new constructor for `PhraseQuery` that allows custom offset

* fix the method name as per pr comment

* Closes #291

Added unit test.
Using offsets from the analyzer in QueryParser.
 2018-08-13 10:14:50 +09:00
Paul Masurel
290620fdee Added slashes 2018-08-13 09:13:01 +09:00
petr-tik
f0d1b85bd8 N370 pr fix num searchers (#371) 
* Change ordering to Acquire

* set_num_searchers now uses AtomicUsize.store
 2018-08-13 08:56:30 +09:00
petr-tik
aaef546f91 Moved NUM_SEARCHERS into a local variable (#369) 
* Moved NUM_SEARCHERS into a local variable

dynamically determined as the number of available cpus.

var name in lowercase (not a constant anymore).

updated it in docstring

* lowercased the varnames

* User can set number of logical cores in create_from_metas

* cargo fmt

* Num_searchers as Arc<AtomicUsize>

Retrieving the value with Relaxed ordering

Reverted create_from_metas signature. However, it calls num_cpus and
sets the Arc val
 2018-08-12 20:08:14 +09:00
Paul Masurel
811ddf2226 Closes #364 (#365) 
* Closes #364

* Trying to raise the recursion limit

* Better unit test and bug fix on token offsets
 2018-08-08 11:15:20 +09:00
Paul Masurel
79a339d353 Removing env_logger dependency 2018-08-02 19:29:09 +09:00
Paul Masurel
e45e4c79d9 update crossbeam 2018-08-02 19:24:08 +09:00
Paul Masurel
848bf41bc9 Updating rand to 0.5 (#363) 2018-08-02 19:19:04 +09:00
Paul Masurel
d11cb087a7 Updated to combine-0.3 (#362) 2018-08-02 18:29:58 +09:00
Jacob Brown
2dd7422f42 replace chan with crossbeam-channel (#361) 
* replace chan with crossbeam-channel

* Update Cargo.toml
 2018-08-02 12:47:22 +09:00
Paul Masurel
e8707c02c0 Issue/333 (#335) 
* Add skip information for posting list (skip to doc ids) 
* Separate num bits from data for positions (skip n positions)
* Address in the position using a n-position offset
* Added a long skip structure to allow efficient opening of the position for a given term.
 2018-07-31 10:51:53 +09:00
dependabot[bot]
55928d756a Update rust-stemmers requirement to 1.0.2 (#350) 
* Update rust-stemmers requirement to 1.0.2

Updates the requirements on [rust-stemmers](https://github.com/CurrySoftware/rust-stemmers) to permit the latest version.
- [Release notes](https://github.com/CurrySoftware/rust-stemmers/releases)
- [Commits](https://github.com/CurrySoftware/rust-stemmers/commits)

Signed-off-by: dependabot[bot] <support@dependabot.com>

* Update Cargo.toml
 2018-07-31 09:32:57 +09:00
dependabot[bot]
a4370bca64 Update owned-read requirement to 0.4 (#352) 
Updates the requirements on [owned-read](https://github.com/tantivy-search/owned-read) to permit the latest version.
- [Release notes](https://github.com/tantivy-search/owned-read/releases)
- [Commits](https://github.com/tantivy-search/owned-read/commits)

Signed-off-by: dependabot[bot] <support@dependabot.com>
 2018-07-31 09:32:01 +09:00
dependabot[bot]
5a5c5a8ca5 Update bit-set requirement to 0.5.0 (#351) 
* Update bit-set requirement to 0.5.0

Updates the requirements on [bit-set](https://github.com/contain-rs/bit-set) to permit the latest version.
- [Release notes](https://github.com/contain-rs/bit-set/releases)
- [Commits](https://github.com/contain-rs/bit-set/commits)

Signed-off-by: dependabot[bot] <support@dependabot.com>

* Update Cargo.toml

* Update Cargo.toml
 2018-07-31 09:31:41 +09:00
dependabot[bot]
1b470dd474 Update log requirement to 0.4.3 (#353) 
* Update log requirement to 0.4.3

Updates the requirements on [log](https://github.com/rust-lang/log) to permit the latest version.
- [Release notes](https://github.com/rust-lang/log/releases)
- [Changelog](https://github.com/rust-lang-nursery/log/blob/master/CHANGELOG.md)
- [Commits](https://github.com/rust-lang/log/commits/env_logger-0.4.3)

Signed-off-by: dependabot[bot] <support@dependabot.com>

* Update Cargo.toml
 2018-07-31 09:31:19 +09:00
Paul Masurel
52b4575245 Issue/355 (#358) 
* issue with top_k sorting (#356)

* Closes #355
 2018-07-31 08:24:55 +09:00
dependabot[bot]
ddd2d5b04c Update lazy_static requirement to 1.0.2 (#349) 
* Update lazy_static requirement to 1.0.2

Updates the requirements on [lazy_static](https://github.com/rust-lang-nursery/lazy-static.rs) to permit the latest version.
- [Release notes](https://github.com/rust-lang-nursery/lazy-static.rs/releases)
- [Commits](https://github.com/rust-lang-nursery/lazy-static.rs/commits/v1.0.2)

Signed-off-by: dependabot[bot] <support@dependabot.com>

* Update Cargo.toml
 2018-07-30 12:34:06 +09:00
dependabot[bot]
fa22b4041a Update itertools requirement to 0.7.8 (#346) 
* Update itertools requirement to 0.7.8

Updates the requirements on [itertools](https://github.com/bluss/rust-itertools) to permit the latest version.
- [Release notes](https://github.com/bluss/rust-itertools/releases)
- [Commits](https://github.com/bluss/rust-itertools/commits/0.7.8)

Signed-off-by: dependabot[bot] <support@dependabot.com>

* Update Cargo.toml
 2018-07-30 11:32:12 +09:00
dependabot[bot]
8faee143fa Update regex requirement to 1.0 (#347) 
Updates the requirements on [regex](https://github.com/rust-lang/regex) to permit the latest version.
- [Release notes](https://github.com/rust-lang/regex/releases)
- [Changelog](https://github.com/rust-lang/regex/blob/master/CHANGELOG.md)
- [Commits](https://github.com/rust-lang/regex/commits/1.0.2)

Signed-off-by: dependabot[bot] <support@dependabot.com>
 2018-07-30 09:59:19 +09:00
dependabot[bot]
366ce98f08 Update tempfile requirement to 3.0 (#348) 
Updates the requirements on [tempfile](https://github.com/Stebalien/tempfile) to permit the latest version.
- [Release notes](https://github.com/Stebalien/tempfile/releases)
- [Changelog](https://github.com/Stebalien/tempfile/blob/master/NEWS)
- [Commits](https://github.com/Stebalien/tempfile/commits/v3.0.3)

Signed-off-by: dependabot[bot] <support@dependabot.com>
 2018-07-30 09:58:56 +09:00
Paul Masurel
190e60a41c Closes #339. (#340) 
As required per the FacetCollector,
facet values needs to be sorted before being encoded in the
multivalued field.
 2018-07-25 18:21:48 +09:00
Vignesh Sarma K
b9558801a1 Declare and implement separate Clone Traits (#336) 
For traits, `Directory` and `MergePolicy`.

refer #306
 2018-07-18 12:36:43 +09:00
Paul Masurel
36728215ac Using the codecov badge 2018-07-10 21:19:59 +09:00
Paul Masurel
39551a0418 fix travis 2018-07-10 13:08:22 +09:00
Paul Masurel
39b98b2e76 fix travis 2018-07-10 13:07:15 +09:00
Paul Masurel
616162400d Add missing space 2018-07-10 12:49:32 +09:00
Paul Masurel
694d164db6 fix travis.yml 2018-07-10 09:39:39 +09:00
Paul Masurel
ef442cefb1 codecov 2018-07-10 09:38:59 +09:00
Paul Masurel
14da241f35 Readed cov 2018-07-10 09:25:24 +09:00
Paul Masurel
346a9e4287 Set dev version 2018-07-10 09:20:21 +09:00
Paul Masurel
31655e92d7 Preparing release 0.6.1 2018-07-10 09:12:26 +09:00
Paul Masurel
6b8d76685a Tiny refactoring 2018-07-05 09:11:55 +09:00
Paul Masurel
ce5683fc6a Removed useless counting_writer 2018-07-04 16:13:19 +09:00
Paul Masurel
5205579db6 Merge branch 'master' of github.com:tantivy-search/tantivy 2018-07-04 16:09:59 +09:00
Paul Masurel
d056ae60dc Removed SourceRead. Relying on the new owned-read crate instead (#332) 2018-07-04 16:08:52 +09:00
Paul Masurel
af9280c95f Removed SourceRead. Relying on the new owned-read crate instead 2018-07-04 12:47:25 +09:00
David Hewson
2e538ce6e6 remove extra space in name (#331) 
the extra space that appeared breaks using the package
 2018-07-02 05:32:19 +09:00
Jason Wolfe
00466d2b08 #328: Support parsing unbounded range queries (#329) 
* #328: Support parsing unbounded range queries. Update CHANGELOG.md for query parser changes.

* Set version to 0.7-dev
 2018-06-30 13:24:02 +09:00
Paul Masurel
8ebbf6b336 Issue/325 (#330) 
* Introducing a SegmentMea inventory.
* Depending on census=0.1
* Cargo fmt
 2018-06-30 13:11:41 +09:00
Paul Masurel
1ce36bb211 Merge branch 'master' of github.com:tantivy-search/tantivy 2018-06-27 16:58:47 +09:00
Jason Wolfe
2ac43bf21b Support parsing RangeQuery and AllQuery in Queryparser (#323) 
* (#321) Add support for range query parsing to grammar / parser. Still needs to be wired through the rest of the way.

* (321) Finish wiring RangeQuery parsing through

* (#321) Add logical AST query parser tests for RangeQuery

* (#321) Support parsing AllQuery

* (#321) Update documentation of QueryParser

* (#321) Support negative numbers in range query parsing
 2018-06-25 08:29:47 +09:00
Paul Masurel
3fd8c2aa5a Removed one keywoard 2018-06-22 14:47:21 +09:00
Paul Masurel
c1022e23d2 Switching to stable rust in AppVeyor. 2018-06-22 14:33:42 +09:00
Paul Masurel
8ccbfdea5d Preparing for release 2018-06-22 14:27:46 +09:00
Paul Masurel
badfce3a23 Preparing for release. 2018-06-22 14:09:14 +09:00
Dru Sellers
e301e0bc87 Add some simple doc tests (#320) 
* Add TopCollector doc test

* Add CountCollector Doc Test

* Add Doc Test for MultiCollector

* Add ChainedCollector Doc Test

* Expose Fuzzy Query where it should be

* Add FuzzyTermQuery Doc Test

* Expose RegexQuery

* Regex Query Doc Test

* Add TermQuery Doc Test

* Add doc comments

* fix test 🤦

* Added explanation about the complexity variables

* Fixing unit tests

* Single threads if you check docids
 2018-06-19 10:45:20 +09:00
Dru Sellers
317baf4e75 Add in simple regex query support (#319) 
* Add fst_regex crate in

* Reduce API surface area

This doesn't need to be public

* better test name

* Pull Automaton weight out so it can be shared

* Implement Regex Query
 2018-06-16 14:08:30 +09:00
Paul Masurel
24398d94e4 Exposing the 2018-06-15 21:40:57 +09:00
Dru Sellers
360f4132eb Standardizes the Index::open_* APIs (#318) 
* Relocate `from_directory` closer to its usage

* Specific methods come before the generic method

* Rename open methods to follow the lead of the create methods
 2018-06-15 12:16:41 +09:00
Dru Sellers
2b8f02764b Standardizes the Index::create_* APIs (#317) 
* Pull all creation methods next to each other

The goal here is to make it clear which methods are performing the
same function, and to assist with standardizing the API calls.

* Make `from_directory` private

This seems to be an internal function, so lets make it internal.

* Rename `create` to `create_in_dir`

This lets the name match the `create_in_ram` pattern and opens up
`create` for the generic implementation.

* Implement the generic create function

All of the create methods now delegate to the common create function
and future `create_in_*` functions now have a clear pattern
to follow as well
 2018-06-14 11:08:42 +09:00
Paul Masurel
0465876854 Issue/257 (#310) 
* Replaced lz4 by a pure rust implementation of snappy.

Closes #257

* snappy is the default compression. One can use lz4 by enabling the lz4 feature flag.

* Removed Compression trait
 2018-06-12 19:02:57 +09:00
Dru Sellers
6f7b099370 Add AutomatonWeight to a fuzzy_search module and FuzzyQuery (#300) 
* Add AutomatonWeight to a fuzzy_search module

* Hacking around ownership issues

* Working through lifetime issues

* Working through tests

* fix test by lower casing the words (reducing distance)

* code review changes

* Suggestion on how to solve the borrow problem

* clean up
 2018-06-11 22:23:03 +09:00
Paul Masurel
84f5cc4388 Added an AUTHORS file. Closes #315 (#316) 2018-06-11 22:21:58 +09:00
Paul Masurel
75aae0d2c2 Update README 2018-06-08 13:05:57 +09:00
Paul Masurel
009a3559be atomicwrites 2.2.0 for ARM compilation 2018-06-06 07:13:09 +09:00
Paul Masurel
7a31669e9d Disabling ARM targets 2018-06-05 12:22:00 +09:00
Paul Masurel
5185eb790b Reduced heap usage in unit test 2018-06-05 10:02:10 +09:00
Paul Masurel
a3dffbf1c6 Added more ARM target. 2018-06-05 09:06:33 +09:00
Paul Masurel
857a5794d8 Updated nix version 2018-06-05 09:02:40 +09:00
Paul Masurel
b0a6fc1448 Reduce RAM usage 2018-06-04 11:20:24 +09:00
Paul Masurel
989d52bea4 Updated atomicwrites version. 2018-06-04 10:00:21 +09:00
Paul Masurel
09661ea7ec Added cross testing on different platforms 2018-06-04 09:47:53 +09:00
Paul Masurel
b59132966f Better heap (#311) 
* Changed the heap to a paged memory arena.
* Trying to simplify the indexing term hashmap
* Exploding datastruct
* Removed some complexity in bitpacker
 2018-06-04 09:39:18 +09:00
Paul Masurel
863d3411bc Update Cargo.toml 2018-05-31 15:54:34 +09:00
Paul Masurel
8a55d133ab Showing Appveyor CI badge for the master branch 
.. before the last build was shown.
 2018-05-28 13:44:53 +09:00
Jason Wolfe
432d49d814 Expose parameters of RangeQuery for external usage (#309) 2018-05-19 14:29:25 +09:00
Jason Wolfe
0cea706f10 Add docs to new Query methods (#307) 2018-05-18 13:53:29 +09:00
Paul Masurel
71d41ca209 Added Google to the license 2018-05-18 10:13:23 +09:00
Paul Masurel
bc69dab822 cargo fmt 2018-05-18 10:08:05 +09:00
Jason Wolfe
72acad0921 Add box_clone() and downcast::Any to Query (#303) 2018-05-18 09:53:11 +09:00
Paul Masurel
c9459f74e8 Update docs about TermDict. 2018-05-18 09:20:39 +09:00
Dru Sellers
08d2cc6c7b Make it possible to stream the terms matching an Automaton (#297) 
* rustfmt and some English grammar

* sort cargo.toml crates

* WIP: something to show

* Remove example for now

* Implement desired method

* Resolving Generic Type Arguments

* Resolve Generic Types

* Banging around on the tests

* DANGER! Change unsafe usage based on compiler warnings

* Unscrew up my rebase

* Clean Up Type Spam

Default Types FTW

* typo

* better variable names

* Remove Duplicate Levenshtein crate
 2018-05-11 12:41:14 -07:00
Dru Sellers
82d87416c2 Implement StopWords Filter (#292) 
* Implement StopWords Filter

- added example doctest for alphanum_only.rs so that I could
drive my own test of the stopword filter

* Style Cop

* Switch HashSet Hasher to FNV for speed

* Update Change Log

* fix missed location renaming
 2018-05-09 18:40:41 -07:00
Paul Masurel
96b2c2971e Testing actual doc ids in unit test 2018-05-09 09:14:22 -07:00
Dru Sellers
162afd73f6 Alive docs iterator (#293) 
* Add non-deleted DocId iterator to SegmentReader

Closes #287

* Add Todo

* Add Unit Test

* Improving test based on feedback

- found bug and fixed it. :)

* Reestablish changes post rebase for clean merge
 2018-05-09 09:03:27 -07:00
Paul Masurel
ddfd87fa59 Merge branch 'master' of github.com:tantivy-search/tantivy 2018-05-08 00:08:17 -07:00
Paul Masurel
24050d0eb5 Remove some unsafe stuff, justified some of it. 2018-05-07 23:57:53 -07:00
Jason Wolfe
89eb209ece #294: Make fieldnorm module public, add documentation (#295) 2018-05-07 20:20:38 -07:00
Paul Masurel
9a0b7f9855 Rustfmt 2018-05-07 19:50:35 -07:00
Jason Wolfe
8e343b1ca3 Add fast field for associating arbitrary bytes to a document (#275) 
* Add fast field for associating arbitrary bytes to a document

* Fix unused macro_use warning

* Improvements from code review

* Make BytesFastFieldWriter public

* Fix json parsing validation failure

* Add bytes fast field to CHANGELOG.md

* Fix compile errors from merge

* Support merging

* Address misc code review comments

* Fix comments from CR
 2018-05-07 19:30:31 -07:00
Paul Masurel
99c0b84036 Integrating #274, #280, #289 into master (#290) 
* Integrating bugfixes into master

Closes #274
Closes #280
Closes #289

* Next version will be 0.6
 2018-05-06 09:48:25 -07:00
Dru Sellers
ca74c14647 Simple Implementation of NGram Tokenizer (#278) 
* Simple Implementation of NGram Tokenizer

It does not yet support edges
It could probably be better in many "rusty" ways
But the test is passing, so I'll call this a good stopping point for
the day.

* Remove Ngram from manager. Too many variations

* Basic configuration model

Should the extensive tests exist here?

* Add Sample to provide an End to End testing

* Basic Edgegram support

* cleanup

* code feedback

* More code review feedback processed
 2018-05-06 09:47:49 -07:00
Dru Sellers
68ee18e4e8 Add Index::open_directory function (#285) 
* Add Index::open_directory function

* dry
 2018-05-03 00:07:46 -07:00
Paul Masurel
5637657c2f Removed ptr dereference for explicit ptr::read_unaligned 2018-04-25 19:15:32 +09:00
Paul Masurel
2e3c9a8878 Bugfix in murmurhash. 2018-04-25 19:06:31 +09:00
Paul Masurel
78673172d0 Cargo fmt 2018-04-21 20:05:36 +09:00
Paul Masurel
175b76f119 Removed streamdict 
Closes #271
 2018-04-21 19:55:41 +09:00
Paul Masurel
9b79e21bd7 Returning error when schema is not valid for a given query. 2018-04-19 13:02:30 +09:00
Paul Masurel
5e38ae336f Bump tantivy version and readded win deps 2018-04-17 18:27:57 +09:00
Paul Masurel
8604351f59 Hide some of the API 
Added some doc.
 2018-04-17 13:31:22 +09:00
Paul Masurel
6a48953d8a Closes #266 (#268) 
PhraseQuery panics with a nice error message when the underlying field does not have any positions.
The `QueryParser` fails as well with a dedicated error.
 2018-04-17 10:03:15 +09:00
pmasurel
0804b42afa Checking the type of range queries 2018-04-16 14:01:10 +09:00
Paul Masurel
8083bc6eef bench working 2018-04-15 12:25:38 +09:00
Paul Masurel
0156f88265 Compiles in stable rust 2018-04-15 11:03:44 +09:00
Paul Masurel
a1c07bf457 Added iterator for facet collector 2018-04-14 20:22:02 +09:00
Paul Masurel
9de74b68d1 Remove range argument 2018-04-13 18:34:23 +09:00
Paul Masurel
57c7073867 Removed 2018-04-13 09:43:36 +09:00
Paul Masurel
121374b89b Removed the need for AtomicU64 2018-04-12 22:08:15 +09:00
Paul Masurel
e44782bf14 No more 2018-04-12 13:01:11 +09:00
Paul Masurel
dfafb24fa6 Bumped bitpacker's version 2018-04-10 21:21:47 +09:00
jason-wolfe
4c6f9541e9 #263: Make MultiValueIntFastFieldWriter public, expose via FastFieldsWriter (#264) 2018-04-10 12:27:34 +09:00
253 changed files with 23179 additions and 10095 deletions
Expand all files Collapse all files

19
.github/ISSUE_TEMPLATE/bug_report.md vendored Normal file
View File

@@ -0,0 +1,19 @@
---
name: Bug report
about: Create a report to help us improve
---
**Describe the bug**
- What did you do?
- What happened?
- What was expected?
**Which version of tantivy are you using?**
If "master", ideally give the specific sha1 revision.
**To Reproduce**
If your bug is deterministic, can you give a minimal reproducing code?
Some bugs are not deterministic. Can you describe with precision in which context it happened?
If this is possible, can you share your code?

14
.github/ISSUE_TEMPLATE/feature_request.md vendored Normal file
View File

@@ -0,0 +1,14 @@
---
name: Feature request
about: Suggest an idea for this project
---
**Is your feature request related to a problem? Please describe.**
A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]
**Describe the solution you'd like**
A clear and concise description of what you want to happen.
**[Optional] describe alternatives you've considered**
A clear and concise description of any alternative solutions or features you've considered.

7
.github/ISSUE_TEMPLATE/question.md vendored Normal file
View File

@@ -0,0 +1,7 @@
---
name: Question
about: Ask any question about tantivy's usage...
---
Try to be specific about your use case...

1
.gitignore vendored
View File

@@ -1,3 +1,4 @@
tantivy.iml
*.swp
target
target/debug

73
.travis.yml
View File

@@ -1,14 +1,17 @@
# Based on the "trust" template v0.1.2
# https://github.com/japaric/trust/tree/v0.1.2
dist: trusty
language: rust
services: docker
sudo: required
cache: cargo
rust:
- nightly
env:
global:
- CC=gcc-4.8
- CXX=g++-4.8
- CRATE_NAME=tantivy
- TRAVIS_CARGO_NIGHTLY_FEATURE=""
- secure: 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
addons:
apt:
sources:
@@ -22,16 +25,64 @@ addons:
- libdw-dev
- binutils-dev
- cmake
matrix:
include:
# Android
- env: TARGET=aarch64-linux-android DISABLE_TESTS=1
#- env: TARGET=arm-linux-androideabi DISABLE_TESTS=1
#- env: TARGET=armv7-linux-androideabi DISABLE_TESTS=1
#- env: TARGET=i686-linux-android DISABLE_TESTS=1
#- env: TARGET=x86_64-linux-android DISABLE_TESTS=1
# Linux
#- env: TARGET=aarch64-unknown-linux-gnu
#- env: TARGET=i686-unknown-linux-gnu
- env: TARGET=x86_64-unknown-linux-gnu CODECOV=1
# - env: TARGET=x86_64-unknown-linux-musl CODECOV=1
# OSX
- env: TARGET=x86_64-apple-darwin
os: osx
before_install:
- set -e
- rustup self update
install:
- sh ci/install.sh
- source ~/.cargo/env || true
before_script:
- export PATH=$HOME/.cargo/bin:$PATH
- cargo install cargo-update || echo "cargo-update already installed"
- cargo install cargo-travis || echo "cargo-travis already installed"
script:
- cargo build
- cargo test
- cargo test -- --ignored
- cargo run --example simple_search
- cargo doc
- bash ci/script.sh
after_success:
- cargo coveralls --exclude-pattern src/functional_test.rs
- cargo doc-upload
before_deploy:
- sh ci/before_deploy.sh
cache: cargo
before_cache:
# Travis can't cache files that are not readable by "others"
- chmod -R a+r $HOME/.cargo
- find ./target/debug -type f -maxdepth 1 -delete
- rm -f ./target/.rustc_info.json
- rm -fr ./target/debug/{deps,.fingerprint}/tantivy*
- rm -r target/debug/examples/
- ls -1 examples/ | sed -e 's/\.rs$//' | xargs -I "{}" find target/* -name "*{}*" -type f -delete
#branches:
# only:
# # release tags
# - /^v\d+\.\d+\.\d+.*$/
# - master
notifications:
email:
on_success: never

11
AUTHORS Normal file
View File

@@ -0,0 +1,11 @@
# This is the list of authors of tantivy for copyright purposes.
Paul Masurel
Laurentiu Nicola
Dru Sellers
Ashley Mannix
Michael J. Curry
Jason Wolfe
# As an employee of Google I am required to add Google LLC
# in the list of authors, but this project is not affiliated to Google
# in any other way.
Google LLC

171
CHANGELOG.md
View File

@@ -1,9 +1,164 @@
Tantivy 0.5.2
Tantivy 0.10.0
=====================
*Tantivy 0.10.0 index format is compatible with the index format in 0.9.0.*
- Added an ASCII folding filter (@drusellers)
- Bugfix in `query.count` in presence of deletes (@pmasurel)
Minor
---------
- Small simplification of the code.
Calling .freq() or .doc() when .advance() has never been called
on segment postings should panic from now on.
- Tokens exceeding `u16::max_value() - 4` chars are discarded silently instead of panicking.
- Fast fields are now preloaded when the `SegmentReader` is created.
- `IndexMeta` is now public. (@hntd187)
- `IndexWriter` `add_document`, `delete_term`. `IndexWriter` is `Sync`, making it possible to use it with a `
Arc<RwLock<IndexWriter>>`. `add_document` and `delete_term` can
only require a read lock. (@pmasurel)
- Introducing `Opstamp` as an expressive type alias for `u64`. (@petr-tik)
- Stamper now relies on `AtomicU64` on all platforms (@petr-tik)
## How to update?
Your existing indexes are usable as is, but you may need some
trivial updates.
### Fast fields
Fast fields used to be accessed directly from the `SegmentReader`.
The API changed, you are now required to acquire your fast field reader via the
`segment_reader.fast_fields()`, and use one of the typed method:
- `.u64()`, `.i64()` if your field is single-valued ;
- `.u64s()`, `.i64s()` if your field is multi-valued ;
- `.bytes()` if your field is bytes fast field.
Tantivy 0.9.0
=====================
*0.9.0 index format is not compatible with the
previous index format.*
- MAJOR BUGFIX :
Some `Mmap` objects were being leaked, and would never get released. (@fulmicoton)
- Removed most unsafe (@fulmicoton)
- Indexer memory footprint improved. (VInt comp, inlining the first block. (@fulmicoton)
- Stemming in other language possible (@pentlander)
- Segments with no docs are deleted earlier (@barrotsteindev)
- Added grouped add and delete operations.
They are guaranteed to happen together (i.e. they cannot be split by a commit).
In addition, adds are guaranteed to happen on the same segment. (@elbow-jason)
- Removed `INT_STORED` and `INT_INDEXED`. It is now possible to use `STORED` and `INDEXED`
for int fields. (@fulmicoton)
- Added DateTime field (@barrotsteindev)
- Added IndexReader. By default, index is reloaded automatically upon new commits (@fulmicoton)
- SIMD linear search within blocks (@fulmicoton)
## How to update ?
tantivy 0.9 brought some API breaking change.
To update from tantivy 0.8, you will need to go through the following steps.
- `schema::INT_INDEXED` and `schema::INT_STORED` should be replaced by `schema::INDEXED` and `schema::INT_STORED`.
- The index now does not hold the pool of searcher anymore. You are required to create an intermediary object called
`IndexReader` for this.
```rust
// create the reader. You typically need to create 1 reader for the entire
// lifetime of you program.
let reader = index.reader()?;
// Acquire a searcher (previously `index.searcher()`) is now written:
let searcher = reader.searcher();
// With the default setting of the reader, you are not required to
// call `index.load_searchers()` anymore.
//
// The IndexReader will pick up that change automatically, regardless
// of whether the update was done in a different process or not.
// If this behavior is not wanted, you can create your reader with
// the `ReloadPolicy::Manual`, and manually decide when to reload the index
// by calling `reader.reload()?`.
```
Tantivy 0.8.2
=====================
Fixing build for x86_64 platforms. (#496)
No need to update from 0.8.1 if tantivy
is building on your platform.
Tantivy 0.8.1
=====================
Hotfix of #476.
Merge was reflecting deletes before commit was passed.
Thanks @barrotsteindev for reporting the bug.
Tantivy 0.8.0
=====================
*No change in the index format*
- API Breaking change in the collector API. (@jwolfe, @fulmicoton)
- Multithreaded search (@jwolfe, @fulmicoton)
Tantivy 0.7.1
=====================
*No change in the index format*
- Bugfix: NGramTokenizer panics on non ascii chars
- Added a space usage API
Tantivy 0.7
=====================
- Skip data for doc ids and positions (@fulmicoton),
greatly improving performance
- Tantivy error now rely on the failure crate (@drusellers)
- Added support for `AND`, `OR`, `NOT` syntax in addition to the `+`,`-` syntax
- Added a snippet generator with highlight (@vigneshsarma, @fulmicoton)
- Added a `TopFieldCollector` (@pentlander)
Tantivy 0.6.1
=========================
- Bugfix #324. GC removing was removing file that were still in useful
- Added support for parsing AllQuery and RangeQuery via QueryParser
- AllQuery: `*`
- RangeQuery:
- Inclusive `field:[startIncl to endIncl]`
- Exclusive `field:{startExcl to endExcl}`
- Mixed `field:[startIncl to endExcl}` and vice versa
- Unbounded `field:[start to *]`, `field:[* to end]`
Tantivy 0.6
==========================
- Removed C code. Tantivy is now pure Rust.
- BM25
- Approximate field norms encoded over 1 byte.
Special thanks to @drusellers and @jason-wolfe for their contributions
to this release!
- Removed C code. Tantivy is now pure Rust. (@pmasurel)
- BM25 (@pmasurel)
- Approximate field norms encoded over 1 byte. (@pmasurel)
- Compiles on stable rust (@pmasurel)
- Add &[u8] fastfield for associating arbitrary bytes to each document (@jason-wolfe) (#270)
- Completely uncompressed
- Internally: One u64 fast field for indexes, one fast field for the bytes themselves.
- Add NGram token support (@drusellers)
- Add Stopword Filter support (@drusellers)
- Add a FuzzyTermQuery (@drusellers)
- Add a RegexQuery (@drusellers)
- Various performance improvements (@pmasurel)_
Tantivy 0.5.2
===========================
- bugfix #274
- bugfix #280
- bugfix #289
Tantivy 0.5.1
==========================
@@ -81,7 +236,7 @@ Tantivy 0.3
Special thanks to @Kodraus @lnicola @Ameobea @manuel-woelker @celaus
for their contribution to this release.
Thanks also to everyone in tantivy gitter chat
Thanks also to everyone in tantivy gitter chat
for their advise and company :)
https://gitter.im/tantivy-search/tantivy
@@ -89,9 +244,9 @@ https://gitter.im/tantivy-search/tantivy
Warning:
Tantivy 0.3 is NOT backward compatible with tantivy 0.2
Tantivy 0.3 is NOT backward compatible with tantivy 0.2
code and index format.
You should not expect backward compatibility before
You should not expect backward compatibility before
tantivy 1.0.
@@ -117,7 +272,7 @@ Thanks to @KodrAus ! (#108)
the natural ordering.
- Building binary targets for tantivy-cli (Thanks to @KodrAus)
- Misc invisible bug fixes, and code cleanup.
- Use
- Use

89
Cargo.toml
View File

@@ -1,10 +1,10 @@
[package]
name = "tantivy"
version = "0.5.1"
version = "0.10.0-dev"
authors = ["Paul Masurel <paul.masurel@gmail.com>"]
license = "MIT"
categories = ["database-implementations", "data-structures"]
description = """Tantivy is a search engine library."""
description = """Search engine library"""
documentation = "https://tantivy-search.github.io/tantivy/tantivy/index.html"
homepage = "https://github.com/tantivy-search/tantivy"
repository = "https://github.com/tantivy-search/tantivy"
@@ -12,68 +12,73 @@ readme = "README.md"
keywords = ["search", "information", "retrieval"]
[dependencies]
base64 = "0.10.0"
byteorder = "1.0"
lazy_static = "0.2.1"
tinysegmenter = "0.1.0"
regex = "0.2"
fst = {version="0.2", default-features=false}
atomicwrites = {version="0.1", optional=true}
log = "0.3.6"
combine = "2.2"
lazy_static = "1"
regex = "1.0"
tantivy-fst = "0.1"
memmap = {version = "0.7", optional=true}
lz4 = {version="1.20", optional=true}
snap = {version="0.2"}
atomicwrites = {version="0.2.2", optional=true}
tempfile = "3.0"
log = "0.4"
combine = ">=3.6.0,<4.0.0"
tempdir = "0.3"
serde = "1.0"
serde_derive = "1.0"
serde_json = "1.0"
num_cpus = "1.2"
itertools = "0.5.9"
bit-set = "0.4.0"
uuid = { version = "0.6", features = ["v4", "serde"] }
chan = "0.1"
crossbeam = "0.3"
fs2={version="0.4", optional=true}
itertools = "0.8"
levenshtein_automata = {version="0.1", features=["fst_automaton"]}
notify = {version="4", optional=true}
bit-set = "0.5"
uuid = { version = "0.7.2", features = ["v4", "serde"] }
crossbeam = "0.5"
futures = "0.1"
futures-cpupool = "0.1"
error-chain = "0.8"
owning_ref = "0.3"
owning_ref = "0.4"
stable_deref_trait = "1.0.0"
rust-stemmers = "0.1.0"
downcast = { version="0.9", features = ["nightly"]}
matches = "0.1"
snap = "0.2"
bitpacking = {path = "../bitpacking"}
rust-stemmers = "1.1"
downcast-rs = { version="1.0" }
bitpacking = "0.6"
census = "0.2"
fnv = "1.0.6"
owned-read = "0.4"
failure = "0.1"
htmlescape = "0.3.1"
fail = "0.2"
scoped-pool = "1.0"
murmurhash32 = "0.2"
chrono = "0.4"
[target.'cfg(windows)'.dependencies]
winapi = "0.2"
[dev-dependencies]
rand = "0.3"
tempfile = "2.1"
env_logger = "0.4"
rand = "0.6"
maplit = "1"
matches = "0.1.8"
time = "0.1.42"
[profile.release]
opt-level = 3
debug = false
lto = true
debug-assertions = false
[profile.test]
debug-assertions = true
overflow-checks = true
[features]
default = ["mmap"]
streamdict = []
mmap = ["fst/mmap", "atomicwrites"]
# by default no-fail is disabled. We manually enable it when running test.
default = ["mmap", "no_fail"]
mmap = ["atomicwrites", "fs2", "memmap", "notify"]
lz4-compression = ["lz4"]
no_fail = ["fail/no_fail"]
unstable = [] # useful for benches.
wasm-bindgen = ["uuid/wasm-bindgen"]
[badges]
travis-ci = { repository = "tantivy-search/tantivy" }
[[example]]
name = "simple_search"
required-features = ["mmap"]
[[bin]]
name = "convert_to_static"
path = "./bin/convert_to_static.rs"
[[bin]]
name = "test_static_dir"
path = "./bin/test_static_dir.rs"

2
LICENSE
View File

@@ -1,4 +1,4 @@
Copyright (c) 2016 Paul Masurel
Copyright (c) 2018 by the project authors, as listed in the AUTHORS file.
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

114
README.md
View File

@@ -1,39 +1,80 @@
![Tantivy](https://tantivy-search.github.io/logo/tantivy-logo.png)
[![Build Status](https://travis-ci.org/tantivy-search/tantivy.svg?branch=master)](https://travis-ci.org/tantivy-search/tantivy)
[![Coverage Status](https://coveralls.io/repos/github/tantivy-search/tantivy/badge.svg?branch=master&refresh1)](https://coveralls.io/github/tantivy-search/tantivy?branch=master)
[![codecov](https://codecov.io/gh/tantivy-search/tantivy/branch/master/graph/badge.svg)](https://codecov.io/gh/tantivy-search/tantivy)
[![Join the chat at https://gitter.im/tantivy-search/tantivy](https://badges.gitter.im/tantivy-search/tantivy.svg)](https://gitter.im/tantivy-search/tantivy?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge)
[![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?svg=true)](https://ci.appveyor.com/project/fulmicoton/tantivy)
[![Build status](https://ci.appveyor.com/api/projects/status/r7nb13kj23u8m9pj/branch/master?svg=true)](https://ci.appveyor.com/project/fulmicoton/tantivy/branch/master)
[![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)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/0)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/0)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/1)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/1)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/2)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/2)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/3)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/3)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/4)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/4)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/5)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/5)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/6)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/6)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/7)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/7)
[![Become a patron](https://c5.patreon.com/external/logo/become_a_patron_button.png)](https://www.patreon.com/fulmicoton)
**Tantivy** is a **full text search engine library** written in rust.
It is strongly inspired by Lucene's design.
It is closer to [Apache Lucene](https://lucene.apache.org/) than to [Elasticsearch](https://www.elastic.co/products/elasticsearch) and [Apache Solr](https://lucene.apache.org/solr/) in the sense it is not
an off-the-shelf search engine server, but rather a crate that can be used
to build such a search engine.
Tantivy is, in fact, strongly inspired by Lucene's design.
# Benchmark
Tantivy is typically faster than Lucene, but the results will 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.
# Features
- Full-text search
- Configurable tokenizer. (stemming available for 17 latin languages. Third party support for Chinese ([tantivy-jieba](https://crates.io/crates/tantivy-jieba) and [cang-jie](https://crates.io/crates/cang-jie)) and [Japanese](https://crates.io/crates/tantivy-tokenizer-tiny-segmenter)
- Fast (check out the :racehorse: :sparkles: [benchmark](https://tantivy-search.github.io/bench/) :sparkles: :racehorse:)
- Tiny startup time (<10ms), perfect for command line tools
- tf-idf scoring
- Basic query language
- Phrase queries
- BM25 scoring (the same as lucene)
- Natural query language `(michael AND jackson) OR "king of pop"`
- Phrase queries search (`"michael jackson"`)
- Incremental indexing
- Multithreaded indexing (indexing English Wikipedia takes < 3 minutes on my desktop)
- Mmap directory
- optional SIMD integer compression
- SIMD integer compression when the platform/CPU includes the SSE2 instruction set.
- Single valued and multivalued u64 and i64 fast fields (equivalent of doc values in Lucene)
- `&[u8]` fast fields
- Text, i64, u64, dates and hierarchical facet fields
- LZ4 compressed document store
- Range queries
- Faceting
- configurable indexing (optional term frequency and position indexing
- Faceted search
- Configurable indexing (optional term frequency and position indexing)
- Cheesy logo with a horse
Tantivy supports Linux, MacOS and Windows.
# Non-features
- Distributed search is out of the scope of tantivy. That being said, tantivy is meant as a
library upon which one could build a distributed search. Serializable/mergeable collector state for instance,
are within the scope of tantivy.
# Supported OS and compiler
Tantivy works on stable rust (>= 1.27) and supports Linux, MacOS and Windows.
# Getting started
- [tantivy's usage example](http://fulmicoton.com/tantivy-examples/simple_search.html)
- [tantivy's simple search example](http://fulmicoton.com/tantivy-examples/simple_search.html)
- [tantivy-cli and its tutorial](https://github.com/tantivy-search/tantivy-cli).
`tantivy-cli` is an actual command line interface that makes it easy for you to create a search engine,
index documents and search via the CLI or a small server with a REST API.
It will walk you through getting a wikipedia search engine up and running in a few minutes.
- [reference doc]
- [For the last released version](https://docs.rs/tantivy/)
@@ -43,43 +84,26 @@ It will walk you through getting a wikipedia search engine up and running in a f
## Development
Tantivy requires Rust Nightly because it uses requires the features [`box_syntax`](https://doc.rust-lang.org/stable/unstable-book/language-features/box-syntax.html), [`optin_builtin_traits`](https://github.com/rust-lang/rfcs/blob/master/text/0019-opt-in-builtin-traits.md), [`conservative_impl_trait`](https://github.com/rust-lang/rfcs/blob/master/text/1522-conservative-impl-trait.md),
and [simd](https://github.com/rust-lang/rust/issues/27731).
Tantivy compiles on stable rust but requires `Rust >= 1.27`.
To check out and run tests, you can simply run :
To check out and run test, you can simply run :
git clone git@github.com:tantivy-search/tantivy.git
git clone https://github.com/tantivy-search/tantivy.git
cd tantivy
cargo +nightly build
cargo build
## Running tests
## Note on release build and performance
Some tests will not run with just `cargo test` because of `fail-rs`.
To run the tests exhaustively, run `./run-tests.sh`.
If your project depends on `tantivy`, for better performance, make sure to enable
`sse3` instructions using a RUSTFLAGS. (This instruction set is likely to
be available on most `x86_64` CPUs you will encounter).
# How can I support this project ?
For instance,
There are many ways to support this project.
RUSTFLAGS='-C target-feature=+sse3'
Or, if you are targetting a specific cpu
RUSTFLAGS='-C target-cpu=native' build --release
Regardless of the flags you pass, by default `tantivy` will contain `SSE3` instructions.
If you want to disable those, you can run the following command :
cargo build --no-default-features
Alternatively, if you are trying to compile `tantivy` without simd compression,
you can disable this functionality. In this case, this submodule is not required
and you can compile tantivy by using the `--no-default-features` flag.
cargo build --no-default-features
# Contribute
Send me an email (paul.masurel at gmail.com) if you want to contribute to tantivy.
- If you use tantivy, tell us about your experience on [gitter](https://gitter.im/tantivy-search/tantivy) or by email (paul.masurel@gmail.com)
- Report bugs
- Write a blog post
- Complete documentation
- Contribute code (you can join [our gitter](https://gitter.im/tantivy-search/tantivy) )
- Talk about tantivy around you
- 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)

9
appveyor.yml
View File

@@ -4,11 +4,8 @@
os: Visual Studio 2015
environment:
matrix:
- channel: nightly
- channel: stable
target: x86_64-pc-windows-msvc
- channel: nightly
target: x86_64-pc-windows-gnu
msys_bits: 64
install:
- appveyor DownloadFile https://win.rustup.rs/ -FileName rustup-init.exe
@@ -21,5 +18,5 @@ install:
build: false
test_script:
- REM SET RUST_LOG=tantivy,test & cargo test --verbose
- REM SET RUST_BACKTRACE=1 & cargo run --example simple_search
- REM SET RUST_LOG=tantivy,test & cargo test --verbose --no-default-features --features mmap -- --test-threads 1
- REM SET RUST_BACKTRACE=1 & cargo build --examples

20
bin/convert_to_static.rs
View File

@@ -1,20 +0,0 @@
use std::env;
use std::path::PathBuf;
use std::fs::File;
use std::io::Write;
extern crate tantivy;
use tantivy::directory::write_static_from_directory;
fn main() {
// Prints each argument on a separate line
let mut args = env::args();
args.next().unwrap();
let directory_path= args.next().expect("Expect 2 args.<directory_path> <outputfile>");
let output_path = args.next().expect("Expect 2 args.<directory_path> <outputfile>");
println!("{} => {}", directory_path, output_path);
let buffer = write_static_from_directory(&PathBuf::from(directory_path)).unwrap();
println!("Read all");
let mut output = File::create(output_path).unwrap();
output.write_all(&buffer[..]).unwrap();
output.flush().unwrap();
}

51
bin/test_static_dir.rs
View File

@@ -1,51 +0,0 @@
use std::env;
use std::path::PathBuf;
use std::fs::File;
use std::io::Write;
extern crate tantivy;
use tantivy::directory::{StaticDirectory, write_static_from_directory};
use tantivy::Index;
use tantivy::query::QueryParser;
use tantivy::collector::TopCollector;
static DATA: &'static [u8] = include_bytes!("output.bin");
fn run() -> tantivy::Result<()> {
// Prints each argument on a separate line
let directory = StaticDirectory::open(DATA).unwrap();
let index = Index::open_directory(directory).unwrap();
index.load_searchers().unwrap();
let searcher = index.searcher();
let schema = index.schema();
let title = schema.get_field("title").unwrap();
let body = schema.get_field("body").unwrap();
let query_parser = QueryParser::for_index(&index, vec![title, body]);
let query = query_parser.parse_query("sea whale")?;
let mut top_collector = TopCollector::with_limit(10);
searcher.search(&*query, &mut top_collector)?;
let doc_addresses = top_collector.docs();
// The actual documents still need to be
// retrieved from Tantivy's store.
//
// Since the body field was not configured as stored,
// the document returned will only contain
// a title.
for doc_address in doc_addresses {
let retrieved_doc = searcher.doc(&doc_address)?;
println!("{}", schema.to_json(&retrieved_doc));
}
Ok(())
}
fn main() {
run().unwrap();
}

23
ci/before_deploy.ps1 Normal file
View File

@@ -0,0 +1,23 @@
# This script takes care of packaging the build artifacts that will go in the
# release zipfile
$SRC_DIR = $PWD.Path
$STAGE = [System.Guid]::NewGuid().ToString()
Set-Location $ENV:Temp
New-Item -Type Directory -Name $STAGE
Set-Location $STAGE
$ZIP = "$SRC_DIR\$($Env:CRATE_NAME)-$($Env:APPVEYOR_REPO_TAG_NAME)-$($Env:TARGET).zip"
# TODO Update this to package the right artifacts
Copy-Item "$SRC_DIR\target\$($Env:TARGET)\release\hello.exe" '.\'
7z a "$ZIP" *
Push-AppveyorArtifact "$ZIP"
Remove-Item *.* -Force
Set-Location ..
Remove-Item $STAGE
Set-Location $SRC_DIR

33
ci/before_deploy.sh Normal file
View File

@@ -0,0 +1,33 @@
# This script takes care of building your crate and packaging it for release
set -ex
main() {
local src=$(pwd) \
stage=
case $TRAVIS_OS_NAME in
linux)
stage=$(mktemp -d)
;;
osx)
stage=$(mktemp -d -t tmp)
;;
esac
test -f Cargo.lock || cargo generate-lockfile
# TODO Update this to build the artifacts that matter to you
cross rustc --bin hello --target $TARGET --release -- -C lto
# TODO Update this to package the right artifacts
cp target/$TARGET/release/hello $stage/
cd $stage
tar czf $src/$CRATE_NAME-$TRAVIS_TAG-$TARGET.tar.gz *
cd $src
rm -rf $stage
}
main

47
ci/install.sh Normal file
View File

@@ -0,0 +1,47 @@
set -ex
main() {
local target=
if [ $TRAVIS_OS_NAME = linux ]; then
target=x86_64-unknown-linux-musl
sort=sort
else
target=x86_64-apple-darwin
sort=gsort # for `sort --sort-version`, from brew's coreutils.
fi
# Builds for iOS are done on OSX, but require the specific target to be
# installed.
case $TARGET in
aarch64-apple-ios)
rustup target install aarch64-apple-ios
;;
armv7-apple-ios)
rustup target install armv7-apple-ios
;;
armv7s-apple-ios)
rustup target install armv7s-apple-ios
;;
i386-apple-ios)
rustup target install i386-apple-ios
;;
x86_64-apple-ios)
rustup target install x86_64-apple-ios
;;
esac
# This fetches latest stable release
local tag=$(git ls-remote --tags --refs --exit-code https://github.com/japaric/cross \
| cut -d/ -f3 \
| grep -E '^v[0.1.0-9.]+$' \
| $sort --version-sort \
| tail -n1)
curl -LSfs https://japaric.github.io/trust/install.sh | \
sh -s -- \
--force \
--git japaric/cross \
--tag $tag \
--target $target
}
main

29
ci/script.sh Normal file
View File

@@ -0,0 +1,29 @@
#!/usr/bin/env bash
# This script takes care of testing your crate
set -ex
main() {
if [ ! -z $CODECOV ]; then
echo "Codecov"
cargo build --verbose && cargo coverage --verbose && bash <(curl -s https://codecov.io/bash) -s target/kcov
else
echo "Build"
cross build --target $TARGET
if [ ! -z $DISABLE_TESTS ]; then
return
fi
echo "Test"
cross test --target $TARGET --no-default-features --features mmap -- --test-threads 1
fi
for example in $(ls examples/*.rs)
do
cargo run --example $(basename $example .rs)
done
}
# we don't run the "test phase" when doing deploys
if [ -z $TRAVIS_TAG ]; then
main
fi

1
doc/.gitignore vendored Normal file
View File

@@ -0,0 +1 @@
book

5
doc/book.toml Normal file
View File

@@ -0,0 +1,5 @@
[book]
authors = ["Paul Masurel"]
multilingual = false
src = "src"
title = "Tantivy, the user guide"

15
doc/src/SUMMARY.md Normal file
View File

@@ -0,0 +1,15 @@
# Summary
[Avant Propos](./avant-propos.md)
- [Segments](./basis.md)
- [Defining your schema](./schema.md)
- [Facetting](./facetting.md)
- [Innerworkings](./innerworkings.md)
- [Inverted index](./inverted_index.md)
- [Best practise](./inverted_index.md)
[Frequently Asked Questions](./faq.md)
[Examples](./examples.md)

34
doc/src/avant-propos.md Normal file
View File

@@ -0,0 +1,34 @@
# Foreword, what is the scope of tantivy?
> Tantivy is a **search** engine **library** for Rust.
If you are familiar with Lucene, it's an excellent approximation to consider tantivy as Lucene for rust. tantivy is heavily inspired by Lucene's design and
they both have the same scope and targetted use cases.
If you are not familiar with Lucene, let's break down our little tagline.
- **Search** here means full-text search : fundamentally, tantivy is here to help you
identify efficiently what are the documents matching a given query in your corpus.
But modern search UI are so much more : text processing, facetting, autocomplete, fuzzy search, good
relevancy, collapsing, highlighting, spatial search.
While some of these features are not available in tantivy yet, all of these are relevant
feature requests. Tantivy's objective is to offer a solid toolbox to create the best search
experience. But keep in mind this is just a toolbox.
Which bring us to the second keyword...
- **Library** means that you will have to write code. tantivy is not an *all-in-one* server solution like elastic search for instance.
Sometimes a functionality will not be available in tantivy because it is too
specific to your use case. By design, tantivy should make it possible to extend
the available set of features using the existing rock-solid datastructures.
Most frequently this will mean writing your own `Collector`, your own `Scorer` or your own
`TokenFilter`... Some of your requirements may also be related to
something closer to architecture or operations. For instance, you may
want to build a large corpus on Hadoop, fine-tune the merge policy to keep your
index sharded in a time-wise fashion, or you may want to convert and existing
index from a different format.
Tantivy exposes a lot of low level API to do all of these things.

77
doc/src/basis.md Normal file
View File

@@ -0,0 +1,77 @@
# Anatomy of an index
## Straight from disk
Tantivy accesses its data using an abstracting trait called `Directory`.
In theory, one can come and override the data access logic. In practise, the
trait somewhat assumes that your data can be mapped to memory, and tantivy
seems deeply married to using `mmap` for its io [^1], and the only persisting
directory shipped with tantivy is the `MmapDirectory`.
While this design has some downsides, this greatly simplifies the source code of
tantivy. Caching is also entirely delegated to the OS.
`tantivy` works entirely (or almost) by directly reading the datastructures as they are layed on disk. As a result, the act of opening an indexing does not involve loading different datastructures from the disk into random access memory : starting a process, opening an index, and performing your first query can typically be done in a matter of milliseconds.
This is an interesting property for a command line search engine, or for some multi-tenant log search engine : spawning a new process for each new query can be a perfectly sensible solution in some use case.
In later chapters, we will discuss tantivy's inverted index data layout.
One key take away is that to achieve great performance, search indexes are extremely compact.
Of course this is crucial to reduce IO, and ensure that as much of our index can sit in RAM.
Also, whenever possible its data is accessed sequentially. Of course, this is an amazing property when tantivy needs to access the data from your spinning hard disk, but this is also
critical for performance, if your data is read from and an `SSD` or even already in your pagecache.
## Segments, and the log method
That kind of compact layout comes at one cost: it prevents our datastructures from being dynamic.
In fact, the `Directory` trait does not even allow you to modify part of a file.
To allow the addition / deletion of documents, and create the illusion that
your index is dynamic (i.e.: adding and deleting documents), tantivy uses a common database trick sometimes referred to as the *log method*.
Let's forget about deletes for a moment.
As you add documents, these documents are processed and stored in a dedicated datastructure, in a `RAM` buffer. This datastructure is not ready for search, but it is useful to receive your data and rearrange it very rapidly.
As you add documents, this buffer will reach its capacity and tantivy will transparently stop adding document to it and start converting this datastructure to its final read-only format on disk. Once written, an brand empty buffer is available to resume adding documents.
The resulting chunk of index obtained after this serialization is called a `Segment`.
> A segment is a self-contained atomic piece of index. It is identified with a UUID, and all of its files are identified using the naming scheme : `<UUID>.*`.
Which brings us to the nature of a tantivy `Index`.
> A tantivy `Index` is a collection of `Segments`.
Physically, this really just means and index is a bunch of segment files in a given `Directory`,
linked together by a `meta.json` file. This transparency can become extremely handy
to get tantivy to fit your use case:
*Example 1* You could for instance use hadoop to build a very large search index in a timely manner, copy all of the resulting segment files in the same directory and edit the `meta.json` to get a functional index.[^2]
*Example 2* You could also disable your merge policy and enforce daily segments. Removing data after one week can then be done very efficiently by just editing the `meta.json` and deleting the files associated to segment `D-7`.
# Merging
As you index more and more data, your index will accumulate more and more segments.
Having a lot of small segments is not really optimal. There is a bit of redundancy in having
all these term dictionary. Also when searching, we will need to do term lookups as many times as we have segments. It can hurt search performance a bit.
That's where merging or compacting comes into place. Tantivy will continuously consider merge
opportunities and start merging segments in the background.
# Indexing throughput, number of indexing threads
[^1]: This may eventually change.
[^2]: Be careful however. By default these files will not be considered as *managed* by tantivy. This means they will never be garbage collected by tantivy, regardless of whether they become obsolete or not.

0
doc/src/best_practise.md.rs Normal file
View File

3
doc/src/examples.md Normal file
View File

@@ -0,0 +1,3 @@
# Examples
- [Basic search](/examples/basic_search.html)

5
doc/src/facetting.md Normal file
View File

@@ -0,0 +1,5 @@
# Facetting
wewew
## weeewe

0
doc/src/faq.md Normal file
View File

1
doc/src/innerworkings.md Normal file
View File

@@ -0,0 +1 @@
# Innerworkings

1
doc/src/inverted_index.md Normal file
View File

@@ -0,0 +1 @@
# Inverted index

1
doc/src/schema.md Normal file
View File

@@ -0,0 +1 @@
# Defining your schema

176
examples/simple_search.rs → examples/basic_search.rs
View File

@@ -1,26 +1,33 @@
extern crate tantivy;
// # Basic Example
//
// This example covers the basic functionalities of
// tantivy.
//
// We will :
// - define our schema
// = create an index in a directory
// - index few documents in our index
// - search for the best document matchings "sea whale"
// - retrieve the best document original content.
extern crate tempdir;
// ---
// Importing tantivy...
#[macro_use]
extern crate serde_json;
use std::path::Path;
use tempdir::TempDir;
use tantivy::Index;
use tantivy::schema::*;
use tantivy::collector::TopCollector;
extern crate tantivy;
use tantivy::collector::TopDocs;
use tantivy::query::QueryParser;
use tantivy::schema::*;
use tantivy::Index;
use tantivy::ReloadPolicy;
use tempdir::TempDir;
fn main() {
fn main() -> tantivy::Result<()> {
// Let's create a temporary directory for the
// sake of this example
if let Ok(dir) = TempDir::new("tantivy_example_dir") {
run_example(dir.path()).unwrap();
dir.close().unwrap();
}
}
let index_path = TempDir::new("tantivy_example_dir")?;
fn run_example(index_path: &Path) -> tantivy::Result<()> {
// # Defining the schema
//
// The Tantivy index requires a very strict schema.
@@ -29,13 +36,13 @@ fn run_example(index_path: &Path) -> tantivy::Result<()> {
// be indexed".
// first we need to define a schema ...
let mut schema_builder = SchemaBuilder::default();
let mut schema_builder = Schema::builder();
// Our first field is title.
// We want full-text search for it, and we also want
// to be able to retrieve the document after the search.
//
// TEXT | STORED is some syntactic sugar to describe
// `TEXT | STORED` is some syntactic sugar to describe
// that.
//
// `TEXT` means the field should be tokenized and indexed,
@@ -64,21 +71,22 @@ fn run_example(index_path: &Path) -> tantivy::Result<()> {
//
// This will actually just save a meta.json
// with our schema in the directory.
let index = Index::create(index_path, schema.clone())?;
let index = Index::create_in_dir(&index_path, schema.clone())?;
// To insert document we need an index writer.
// There must be only one writer at a time.
// This single `IndexWriter` is already
// multithreaded.
//
// Here we use a buffer of 50MB per thread. Using a bigger
// heap for the indexer can increase its throughput.
// Here we give tantivy a budget of `50MB`.
// Using a bigger heap for the indexer may increase
// throughput, but 50 MB is already plenty.
let mut index_writer = index.writer(50_000_000)?;
// Let's index our documents!
// We first need a handle on the title and the body field.
// ### Create a document "manually".
// ### Adding documents
//
// We can create a document manually, by setting the fields
// one by one in a Document object.
@@ -96,46 +104,47 @@ fn run_example(index_path: &Path) -> tantivy::Result<()> {
// ... and add it to the `IndexWriter`.
index_writer.add_document(old_man_doc);
// ### Create a document directly from json.
//
// Alternatively, we can use our schema to parse a
// document object directly from json.
// The document is a string, but we use the `json` macro
// from `serde_json` for the convenience of multi-line support.
let json = json!({
"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"
});
let mice_and_men_doc = schema.parse_document(&json.to_string())?;
// For convenience, tantivy also comes with a macro to
// reduce the boilerplate above.
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"
));
index_writer.add_document(mice_and_men_doc);
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"
));
// Multi-valued field are allowed, they are
// expressed in JSON by an array.
// The following document has two titles.
let json = json!({
"title": ["Frankenstein", "The Modern Prometheus"],
"body": "You will rejoice to hear that no disaster has accompanied the commencement of an \
enterprise which you have regarded with such evil forebodings. I arrived here \
yesterday, and my first task is to assure my dear sister of my welfare and \
increasing confidence in the success of my undertaking."
});
let frankenstein_doc = schema.parse_document(&json.to_string())?;
index_writer.add_document(frankenstein_doc);
// Multivalued field just need to be repeated.
index_writer.add_document(doc!(
title => "Frankenstein",
title => "The Modern Prometheus",
body => "You will rejoice to hear that no disaster has accompanied the commencement of an \
enterprise which you have regarded with such evil forebodings. I arrived here \
yesterday, and my first task is to assure my dear sister of my welfare and \
increasing confidence in the success of my undertaking."
));
// This is an example, so we will only index 3 documents
// here. You can check out tantivy's tutorial to index
// the English wikipedia. Tantivy's indexing is rather fast.
// Indexing 5 million articles of the English wikipedia takes
// around 4 minutes on my computer!
// around 3 minutes on my computer!
// ### Committing
//
@@ -160,16 +169,37 @@ fn run_example(index_path: &Path) -> tantivy::Result<()> {
// # Searching
//
// Let's search our index. Start by reloading
// searchers in the index. This should be done
// after every commit().
index.load_searchers()?;
// Afterwards create one (or more) searchers.
// ### Searcher
//
// You should create a searcher
// every time you start a "search query".
let searcher = index.searcher();
// A reader is required to get search the index.
// It acts as a `Searcher` pool that reloads itself,
// depending on a `ReloadPolicy`.
//
// For a search server you will typically create one reader for the entire lifetime of your
// program, and acquire a new searcher for every single request.
//
// In the code below, we rely on the 'ON_COMMIT' policy: the reader
// will reload the index automatically after each commit.
let reader = index
.reader_builder()
.reload_policy(ReloadPolicy::OnCommit)
.try_into()?;
// We now need to acquire a searcher.
//
// A searcher points to snapshotted, immutable version of the index.
//
// Some search experience might require more than
// one query. Using the same searcher ensures that all of these queries will run on the
// same version of the index.
//
// Acquiring a `searcher` is very cheap.
//
// You should acquire a searcher every time you start processing a request and
// and release it right after your query is finished.
let searcher = reader.searcher();
// ### Query
// The query parser can interpret human queries.
// Here, if the user does not specify which
@@ -193,15 +223,10 @@ fn run_example(index_path: &Path) -> tantivy::Result<()> {
//
// We are not interested in all of the documents but
// only in the top 10. Keeping track of our top 10 best documents
// is the role of the TopCollector.
let mut top_collector = TopCollector::with_limit(10);
// is the role of the TopDocs.
// We can now perform our query.
searcher.search(&*query, &mut top_collector)?;
// Our top collector now contains the 10
// most relevant doc ids...
let doc_addresses = top_collector.docs();
let top_docs = searcher.search(&query, &TopDocs::with_limit(10))?;
// The actual documents still need to be
// retrieved from Tantivy's store.
@@ -209,17 +234,10 @@ fn run_example(index_path: &Path) -> tantivy::Result<()> {
// Since the body field was not configured as stored,
// the document returned will only contain
// a title.
for doc_address in doc_addresses {
let retrieved_doc = searcher.doc(&doc_address)?;
for (_score, doc_address) in top_docs {
let retrieved_doc = searcher.doc(doc_address)?;
println!("{}", schema.to_json(&retrieved_doc));
}
// Wait for indexing and merging threads to shut down.
// Usually this isn't needed, but in `main` we try to
// delete the temporary directory and that fails on
// Windows if the files are still open.
index_writer.wait_merging_threads()?;
Ok(())
}

196
examples/custom_collector.rs Normal file
View File

@@ -0,0 +1,196 @@
// # Custom collector example
//
// This example shows how you can implement your own
// collector. As an example, we will compute a collector
// that computes the standard deviation of a given fast field.
//
// Of course, you can have a look at the tantivy's built-in collectors
// such as the `CountCollector` for more examples.
extern crate tempdir;
// ---
// Importing tantivy...
#[macro_use]
extern crate tantivy;
use tantivy::collector::{Collector, SegmentCollector};
use tantivy::fastfield::FastFieldReader;
use tantivy::query::QueryParser;
use tantivy::schema::Field;
use tantivy::schema::{Schema, FAST, INDEXED, TEXT};
use tantivy::SegmentReader;
use tantivy::{Index, TantivyError};
#[derive(Default)]
struct Stats {
count: usize,
sum: f64,
squared_sum: f64,
}
impl Stats {
pub fn count(&self) -> usize {
self.count
}
pub fn mean(&self) -> f64 {
self.sum / (self.count as f64)
}
fn square_mean(&self) -> f64 {
self.squared_sum / (self.count as f64)
}
pub fn standard_deviation(&self) -> f64 {
let mean = self.mean();
(self.square_mean() - mean * mean).sqrt()
}
fn non_zero_count(self) -> Option<Stats> {
if self.count == 0 {
None
} else {
Some(self)
}
}
}
struct StatsCollector {
field: Field,
}
impl StatsCollector {
fn with_field(field: Field) -> StatsCollector {
StatsCollector { field }
}
}
impl Collector for StatsCollector {
// That's the type of our result.
// Our standard deviation will be a float.
type Fruit = Option<Stats>;
type Child = StatsSegmentCollector;
fn for_segment(
&self,
_segment_local_id: u32,
segment_reader: &SegmentReader,
) -> tantivy::Result<StatsSegmentCollector> {
let fast_field_reader = segment_reader
.fast_fields()
.u64(self.field)
.ok_or_else(|| {
let field_name = segment_reader.schema().get_field_name(self.field);
TantivyError::SchemaError(format!(
"Field {:?} is not a u64 fast field.",
field_name
))
})?;
Ok(StatsSegmentCollector {
fast_field_reader,
stats: Stats::default(),
})
}
fn requires_scoring(&self) -> bool {
// this collector does not care about score.
false
}
fn merge_fruits(&self, segment_stats: Vec<Option<Stats>>) -> tantivy::Result<Option<Stats>> {
let mut stats = Stats::default();
for segment_stats_opt in segment_stats {
if let Some(segment_stats) = segment_stats_opt {
stats.count += segment_stats.count;
stats.sum += segment_stats.sum;
stats.squared_sum += segment_stats.squared_sum;
}
}
Ok(stats.non_zero_count())
}
}
struct StatsSegmentCollector {
fast_field_reader: FastFieldReader<u64>,
stats: Stats,
}
impl SegmentCollector for StatsSegmentCollector {
type Fruit = Option<Stats>;
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;
self.stats.squared_sum += value * value;
}
fn harvest(self) -> <Self as SegmentCollector>::Fruit {
self.stats.non_zero_count()
}
}
fn main() -> tantivy::Result<()> {
// # Defining the schema
//
// The Tantivy index requires a very strict schema.
// The schema declares which fields are in the index,
// and for each field, its type and "the way it should
// be indexed".
// first we need to define a schema ...
let mut schema_builder = Schema::builder();
// We'll assume a fictional index containing
// products, and with a name, a description, and a price.
let product_name = schema_builder.add_text_field("name", TEXT);
let product_description = schema_builder.add_text_field("description", TEXT);
let price = schema_builder.add_u64_field("price", INDEXED | FAST);
let schema = schema_builder.build();
// # Indexing documents
//
// Lets index a bunch of fake documents for the sake of
// this example.
let index = Index::create_in_ram(schema.clone());
let mut index_writer = index.writer(50_000_000)?;
index_writer.add_document(doc!(
product_name => "Super Broom 2000",
product_description => "While it is ok for short distance travel, this broom \
was designed quiditch. It will up your game.",
price => 30_200u64
));
index_writer.add_document(doc!(
product_name => "Turbulobroom",
product_description => "You might have heard of this broom before : it is the sponsor of the Wales team.\
You'll enjoy its sharp turns, and rapid acceleration",
price => 29_240u64
));
index_writer.add_document(doc!(
product_name => "Broomio",
product_description => "Great value for the price. This broom is a market favorite",
price => 21_240u64
));
index_writer.add_document(doc!(
product_name => "Whack a Mole",
product_description => "Prime quality bat.",
price => 5_200u64
));
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let query_parser = QueryParser::for_index(&index, vec![product_name, product_description]);
// here we want to get a hit on the 'ken' in Frankenstein
let query = query_parser.parse_query("broom")?;
if let Some(stats) = searcher.search(&query, &StatsCollector::with_field(price))? {
println!("count: {}", stats.count());
println!("mean: {}", stats.mean());
println!("standard deviation: {}", stats.standard_deviation());
}
Ok(())
}

115
examples/custom_tokenizer.rs Normal file
View File

@@ -0,0 +1,115 @@
// # Defining a tokenizer pipeline
//
// In this example, we'll see how to define a tokenizer pipeline
// by aligning a bunch of `TokenFilter`.
#[macro_use]
extern crate tantivy;
use tantivy::collector::TopDocs;
use tantivy::query::QueryParser;
use tantivy::schema::*;
use tantivy::tokenizer::NgramTokenizer;
use tantivy::Index;
fn main() -> tantivy::Result<()> {
// # Defining the schema
//
// The Tantivy index requires a very strict schema.
// The schema declares which fields are in the index,
// and for each field, its type and "the way it should
// be indexed".
// first we need to define a schema ...
let mut schema_builder = Schema::builder();
// Our first field is title.
// In this example we want to use NGram searching
// we will set that to 3 characters, so any three
// char in the title should be findable.
let text_field_indexing = TextFieldIndexing::default()
.set_tokenizer("ngram3")
.set_index_option(IndexRecordOption::WithFreqsAndPositions);
let text_options = TextOptions::default()
.set_indexing_options(text_field_indexing)
.set_stored();
let title = schema_builder.add_text_field("title", text_options);
// Our second field is body.
// We want full-text search for it, but we do not
// need to be able to be able to retrieve it
// for our application.
//
// We can make our index lighter and
// by omitting `STORED` flag.
let body = schema_builder.add_text_field("body", TEXT);
let schema = schema_builder.build();
// # Indexing documents
//
// Let's create a brand new index.
// To simplify we will work entirely in RAM.
// This is not what you want in reality, but it is very useful
// for your unit tests... Or this example.
let index = Index::create_in_ram(schema.clone());
// here we are registering our custome tokenizer
// this will store tokens of 3 characters each
index
.tokenizers()
.register("ngram3", NgramTokenizer::new(3, 3, false));
// To insert document we need an index writer.
// There must be only one writer at a time.
// This single `IndexWriter` is already
// multithreaded.
//
// Here we use a buffer of 50MB per thread. Using a bigger
// heap for the indexer can increase its throughput.
let mut index_writer = index.writer(50_000_000)?;
index_writer.add_document(doc!(
title => "The Old Man and the Sea",
body => "He was an old man who fished alone in a skiff in the Gulf Stream and \
he had gone eighty-four days now without taking a fish."
));
index_writer.add_document(doc!(
title => "Of Mice and Men",
body => r#"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"#
));
index_writer.add_document(doc!(
title => "Frankenstein",
body => r#"You will rejoice to hear that no disaster has accompanied the commencement of an
enterprise which you have regarded with such evil forebodings. I arrived here
yesterday, and my first task is to assure my dear sister of my welfare and
increasing confidence in the success of my undertaking."#
));
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
// The query parser can interpret human queries.
// Here, if the user does not specify which
// field they want to search, tantivy will search
// in both title and body.
let query_parser = QueryParser::for_index(&index, vec![title, body]);
// here we want to get a hit on the 'ken' in Frankenstein
let query = query_parser.parse_query("ken")?;
let top_docs = searcher.search(&query, &TopDocs::with_limit(10))?;
for (_, doc_address) in top_docs {
let retrieved_doc = searcher.doc(doc_address)?;
println!("{}", schema.to_json(&retrieved_doc));
}
Ok(())
}

146
examples/deleting_updating_documents.rs Normal file
View File

@@ -0,0 +1,146 @@
// # Deleting and Updating (?) documents
//
// This example explains how to delete and update documents.
// In fact there is actually no such thing as an update in tantivy.
//
// To update a document, you need to delete a document and then reinsert
// its new version.
//
// ---
// Importing tantivy...
#[macro_use]
extern crate tantivy;
use tantivy::collector::TopDocs;
use tantivy::query::TermQuery;
use tantivy::schema::*;
use tantivy::Index;
use tantivy::IndexReader;
// A simple helper function to fetch a single document
// given its id from our index.
// It will be helpful to check our work.
fn extract_doc_given_isbn(
reader: &IndexReader,
isbn_term: &Term,
) -> tantivy::Result<Option<Document>> {
let searcher = reader.searcher();
// This is the simplest query you can think of.
// It matches all of the documents containing a specific term.
//
// The second argument is here to tell we don't care about decoding positions,
// or term frequencies.
let term_query = TermQuery::new(isbn_term.clone(), IndexRecordOption::Basic);
let top_docs = searcher.search(&term_query, &TopDocs::with_limit(1))?;
if let Some((_score, doc_address)) = top_docs.first() {
let doc = searcher.doc(*doc_address)?;
Ok(Some(doc))
} else {
// no doc matching this ID.
Ok(None)
}
}
fn main() -> tantivy::Result<()> {
// # Defining the schema
//
// Check out the *basic_search* example if this makes
// small sense to you.
let mut schema_builder = Schema::builder();
// Tantivy does not really have a notion of primary id.
// This may change in the future.
//
// Still, we can create a `isbn` field and use it as an id. This
// field can be `u64` or a `text`, depending on your use case.
// It just needs to be indexed.
//
// If it is `text`, let's make sure to keep it `raw` and let's avoid
// running any text processing on it.
// This is done by associating this field to the tokenizer named `raw`.
// Rather than building our [`TextOptions`](//docs.rs/tantivy/~0/tantivy/schema/struct.TextOptions.html) manually,
// We use the `STRING` shortcut. `STRING` stands for indexed (without term frequency or positions)
// and untokenized.
//
// Because we also want to be able to see this `id` in our returned documents,
// we also mark the field as stored.
let isbn = schema_builder.add_text_field("isbn", STRING | STORED);
let title = schema_builder.add_text_field("title", TEXT | STORED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema.clone());
let mut index_writer = index.writer(50_000_000)?;
// Let's add a couple of documents, for the sake of the example.
let mut old_man_doc = Document::default();
old_man_doc.add_text(title, "The Old Man and the Sea");
index_writer.add_document(doc!(
isbn => "978-0099908401",
title => "The old Man and the see"
));
index_writer.add_document(doc!(
isbn => "978-0140177398",
title => "Of Mice and Men",
));
index_writer.add_document(doc!(
title => "Frankentein", //< Oops there is a typo here.
isbn => "978-9176370711",
));
index_writer.commit()?;
let reader = index.reader()?;
let frankenstein_isbn = Term::from_field_text(isbn, "978-9176370711");
// Oops our frankenstein doc seems mispelled
let frankenstein_doc_misspelled = extract_doc_given_isbn(&reader, &frankenstein_isbn)?.unwrap();
assert_eq!(
schema.to_json(&frankenstein_doc_misspelled),
r#"{"isbn":["978-9176370711"],"title":["Frankentein"]}"#,
);
// # Update = Delete + Insert
//
// Here we will want to update the typo in the `Frankenstein` book.
//
// Tantivy does not handle updates directly, we need to delete
// and reinsert the document.
//
// This can be complicated as it means you need to have access
// to the entire document. It is good practise to integrate tantivy
// with a key value store for this reason.
//
// To remove one of the document, we just call `delete_term`
// on its id.
//
// Note that `tantivy` does nothing to enforce the idea that
// there is only one document associated to this id.
//
// Also you might have noticed that we apply the delete before
// having committed. This does not matter really...
index_writer.delete_term(frankenstein_isbn.clone());
// We now need to reinsert our document without the typo.
index_writer.add_document(doc!(
title => "Frankenstein",
isbn => "978-9176370711",
));
// You are guaranteed that your clients will only observe your index in
// the state it was in after a commit.
// In this example, your search engine will at no point be missing the *Frankenstein* document.
// Everything happened as if the document was updated.
index_writer.commit()?;
// We reload our searcher to make our change available to clients.
reader.reload()?;
// No more typo!
let frankenstein_new_doc = extract_doc_given_isbn(&reader, &frankenstein_isbn)?.unwrap();
assert_eq!(
schema.to_json(&frankenstein_new_doc),
r#"{"isbn":["978-9176370711"],"title":["Frankenstein"]}"#,
);
Ok(())
}

80
examples/faceted_search.rs Normal file
View File

@@ -0,0 +1,80 @@
// # Basic Example
//
// This example covers the basic functionalities of
// tantivy.
//
// We will :
// - define our schema
// = create an index in a directory
// - index few documents in our index
// - search for the best document matchings "sea whale"
// - retrieve the best document original content.
extern crate tempdir;
// ---
// Importing tantivy...
#[macro_use]
extern crate tantivy;
use tantivy::collector::FacetCollector;
use tantivy::query::AllQuery;
use tantivy::schema::*;
use tantivy::Index;
fn main() -> tantivy::Result<()> {
// Let's create a temporary directory for the
// sake of this example
let index_path = TempDir::new("tantivy_facet_example_dir")?;
let mut schema_builder = Schema::builder();
schema_builder.add_text_field("name", TEXT | STORED);
// this is our faceted field
schema_builder.add_facet_field("tags");
let schema = schema_builder.build();
let index = Index::create_in_dir(&index_path, schema.clone())?;
let mut index_writer = index.writer(50_000_000)?;
let name = schema.get_field("name").unwrap();
let tags = schema.get_field("tags").unwrap();
// For convenience, tantivy also comes with a macro to
// reduce the boilerplate above.
index_writer.add_document(doc!(
name => "the ditch",
tags => Facet::from("/pools/north")
));
index_writer.add_document(doc!(
name => "little stacey",
tags => Facet::from("/pools/south")
));
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let mut facet_collector = FacetCollector::for_field(tags);
facet_collector.add_facet("/pools");
let facet_counts = searcher.search(&AllQuery, &facet_collector).unwrap();
// This lists all of the facet counts
let facets: Vec<(&Facet, u64)> = facet_counts.get("/pools").collect();
assert_eq!(
facets,
vec![
(&Facet::from("/pools/north"), 1),
(&Facet::from("/pools/south"), 1),
]
);
Ok(())
}
use tempdir::TempDir;

2
examples/generate_html.sh
View File

@@ -1,2 +0,0 @@
#!/bin/bash
docco simple_search.rs -o html

518
examples/html/docco.css
View File

@@ -1,518 +0,0 @@
/*--------------------- Typography ----------------------------*/
@font-face {
font-family: 'aller-light';
src: url('public/fonts/aller-light.eot');
src: url('public/fonts/aller-light.eot?#iefix') format('embedded-opentype'),
url('public/fonts/aller-light.woff') format('woff'),
url('public/fonts/aller-light.ttf') format('truetype');
font-weight: normal;
font-style: normal;
}
@font-face {
font-family: 'aller-bold';
src: url('public/fonts/aller-bold.eot');
src: url('public/fonts/aller-bold.eot?#iefix') format('embedded-opentype'),
url('public/fonts/aller-bold.woff') format('woff'),
url('public/fonts/aller-bold.ttf') format('truetype');
font-weight: normal;
font-style: normal;
}
@font-face {
font-family: 'roboto-black';
src: url('public/fonts/roboto-black.eot');
src: url('public/fonts/roboto-black.eot?#iefix') format('embedded-opentype'),
url('public/fonts/roboto-black.woff') format('woff'),
url('public/fonts/roboto-black.ttf') format('truetype');
font-weight: normal;
font-style: normal;
}
/*--------------------- Layout ----------------------------*/
html { height: 100%; }
body {
font-family: "aller-light";
font-size: 14px;
line-height: 18px;
color: #30404f;
margin: 0; padding: 0;
height:100%;
}
#container { min-height: 100%; }
a {
color: #000;
}
b, strong {
font-weight: normal;
font-family: "aller-bold";
}
p {
margin: 15px 0 0px;
}
.annotation ul, .annotation ol {
margin: 25px 0;
}
.annotation ul li, .annotation ol li {
font-size: 14px;
line-height: 18px;
margin: 10px 0;
}
h1, h2, h3, h4, h5, h6 {
color: #112233;
line-height: 1em;
font-weight: normal;
font-family: "roboto-black";
text-transform: uppercase;
margin: 30px 0 15px 0;
}
h1 {
margin-top: 40px;
}
h2 {
font-size: 1.26em;
}
hr {
border: 0;
background: 1px #ddd;
height: 1px;
margin: 20px 0;
}
pre, tt, code {
font-size: 12px; line-height: 16px;
font-family: Menlo, Monaco, Consolas, "Lucida Console", monospace;
margin: 0; padding: 0;
}
.annotation pre {
display: block;
margin: 0;
padding: 7px 10px;
background: #fcfcfc;
-moz-box-shadow: inset 0 0 10px rgba(0,0,0,0.1);
-webkit-box-shadow: inset 0 0 10px rgba(0,0,0,0.1);
box-shadow: inset 0 0 10px rgba(0,0,0,0.1);
overflow-x: auto;
}
.annotation pre code {
border: 0;
padding: 0;
background: transparent;
}
blockquote {
border-left: 5px solid #ccc;
margin: 0;
padding: 1px 0 1px 1em;
}
.sections blockquote p {
font-family: Menlo, Consolas, Monaco, monospace;
font-size: 12px; line-height: 16px;
color: #999;
margin: 10px 0 0;
white-space: pre-wrap;
}
ul.sections {
list-style: none;
padding:0 0 5px 0;;
margin:0;
}
/*
Force border-box so that % widths fit the parent
container without overlap because of margin/padding.
More Info : http://www.quirksmode.org/css/box.html
*/
ul.sections > li > div {
-moz-box-sizing: border-box; /* firefox */
-ms-box-sizing: border-box; /* ie */
-webkit-box-sizing: border-box; /* webkit */
-khtml-box-sizing: border-box; /* konqueror */
box-sizing: border-box; /* css3 */
}
/*---------------------- Jump Page -----------------------------*/
#jump_to, #jump_page {
margin: 0;
background: white;
-webkit-box-shadow: 0 0 25px #777; -moz-box-shadow: 0 0 25px #777;
-webkit-border-bottom-left-radius: 5px; -moz-border-radius-bottomleft: 5px;
font: 16px Arial;
cursor: pointer;
text-align: right;
list-style: none;
}
#jump_to a {
text-decoration: none;
}
#jump_to a.large {
display: none;
}
#jump_to a.small {
font-size: 22px;
font-weight: bold;
color: #676767;
}
#jump_to, #jump_wrapper {
position: fixed;
right: 0; top: 0;
padding: 10px 15px;
margin:0;
}
#jump_wrapper {
display: none;
padding:0;
}
#jump_to:hover #jump_wrapper {
display: block;
}
#jump_page_wrapper{
position: fixed;
right: 0;
top: 0;
bottom: 0;
}
#jump_page {
padding: 5px 0 3px;
margin: 0 0 25px 25px;
max-height: 100%;
overflow: auto;
}
#jump_page .source {
display: block;
padding: 15px;
text-decoration: none;
border-top: 1px solid #eee;
}
#jump_page .source:hover {
background: #f5f5ff;
}
#jump_page .source:first-child {
}
/*---------------------- Low resolutions (> 320px) ---------------------*/
@media only screen and (min-width: 320px) {
.pilwrap { display: none; }
ul.sections > li > div {
display: block;
padding:5px 10px 0 10px;
}
ul.sections > li > div.annotation ul, ul.sections > li > div.annotation ol {
padding-left: 30px;
}
ul.sections > li > div.content {
overflow-x:auto;
-webkit-box-shadow: inset 0 0 5px #e5e5ee;
box-shadow: inset 0 0 5px #e5e5ee;
border: 1px solid #dedede;
margin:5px 10px 5px 10px;
padding-bottom: 5px;
}
ul.sections > li > div.annotation pre {
margin: 7px 0 7px;
padding-left: 15px;
}
ul.sections > li > div.annotation p tt, .annotation code {
background: #f8f8ff;
border: 1px solid #dedede;
font-size: 12px;
padding: 0 0.2em;
}
}
/*---------------------- (> 481px) ---------------------*/
@media only screen and (min-width: 481px) {
#container {
position: relative;
}
body {
background-color: #F5F5FF;
font-size: 15px;
line-height: 21px;
}
pre, tt, code {
line-height: 18px;
}
p, ul, ol {
margin: 0 0 15px;
}
#jump_to {
padding: 5px 10px;
}
#jump_wrapper {
padding: 0;
}
#jump_to, #jump_page {
font: 10px Arial;
text-transform: uppercase;
}
#jump_page .source {
padding: 5px 10px;
}
#jump_to a.large {
display: inline-block;
}
#jump_to a.small {
display: none;
}
#background {
position: absolute;
top: 0; bottom: 0;
width: 350px;
background: #fff;
border-right: 1px solid #e5e5ee;
z-index: -1;
}
ul.sections > li > div.annotation ul, ul.sections > li > div.annotation ol {
padding-left: 40px;
}
ul.sections > li {
white-space: nowrap;
}
ul.sections > li > div {
display: inline-block;
}
ul.sections > li > div.annotation {
max-width: 350px;
min-width: 350px;
min-height: 5px;
padding: 13px;
overflow-x: hidden;
white-space: normal;
vertical-align: top;
text-align: left;
}
ul.sections > li > div.annotation pre {
margin: 15px 0 15px;
padding-left: 15px;
}
ul.sections > li > div.content {
padding: 13px;
vertical-align: top;
border: none;
-webkit-box-shadow: none;
box-shadow: none;
}
.pilwrap {
position: relative;
display: inline;
}
.pilcrow {
font: 12px Arial;
text-decoration: none;
color: #454545;
position: absolute;
top: 3px; left: -20px;
padding: 1px 2px;
opacity: 0;
-webkit-transition: opacity 0.2s linear;
}
.for-h1 .pilcrow {
top: 47px;
}
.for-h2 .pilcrow, .for-h3 .pilcrow, .for-h4 .pilcrow {
top: 35px;
}
ul.sections > li > div.annotation:hover .pilcrow {
opacity: 1;
}
}
/*---------------------- (> 1025px) ---------------------*/
@media only screen and (min-width: 1025px) {
body {
font-size: 16px;
line-height: 24px;
}
#background {
width: 525px;
}
ul.sections > li > div.annotation {
max-width: 525px;
min-width: 525px;
padding: 10px 25px 1px 50px;
}
ul.sections > li > div.content {
padding: 9px 15px 16px 25px;
}
}
/*---------------------- Syntax Highlighting -----------------------------*/
td.linenos { background-color: #f0f0f0; padding-right: 10px; }
span.lineno { background-color: #f0f0f0; padding: 0 5px 0 5px; }
/*
github.com style (c) Vasily Polovnyov <vast@whiteants.net>
*/
pre code {
display: block; padding: 0.5em;
color: #000;
background: #f8f8ff
}
pre .hljs-comment,
pre .hljs-template_comment,
pre .hljs-diff .hljs-header,
pre .hljs-javadoc {
color: #408080;
font-style: italic
}
pre .hljs-keyword,
pre .hljs-assignment,
pre .hljs-literal,
pre .hljs-css .hljs-rule .hljs-keyword,
pre .hljs-winutils,
pre .hljs-javascript .hljs-title,
pre .hljs-lisp .hljs-title,
pre .hljs-subst {
color: #954121;
/*font-weight: bold*/
}
pre .hljs-number,
pre .hljs-hexcolor {
color: #40a070
}
pre .hljs-string,
pre .hljs-tag .hljs-value,
pre .hljs-phpdoc,
pre .hljs-tex .hljs-formula {
color: #219161;
}
pre .hljs-title,
pre .hljs-id {
color: #19469D;
}
pre .hljs-params {
color: #00F;
}
pre .hljs-javascript .hljs-title,
pre .hljs-lisp .hljs-title,
pre .hljs-subst {
font-weight: normal
}
pre .hljs-class .hljs-title,
pre .hljs-haskell .hljs-label,
pre .hljs-tex .hljs-command {
color: #458;
font-weight: bold
}
pre .hljs-tag,
pre .hljs-tag .hljs-title,
pre .hljs-rules .hljs-property,
pre .hljs-django .hljs-tag .hljs-keyword {
color: #000080;
font-weight: normal
}
pre .hljs-attribute,
pre .hljs-variable,
pre .hljs-instancevar,
pre .hljs-lisp .hljs-body {
color: #008080
}
pre .hljs-regexp {
color: #B68
}
pre .hljs-class {
color: #458;
font-weight: bold
}
pre .hljs-symbol,
pre .hljs-ruby .hljs-symbol .hljs-string,
pre .hljs-ruby .hljs-symbol .hljs-keyword,
pre .hljs-ruby .hljs-symbol .hljs-keymethods,
pre .hljs-lisp .hljs-keyword,
pre .hljs-tex .hljs-special,
pre .hljs-input_number {
color: #990073
}
pre .hljs-builtin,
pre .hljs-constructor,
pre .hljs-built_in,
pre .hljs-lisp .hljs-title {
color: #0086b3
}
pre .hljs-preprocessor,
pre .hljs-pi,
pre .hljs-doctype,
pre .hljs-shebang,
pre .hljs-cdata {
color: #999;
font-weight: bold
}
pre .hljs-deletion {
background: #fdd
}
pre .hljs-addition {
background: #dfd
}
pre .hljs-diff .hljs-change {
background: #0086b3
}
pre .hljs-chunk {
color: #aaa
}
pre .hljs-tex .hljs-formula {
opacity: 0.5;
}

BIN
examples/html/public/fonts/aller-bold.eot
View File

Binary file not shown.

BIN
examples/html/public/fonts/aller-bold.ttf
View File

Binary file not shown.

BIN
examples/html/public/fonts/aller-bold.woff
View File

Binary file not shown.

BIN
examples/html/public/fonts/aller-light.eot
View File

Binary file not shown.

BIN
examples/html/public/fonts/aller-light.ttf
View File

Binary file not shown.

BIN
examples/html/public/fonts/aller-light.woff
View File

Binary file not shown.

BIN
examples/html/public/fonts/fleurons.eot
View File

Binary file not shown.

BIN
examples/html/public/fonts/fleurons.ttf
View File

Binary file not shown.

BIN
examples/html/public/fonts/fleurons.woff
View File

Binary file not shown.

BIN
examples/html/public/fonts/roboto-black.eot
View File

Binary file not shown.

BIN
examples/html/public/fonts/roboto-black.ttf
View File

Binary file not shown.

BIN
examples/html/public/fonts/roboto-black.woff
View File

Binary file not shown.

BIN
examples/html/public/images/gray.png
View File

Binary file not shown.
Side by Side

Before

Width:  |  Height:  |  Size: 56 KiB

375
examples/html/public/stylesheets/normalize.css vendored
View File

@@ -1,375 +0,0 @@
/*! normalize.css v2.0.1 | MIT License | git.io/normalize */
/* ==========================================================================
HTML5 display definitions
========================================================================== */
/*
* Corrects `block` display not defined in IE 8/9.
*/
article,
aside,
details,
figcaption,
figure,
footer,
header,
hgroup,
nav,
section,
summary {
display: block;
}
/*
* Corrects `inline-block` display not defined in IE 8/9.
*/
audio,
canvas,
video {
display: inline-block;
}
/*
* Prevents modern browsers from displaying `audio` without controls.
* Remove excess height in iOS 5 devices.
*/
audio:not([controls]) {
display: none;
height: 0;
}
/*
* Addresses styling for `hidden` attribute not present in IE 8/9.
*/
[hidden] {
display: none;
}
/* ==========================================================================
Base
========================================================================== */
/*
* 1. Sets default font family to sans-serif.
* 2. Prevents iOS text size adjust after orientation change, without disabling
* user zoom.
*/
html {
font-family: sans-serif; /* 1 */
-webkit-text-size-adjust: 100%; /* 2 */
-ms-text-size-adjust: 100%; /* 2 */
}
/*
* Removes default margin.
*/
body {
margin: 0;
}
/* ==========================================================================
Links
========================================================================== */
/*
* Addresses `outline` inconsistency between Chrome and other browsers.
*/
a:focus {
outline: thin dotted;
}
/*
* Improves readability when focused and also mouse hovered in all browsers.
*/
a:active,
a:hover {
outline: 0;
}
/* ==========================================================================
Typography
========================================================================== */
/*
* Addresses `h1` font sizes within `section` and `article` in Firefox 4+,
* Safari 5, and Chrome.
*/
h1 {
font-size: 2em;
}
/*
* Addresses styling not present in IE 8/9, Safari 5, and Chrome.
*/
abbr[title] {
border-bottom: 1px dotted;
}
/*
* Addresses style set to `bolder` in Firefox 4+, Safari 5, and Chrome.
*/
b,
strong {
font-weight: bold;
}
/*
* Addresses styling not present in Safari 5 and Chrome.
*/
dfn {
font-style: italic;
}
/*
* Addresses styling not present in IE 8/9.
*/
mark {
background: #ff0;
color: #000;
}
/*
* Corrects font family set oddly in Safari 5 and Chrome.
*/
code,
kbd,
pre,
samp {
font-family: monospace, serif;
font-size: 1em;
}
/*
* Improves readability of pre-formatted text in all browsers.
*/
pre {
white-space: pre;
white-space: pre-wrap;
word-wrap: break-word;
}
/*
* Sets consistent quote types.
*/
q {
quotes: "\201C" "\201D" "\2018" "\2019";
}
/*
* Addresses inconsistent and variable font size in all browsers.
*/
small {
font-size: 80%;
}
/*
* Prevents `sub` and `sup` affecting `line-height` in all browsers.
*/
sub,
sup {
font-size: 75%;
line-height: 0;
position: relative;
vertical-align: baseline;
}
sup {
top: -0.5em;
}
sub {
bottom: -0.25em;
}
/* ==========================================================================
Embedded content
========================================================================== */
/*
* Removes border when inside `a` element in IE 8/9.
*/
img {
border: 0;
}
/*
* Corrects overflow displayed oddly in IE 9.
*/
svg:not(:root) {
overflow: hidden;
}
/* ==========================================================================
Figures
========================================================================== */
/*
* Addresses margin not present in IE 8/9 and Safari 5.
*/
figure {
margin: 0;
}
/* ==========================================================================
Forms
========================================================================== */
/*
* Define consistent border, margin, and padding.
*/
fieldset {
border: 1px solid #c0c0c0;
margin: 0 2px;
padding: 0.35em 0.625em 0.75em;
}
/*
* 1. Corrects color not being inherited in IE 8/9.
* 2. Remove padding so people aren't caught out if they zero out fieldsets.
*/
legend {
border: 0; /* 1 */
padding: 0; /* 2 */
}
/*
* 1. Corrects font family not being inherited in all browsers.
* 2. Corrects font size not being inherited in all browsers.
* 3. Addresses margins set differently in Firefox 4+, Safari 5, and Chrome
*/
button,
input,
select,
textarea {
font-family: inherit; /* 1 */
font-size: 100%; /* 2 */
margin: 0; /* 3 */
}
/*
* Addresses Firefox 4+ setting `line-height` on `input` using `!important` in
* the UA stylesheet.
*/
button,
input {
line-height: normal;
}
/*
* 1. Avoid the WebKit bug in Android 4.0.* where (2) destroys native `audio`
* and `video` controls.
* 2. Corrects inability to style clickable `input` types in iOS.
* 3. Improves usability and consistency of cursor style between image-type
* `input` and others.
*/
button,
html input[type="button"], /* 1 */
input[type="reset"],
input[type="submit"] {
-webkit-appearance: button; /* 2 */
cursor: pointer; /* 3 */
}
/*
* Re-set default cursor for disabled elements.
*/
button[disabled],
input[disabled] {
cursor: default;
}
/*
* 1. Addresses box sizing set to `content-box` in IE 8/9.
* 2. Removes excess padding in IE 8/9.
*/
input[type="checkbox"],
input[type="radio"] {
box-sizing: border-box; /* 1 */
padding: 0; /* 2 */
}
/*
* 1. Addresses `appearance` set to `searchfield` in Safari 5 and Chrome.
* 2. Addresses `box-sizing` set to `border-box` in Safari 5 and Chrome
* (include `-moz` to future-proof).
*/
input[type="search"] {
-webkit-appearance: textfield; /* 1 */
-moz-box-sizing: content-box;
-webkit-box-sizing: content-box; /* 2 */
box-sizing: content-box;
}
/*
* Removes inner padding and search cancel button in Safari 5 and Chrome
* on OS X.
*/
input[type="search"]::-webkit-search-cancel-button,
input[type="search"]::-webkit-search-decoration {
-webkit-appearance: none;
}
/*
* Removes inner padding and border in Firefox 4+.
*/
button::-moz-focus-inner,
input::-moz-focus-inner {
border: 0;
padding: 0;
}
/*
* 1. Removes default vertical scrollbar in IE 8/9.
* 2. Improves readability and alignment in all browsers.
*/
textarea {
overflow: auto; /* 1 */
vertical-align: top; /* 2 */
}
/* ==========================================================================
Tables
========================================================================== */
/*
* Remove most spacing between table cells.
*/
table {
border-collapse: collapse;
border-spacing: 0;
}

542
examples/html/simple_search.html
View File

@@ -1,542 +0,0 @@
<!DOCTYPE html>
<html>
<head>
<title>simple_search.rs</title>
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
<meta name="viewport" content="width=device-width, target-densitydpi=160dpi, initial-scale=1.0; maximum-scale=1.0; user-scalable=0;">
<link rel="stylesheet" media="all" href="docco.css" />
</head>
<body>
<div id="container">
<div id="background"></div>
<ul class="sections">
<li id="title">
<div class="annotation">
<h1>simple_search.rs</h1>
</div>
</li>
<li id="section-1">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-1">&#182;</a>
</div>
</div>
<div class="content"><div class='highlight'><pre><span class="hljs-keyword">extern</span> <span class="hljs-keyword">crate</span> tantivy;
<span class="hljs-keyword">extern</span> <span class="hljs-keyword">crate</span> tempdir;
<span class="hljs-meta">#[macro_use]</span>
<span class="hljs-keyword">extern</span> <span class="hljs-keyword">crate</span> serde_json;
<span class="hljs-keyword">use</span> std::path::Path;
<span class="hljs-keyword">use</span> tempdir::TempDir;
<span class="hljs-keyword">use</span> tantivy::Index;
<span class="hljs-keyword">use</span> tantivy::schema::*;
<span class="hljs-keyword">use</span> tantivy::collector::TopCollector;
<span class="hljs-keyword">use</span> tantivy::query::QueryParser;
<span class="hljs-function"><span class="hljs-keyword">fn</span> <span class="hljs-title">main</span></span>() {</pre></div></div>
</li>
<li id="section-2">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-2">&#182;</a>
</div>
<p>Let’s create a temporary directory for the
sake of this example</p>
</div>
<div class="content"><div class='highlight'><pre> <span class="hljs-keyword">if</span> <span class="hljs-keyword">let</span> <span class="hljs-literal">Ok</span>(dir) = TempDir::new(<span class="hljs-string">"tantivy_example_dir"</span>) {
run_example(dir.path()).unwrap();
dir.close().unwrap();
}
}
<span class="hljs-function"><span class="hljs-keyword">fn</span> <span class="hljs-title">run_example</span></span>(index_path: &amp;Path) -&gt; tantivy::<span class="hljs-built_in">Result</span>&lt;()&gt; {</pre></div></div>
</li>
<li id="section-3">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-3">&#182;</a>
</div>
<h1 id="defining-the-schema">Defining the schema</h1>
<p>The Tantivy index requires a very strict schema.
The schema declares which fields are in the index,
and for each field, its type and “the way it should
be indexed”.</p>
</div>
</li>
<li id="section-4">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-4">&#182;</a>
</div>
<p>first we need to define a schema …</p>
</div>
<div class="content"><div class='highlight'><pre> <span class="hljs-keyword">let</span> <span class="hljs-keyword">mut</span> schema_builder = SchemaBuilder::<span class="hljs-keyword">default</span>();</pre></div></div>
</li>
<li id="section-5">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-5">&#182;</a>
</div>
<p>Our first field is title.
We want full-text search for it, and we also want
to be able to retrieve the document after the search.</p>
<p>TEXT | STORED is some syntactic sugar to describe
that.</p>
<p><code>TEXT</code> means the field should be tokenized and indexed,
along with its term frequency and term positions.</p>
<p><code>STORED</code> means that the field will also be saved
in a compressed, row-oriented key-value store.
This store is useful to reconstruct the
documents that were selected during the search phase.</p>
</div>
<div class="content"><div class='highlight'><pre> schema_builder.add_text_field(<span class="hljs-string">"title"</span>, TEXT | STORED);</pre></div></div>
</li>
<li id="section-6">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-6">&#182;</a>
</div>
<p>Our second field is body.
We want full-text search for it, but we do not
need to be able to be able to retrieve it
for our application. </p>
<p>We can make our index lighter and
by omitting <code>STORED</code> flag.</p>
</div>
<div class="content"><div class='highlight'><pre> schema_builder.add_text_field(<span class="hljs-string">"body"</span>, TEXT);
<span class="hljs-keyword">let</span> schema = schema_builder.build();</pre></div></div>
</li>
<li id="section-7">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-7">&#182;</a>
</div>
<h1 id="indexing-documents">Indexing documents</h1>
<p>Let’s create a brand new index.</p>
<p>This will actually just save a meta.json
with our schema in the directory.</p>
</div>
<div class="content"><div class='highlight'><pre> <span class="hljs-keyword">let</span> index = Index::create(index_path, schema.clone())?;</pre></div></div>
</li>
<li id="section-8">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-8">&#182;</a>
</div>
<p>To insert document we need an index writer.
There must be only one writer at a time.
This single <code>IndexWriter</code> is already
multithreaded.</p>
<p>Here we use a buffer of 50MB per thread. Using a bigger
heap for the indexer can increase its throughput.</p>
</div>
<div class="content"><div class='highlight'><pre> <span class="hljs-keyword">let</span> <span class="hljs-keyword">mut</span> index_writer = index.writer(<span class="hljs-number">50_000_000</span>)?;</pre></div></div>
</li>
<li id="section-9">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-9">&#182;</a>
</div>
<p>Let’s index our documents!
We first need a handle on the title and the body field.</p>
</div>
</li>
<li id="section-10">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-10">&#182;</a>
</div>
<h3 id="create-a-document-manually-">Create a document “manually”.</h3>
<p>We can create a document manually, by setting the fields
one by one in a Document object.</p>
</div>
<div class="content"><div class='highlight'><pre> <span class="hljs-keyword">let</span> title = schema.get_field(<span class="hljs-string">"title"</span>).unwrap();
<span class="hljs-keyword">let</span> body = schema.get_field(<span class="hljs-string">"body"</span>).unwrap();
<span class="hljs-keyword">let</span> <span class="hljs-keyword">mut</span> old_man_doc = Document::<span class="hljs-keyword">default</span>();
old_man_doc.add_text(title, <span class="hljs-string">"The Old Man and the Sea"</span>);
old_man_doc.add_text(
body,
<span class="hljs-string">"He was an old man who fished alone in a skiff in the Gulf Stream and \
he had gone eighty-four days now without taking a fish."</span>,
);</pre></div></div>
</li>
<li id="section-11">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-11">&#182;</a>
</div>
<p>… and add it to the <code>IndexWriter</code>.</p>
</div>
<div class="content"><div class='highlight'><pre> index_writer.add_document(old_man_doc);</pre></div></div>
</li>
<li id="section-12">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-12">&#182;</a>
</div>
<h3 id="create-a-document-directly-from-json-">Create a document directly from json.</h3>
<p>Alternatively, we can use our schema to parse a
document object directly from json.
The document is a string, but we use the <code>json</code> macro
from <code>serde_json</code> for the convenience of multi-line support.</p>
</div>
<div class="content"><div class='highlight'><pre> <span class="hljs-keyword">let</span> json = json!({
<span class="hljs-string">"title"</span>: <span class="hljs-string">"Of Mice and Men"</span>,
<span class="hljs-string">"body"</span>: <span class="hljs-string">"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"</span>
});
<span class="hljs-keyword">let</span> mice_and_men_doc = schema.parse_document(&amp;json.to_string())?;
index_writer.add_document(mice_and_men_doc);</pre></div></div>
</li>
<li id="section-13">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-13">&#182;</a>
</div>
<p>Multi-valued field are allowed, they are
expressed in JSON by an array.
The following document has two titles.</p>
</div>
<div class="content"><div class='highlight'><pre> <span class="hljs-keyword">let</span> json = json!({
<span class="hljs-string">"title"</span>: [<span class="hljs-string">"Frankenstein"</span>, <span class="hljs-string">"The Modern Prometheus"</span>],
<span class="hljs-string">"body"</span>: <span class="hljs-string">"You will rejoice to hear that no disaster has accompanied the commencement of an \
enterprise which you have regarded with such evil forebodings. I arrived here \
yesterday, and my first task is to assure my dear sister of my welfare and \
increasing confidence in the success of my undertaking."</span>
});
<span class="hljs-keyword">let</span> frankenstein_doc = schema.parse_document(&amp;json.to_string())?;
index_writer.add_document(frankenstein_doc);</pre></div></div>
</li>
<li id="section-14">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-14">&#182;</a>
</div>
<p>This is an example, so we will only index 3 documents
here. You can check out tantivy’s tutorial to index
the English wikipedia. Tantivy’s indexing is rather fast.
Indexing 5 million articles of the English wikipedia takes
around 4 minutes on my computer!</p>
</div>
</li>
<li id="section-15">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-15">&#182;</a>
</div>
<h3 id="committing">Committing</h3>
<p>At this point our documents are not searchable.</p>
<p>We need to call .commit() explicitly to force the
index_writer to finish processing the documents in the queue,
flush the current index to the disk, and advertise
the existence of new documents.</p>
<p>This call is blocking.</p>
</div>
<div class="content"><div class='highlight'><pre> index_writer.commit()?;</pre></div></div>
</li>
<li id="section-16">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-16">&#182;</a>
</div>
<p>If <code>.commit()</code> returns correctly, then all of the
documents that have been added are guaranteed to be
persistently indexed.</p>
<p>In the scenario of a crash or a power failure,
tantivy behaves as if has rolled back to its last
commit.</p>
</div>
</li>
<li id="section-17">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-17">&#182;</a>
</div>
<h1 id="searching">Searching</h1>
<p>Let’s search our index. Start by reloading
searchers in the index. This should be done
after every commit().</p>
</div>
<div class="content"><div class='highlight'><pre> index.load_searchers()?;</pre></div></div>
</li>
<li id="section-18">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-18">&#182;</a>
</div>
<p>Afterwards create one (or more) searchers.</p>
<p>You should create a searcher
every time you start a “search query”.</p>
</div>
<div class="content"><div class='highlight'><pre> <span class="hljs-keyword">let</span> searcher = index.searcher();</pre></div></div>
</li>
<li id="section-19">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-19">&#182;</a>
</div>
<p>The query parser can interpret human queries.
Here, if the user does not specify which
field they want to search, tantivy will search
in both title and body.</p>
</div>
<div class="content"><div class='highlight'><pre> <span class="hljs-keyword">let</span> <span class="hljs-keyword">mut</span> query_parser = QueryParser::for_index(index, <span class="hljs-built_in">vec!</span>[title, body]);</pre></div></div>
</li>
<li id="section-20">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-20">&#182;</a>
</div>
<p>QueryParser may fail if the query is not in the right
format. For user facing applications, this can be a problem.
A ticket has been opened regarding this problem.</p>
</div>
<div class="content"><div class='highlight'><pre> <span class="hljs-keyword">let</span> query = query_parser.parse_query(<span class="hljs-string">"sea whale"</span>)?;</pre></div></div>
</li>
<li id="section-21">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-21">&#182;</a>
</div>
<p>A query defines a set of documents, as
well as the way they should be scored.</p>
<p>A query created by the query parser is scored according
to a metric called Tf-Idf, and will consider
any document matching at least one of our terms.</p>
</div>
</li>
<li id="section-22">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-22">&#182;</a>
</div>
<h3 id="collectors">Collectors</h3>
<p>We are not interested in all of the documents but
only in the top 10. Keeping track of our top 10 best documents
is the role of the TopCollector.</p>
</div>
<div class="content"><div class='highlight'><pre> <span class="hljs-keyword">let</span> <span class="hljs-keyword">mut</span> top_collector = TopCollector::with_limit(<span class="hljs-number">10</span>);</pre></div></div>
</li>
<li id="section-23">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-23">&#182;</a>
</div>
<p>We can now perform our query.</p>
</div>
<div class="content"><div class='highlight'><pre> searcher.search(&amp;*query, &amp;<span class="hljs-keyword">mut</span> top_collector)?;</pre></div></div>
</li>
<li id="section-24">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-24">&#182;</a>
</div>
<p>Our top collector now contains the 10
most relevant doc ids…</p>
</div>
<div class="content"><div class='highlight'><pre> <span class="hljs-keyword">let</span> doc_addresses = top_collector.docs();</pre></div></div>
</li>
<li id="section-25">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-25">&#182;</a>
</div>
<p>The actual documents still need to be
retrieved from Tantivy’s store.</p>
<p>Since the body field was not configured as stored,
the document returned will only contain
a title.</p>
</div>
<div class="content"><div class='highlight'><pre>
<span class="hljs-keyword">for</span> doc_address <span class="hljs-keyword">in</span> doc_addresses {
<span class="hljs-keyword">let</span> retrieved_doc = searcher.doc(&amp;doc_address)?;
<span class="hljs-built_in">println!</span>(<span class="hljs-string">"{}"</span>, schema.to_json(&amp;retrieved_doc));
}</pre></div></div>
</li>
<li id="section-26">
<div class="annotation">
<div class="pilwrap ">
<a class="pilcrow" href="#section-26">&#182;</a>
</div>
<p>Wait for indexing and merging threads to shut down.
Usually this isn’t needed, but in <code>main</code> we try to
delete the temporary directory and that fails on
Windows if the files are still open.</p>
</div>
<div class="content"><div class='highlight'><pre> index_writer.wait_merging_threads()?;
<span class="hljs-literal">Ok</span>(())
}</pre></div></div>
</li>
</ul>
</div>
</body>
</html>

43
examples/integer_range_search.rs Normal file
View File

@@ -0,0 +1,43 @@
// # Searching a range on an indexed int field.
//
// Below is an example of creating an indexed integer field in your schema
// You can use RangeQuery to get a Count of all occurrences in a given range.
#[macro_use]
extern crate tantivy;
use tantivy::collector::Count;
use tantivy::query::RangeQuery;
use tantivy::schema::{Schema, INDEXED};
use tantivy::Index;
use tantivy::Result;
fn run() -> Result<()> {
// For the sake of simplicity, this schema will only have 1 field
let mut schema_builder = Schema::builder();
// `INDEXED` is a short-hand to indicate that our field should be "searchable".
let year_field = schema_builder.add_u64_field("year", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let reader = index.reader()?;
{
let mut index_writer = index.writer_with_num_threads(1, 6_000_000)?;
for year in 1950u64..2019u64 {
index_writer.add_document(doc!(year_field => year));
}
index_writer.commit()?;
// The index will be a range of years
}
reader.reload()?;
let searcher = reader.searcher();
// The end is excluded i.e. here we are searching up to 1969
let docs_in_the_sixties = RangeQuery::new_u64(year_field, 1960..1970);
// Uses a Count collector to sum the total number of docs in the range
let num_60s_books = searcher.search(&docs_in_the_sixties, &Count)?;
assert_eq!(num_60s_books, 10);
Ok(())
}
fn main() {
run().unwrap()
}

133
examples/iterating_docs_and_positions.rs Normal file
View File

@@ -0,0 +1,133 @@
// # Iterating docs and positioms.
//
// At its core of tantivy, relies on a data structure
// called an inverted index.
//
// This example shows how to manually iterate through
// the list of documents containing a term, getting
// its term frequency, and accessing its positions.
// ---
// Importing tantivy...
#[macro_use]
extern crate tantivy;
use tantivy::schema::*;
use tantivy::Index;
use tantivy::{DocId, DocSet, Postings};
fn main() -> tantivy::Result<()> {
// We first create a schema for the sake of the
// example. Check the `basic_search` example for more information.
let mut schema_builder = Schema::builder();
// For this example, we need to make sure to index positions for our title
// field. `TEXT` precisely does this.
let title = schema_builder.add_text_field("title", TEXT | STORED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema.clone());
let mut index_writer = index.writer_with_num_threads(1, 50_000_000)?;
index_writer.add_document(doc!(title => "The Old Man and the Sea"));
index_writer.add_document(doc!(title => "Of Mice and Men"));
index_writer.add_document(doc!(title => "The modern Promotheus"));
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
// A tantivy index is actually a collection of segments.
// Similarly, a searcher just wraps a list `segment_reader`.
//
// (Because we indexed a very small number of documents over one thread
// there is actually only one segment here, but let's iterate through the list
// anyway)
for segment_reader in searcher.segment_readers() {
// A segment contains different data structure.
// 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);
// 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
let term_the = Term::from_field_text(title, "the");
// 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
// and positions.
//
// 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)
{
// this buffer will be used to request for positions
let mut positions: Vec<u32> = Vec::with_capacity(100);
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) {
continue;
}
// the number of time the term appears in the document.
let term_freq: u32 = segment_postings.term_freq();
// accessing positions is slightly expensive and lazy, do not request
// for them if you don't need them for some documents.
segment_postings.positions(&mut positions);
// By definition we should have `term_freq` positions.
assert_eq!(positions.len(), term_freq as usize);
// This prints:
// ```
// Doc 0: TermFreq 2: [0, 4]
// Doc 2: TermFreq 1: [0]
// ```
println!("Doc {}: TermFreq {}: {:?}", doc_id, term_freq, positions);
}
}
}
// 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
let term_the = Term::from_field_text(title, "the");
// Some other powerful operations (especially `.skip_to`) may be useful to consume these
// posting lists rapidly.
// You can check for them in the [`DocSet`](https://docs.rs/tantivy/~0/tantivy/trait.DocSet.html) trait
// and the [`Postings`](https://docs.rs/tantivy/~0/tantivy/trait.Postings.html) trait
// Also, for some VERY specific high performance use case like an OLAP analysis of logs,
// you can get better performance by accessing directly the blocks of doc ids.
for segment_reader in searcher.segment_readers() {
// A segment contains different data structure.
// 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);
// 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
// and positions.
//
// 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)
{
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);
}
}
}
Ok(())
}

107
examples/multiple_producer.rs Normal file
View File

@@ -0,0 +1,107 @@
// # Indexing from different threads.
//
// It is fairly common to have to index from different threads.
// Tantivy forbids to create more than one `IndexWriter` at a time.
//
// This `IndexWriter` itself has its own multithreaded layer, so managing your own
// indexing threads will not help. However, it can still be useful for some applications.
//
// For instance, if preparing documents to send to tantivy before indexing is the bottleneck of
// your application, it is reasonable to have multiple threads.
//
// Another very common reason to want to index from multiple threads, is implementing a webserver
// with CRUD capabilities. The server framework will most likely handle request from
// different threads.
//
// The recommended way to address both of these use case is to wrap your `IndexWriter` into a
// `Arc<RwLock<IndexWriter>>`.
//
// While this is counterintuitive, adding and deleting documents do not require mutability
// over the `IndexWriter`, so several threads will be able to do this operation concurrently.
//
// The example below does not represent an actual real-life use case (who would spawn thread to
// index a single document?), but aims at demonstrating the mechanism that makes indexing
// from several threads possible.
extern crate tempdir;
// ---
// Importing tantivy...
#[macro_use]
extern crate tantivy;
use std::sync::{Arc, RwLock};
use std::thread;
use std::time::Duration;
use tantivy::schema::{Schema, STORED, TEXT};
use tantivy::Opstamp;
use tantivy::{Index, IndexWriter};
fn main() -> tantivy::Result<()> {
// # Defining the schema
let mut schema_builder = Schema::builder();
let title = schema_builder.add_text_field("title", TEXT | STORED);
let body = schema_builder.add_text_field("body", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let index_writer: Arc<RwLock<IndexWriter>> = Arc::new(RwLock::new(index.writer(50_000_000)?));
// # First indexing thread.
let index_writer_clone_1 = index_writer.clone();
thread::spawn(move || {
// we index 100 times the document... for the sake of the example.
for i in 0..100 {
let opstamp = {
// A read lock is sufficient here.
let index_writer_rlock = index_writer_clone_1.read().unwrap();
index_writer_rlock.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"
))
};
println!("add doc {} from thread 1 - opstamp {}", i, opstamp);
thread::sleep(Duration::from_millis(20));
}
});
// # Second indexing thread.
let index_writer_clone_2 = index_writer.clone();
// For convenience, tantivy also comes with a macro to
// reduce the boilerplate above.
thread::spawn(move || {
// we index 100 times the document... for the sake of the example.
for i in 0..100 {
// A read lock is sufficient here.
let opstamp = {
let index_writer_rlock = index_writer_clone_2.read().unwrap();
index_writer_rlock.add_document(doc!(
title => "Manufacturing consent",
body => "Some great book description..."
))
};
println!("add doc {} from thread 2 - opstamp {}", i, opstamp);
thread::sleep(Duration::from_millis(10));
}
});
// # In the main thread, we commit 10 times, once every 500ms.
for _ in 0..10 {
let opstamp: Opstamp = {
// Committing or rollbacking on the other hand requires write lock. This will block other threads.
let mut index_writer_wlock = index_writer.write().unwrap();
index_writer_wlock.commit().unwrap()
};
println!("committed with opstamp {}", opstamp);
thread::sleep(Duration::from_millis(500));
}
Ok(())
}

86
examples/snippet.rs Normal file
View File

@@ -0,0 +1,86 @@
// # Snippet example
//
// This example shows how to return a representative snippet of
// your hit result.
// Snippet are an extracted of a target document, and returned in HTML format.
// The keyword searched by the user are highlighted with a `<b>` tag.
extern crate tempdir;
// ---
// Importing tantivy...
#[macro_use]
extern crate tantivy;
use tantivy::collector::TopDocs;
use tantivy::query::QueryParser;
use tantivy::schema::*;
use tantivy::Index;
use tantivy::{Snippet, SnippetGenerator};
use tempdir::TempDir;
fn main() -> tantivy::Result<()> {
// Let's create a temporary directory for the
// sake of this example
let index_path = TempDir::new("tantivy_example_dir")?;
// # Defining the schema
let mut schema_builder = Schema::builder();
let title = schema_builder.add_text_field("title", TEXT | STORED);
let body = schema_builder.add_text_field("body", TEXT | STORED);
let schema = schema_builder.build();
// # Indexing documents
let index = Index::create_in_dir(&index_path, schema.clone())?;
let mut index_writer = index.writer(50_000_000)?;
// we'll only need one doc for this example.
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"
));
// ...
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let query_parser = QueryParser::for_index(&index, vec![title, body]);
let query = query_parser.parse_query("sycamore spring")?;
let top_docs = searcher.search(&query, &TopDocs::with_limit(10))?;
let snippet_generator = SnippetGenerator::create(&searcher, &*query, body)?;
for (score, doc_address) in top_docs {
let doc = searcher.doc(doc_address)?;
let snippet = snippet_generator.snippet_from_doc(&doc);
println!("Document score {}:", score);
println!("title: {}", doc.get_first(title).unwrap().text().unwrap());
println!("snippet: {}", snippet.to_html());
println!("custom highlighting: {}", highlight(snippet));
}
Ok(())
}
fn highlight(snippet: Snippet) -> String {
let mut result = String::new();
let mut start_from = 0;
for (start, end) in snippet.highlighted().iter().map(|h| h.bounds()) {
result.push_str(&snippet.fragments()[start_from..start]);
result.push_str(" --> ");
result.push_str(&snippet.fragments()[start..end]);
result.push_str(" <-- ");
start_from = end;
}
result.push_str(&snippet.fragments()[start_from..]);
result
}

117
examples/stop_words.rs Normal file
View File

@@ -0,0 +1,117 @@
// # Stop Words Example
//
// This example covers the basic usage of stop words
// with tantivy
//
// We will :
// - define our schema
// - create an index in a directory
// - add a few stop words
// - index few documents in our index
extern crate tempdir;
// ---
// Importing tantivy...
#[macro_use]
extern crate tantivy;
use tantivy::collector::TopDocs;
use tantivy::query::QueryParser;
use tantivy::schema::*;
use tantivy::tokenizer::*;
use tantivy::Index;
fn main() -> tantivy::Result<()> {
// this example assumes you understand the content in `basic_search`
let mut schema_builder = Schema::builder();
// This configures your custom options for how tantivy will
// store and process your content in the index; The key
// to note is that we are setting the tokenizer to `stoppy`
// which will be defined and registered below.
let text_field_indexing = TextFieldIndexing::default()
.set_tokenizer("stoppy")
.set_index_option(IndexRecordOption::WithFreqsAndPositions);
let text_options = TextOptions::default()
.set_indexing_options(text_field_indexing)
.set_stored();
// Our first field is title.
schema_builder.add_text_field("title", text_options);
// Our second field is body.
let text_field_indexing = TextFieldIndexing::default()
.set_tokenizer("stoppy")
.set_index_option(IndexRecordOption::WithFreqsAndPositions);
let text_options = TextOptions::default()
.set_indexing_options(text_field_indexing)
.set_stored();
schema_builder.add_text_field("body", text_options);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema.clone());
// This tokenizer lowers all of the text (to help with stop word matching)
// then removes all instances of `the` and `and` from the corpus
let tokenizer = SimpleTokenizer
.filter(LowerCaser)
.filter(StopWordFilter::remove(vec![
"the".to_string(),
"and".to_string(),
]));
index.tokenizers().register("stoppy", tokenizer);
let mut index_writer = index.writer(50_000_000)?;
let title = schema.get_field("title").unwrap();
let body = schema.get_field("body").unwrap();
index_writer.add_document(doc!(
title => "The Old Man and the Sea",
body => "He was an old man who fished alone in a skiff in the Gulf Stream and \
he had gone eighty-four days now without taking a fish."
));
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"
));
index_writer.add_document(doc!(
title => "Frankenstein",
body => "You will rejoice to hear that no disaster has accompanied the commencement of an \
enterprise which you have regarded with such evil forebodings. I arrived here \
yesterday, and my first task is to assure my dear sister of my welfare and \
increasing confidence in the success of my undertaking."
));
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let query_parser = QueryParser::for_index(&index, vec![title, body]);
// stop words are applied on the query as well.
// The following will be equivalent to `title:frankenstein`
let query = query_parser.parse_query("title:\"the Frankenstein\"")?;
let top_docs = searcher.search(&query, &TopDocs::with_limit(10))?;
for (score, doc_address) in top_docs {
let retrieved_doc = searcher.doc(doc_address)?;
println!("\n==\nDocument score {}:", score);
println!("{}", schema.to_json(&retrieved_doc));
}
Ok(())
}

41
examples/working_with_json.rs Normal file
View File

@@ -0,0 +1,41 @@
extern crate tantivy;
use tantivy::schema::*;
// # Document from json
//
// For convenience, `Document` can be parsed directly from json.
fn main() -> tantivy::Result<()> {
// Let's first define a schema and an index.
// Check out the basic example if this is confusing to you.
//
// first we need to define a schema ...
let mut schema_builder = Schema::builder();
schema_builder.add_text_field("title", TEXT | STORED);
schema_builder.add_text_field("body", TEXT);
schema_builder.add_u64_field("year", INDEXED);
let schema = schema_builder.build();
// Let's assume we have a json-serialized document.
let mice_and_men_doc_json = r#"{
"title": "Of Mice and Men",
"year": 1937
}"#;
// We can parse our document
let _mice_and_men_doc = schema.parse_document(&mice_and_men_doc_json)?;
// Multi-valued field are allowed, they are
// expressed in JSON by an array.
// The following document has two titles.
let frankenstein_json = r#"{
"title": ["Frankenstein", "The Modern Prometheus"],
"year": 1818
}"#;
let _frankenstein_doc = schema.parse_document(&frankenstein_json)?;
// Note that the schema is saved in your index directory.
//
// As a result, Indexes are aware of their schema, and you can use this feature
// just by opening an existing `Index`, and calling `index.schema()..parse_document(json)`.
Ok(())
}

2
run-tests.sh Executable file
View File

@@ -0,0 +1,2 @@
#!/bin/bash
cargo test --no-default-features --features mmap -- --test-threads 1

91
src/collector/chained_collector.rs
View File

@@ -1,91 +0,0 @@
use Result;
use collector::Collector;
use SegmentLocalId;
use SegmentReader;
use DocId;
use Score;
/// Collector that does nothing.
/// This is used in the chain Collector and will hopefully
/// be optimized away by the compiler.
pub struct DoNothingCollector;
impl Collector for DoNothingCollector {
#[inline]
fn set_segment(&mut self, _: SegmentLocalId, _: &SegmentReader) -> Result<()> {
Ok(())
}
#[inline]
fn collect(&mut self, _doc: DocId, _score: Score) {}
#[inline]
fn requires_scoring(&self) -> bool {
false
}
}
/// Zero-cost abstraction used to collect on multiple collectors.
/// This contraption is only usable if the type of your collectors
/// are known at compile time.
pub struct ChainedCollector<Left: Collector, Right: Collector> {
left: Left,
right: Right,
}
impl<Left: Collector, Right: Collector> ChainedCollector<Left, Right> {
/// Adds a collector
pub fn push<C: Collector>(self, new_collector: &mut C) -> ChainedCollector<Self, &mut C> {
ChainedCollector {
left: self,
right: new_collector,
}
}
}
impl<Left: Collector, Right: Collector> Collector for ChainedCollector<Left, Right> {
fn set_segment(
&mut self,
segment_local_id: SegmentLocalId,
segment: &SegmentReader,
) -> Result<()> {
self.left.set_segment(segment_local_id, segment)?;
self.right.set_segment(segment_local_id, segment)?;
Ok(())
}
fn collect(&mut self, doc: DocId, score: Score) {
self.left.collect(doc, score);
self.right.collect(doc, score);
}
fn requires_scoring(&self) -> bool {
self.left.requires_scoring() || self.right.requires_scoring()
}
}
/// Creates a `ChainedCollector`
pub fn chain() -> ChainedCollector<DoNothingCollector, DoNothingCollector> {
ChainedCollector {
left: DoNothingCollector,
right: DoNothingCollector,
}
}
#[cfg(test)]
mod tests {
use super::*;
use collector::{Collector, CountCollector, TopCollector};
#[test]
fn test_chained_collector() {
let mut top_collector = TopCollector::with_limit(2);
let mut count_collector = CountCollector::default();
{
let mut collectors = chain().push(&mut top_collector).push(&mut count_collector);
collectors.collect(1, 0.2);
collectors.collect(2, 0.1);
collectors.collect(3, 0.5);
}
assert_eq!(count_collector.count(), 3);
assert!(top_collector.at_capacity());
}
}

135
src/collector/count_collector.rs
View File

@@ -1,54 +1,129 @@
use super::Collector;
use collector::SegmentCollector;
use DocId;
use Score;
use Result;
use SegmentReader;
use Score;
use SegmentLocalId;
use SegmentReader;
/// `CountCollector` collector only counts how many
/// documents match the query.
#[derive(Default)]
pub struct CountCollector {
count: usize,
}
///
/// ```rust
/// #[macro_use]
/// extern crate tantivy;
/// use tantivy::schema::{Schema, TEXT};
/// use tantivy::{Index, Result};
/// use tantivy::collector::Count;
/// use tantivy::query::QueryParser;
///
/// # fn main() { example().unwrap(); }
/// fn example() -> Result<()> {
/// 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(3_000_000)?;
/// 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.commit().unwrap();
/// }
///
/// let reader = index.reader()?;
/// let searcher = reader.searcher();
///
/// {
/// let query_parser = QueryParser::for_index(&index, vec![title]);
/// let query = query_parser.parse_query("diary")?;
/// let count = searcher.search(&query, &Count).unwrap();
///
/// assert_eq!(count, 2);
/// }
///
/// Ok(())
/// }
/// ```
pub struct Count;
impl CountCollector {
/// Returns the count of documents that were
/// collected.
pub fn count(&self) -> usize {
self.count
}
}
impl Collector for Count {
type Fruit = usize;
impl Collector for CountCollector {
fn set_segment(&mut self, _: SegmentLocalId, _: &SegmentReader) -> Result<()> {
Ok(())
}
type Child = SegmentCountCollector;
fn collect(&mut self, _: DocId, _: Score) {
self.count += 1;
fn for_segment(&self, _: SegmentLocalId, _: &SegmentReader) -> Result<SegmentCountCollector> {
Ok(SegmentCountCollector::default())
}
fn requires_scoring(&self) -> bool {
false
}
fn merge_fruits(&self, segment_counts: Vec<usize>) -> Result<usize> {
Ok(segment_counts.into_iter().sum())
}
}
#[derive(Default)]
pub struct SegmentCountCollector {
count: usize,
}
impl SegmentCollector for SegmentCountCollector {
type Fruit = usize;
fn collect(&mut self, _: DocId, _: Score) {
self.count += 1;
}
fn harvest(self) -> usize {
self.count
}
}
#[cfg(test)]
mod tests {
use collector::{Collector, CountCollector};
use super::{Count, SegmentCountCollector};
use collector::Collector;
use collector::SegmentCollector;
#[test]
fn test_count_collector() {
let mut count_collector = CountCollector::default();
assert_eq!(count_collector.count(), 0);
count_collector.collect(0u32, 1f32);
assert_eq!(count_collector.count(), 1);
assert_eq!(count_collector.count(), 1);
count_collector.collect(1u32, 1f32);
assert_eq!(count_collector.count(), 2);
assert!(!count_collector.requires_scoring());
fn test_count_collect_does_not_requires_scoring() {
assert!(!Count.requires_scoring());
}
#[test]
fn test_segment_count_collector() {
{
let count_collector = SegmentCountCollector::default();
assert_eq!(count_collector.harvest(), 0);
}
{
let mut count_collector = SegmentCountCollector::default();
count_collector.collect(0u32, 1f32);
assert_eq!(count_collector.harvest(), 1);
}
{
let mut count_collector = SegmentCountCollector::default();
count_collector.collect(0u32, 1f32);
assert_eq!(count_collector.harvest(), 1);
}
{
let mut count_collector = SegmentCountCollector::default();
count_collector.collect(0u32, 1f32);
count_collector.collect(1u32, 1f32);
assert_eq!(count_collector.harvest(), 2);
}
}
}

479
src/collector/facet_collector.rs
View File

@@ -1,27 +1,23 @@
use std::mem;
use collector::Collector;
use collector::SegmentCollector;
use docset::SkipResult;
use fastfield::FacetReader;
use schema::Field;
use std::cell::UnsafeCell;
use schema::Facet;
use schema::Field;
use std::cmp::Ordering;
use std::collections::btree_map;
use std::collections::BTreeMap;
use std::collections::BTreeSet;
use std::collections::BinaryHeap;
use std::collections::Bound;
use termdict::TermDictionary;
use termdict::TermStreamer;
use termdict::TermStreamerBuilder;
use std::collections::BTreeSet;
use termdict::TermMerger;
use docset::SkipResult;
use std::{usize, u64};
use std::iter::Peekable;
use std::{u64, usize};
use DocId;
use Result;
use Score;
use SegmentReader;
use SegmentLocalId;
use std::cmp::Ordering;
use SegmentReader;
use TantivyError;
struct Hit<'a> {
count: u64,
@@ -48,12 +44,6 @@ impl<'a> Ord for Hit<'a> {
}
}
struct SegmentFacetCounter {
pub facet_reader: FacetReader,
pub facet_ords: Vec<u64>,
pub facet_counts: Vec<u64>,
}
fn facet_depth(facet_bytes: &[u8]) -> usize {
if facet_bytes.is_empty() {
0
@@ -93,14 +83,14 @@ fn facet_depth(facet_bytes: &[u8]) -> usize {
/// ```rust
/// #[macro_use]
/// extern crate tantivy;
/// use tantivy::schema::{Facet, SchemaBuilder, TEXT};
/// use tantivy::schema::{Facet, Schema, TEXT};
/// use tantivy::{Index, Result};
/// use tantivy::collector::FacetCollector;
/// use tantivy::query::AllQuery;
///
/// # fn main() { example().unwrap(); }
/// fn example() -> Result<()> {
/// let mut schema_builder = SchemaBuilder::new();
/// let mut schema_builder = Schema::builder();
///
/// // Facet have their own specific type.
/// // It is not a bad practise to put all of your
@@ -133,23 +123,19 @@ fn facet_depth(facet_bytes: &[u8]) -> usize {
/// facet => Facet::from("/lang/en"),
/// facet => Facet::from("/category/biography")
/// ));
/// index_writer.commit().unwrap();
/// index_writer.commit()?;
/// }
///
/// index.load_searchers()?;
/// let searcher = index.searcher();
/// let reader = index.reader()?;
/// let searcher = reader.searcher();
///
/// {
/// let mut facet_collector = FacetCollector::for_field(facet);
/// facet_collector.add_facet("/lang");
/// facet_collector.add_facet("/category");
/// searcher.search(&AllQuery, &mut facet_collector).unwrap();
///
/// // this object contains count aggregate for all of the facets.
/// let counts = facet_collector.harvest();
/// let facet_counts = searcher.search(&AllQuery, &facet_collector)?;
///
/// // This lists all of the facet counts
/// let facets: Vec<(&Facet, u64)> = counts
/// let facets: Vec<(&Facet, u64)> = facet_counts
/// .get("/category")
/// .collect();
/// assert_eq!(facets, vec![
@@ -161,13 +147,10 @@ fn facet_depth(facet_bytes: &[u8]) -> usize {
/// {
/// let mut facet_collector = FacetCollector::for_field(facet);
/// facet_collector.add_facet("/category/fiction");
/// searcher.search(&AllQuery, &mut facet_collector).unwrap();
///
/// // this object contains count aggregate for all of the facets.
/// let counts = facet_collector.harvest();
/// let facet_counts = searcher.search(&AllQuery, &facet_collector)?;
///
/// // This lists all of the facet counts
/// let facets: Vec<(&Facet, u64)> = counts
/// let facets: Vec<(&Facet, u64)> = facet_counts
/// .get("/category/fiction")
/// .collect();
/// assert_eq!(facets, vec![
@@ -180,13 +163,10 @@ fn facet_depth(facet_bytes: &[u8]) -> usize {
/// {
/// let mut facet_collector = FacetCollector::for_field(facet);
/// facet_collector.add_facet("/category/fiction");
/// searcher.search(&AllQuery, &mut facet_collector).unwrap();
///
/// // this object contains count aggregate for all of the facets.
/// let counts = facet_collector.harvest();
/// let facet_counts = searcher.search(&AllQuery, &facet_collector)?;
///
/// // This lists all of the facet counts
/// let facets: Vec<(&Facet, u64)> = counts.top_k("/category/fiction", 1);
/// let facets: Vec<(&Facet, u64)> = facet_counts.top_k("/category/fiction", 1);
/// assert_eq!(facets, vec![
/// (&Facet::from("/category/fiction/fantasy"), 2)
/// ]);
@@ -196,28 +176,28 @@ fn facet_depth(facet_bytes: &[u8]) -> usize {
/// }
/// ```
pub struct FacetCollector {
facet_ords: Vec<u64>,
field: Field,
ff_reader: Option<UnsafeCell<FacetReader>>,
segment_counters: Vec<SegmentFacetCounter>,
// facet_ord -> collapse facet_id
current_segment_collapse_mapping: Vec<usize>,
// collapse facet_id -> count
current_segment_counts: Vec<u64>,
// collapse facet_id -> facet_ord
current_collapse_facet_ords: Vec<u64>,
facets: BTreeSet<Facet>,
}
pub struct FacetSegmentCollector {
reader: FacetReader,
facet_ords_buf: Vec<u64>,
// facet_ord -> collapse facet_id
collapse_mapping: Vec<usize>,
// collapse facet_id -> count
counts: Vec<u64>,
// collapse facet_id -> facet_ord
collapse_facet_ords: Vec<u64>,
}
fn skip<'a, I: Iterator<Item = &'a Facet>>(
target: &[u8],
collapse_it: &mut Peekable<I>,
) -> SkipResult {
loop {
match collapse_it.peek() {
Some(facet_bytes) => match facet_bytes.encoded_bytes().cmp(target) {
Some(facet_bytes) => match facet_bytes.encoded_str().as_bytes().cmp(target) {
Ordering::Less => {}
Ordering::Greater => {
return SkipResult::OverStep;
@@ -242,15 +222,8 @@ impl FacetCollector {
/// is of the proper type.
pub fn for_field(field: Field) -> FacetCollector {
FacetCollector {
facet_ords: Vec::with_capacity(255),
segment_counters: Vec::new(),
field,
ff_reader: None,
facets: BTreeSet::new(),
current_segment_collapse_mapping: Vec::new(),
current_collapse_facet_ords: Vec::new(),
current_segment_counts: Vec::new(),
facets: BTreeSet::default(),
}
}
@@ -280,139 +253,103 @@ impl FacetCollector {
}
self.facets.insert(facet);
}
fn set_collapse_mapping(&mut self, facet_reader: &FacetReader) {
self.current_segment_collapse_mapping.clear();
self.current_collapse_facet_ords.clear();
self.current_segment_counts.clear();
let mut collapse_facet_it = self.facets.iter().peekable();
self.current_collapse_facet_ords.push(0);
let mut facet_streamer = facet_reader.facet_dict().range().into_stream();
if !facet_streamer.advance() {
return;
}
'outer: loop {
// at the begining of this loop, facet_streamer
// 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::Reached => {
// we reach a facet we decided to collapse.
let collapse_depth = facet_depth(facet_streamer.key());
let mut collapsed_id = 0;
self.current_segment_collapse_mapping.push(0);
while facet_streamer.advance() {
let depth = facet_depth(facet_streamer.key());
if depth <= collapse_depth {
continue 'outer;
}
if depth == collapse_depth + 1 {
collapsed_id = self.current_collapse_facet_ords.len();
self.current_collapse_facet_ords
.push(facet_streamer.term_ord());
self.current_segment_collapse_mapping.push(collapsed_id);
} else {
self.current_segment_collapse_mapping.push(collapsed_id);
}
}
break;
}
SkipResult::End | SkipResult::OverStep => {
self.current_segment_collapse_mapping.push(0);
if !facet_streamer.advance() {
break;
}
}
}
}
}
fn finalize_segment(&mut self) {
if self.ff_reader.is_some() {
self.segment_counters.push(SegmentFacetCounter {
facet_reader: self.ff_reader.take().unwrap().into_inner(),
facet_ords: mem::replace(&mut self.current_collapse_facet_ords, Vec::new()),
facet_counts: mem::replace(&mut self.current_segment_counts, Vec::new()),
});
}
}
/// Returns the results of the collection.
///
/// This method does not just return the counters,
/// it also translates the facet ordinals of the last segment.
pub fn harvest(mut self) -> FacetCounts {
self.finalize_segment();
let collapsed_facet_ords: Vec<&[u64]> = self.segment_counters
.iter()
.map(|segment_counter| &segment_counter.facet_ords[..])
.collect();
let collapsed_facet_counts: Vec<&[u64]> = self.segment_counters
.iter()
.map(|segment_counter| &segment_counter.facet_counts[..])
.collect();
let facet_streams = self.segment_counters
.iter()
.map(|seg_counts| seg_counts.facet_reader.facet_dict().range().into_stream())
.collect::<Vec<_>>();
let mut facet_merger = TermMerger::new(facet_streams);
let mut facet_counts = BTreeMap::new();
while facet_merger.advance() {
let count = facet_merger
.current_kvs()
.iter()
.map(|it| {
let seg_ord = it.segment_ord;
let term_ord = it.streamer.term_ord();
collapsed_facet_ords[seg_ord]
.binary_search(&term_ord)
.map(|collapsed_term_id| {
if collapsed_term_id == 0 {
0
} else {
collapsed_facet_counts[seg_ord][collapsed_term_id]
}
})
.unwrap_or(0)
})
.sum();
if count > 0u64 {
let bytes = facet_merger.key().to_owned();
facet_counts.insert(Facet::from_encoded(bytes), count);
}
}
FacetCounts { facet_counts }
}
}
impl Collector for FacetCollector {
fn set_segment(&mut self, _: SegmentLocalId, reader: &SegmentReader) -> Result<()> {
self.finalize_segment();
let facet_reader = reader.facet_reader(self.field)?;
self.set_collapse_mapping(&facet_reader);
self.current_segment_counts
.resize(self.current_collapse_facet_ords.len(), 0);
self.ff_reader = Some(UnsafeCell::new(facet_reader));
Ok(())
type Fruit = FacetCounts;
type Child = FacetSegmentCollector;
fn for_segment(
&self,
_: SegmentLocalId,
reader: &SegmentReader,
) -> Result<FacetSegmentCollector> {
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();
let mut collapse_facet_ords = Vec::new();
let mut collapse_facet_it = self.facets.iter().peekable();
collapse_facet_ords.push(0);
{
let mut facet_streamer = facet_reader.facet_dict().range().into_stream();
if facet_streamer.advance() {
'outer: loop {
// at the begining of this loop, facet_streamer
// 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::Reached => {
// we reach a facet we decided to collapse.
let collapse_depth = facet_depth(facet_streamer.key());
let mut collapsed_id = 0;
collapse_mapping.push(0);
while facet_streamer.advance() {
let depth = facet_depth(facet_streamer.key());
if depth <= collapse_depth {
continue 'outer;
}
if depth == collapse_depth + 1 {
collapsed_id = collapse_facet_ords.len();
collapse_facet_ords.push(facet_streamer.term_ord());
collapse_mapping.push(collapsed_id);
} else {
collapse_mapping.push(collapsed_id);
}
}
break;
}
SkipResult::End | SkipResult::OverStep => {
collapse_mapping.push(0);
if !facet_streamer.advance() {
break;
}
}
}
}
}
}
counts.resize(collapse_facet_ords.len(), 0);
Ok(FacetSegmentCollector {
reader: facet_reader,
facet_ords_buf: Vec::with_capacity(255),
collapse_mapping,
counts,
collapse_facet_ords,
})
}
fn requires_scoring(&self) -> bool {
false
}
fn merge_fruits(&self, segments_facet_counts: Vec<FacetCounts>) -> Result<FacetCounts> {
let mut facet_counts: BTreeMap<Facet, u64> = BTreeMap::new();
for segment_facet_counts in segments_facet_counts {
for (facet, count) in segment_facet_counts.facet_counts {
*(facet_counts.entry(facet).or_insert(0)) += count;
}
}
Ok(FacetCounts { facet_counts })
}
}
impl SegmentCollector for FacetSegmentCollector {
type Fruit = FacetCounts;
fn collect(&mut self, doc: DocId, _: Score) {
let facet_reader: &mut FacetReader = unsafe {
&mut *self.ff_reader
.as_ref()
.expect("collect() was called before set_segment. This should never happen.")
.get()
};
facet_reader.facet_ords(doc, &mut self.facet_ords);
self.reader.facet_ords(doc, &mut self.facet_ords_buf);
let mut previous_collapsed_ord: usize = usize::MAX;
for &facet_ord in &self.facet_ords {
let collapsed_ord = self.current_segment_collapse_mapping[facet_ord as usize];
self.current_segment_counts[collapsed_ord] += if collapsed_ord == previous_collapsed_ord
{
for &facet_ord in &self.facet_ords_buf {
let collapsed_ord = self.collapse_mapping[facet_ord as usize];
self.counts[collapsed_ord] += if collapsed_ord == previous_collapsed_ord {
0
} else {
1
@@ -421,8 +358,24 @@ impl Collector for FacetCollector {
}
}
fn requires_scoring(&self) -> bool {
false
/// Returns the results of the collection.
///
/// This method does not just return the counters,
/// it also translates the facet ordinals of the last segment.
fn harvest(self) -> FacetCounts {
let mut facet_counts = BTreeMap::new();
let facet_dict = self.reader.facet_dict();
for (collapsed_facet_ord, count) in self.counts.iter().cloned().enumerate() {
if count == 0 {
continue;
}
let mut facet = vec![];
let facet_ord = self.collapse_facet_ords[collapsed_facet_ord];
facet_dict.ord_to_term(facet_ord as u64, &mut facet);
// TODO
facet_counts.insert(Facet::from_encoded(facet).unwrap(), count);
}
FacetCounts { facet_counts }
}
}
@@ -432,9 +385,20 @@ pub struct FacetCounts {
facet_counts: BTreeMap<Facet, u64>,
}
pub struct FacetChildIterator<'a> {
underlying: btree_map::Range<'a, Facet, u64>,
}
impl<'a> Iterator for FacetChildIterator<'a> {
type Item = (&'a Facet, u64);
fn next(&mut self) -> Option<Self::Item> {
self.underlying.next().map(|(facet, count)| (facet, *count))
}
}
impl FacetCounts {
#[allow(needless_lifetimes)] //< compiler fails if we remove the lifetime
pub fn get<'a, T>(&'a self, facet_from: T) -> impl Iterator<Item = (&'a Facet, u64)>
pub fn get<T>(&self, facet_from: T) -> FacetChildIterator
where
Facet: From<T>,
{
@@ -443,15 +407,13 @@ impl FacetCounts {
let right_bound = if facet.is_root() {
Bound::Unbounded
} else {
let mut facet_after_bytes = facet.encoded_bytes().to_owned();
facet_after_bytes.push(1u8);
let facet_after = Facet::from_encoded(facet_after_bytes);
let mut facet_after_bytes: String = facet.encoded_str().to_owned();
facet_after_bytes.push('\u{1}');
let facet_after = Facet::from_encoded_string(facet_after_bytes);
Bound::Excluded(facet_after)
};
self.facet_counts
.range((left_bound, right_bound))
.map(|(facet, count)| (facet, *count))
let underlying: btree_map::Range<_, _> = self.facet_counts.range((left_bound, right_bound));
FacetChildIterator { underlying }
}
pub fn top_k<T>(&self, facet: T, k: usize) -> Vec<(&Facet, u64)>
@@ -461,17 +423,24 @@ impl FacetCounts {
let mut heap = BinaryHeap::with_capacity(k);
let mut it = self.get(facet);
// push the first k elements to first bring the heap
// to capacity
for (facet, count) in (&mut it).take(k) {
heap.push(Hit { count, facet });
}
let mut lowest_count: u64 = heap.peek().map(|hit| hit.count).unwrap_or(u64::MIN);
let mut lowest_count: u64 = heap.peek().map(|hit| hit.count).unwrap_or(u64::MIN); //< the `unwrap_or` case may be triggered but the value
// is never used in that case.
for (facet, count) in it {
if count > lowest_count {
lowest_count = count;
if let Some(mut head) = heap.peek_mut() {
*head = Hit { count, facet };
}
// the heap gets reconstructed at this point
if let Some(head) = heap.peek() {
lowest_count = head.count;
}
}
}
heap.into_sorted_vec()
@@ -483,23 +452,23 @@ impl FacetCounts {
#[cfg(test)]
mod tests {
use test::Bencher;
use core::Index;
use schema::{Document, Facet, SchemaBuilder};
use query::AllQuery;
use super::{FacetCollector, FacetCounts};
use std::iter;
use schema::Field;
use core::Index;
use query::AllQuery;
use rand::distributions::Uniform;
use rand::prelude::SliceRandom;
use rand::{thread_rng, Rng};
use schema::{Document, Facet, Field, Schema};
use std::iter;
#[test]
fn test_facet_collector_drilldown() {
let mut schema_builder = SchemaBuilder::new();
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(3_000_000).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| {
@@ -517,14 +486,12 @@ mod tests {
index_writer.add_document(doc);
}
index_writer.commit().unwrap();
index.load_searchers().unwrap();
let searcher = index.searcher();
let reader = index.reader().unwrap();
let searcher = reader.searcher();
let mut facet_collector = FacetCollector::for_field(facet_field);
facet_collector.add_facet(Facet::from("/top1"));
searcher.search(&AllQuery, &mut facet_collector).unwrap();
let counts = searcher.search(&AllQuery, &facet_collector).unwrap();
let counts: FacetCounts = facet_collector.harvest();
{
let facets: Vec<(String, u64)> = counts
.get("/top1")
@@ -537,9 +504,10 @@ mod tests {
("/top1/mid1", 50),
("/top1/mid2", 50),
("/top1/mid3", 50),
].iter()
.map(|&(facet_str, count)| (String::from(facet_str), count))
.collect::<Vec<_>>()
]
.iter()
.map(|&(facet_str, count)| (String::from(facet_str), count))
.collect::<Vec<_>>()
);
}
}
@@ -553,6 +521,30 @@ mod tests {
facet_collector.add_facet(Facet::from("/country/europe"));
}
#[test]
fn test_doc_unsorted_multifacet() {
let mut schema_builder = Schema::builder();
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_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"),
facet_field => Facet::from_text(&"/subjects/A/b"),
facet_field => Facet::from_text(&"/subjects/B/b"),
));
index_writer.commit().unwrap();
let reader = index.reader().unwrap();
let searcher = reader.searcher();
assert_eq!(searcher.num_docs(), 1);
let mut facet_collector = FacetCollector::for_field(facet_field);
facet_collector.add_facet("/subjects");
let counts = searcher.search(&AllQuery, &facet_collector).unwrap();
let facets: Vec<(&Facet, u64)> = counts.get("/subjects").collect();
assert_eq!(facets[0].1, 1);
}
#[test]
fn test_non_used_facet_collector() {
let mut facet_collector = FacetCollector::for_field(Field(0));
@@ -562,51 +554,69 @@ mod tests {
#[test]
fn test_facet_collector_topk() {
let mut schema_builder = SchemaBuilder::new();
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 uniform = Uniform::new_inclusive(1, 100_000);
let mut docs: Vec<Document> = vec![("a", 10), ("b", 100), ("c", 7), ("d", 12), ("e", 21)]
.into_iter()
.flat_map(|(c, count)| {
let facet = Facet::from(&format!("/facet_{}", c));
let facet = Facet::from(&format!("/facet/{}", c));
let doc = doc!(facet_field => facet);
iter::repeat(doc).take(count)
})
.map(|mut doc| {
doc.add_facet(
facet_field,
&format!("/facet/{}", thread_rng().sample(&uniform)),
);
doc
})
.collect();
thread_rng().shuffle(&mut docs[..]);
docs[..].shuffle(&mut thread_rng());
let mut index_writer = index.writer(3_000_000).unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
for doc in docs {
index_writer.add_document(doc);
}
index_writer.commit().unwrap();
index.load_searchers().unwrap();
let searcher = index.searcher();
let searcher = index.reader().unwrap().searcher();
let mut facet_collector = FacetCollector::for_field(facet_field);
facet_collector.add_facet("/");
searcher.search(&AllQuery, &mut facet_collector).unwrap();
facet_collector.add_facet("/facet");
let counts: FacetCounts = searcher.search(&AllQuery, &facet_collector).unwrap();
let counts: FacetCounts = facet_collector.harvest();
{
let facets: Vec<(&Facet, u64)> = counts.top_k("/", 3);
let facets: Vec<(&Facet, u64)> = counts.top_k("/facet", 3);
assert_eq!(
facets,
vec![
(&Facet::from("/facet_b"), 100),
(&Facet::from("/facet_e"), 21),
(&Facet::from("/facet_d"), 12),
(&Facet::from("/facet/b"), 100),
(&Facet::from("/facet/e"), 21),
(&Facet::from("/facet/d"), 12),
]
);
}
}
}
#[cfg(all(test, feature = "unstable"))]
mod bench {
use collector::FacetCollector;
use query::AllQuery;
use rand::{thread_rng, Rng};
use schema::Facet;
use schema::Schema;
use test::Bencher;
use Index;
#[bench]
fn bench_facet_collector(b: &mut Bencher) {
let mut schema_builder = SchemaBuilder::new();
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);
@@ -621,17 +631,16 @@ mod tests {
// 40425 docs
thread_rng().shuffle(&mut docs[..]);
let mut index_writer = index.writer(3_000_000).unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
for doc in docs {
index_writer.add_document(doc);
}
index_writer.commit().unwrap();
index.load_searchers().unwrap();
let reader = index.reader().unwrap();
b.iter(|| {
let searcher = index.searcher();
let mut facet_collector = FacetCollector::for_field(facet_field);
searcher.search(&AllQuery, &mut facet_collector).unwrap();
let facet_collector = FacetCollector::for_field(facet_field);
searcher.search(&AllQuery, &facet_collector).unwrap();
});
}
}

7
src/collector/int_facet_collector.rs
View File

@@ -79,7 +79,7 @@ mod tests {
// make sure we have facet counters correctly filled
fn test_facet_collector_results() {
let mut schema_builder = schema::SchemaBuilder::new();
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 text_field = schema_builder.add_text_field("text", STRING);
@@ -88,7 +88,7 @@ mod tests {
let index = Index::create_in_ram(schema.clone());
{
let mut index_writer = index.writer_with_num_threads(1, 40_000_000).unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
{
for i in 0u64..10u64 {
index_writer.add_document(doc!(
@@ -101,8 +101,7 @@ mod tests {
assert_eq!(index_writer.commit().unwrap(), 10u64);
}
index.load_searchers().unwrap();
let searcher = index.searcher();
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);

472
src/collector/mod.rs
View File

@@ -1,200 +1,364 @@
/*!
Defines how the documents matching a search query should be processed.
# Collectors
Collectors define the information you want to extract from the documents matching the queries.
In tantivy jargon, we call this information your search "fruit".
Your fruit could for instance be :
- [the count of matching documents](./struct.Count.html)
- [the top 10 documents, by relevancy or by a fast field](./struct.TopDocs.html)
- [facet counts](./struct.FacetCollector.html)
At one point in your code, you will trigger the actual search operation by calling
[the `search(...)` method of your `Searcher` object](../struct.Searcher.html#method.search).
This call will look like this.
```verbatim
let fruit = searcher.search(&query, &collector)?;
```
Here the type of fruit is actually determined as an associated type of the collector (`Collector::Fruit`).
# Combining several collectors
A rich search experience often requires to run several collectors on your search query.
For instance,
- selecting the top-K products matching your query
- counting the matching documents
- computing several facets
- computing statistics about the matching product prices
A simple and efficient way to do that is to pass your collectors as one tuple.
The resulting `Fruit` will then be a typed tuple with each collector's original fruits
in their respective position.
```rust
# extern crate tantivy;
# use tantivy::schema::*;
# use tantivy::*;
# use tantivy::query::*;
use tantivy::collector::{Count, TopDocs};
#
# fn main() -> tantivy::Result<()> {
# 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(3_000_000)?;
# index_writer.add_document(doc!(
# title => "The Name of the Wind",
# ));
# index_writer.add_document(doc!(
# title => "The Diary of Muadib",
# ));
# index_writer.commit()?;
# let reader = index.reader()?;
# let searcher = reader.searcher();
# let query_parser = QueryParser::for_index(&index, vec![title]);
# let query = query_parser.parse_query("diary")?;
let (doc_count, top_docs): (usize, Vec<(Score, DocAddress)>) =
searcher.search(&query, &(Count, TopDocs::with_limit(2)))?;
# Ok(())
# }
```
The `Collector` trait is implemented for up to 4 collectors.
If you have more than 4 collectors, you can either group them into
tuples of tuples `(a,(b,(c,d)))`, or rely on `MultiCollector`'s.
# Combining several collectors dynamically
Combining collectors into a tuple is a zero-cost abstraction: everything
happens as if you had manually implemented a single collector
combining all of our features.
Unfortunately it requires you to know at compile time your collector types.
If on the other hand, the collectors depend on some query parameter,
you can rely on `MultiCollector`'s.
# Implementing your own collectors.
See the `custom_collector` example.
*/
use SegmentReader;
use SegmentLocalId;
use downcast_rs;
use DocId;
use Score;
use Result;
use Score;
use SegmentLocalId;
use SegmentReader;
mod count_collector;
pub use self::count_collector::CountCollector;
pub use self::count_collector::Count;
mod multi_collector;
pub use self::multi_collector::MultiCollector;
mod top_collector;
pub use self::top_collector::TopCollector;
mod top_score_collector;
pub use self::top_score_collector::TopDocs;
mod top_field_collector;
pub use self::top_field_collector::TopDocsByField;
mod facet_collector;
pub use self::facet_collector::FacetCollector;
mod chained_collector;
pub use self::chained_collector::chain;
/// `Fruit` is the type for the result of our collection.
/// e.g. `usize` for the `Count` collector.
pub trait Fruit: Send + downcast_rs::Downcast {}
impl<T> Fruit for T where T: Send + downcast_rs::Downcast {}
/// Collectors are in charge of collecting and retaining relevant
/// information from the document found and scored by the query.
///
///
/// For instance,
///
/// - keeping track of the top 10 best documents
/// - computing a breakdown over a fast field
/// - computing the number of documents matching the query
///
/// Queries are in charge of pushing the `DocSet` to the collector.
/// Our search index is in fact a collection of segments, so
/// a `Collector` trait is actually more of a factory to instance
/// `SegmentCollector`s for each segments.
///
/// As they work on multiple segments, they first inform
/// the collector of a change in a segment and then
/// call the `collect` method to push the document to the collector.
///
/// Temporally, our collector will receive calls
/// - `.set_segment(0, segment_reader_0)`
/// - `.collect(doc0_of_segment_0)`
/// - `.collect(...)`
/// - `.collect(last_doc_of_segment_0)`
/// - `.set_segment(1, segment_reader_1)`
/// - `.collect(doc0_of_segment_1)`
/// - `.collect(...)`
/// - `.collect(last_doc_of_segment_1)`
/// - `...`
/// - `.collect(last_doc_of_last_segment)`
/// The collection logic itself is in the `SegmentCollector`.
///
/// Segments are not guaranteed to be visited in any specific order.
pub trait Collector {
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<Fruit = Self::Fruit>;
/// `set_segment` is called before beginning to enumerate
/// on this segment.
fn set_segment(
&mut self,
fn for_segment(
&self,
segment_local_id: SegmentLocalId,
segment: &SegmentReader,
) -> Result<()>;
/// The query pushes the scored document to the collector via this method.
fn collect(&mut self, doc: DocId, score: Score);
) -> Result<Self::Child>;
/// Returns true iff the collector requires to compute scores for documents.
fn requires_scoring(&self) -> bool;
/// Combines the fruit associated to the collection of each segments
/// into one fruit.
fn merge_fruits(&self, segment_fruits: Vec<Self::Fruit>) -> Result<Self::Fruit>;
}
impl<'a, C: Collector> Collector for &'a mut C {
fn set_segment(
&mut self,
segment_local_id: SegmentLocalId,
segment: &SegmentReader,
) -> Result<()> {
(*self).set_segment(segment_local_id, segment)
}
/// The `SegmentCollector` is the trait in charge of defining the
/// collect operation at the scale of the segment.
///
/// `.collect(doc, score)` will be called for every documents
/// matching the query.
pub trait SegmentCollector: 'static {
/// `Fruit` is the type for the result of our collection.
/// e.g. `usize` for the `Count` collector.
type Fruit: Fruit;
/// The query pushes the scored document to the collector via this method.
fn collect(&mut self, doc: DocId, score: Score) {
C::collect(self, doc, score)
fn collect(&mut self, doc: DocId, score: Score);
/// Extract the fruit of the collection from the `SegmentCollector`.
fn harvest(self) -> Self::Fruit;
}
// -----------------------------------------------
// Tuple implementations.
impl<Left, Right> Collector for (Left, Right)
where
Left: Collector,
Right: Collector,
{
type Fruit = (Left::Fruit, Right::Fruit);
type Child = (Left::Child, Right::Child);
fn for_segment(&self, segment_local_id: u32, segment: &SegmentReader) -> Result<Self::Child> {
let left = self.0.for_segment(segment_local_id, segment)?;
let right = self.1.for_segment(segment_local_id, segment)?;
Ok((left, right))
}
fn requires_scoring(&self) -> bool {
C::requires_scoring(self)
self.0.requires_scoring() || self.1.requires_scoring()
}
fn merge_fruits(
&self,
children: Vec<(Left::Fruit, Right::Fruit)>,
) -> Result<(Left::Fruit, Right::Fruit)> {
let mut left_fruits = vec![];
let mut right_fruits = vec![];
for (left_fruit, right_fruit) in children {
left_fruits.push(left_fruit);
right_fruits.push(right_fruit);
}
Ok((
self.0.merge_fruits(left_fruits)?,
self.1.merge_fruits(right_fruits)?,
))
}
}
impl<Left, Right> SegmentCollector for (Left, Right)
where
Left: SegmentCollector,
Right: SegmentCollector,
{
type Fruit = (Left::Fruit, Right::Fruit);
fn collect(&mut self, doc: DocId, score: Score) {
self.0.collect(doc, score);
self.1.collect(doc, score);
}
fn harvest(self) -> <Self as SegmentCollector>::Fruit {
(self.0.harvest(), self.1.harvest())
}
}
// 3-Tuple
impl<One, Two, Three> Collector for (One, Two, Three)
where
One: Collector,
Two: Collector,
Three: Collector,
{
type Fruit = (One::Fruit, Two::Fruit, Three::Fruit);
type Child = (One::Child, Two::Child, Three::Child);
fn for_segment(&self, segment_local_id: u32, segment: &SegmentReader) -> Result<Self::Child> {
let one = self.0.for_segment(segment_local_id, segment)?;
let two = self.1.for_segment(segment_local_id, segment)?;
let three = self.2.for_segment(segment_local_id, segment)?;
Ok((one, two, three))
}
fn requires_scoring(&self) -> bool {
self.0.requires_scoring() || self.1.requires_scoring() || self.2.requires_scoring()
}
fn merge_fruits(&self, children: Vec<Self::Fruit>) -> Result<Self::Fruit> {
let mut one_fruits = vec![];
let mut two_fruits = vec![];
let mut three_fruits = vec![];
for (one_fruit, two_fruit, three_fruit) in children {
one_fruits.push(one_fruit);
two_fruits.push(two_fruit);
three_fruits.push(three_fruit);
}
Ok((
self.0.merge_fruits(one_fruits)?,
self.1.merge_fruits(two_fruits)?,
self.2.merge_fruits(three_fruits)?,
))
}
}
impl<One, Two, Three> SegmentCollector for (One, Two, Three)
where
One: SegmentCollector,
Two: SegmentCollector,
Three: SegmentCollector,
{
type Fruit = (One::Fruit, Two::Fruit, Three::Fruit);
fn collect(&mut self, doc: DocId, score: Score) {
self.0.collect(doc, score);
self.1.collect(doc, score);
self.2.collect(doc, score);
}
fn harvest(self) -> <Self as SegmentCollector>::Fruit {
(self.0.harvest(), self.1.harvest(), self.2.harvest())
}
}
// 4-Tuple
impl<One, Two, Three, Four> Collector for (One, Two, Three, Four)
where
One: Collector,
Two: Collector,
Three: Collector,
Four: Collector,
{
type Fruit = (One::Fruit, Two::Fruit, Three::Fruit, Four::Fruit);
type Child = (One::Child, Two::Child, Three::Child, Four::Child);
fn for_segment(&self, segment_local_id: u32, segment: &SegmentReader) -> Result<Self::Child> {
let one = self.0.for_segment(segment_local_id, segment)?;
let two = self.1.for_segment(segment_local_id, segment)?;
let three = self.2.for_segment(segment_local_id, segment)?;
let four = self.3.for_segment(segment_local_id, segment)?;
Ok((one, two, three, four))
}
fn requires_scoring(&self) -> bool {
self.0.requires_scoring()
|| self.1.requires_scoring()
|| self.2.requires_scoring()
|| self.3.requires_scoring()
}
fn merge_fruits(&self, children: Vec<Self::Fruit>) -> Result<Self::Fruit> {
let mut one_fruits = vec![];
let mut two_fruits = vec![];
let mut three_fruits = vec![];
let mut four_fruits = vec![];
for (one_fruit, two_fruit, three_fruit, four_fruit) in children {
one_fruits.push(one_fruit);
two_fruits.push(two_fruit);
three_fruits.push(three_fruit);
four_fruits.push(four_fruit);
}
Ok((
self.0.merge_fruits(one_fruits)?,
self.1.merge_fruits(two_fruits)?,
self.2.merge_fruits(three_fruits)?,
self.3.merge_fruits(four_fruits)?,
))
}
}
impl<One, Two, Three, Four> SegmentCollector for (One, Two, Three, Four)
where
One: SegmentCollector,
Two: SegmentCollector,
Three: SegmentCollector,
Four: SegmentCollector,
{
type Fruit = (One::Fruit, Two::Fruit, Three::Fruit, Four::Fruit);
fn collect(&mut self, doc: DocId, score: Score) {
self.0.collect(doc, score);
self.1.collect(doc, score);
self.2.collect(doc, score);
self.3.collect(doc, score);
}
fn harvest(self) -> <Self as SegmentCollector>::Fruit {
(
self.0.harvest(),
self.1.harvest(),
self.2.harvest(),
self.3.harvest(),
)
}
}
impl_downcast!(Fruit);
#[cfg(test)]
pub mod tests {
use super::*;
use test::Bencher;
use DocId;
use Score;
use core::SegmentReader;
use SegmentLocalId;
use fastfield::FastFieldReader;
use schema::Field;
/// Stores all of the doc ids.
/// This collector is only used for tests.
/// It is unusable in practise, as it does not store
/// the segment ordinals
pub struct TestCollector {
offset: DocId,
segment_max_doc: DocId,
docs: Vec<DocId>,
scores: Vec<Score>,
}
impl TestCollector {
/// Return the exhalist of documents.
pub fn docs(self) -> Vec<DocId> {
self.docs
}
pub fn scores(self) -> Vec<Score> {
self.scores
}
}
impl Default for TestCollector {
fn default() -> TestCollector {
TestCollector {
offset: 0,
segment_max_doc: 0,
docs: Vec::new(),
scores: Vec::new(),
}
}
}
impl Collector for TestCollector {
fn set_segment(&mut self, _: SegmentLocalId, reader: &SegmentReader) -> Result<()> {
self.offset += self.segment_max_doc;
self.segment_max_doc = reader.max_doc();
Ok(())
}
fn collect(&mut self, doc: DocId, score: Score) {
self.docs.push(doc + self.offset);
self.scores.push(score);
}
fn requires_scoring(&self) -> bool {
true
}
}
/// Collects in order all of the fast fields for all of the
/// doc in the `DocSet`
///
/// This collector is mainly useful for tests.
pub struct FastFieldTestCollector {
vals: Vec<u64>,
field: Field,
ff_reader: Option<FastFieldReader<u64>>,
}
impl FastFieldTestCollector {
pub fn for_field(field: Field) -> FastFieldTestCollector {
FastFieldTestCollector {
vals: Vec::new(),
field,
ff_reader: None,
}
}
pub fn vals(self) -> Vec<u64> {
self.vals
}
}
impl Collector for FastFieldTestCollector {
fn set_segment(&mut self, _: SegmentLocalId, reader: &SegmentReader) -> Result<()> {
self.ff_reader = Some(reader.fast_field_reader(self.field)?);
Ok(())
}
fn collect(&mut self, doc: DocId, _score: Score) {
let val = self.ff_reader.as_ref().unwrap().get(doc);
self.vals.push(val);
}
fn requires_scoring(&self) -> bool {
false
}
}
#[bench]
fn build_collector(b: &mut Bencher) {
b.iter(|| {
let mut count_collector = CountCollector::default();
let docs: Vec<u32> = (0..1_000_000).collect();
for doc in docs {
count_collector.collect(doc, 1f32);
}
count_collector.count()
});
}
}
pub mod tests;

290
src/collector/multi_collector.rs
View File

@@ -1,46 +1,255 @@
use super::Collector;
use super::SegmentCollector;
use collector::Fruit;
use std::marker::PhantomData;
use std::ops::Deref;
use DocId;
use Score;
use Result;
use SegmentReader;
use Score;
use SegmentLocalId;
use SegmentReader;
use TantivyError;
pub struct MultiFruit {
sub_fruits: Vec<Option<Box<Fruit>>>,
}
pub struct CollectorWrapper<TCollector: Collector>(TCollector);
impl<TCollector: Collector> Collector for CollectorWrapper<TCollector> {
type Fruit = Box<Fruit>;
type Child = Box<BoxableSegmentCollector>;
fn for_segment(
&self,
segment_local_id: u32,
reader: &SegmentReader,
) -> Result<Box<BoxableSegmentCollector>> {
let child = self.0.for_segment(segment_local_id, reader)?;
Ok(Box::new(SegmentCollectorWrapper(child)))
}
fn requires_scoring(&self) -> bool {
self.0.requires_scoring()
}
fn merge_fruits(&self, children: Vec<<Self as Collector>::Fruit>) -> Result<Box<Fruit>> {
let typed_fruit: Vec<TCollector::Fruit> = children
.into_iter()
.map(|untyped_fruit| {
untyped_fruit
.downcast::<TCollector::Fruit>()
.map(|boxed_but_typed| *boxed_but_typed)
.map_err(|_| {
TantivyError::InvalidArgument("Failed to cast child fruit.".to_string())
})
})
.collect::<Result<_>>()?;
let merged_fruit = self.0.merge_fruits(typed_fruit)?;
Ok(Box::new(merged_fruit))
}
}
impl SegmentCollector for Box<BoxableSegmentCollector> {
type Fruit = Box<Fruit>;
fn collect(&mut self, doc: u32, score: f32) {
self.as_mut().collect(doc, score);
}
fn harvest(self) -> Box<Fruit> {
BoxableSegmentCollector::harvest_from_box(self)
}
}
pub trait BoxableSegmentCollector {
fn collect(&mut self, doc: u32, score: f32);
fn harvest_from_box(self: Box<Self>) -> Box<Fruit>;
}
pub struct SegmentCollectorWrapper<TSegmentCollector: SegmentCollector>(TSegmentCollector);
impl<TSegmentCollector: SegmentCollector> BoxableSegmentCollector
for SegmentCollectorWrapper<TSegmentCollector>
{
fn collect(&mut self, doc: u32, score: f32) {
self.0.collect(doc, score);
}
fn harvest_from_box(self: Box<Self>) -> Box<Fruit> {
Box::new(self.0.harvest())
}
}
pub struct FruitHandle<TFruit: Fruit> {
pos: usize,
_phantom: PhantomData<TFruit>,
}
impl<TFruit: Fruit> FruitHandle<TFruit> {
pub fn extract(self, fruits: &mut MultiFruit) -> TFruit {
let boxed_fruit = fruits.sub_fruits[self.pos].take().expect("");
*boxed_fruit
.downcast::<TFruit>()
.map_err(|_| ())
.expect("Failed to downcast collector fruit.")
}
}
/// Multicollector makes it possible to collect on more than one collector.
/// It should only be used for use cases where the Collector types is unknown
/// at compile time.
/// If the type of the collectors is known, you should prefer to use `ChainedCollector`.
///
/// If the type of the collectors is known, you can just group yours collectors
/// in a tuple. See the
/// [Combining several collectors section of the collector documentation](./index.html#combining-several-collectors).
///
/// ```rust
/// #[macro_use]
/// extern crate tantivy;
/// use tantivy::schema::{Schema, TEXT};
/// use tantivy::{Index, Result};
/// use tantivy::collector::{Count, TopDocs, MultiCollector};
/// use tantivy::query::QueryParser;
///
/// # fn main() { example().unwrap(); }
/// fn example() -> Result<()> {
/// 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(3_000_000)?;
/// 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.commit().unwrap();
/// }
///
/// let reader = index.reader()?;
/// let searcher = reader.searcher();
///
/// let mut collectors = MultiCollector::new();
/// let top_docs_handle = collectors.add_collector(TopDocs::with_limit(2));
/// let count_handle = collectors.add_collector(Count);
/// let query_parser = QueryParser::for_index(&index, vec![title]);
/// let query = query_parser.parse_query("diary")?;
/// let mut multi_fruit = searcher.search(&query, &collectors)?;
///
/// let count = count_handle.extract(&mut multi_fruit);
/// let top_docs = top_docs_handle.extract(&mut multi_fruit);
///
/// # assert_eq!(count, 2);
/// # assert_eq!(top_docs.len(), 2);
///
/// Ok(())
/// }
/// ```
#[allow(clippy::type_complexity)]
#[derive(Default)]
pub struct MultiCollector<'a> {
collectors: Vec<&'a mut Collector>,
collector_wrappers:
Vec<Box<Collector<Child = Box<BoxableSegmentCollector>, Fruit = Box<Fruit>> + 'a>>,
}
impl<'a> MultiCollector<'a> {
/// Constructor
pub fn from(collectors: Vec<&'a mut Collector>) -> MultiCollector {
MultiCollector { collectors }
/// Create a new `MultiCollector`
pub fn new() -> Self {
Default::default()
}
/// Add a new collector to our `MultiCollector`.
pub fn add_collector<'b: 'a, TCollector: Collector + 'b>(
&mut self,
collector: TCollector,
) -> FruitHandle<TCollector::Fruit> {
let pos = self.collector_wrappers.len();
self.collector_wrappers
.push(Box::new(CollectorWrapper(collector)));
FruitHandle {
pos,
_phantom: PhantomData,
}
}
}
impl<'a> Collector for MultiCollector<'a> {
fn set_segment(
&mut self,
type Fruit = MultiFruit;
type Child = MultiCollectorChild;
fn for_segment(
&self,
segment_local_id: SegmentLocalId,
segment: &SegmentReader,
) -> Result<()> {
for collector in &mut self.collectors {
collector.set_segment(segment_local_id, segment)?;
}
Ok(())
) -> Result<MultiCollectorChild> {
let children = self
.collector_wrappers
.iter()
.map(|collector_wrapper| collector_wrapper.for_segment(segment_local_id, segment))
.collect::<Result<Vec<_>>>()?;
Ok(MultiCollectorChild { children })
}
fn requires_scoring(&self) -> bool {
self.collector_wrappers
.iter()
.map(Deref::deref)
.any(Collector::requires_scoring)
}
fn merge_fruits(&self, segments_multifruits: Vec<MultiFruit>) -> Result<MultiFruit> {
let mut segment_fruits_list: Vec<Vec<Box<Fruit>>> = (0..self.collector_wrappers.len())
.map(|_| Vec::with_capacity(segments_multifruits.len()))
.collect::<Vec<_>>();
for segment_multifruit in segments_multifruits {
for (idx, segment_fruit_opt) in segment_multifruit.sub_fruits.into_iter().enumerate() {
if let Some(segment_fruit) = segment_fruit_opt {
segment_fruits_list[idx].push(segment_fruit);
}
}
}
let sub_fruits = self
.collector_wrappers
.iter()
.zip(segment_fruits_list)
.map(|(child_collector, segment_fruits)| {
Ok(Some(child_collector.merge_fruits(segment_fruits)?))
})
.collect::<Result<_>>()?;
Ok(MultiFruit { sub_fruits })
}
}
pub struct MultiCollectorChild {
children: Vec<Box<BoxableSegmentCollector>>,
}
impl SegmentCollector for MultiCollectorChild {
type Fruit = MultiFruit;
fn collect(&mut self, doc: DocId, score: Score) {
for collector in &mut self.collectors {
collector.collect(doc, score);
for child in &mut self.children {
child.collect(doc, score);
}
}
fn requires_scoring(&self) -> bool {
self.collectors
.iter()
.any(|collector| collector.requires_scoring())
fn harvest(self) -> MultiFruit {
MultiFruit {
sub_fruits: self
.children
.into_iter()
.map(|child| Some(child.harvest()))
.collect(),
}
}
}
@@ -48,20 +257,41 @@ impl<'a> Collector for MultiCollector<'a> {
mod tests {
use super::*;
use collector::{Collector, CountCollector, TopCollector};
use collector::{Count, TopDocs};
use query::TermQuery;
use schema::IndexRecordOption;
use schema::{Schema, TEXT};
use Index;
use Term;
#[test]
fn test_multi_collector() {
let mut top_collector = TopCollector::with_limit(2);
let mut count_collector = CountCollector::default();
let mut schema_builder = Schema::builder();
let text = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
let mut collectors =
MultiCollector::from(vec![&mut top_collector, &mut count_collector]);
collectors.collect(1, 0.2);
collectors.collect(2, 0.1);
collectors.collect(3, 0.5);
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"));
index_writer.commit().unwrap();
index_writer.add_document(doc!(text=>""));
index_writer.add_document(doc!(text=>"abc abc abc abc"));
index_writer.add_document(doc!(text=>"abc"));
index_writer.commit().unwrap();
}
assert_eq!(count_collector.count(), 3);
assert!(top_collector.at_capacity());
let searcher = index.reader().unwrap().searcher();
let term = Term::from_field_text(text, "abc");
let query = TermQuery::new(term, IndexRecordOption::Basic);
let mut collectors = MultiCollector::new();
let topdocs_handler = collectors.add_collector(TopDocs::with_limit(2));
let count_handler = collectors.add_collector(Count);
let mut multifruits = searcher.search(&query, &mut collectors).unwrap();
assert_eq!(count_handler.extract(&mut multifruits), 5);
assert_eq!(topdocs_handler.extract(&mut multifruits).len(), 2);
}
}

208
src/collector/tests.rs Normal file
View File

@@ -0,0 +1,208 @@
use super::*;
use core::SegmentReader;
use fastfield::BytesFastFieldReader;
use fastfield::FastFieldReader;
use schema::Field;
use DocAddress;
use DocId;
use Score;
use SegmentLocalId;
/// Stores all of the doc ids.
/// This collector is only used for tests.
/// It is unusable in pr
///
/// actise, as it does not store
/// the segment ordinals
pub struct TestCollector;
pub struct TestSegmentCollector {
segment_id: SegmentLocalId,
fruit: TestFruit,
}
#[derive(Default)]
pub struct TestFruit {
docs: Vec<DocAddress>,
scores: Vec<Score>,
}
impl TestFruit {
/// Return the list of matching documents exhaustively.
pub fn docs(&self) -> &[DocAddress] {
&self.docs[..]
}
pub fn scores(&self) -> &[Score] {
&self.scores[..]
}
}
impl Collector for TestCollector {
type Fruit = TestFruit;
type Child = TestSegmentCollector;
fn for_segment(
&self,
segment_id: SegmentLocalId,
_reader: &SegmentReader,
) -> Result<TestSegmentCollector> {
Ok(TestSegmentCollector {
segment_id,
fruit: TestFruit::default(),
})
}
fn requires_scoring(&self) -> bool {
true
}
fn merge_fruits(&self, mut children: Vec<TestFruit>) -> Result<TestFruit> {
children.sort_by_key(|fruit| {
if fruit.docs().is_empty() {
0
} else {
fruit.docs()[0].segment_ord()
}
});
let mut docs = vec![];
let mut scores = vec![];
for child in children {
docs.extend(child.docs());
scores.extend(child.scores);
}
Ok(TestFruit { docs, scores })
}
}
impl SegmentCollector for TestSegmentCollector {
type Fruit = TestFruit;
fn collect(&mut self, doc: DocId, score: Score) {
self.fruit.docs.push(DocAddress(self.segment_id, doc));
self.fruit.scores.push(score);
}
fn harvest(self) -> <Self as SegmentCollector>::Fruit {
self.fruit
}
}
/// Collects in order all of the fast fields for all of the
/// doc in the `DocSet`
///
/// This collector is mainly useful for tests.
pub struct FastFieldTestCollector {
field: Field,
}
pub struct FastFieldSegmentCollector {
vals: Vec<u64>,
reader: FastFieldReader<u64>,
}
impl FastFieldTestCollector {
pub fn for_field(field: Field) -> FastFieldTestCollector {
FastFieldTestCollector { field }
}
}
impl Collector for FastFieldTestCollector {
type Fruit = Vec<u64>;
type Child = FastFieldSegmentCollector;
fn for_segment(
&self,
_: SegmentLocalId,
segment_reader: &SegmentReader,
) -> Result<FastFieldSegmentCollector> {
let reader = segment_reader
.fast_fields()
.u64(self.field)
.expect("Requested field is not a fast field.");
Ok(FastFieldSegmentCollector {
vals: Vec::new(),
reader,
})
}
fn requires_scoring(&self) -> bool {
false
}
fn merge_fruits(&self, children: Vec<Vec<u64>>) -> Result<Vec<u64>> {
Ok(children.into_iter().flat_map(|v| v.into_iter()).collect())
}
}
impl SegmentCollector for FastFieldSegmentCollector {
type Fruit = Vec<u64>;
fn collect(&mut self, doc: DocId, _score: Score) {
let val = self.reader.get(doc);
self.vals.push(val);
}
fn harvest(self) -> Vec<u64> {
self.vals
}
}
/// Collects in order all of the fast field bytes for all of the
/// docs in the `DocSet`
///
/// This collector is mainly useful for tests.
pub struct BytesFastFieldTestCollector {
field: Field,
}
pub struct BytesFastFieldSegmentCollector {
vals: Vec<u8>,
reader: BytesFastFieldReader,
}
impl BytesFastFieldTestCollector {
pub fn for_field(field: Field) -> BytesFastFieldTestCollector {
BytesFastFieldTestCollector { field }
}
}
impl Collector for BytesFastFieldTestCollector {
type Fruit = Vec<u8>;
type Child = BytesFastFieldSegmentCollector;
fn for_segment(
&self,
_segment_local_id: u32,
segment_reader: &SegmentReader,
) -> Result<BytesFastFieldSegmentCollector> {
Ok(BytesFastFieldSegmentCollector {
vals: Vec::new(),
reader: segment_reader
.fast_fields()
.bytes(self.field)
.expect("Field is not a bytes fast field."),
})
}
fn requires_scoring(&self) -> bool {
false
}
fn merge_fruits(&self, children: Vec<Vec<u8>>) -> Result<Vec<u8>> {
Ok(children.into_iter().flat_map(|c| c.into_iter()).collect())
}
}
impl SegmentCollector for BytesFastFieldSegmentCollector {
type Fruit = Vec<u8>;
fn collect(&mut self, doc: u32, _score: f32) {
let data = self.reader.get_bytes(doc);
self.vals.extend(data);
}
fn harvest(self) -> <Self as SegmentCollector>::Fruit {
self.vals
}
}

257
src/collector/top_collector.rs
View File

@@ -1,189 +1,224 @@
use super::Collector;
use SegmentReader;
use SegmentLocalId;
use DocAddress;
use Result;
use std::collections::BinaryHeap;
use serde::export::PhantomData;
use std::cmp::Ordering;
use std::collections::BinaryHeap;
use DocAddress;
use DocId;
use Score;
use Result;
use SegmentLocalId;
use SegmentReader;
// Rust heap is a max-heap and we need a min heap.
#[derive(Clone, Copy)]
struct GlobalScoredDoc {
score: Score,
doc_address: DocAddress,
/// Contains a feature (field, score, etc.) of a document along with the document address.
///
/// It has a custom implementation of `PartialOrd` that reverses the order. This is because the
/// default Rust heap is a max heap, whereas a min heap is needed.
///
/// WARNING: equality is not what you would expect here.
/// 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.
struct ComparableDoc<T, D> {
feature: T,
doc: D,
}
impl PartialOrd for GlobalScoredDoc {
fn partial_cmp(&self, other: &GlobalScoredDoc) -> Option<Ordering> {
impl<T: PartialOrd, D> PartialOrd for ComparableDoc<T, D> {
fn partial_cmp(&self, other: &Self) -> Option<Ordering> {
Some(self.cmp(other))
}
}
impl Ord for GlobalScoredDoc {
impl<T: PartialOrd, D> Ord for ComparableDoc<T, D> {
#[inline]
fn cmp(&self, other: &GlobalScoredDoc) -> Ordering {
fn cmp(&self, other: &Self) -> Ordering {
other
.score
.partial_cmp(&self.score)
.unwrap_or_else(|| other.doc_address.cmp(&self.doc_address))
.feature
.partial_cmp(&self.feature)
.unwrap_or_else(|| Ordering::Equal)
}
}
impl PartialEq for GlobalScoredDoc {
fn eq(&self, other: &GlobalScoredDoc) -> bool {
impl<T: PartialOrd, D> PartialEq for ComparableDoc<T, D> {
fn eq(&self, other: &Self) -> bool {
self.cmp(other) == Ordering::Equal
}
}
impl Eq for GlobalScoredDoc {}
impl<T: PartialOrd, D> Eq for ComparableDoc<T, D> {}
/// The Top Collector keeps track of the K documents
/// with the best scores.
///
/// The implementation is based on a `BinaryHeap`.
/// The theorical complexity is `O(n log K)`.
pub struct TopCollector {
pub(crate) struct TopCollector<T> {
limit: usize,
heap: BinaryHeap<GlobalScoredDoc>,
segment_id: u32,
_marker: PhantomData<T>,
}
impl TopCollector {
impl<T> TopCollector<T>
where
T: PartialOrd + Clone,
{
/// Creates a top collector, with a number of documents equal to "limit".
///
/// # Panics
/// The method panics if limit is 0
pub fn with_limit(limit: usize) -> TopCollector {
pub fn with_limit(limit: usize) -> TopCollector<T> {
if limit < 1 {
panic!("Limit must be strictly greater than 0.");
}
TopCollector {
limit,
heap: BinaryHeap::with_capacity(limit),
segment_id: 0,
_marker: PhantomData,
}
}
/// Returns K best documents sorted in decreasing order.
///
/// Calling this method triggers the sort.
/// The result of the sort is not cached.
pub fn docs(&self) -> Vec<DocAddress> {
self.score_docs()
.into_iter()
.map(|score_doc| score_doc.1)
.collect()
pub fn limit(&self) -> usize {
self.limit
}
/// Returns K best ScoredDocument sorted in decreasing order.
///
/// Calling this method triggers the sort.
/// The result of the sort is not cached.
pub fn score_docs(&self) -> Vec<(Score, DocAddress)> {
let mut scored_docs: Vec<GlobalScoredDoc> = self.heap.iter().cloned().collect();
scored_docs.sort();
scored_docs
pub fn merge_fruits(
&self,
children: Vec<Vec<(T, DocAddress)>>,
) -> Result<Vec<(T, DocAddress)>> {
if self.limit == 0 {
return Ok(Vec::new());
}
let mut top_collector = BinaryHeap::new();
for child_fruit in children {
for (feature, doc) in child_fruit {
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 {
*head = ComparableDoc { feature, doc };
}
}
}
}
Ok(top_collector
.into_sorted_vec()
.into_iter()
.map(|GlobalScoredDoc { score, doc_address }| (score, doc_address))
.map(|cdoc| (cdoc.feature, cdoc.doc))
.collect())
}
pub(crate) fn for_segment<F: PartialOrd>(
&self,
segment_id: SegmentLocalId,
_: &SegmentReader,
) -> Result<TopSegmentCollector<F>> {
Ok(TopSegmentCollector::new(segment_id, self.limit))
}
}
/// The Top Collector keeps track of the K documents
/// sorted by type `T`.
///
/// The implementation is based on a `BinaryHeap`.
/// The theorical complexity for collecting the top `K` out of `n` documents
/// is `O(n log K)`.
pub(crate) struct TopSegmentCollector<T> {
limit: usize,
heap: BinaryHeap<ComparableDoc<T, DocId>>,
segment_id: u32,
}
impl<T: PartialOrd> TopSegmentCollector<T> {
fn new(segment_id: SegmentLocalId, limit: usize) -> TopSegmentCollector<T> {
TopSegmentCollector {
limit,
heap: BinaryHeap::with_capacity(limit),
segment_id,
}
}
}
impl<T: PartialOrd + Clone> TopSegmentCollector<T> {
pub fn harvest(self) -> Vec<(T, DocAddress)> {
let segment_id = self.segment_id;
self.heap
.into_sorted_vec()
.into_iter()
.map(|comparable_doc| {
(
comparable_doc.feature,
DocAddress(segment_id, comparable_doc.doc),
)
})
.collect()
}
/// Return true iff at least K documents have gone through
/// the collector.
#[inline]
pub fn at_capacity(&self) -> bool {
#[inline(always)]
pub(crate) fn at_capacity(&self) -> bool {
self.heap.len() >= self.limit
}
}
impl Collector for TopCollector {
fn set_segment(&mut self, segment_id: SegmentLocalId, _: &SegmentReader) -> Result<()> {
self.segment_id = segment_id;
Ok(())
}
fn collect(&mut self, doc: DocId, score: Score) {
/// Collects a document scored by the given feature
///
/// It collects documents until it has reached the max capacity. Once it reaches capacity, it
/// will compare the lowest scoring item with the given one and keep whichever is greater.
#[inline(always)]
pub fn collect(&mut self, doc: DocId, feature: T) {
if self.at_capacity() {
// It's ok to unwrap as long as a limit of 0 is forbidden.
let limit_doc: GlobalScoredDoc = *self.heap
.peek()
.expect("Top collector with size 0 is forbidden");
if limit_doc.score < score {
let mut mut_head = self.heap
.peek_mut()
.expect("Top collector with size 0 is forbidden");
mut_head.score = score;
mut_head.doc_address = DocAddress(self.segment_id, doc);
if let Some(limit_feature) = self.heap.peek().map(|head| head.feature.clone()) {
if limit_feature < feature {
if let Some(mut head) = self.heap.peek_mut() {
head.feature = feature;
head.doc = doc;
}
}
}
} else {
let wrapped_doc = GlobalScoredDoc {
score,
doc_address: DocAddress(self.segment_id, doc),
};
self.heap.push(wrapped_doc);
// we have not reached capacity yet, so we can just push the
// element.
self.heap.push(ComparableDoc { feature, doc });
}
}
fn requires_scoring(&self) -> bool {
true
}
}
#[cfg(test)]
mod tests {
use super::*;
use DocId;
use super::{TopCollector, TopSegmentCollector};
use DocAddress;
use Score;
use collector::Collector;
#[test]
fn test_top_collector_not_at_capacity() {
let mut top_collector = TopCollector::with_limit(4);
let mut top_collector = TopSegmentCollector::new(0, 4);
top_collector.collect(1, 0.8);
top_collector.collect(3, 0.2);
top_collector.collect(5, 0.3);
assert!(!top_collector.at_capacity());
let score_docs: Vec<(Score, DocId)> = top_collector
.score_docs()
.into_iter()
.map(|(score, doc_address)| (score, doc_address.doc()))
.collect();
assert_eq!(score_docs, vec![(0.8, 1), (0.3, 5), (0.2, 3)]);
assert_eq!(
top_collector.harvest(),
vec![
(0.8, DocAddress(0, 1)),
(0.3, DocAddress(0, 5)),
(0.2, DocAddress(0, 3))
]
);
}
#[test]
fn test_top_collector_at_capacity() {
let mut top_collector = TopCollector::with_limit(4);
let mut top_collector = TopSegmentCollector::new(0, 4);
top_collector.collect(1, 0.8);
top_collector.collect(3, 0.2);
top_collector.collect(5, 0.3);
top_collector.collect(7, 0.9);
top_collector.collect(9, -0.2);
assert!(top_collector.at_capacity());
{
let score_docs: Vec<(Score, DocId)> = top_collector
.score_docs()
.into_iter()
.map(|(score, doc_address)| (score, doc_address.doc()))
.collect();
assert_eq!(score_docs, vec![(0.9, 7), (0.8, 1), (0.3, 5), (0.2, 3)]);
}
{
let docs: Vec<DocId> = top_collector
.docs()
.into_iter()
.map(|doc_address| doc_address.doc())
.collect();
assert_eq!(docs, vec![7, 1, 5, 3]);
}
assert_eq!(
top_collector.harvest(),
vec![
(0.9, DocAddress(0, 7)),
(0.8, DocAddress(0, 1)),
(0.3, DocAddress(0, 5)),
(0.2, DocAddress(0, 3))
]
);
}
#[test]
#[should_panic]
fn test_top_0() {
TopCollector::with_limit(0);
let _collector: TopCollector<Score> = TopCollector::with_limit(0);
}
}

271
src/collector/top_field_collector.rs Normal file
View File

@@ -0,0 +1,271 @@
use super::Collector;
use collector::top_collector::TopCollector;
use collector::top_collector::TopSegmentCollector;
use collector::SegmentCollector;
use fastfield::FastFieldReader;
use fastfield::FastValue;
use schema::Field;
use std::marker::PhantomData;
use DocAddress;
use Result;
use SegmentLocalId;
use SegmentReader;
use TantivyError;
/// The Top Field Collector keeps track of the K documents
/// sorted by a fast field in the index
///
/// The implementation is based on a `BinaryHeap`.
/// The theorical complexity for collecting the top `K` out of `n` documents
/// is `O(n log K)`.
///
/// ```rust
/// #[macro_use]
/// extern crate tantivy;
/// # use tantivy::schema::{Schema, Field, FAST, TEXT};
/// # use tantivy::{Index, Result, DocAddress};
/// # use tantivy::query::{Query, QueryParser};
/// use tantivy::Searcher;
/// use tantivy::collector::TopDocs;
///
/// # fn main() -> tantivy::Result<()> {
/// # let mut schema_builder = Schema::builder();
/// # let title = schema_builder.add_text_field("title", TEXT);
/// # let rating = schema_builder.add_u64_field("rating", FAST);
/// # 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)?;
/// # 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));
/// # index_writer.commit()?;
/// # let reader = index.reader()?;
/// # 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,
/// # vec![(97u64, DocAddress(0u32, 1)),
/// # (80u64, DocAddress(0u32, 3))]);
/// # Ok(())
/// # }
/// #
/// /// Searches the document matching the given query, and
/// /// collects the top 10 documents, order by the `field`
/// /// given in argument.
/// ///
/// /// `field` is required to be a FAST field.
/// fn docs_sorted_by_rating(searcher: &Searcher,
/// query: &Query,
/// sort_by_field: Field)
/// -> Result<Vec<(u64, DocAddress)>> {
///
/// // This is where we build our collector!
/// let top_docs_by_rating = TopDocs::with_limit(2).order_by_field(sort_by_field);
///
/// // ... and here is our documents. Not this is a simple vec.
/// // The `u64` in the pair is the value of our fast field for each documents.
/// searcher.search(query, &top_docs_by_rating)
/// }
/// ```
pub struct TopDocsByField<T> {
collector: TopCollector<T>,
field: Field,
}
impl<T: FastValue + PartialOrd + Clone> TopDocsByField<T> {
/// Creates a top field collector, with a number of documents equal to "limit".
///
/// The given field name must be a fast field, otherwise the collector have an error while
/// collecting results.
///
/// This constructor is crate-private. Client are supposed to call
/// build `TopDocsByField` object using the `TopDocs` API.
///
/// e.g.:
/// `TopDocs::with_limit(2).order_by_field(sort_by_field)`
///
/// # Panics
/// The method panics if limit is 0
pub(crate) fn new(field: Field, limit: usize) -> TopDocsByField<T> {
TopDocsByField {
collector: TopCollector::with_limit(limit),
field,
}
}
}
impl<T: FastValue + PartialOrd + Send + Sync + 'static> Collector for TopDocsByField<T> {
type Fruit = Vec<(T, DocAddress)>;
type Child = TopFieldSegmentCollector<T>;
fn for_segment(
&self,
segment_local_id: SegmentLocalId,
reader: &SegmentReader,
) -> Result<TopFieldSegmentCollector<T>> {
let collector = self.collector.for_segment(segment_local_id, reader)?;
let reader = reader.fast_fields().u64(self.field).ok_or_else(|| {
let field_name = reader.schema().get_field_name(self.field);
TantivyError::SchemaError(format!("Failed to find fast field reader {:?}", field_name))
})?;
Ok(TopFieldSegmentCollector {
collector,
reader,
_type: PhantomData,
})
}
fn requires_scoring(&self) -> bool {
false
}
fn merge_fruits(
&self,
segment_fruits: Vec<Vec<(T, DocAddress)>>,
) -> Result<Vec<(T, DocAddress)>> {
self.collector.merge_fruits(segment_fruits)
}
}
pub struct TopFieldSegmentCollector<T> {
collector: TopSegmentCollector<u64>,
reader: FastFieldReader<u64>,
_type: PhantomData<T>,
}
impl<T: FastValue + PartialOrd + Send + Sync + 'static> SegmentCollector
for TopFieldSegmentCollector<T>
{
type Fruit = Vec<(T, DocAddress)>;
fn collect(&mut self, doc: u32, _score: f32) {
let field_value = self.reader.get(doc);
self.collector.collect(doc, field_value);
}
fn harvest(self) -> Vec<(T, DocAddress)> {
self.collector
.harvest()
.into_iter()
.map(|(val, doc_address)| (T::from_u64(val), doc_address))
.collect()
}
}
#[cfg(test)]
mod tests {
use super::TopDocsByField;
use collector::Collector;
use collector::TopDocs;
use query::Query;
use query::QueryParser;
use schema::Field;
use schema::IntOptions;
use schema::{Schema, FAST, TEXT};
use DocAddress;
use Index;
use IndexWriter;
use TantivyError;
const TITLE: &str = "title";
const SIZE: &str = "size";
#[test]
fn test_top_collector_not_at_capacity() {
let mut schema_builder = Schema::builder();
let title = schema_builder.add_text_field(TITLE, TEXT);
let size = schema_builder.add_u64_field(SIZE, FAST);
let schema = schema_builder.build();
let (index, query) = index("beer", title, schema, |index_writer| {
index_writer.add_document(doc!(
title => "bottle of beer",
size => 12u64,
));
index_writer.add_document(doc!(
title => "growler of beer",
size => 64u64,
));
index_writer.add_document(doc!(
title => "pint of beer",
size => 16u64,
));
});
let searcher = index.reader().unwrap().searcher();
let top_collector = TopDocs::with_limit(4).order_by_field(size);
let top_docs: Vec<(u64, DocAddress)> = searcher.search(&query, &top_collector).unwrap();
assert_eq!(
top_docs,
vec![
(64, DocAddress(0, 1)),
(16, DocAddress(0, 2)),
(12, DocAddress(0, 0))
]
);
}
#[test]
#[should_panic]
fn test_field_does_not_exist() {
let mut schema_builder = Schema::builder();
let title = schema_builder.add_text_field(TITLE, TEXT);
let size = schema_builder.add_u64_field(SIZE, FAST);
let schema = schema_builder.build();
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 top_collector: TopDocsByField<u64> = TopDocs::with_limit(4).order_by_field(Field(2));
let segment_reader = searcher.segment_reader(0u32);
top_collector
.for_segment(0, segment_reader)
.expect("should panic");
}
#[test]
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, IntOptions::default());
let schema = schema_builder.build();
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: TopDocsByField<u64> = TopDocs::with_limit(4).order_by_field(size);
assert_matches!(
top_collector
.for_segment(0, segment)
.map(|_| ())
.unwrap_err(),
TantivyError::SchemaError(_)
);
}
fn index(
query: &str,
query_field: Field,
schema: Schema,
mut doc_adder: impl FnMut(&mut IndexWriter) -> (),
) -> (Index, Box<Query>) {
let index = Index::create_in_ram(schema);
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]);
let query = query_parser.parse_query(query).unwrap();
(index, query)
}
}

203
src/collector/top_score_collector.rs Normal file
View File

@@ -0,0 +1,203 @@
use super::Collector;
use collector::top_collector::TopCollector;
use collector::top_collector::TopSegmentCollector;
use collector::SegmentCollector;
use collector::TopDocsByField;
use fastfield::FastValue;
use schema::Field;
use DocAddress;
use DocId;
use Result;
use Score;
use SegmentLocalId;
use SegmentReader;
/// The Top Score Collector keeps track of the K documents
/// sorted by their score.
///
/// The implementation is based on a `BinaryHeap`.
/// The theorical complexity for collecting the top `K` out of `n` documents
/// is `O(n log K)`.
///
/// ```rust
/// #[macro_use]
/// extern crate tantivy;
/// use tantivy::DocAddress;
/// use tantivy::schema::{Schema, TEXT};
/// use tantivy::{Index, Result};
/// use tantivy::collector::TopDocs;
/// use tantivy::query::QueryParser;
///
/// # fn main() { example().unwrap(); }
/// fn example() -> Result<()> {
/// 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, 3_000_000)?;
/// 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.commit().unwrap();
/// }
///
/// let reader = index.reader()?;
/// let searcher = reader.searcher();
///
/// let query_parser = QueryParser::for_index(&index, vec![title]);
/// let query = query_parser.parse_query("diary")?;
/// let top_docs = searcher.search(&query, &TopDocs::with_limit(2))?;
///
/// assert_eq!(&top_docs[0], &(0.7261542, DocAddress(0, 1)));
/// assert_eq!(&top_docs[1], &(0.6099695, DocAddress(0, 3)));
///
/// Ok(())
/// }
/// ```
pub struct TopDocs(TopCollector<Score>);
impl TopDocs {
/// Creates a top score collector, with a number of documents equal to "limit".
///
/// # Panics
/// The method panics if limit is 0
pub fn with_limit(limit: usize) -> TopDocs {
TopDocs(TopCollector::with_limit(limit))
}
/// Set top-K to rank documents by a given fast field.
///
/// (By default, `TopDocs` collects the top-K documents sorted by
/// the similarity score.)
pub fn order_by_field<T: PartialOrd + FastValue + Clone>(
self,
field: Field,
) -> TopDocsByField<T> {
TopDocsByField::new(field, self.0.limit())
}
}
impl Collector for TopDocs {
type Fruit = Vec<(Score, DocAddress)>;
type Child = TopScoreSegmentCollector;
fn for_segment(
&self,
segment_local_id: SegmentLocalId,
reader: &SegmentReader,
) -> Result<Self::Child> {
let collector = self.0.for_segment(segment_local_id, reader)?;
Ok(TopScoreSegmentCollector(collector))
}
fn requires_scoring(&self) -> bool {
true
}
fn merge_fruits(&self, child_fruits: Vec<Vec<(Score, DocAddress)>>) -> Result<Self::Fruit> {
self.0.merge_fruits(child_fruits)
}
}
/// Segment Collector associated to `TopDocs`.
pub struct TopScoreSegmentCollector(TopSegmentCollector<Score>);
impl SegmentCollector for TopScoreSegmentCollector {
type Fruit = Vec<(Score, DocAddress)>;
fn collect(&mut self, doc: DocId, score: Score) {
self.0.collect(doc, score)
}
fn harvest(self) -> Vec<(Score, DocAddress)> {
self.0.harvest()
}
}
#[cfg(test)]
mod tests {
use super::TopDocs;
use query::QueryParser;
use schema::Schema;
use schema::TEXT;
use DocAddress;
use Index;
use Score;
fn make_index() -> Index {
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=>"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"));
assert!(index_writer.commit().is_ok());
}
index
}
#[test]
fn test_top_collector_not_at_capacity() {
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))
.unwrap();
assert_eq!(
score_docs,
vec![
(0.81221175, DocAddress(0u32, 1)),
(0.5376842, DocAddress(0u32, 2)),
(0.48527452, DocAddress(0, 0))
]
);
}
#[test]
fn test_top_collector_at_capacity() {
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))
.unwrap();
assert_eq!(
score_docs,
vec![
(0.81221175, DocAddress(0u32, 1)),
(0.5376842, DocAddress(0u32, 2)),
]
);
}
#[test]
#[should_panic]
fn test_top_0() {
TopDocs::with_limit(0);
}
}

95
src/common/bitpacker.rs
View File

@@ -1,9 +1,6 @@
use std::io::Write;
use byteorder::{ByteOrder, LittleEndian, WriteBytesExt};
use std::io;
use common::serialize::BinarySerializable;
use std::mem;
use std::ops::Deref;
use std::ptr;
pub(crate) struct BitPacker {
mini_buffer: u64,
@@ -18,7 +15,7 @@ impl BitPacker {
}
}
pub fn write<TWrite: Write>(
pub fn write<TWrite: io::Write>(
&mut self,
val: u64,
num_bits: u8,
@@ -28,14 +25,14 @@ impl BitPacker {
let num_bits = num_bits as usize;
if self.mini_buffer_written + num_bits > 64 {
self.mini_buffer |= val_u64.wrapping_shl(self.mini_buffer_written as u32);
self.mini_buffer.serialize(output)?;
output.write_u64::<LittleEndian>(self.mini_buffer)?;
self.mini_buffer = val_u64.wrapping_shr((64 - self.mini_buffer_written) as u32);
self.mini_buffer_written = self.mini_buffer_written + num_bits - 64;
} else {
self.mini_buffer |= val_u64 << self.mini_buffer_written;
self.mini_buffer_written += num_bits;
if self.mini_buffer_written == 64 {
self.mini_buffer.serialize(output)?;
output.write_u64::<LittleEndian>(self.mini_buffer)?;
self.mini_buffer_written = 0;
self.mini_buffer = 0u64;
}
@@ -43,17 +40,18 @@ impl BitPacker {
Ok(())
}
pub fn flush<TWrite: Write>(&mut self, output: &mut TWrite) -> io::Result<()> {
pub fn flush<TWrite: io::Write>(&mut self, output: &mut TWrite) -> io::Result<()> {
if self.mini_buffer_written > 0 {
let num_bytes = (self.mini_buffer_written + 7) / 8;
let arr: [u8; 8] = unsafe { mem::transmute::<u64, [u8; 8]>(self.mini_buffer) };
let mut arr: [u8; 8] = [0u8; 8];
LittleEndian::write_u64(&mut arr, self.mini_buffer);
output.write_all(&arr[..num_bytes])?;
self.mini_buffer_written = 0;
}
Ok(())
}
pub fn close<TWrite: Write>(&mut self, output: &mut TWrite) -> io::Result<()> {
pub fn close<TWrite: io::Write>(&mut self, output: &mut TWrite) -> io::Result<()> {
self.flush(output)?;
// Padding the write file to simplify reads.
output.write_all(&[0u8; 7])?;
@@ -66,7 +64,7 @@ pub struct BitUnpacker<Data>
where
Data: Deref<Target = [u8]>,
{
num_bits: usize,
num_bits: u64,
mask: u64,
data: Data,
}
@@ -82,13 +80,13 @@ where
(1u64 << num_bits) - 1u64
};
BitUnpacker {
num_bits: num_bits as usize,
num_bits: u64::from(num_bits),
mask,
data,
}
}
pub fn get(&self, idx: usize) -> u64 {
pub fn get(&self, idx: u64) -> u64 {
if self.num_bits == 0 {
return 0u64;
}
@@ -98,55 +96,13 @@ where
let addr_in_bits = idx * num_bits;
let addr = addr_in_bits >> 3;
let bit_shift = addr_in_bits & 7;
if cfg!(feature = "simdcompression") {
// for simdcompression,
// the bitpacker is only used for fastfields,
// and we expect them to be always padded.
debug_assert!(
addr + 8 <= data.len(),
"The fast field field should have been padded with 7 bytes."
);
let val_unshifted_unmasked: u64 = unsafe { ptr::read_unaligned(data[addr..].as_ptr() as *const u64) };
let val_shifted = (val_unshifted_unmasked >> bit_shift) as u64;
val_shifted & mask
} else {
let val_unshifted_unmasked: u64 = if addr + 8 <= data.len() {
unsafe { ptr::read_unaligned(data[addr..].as_ptr() as *const u64) }
} else {
let mut buffer = [0u8; 8];
for i in addr..data.len() {
buffer[i - addr] += data[i];
}
unsafe { ptr::read_unaligned(buffer[..].as_ptr() as *const u64) }
};
let val_shifted = val_unshifted_unmasked >> (bit_shift as u64);
val_shifted & mask
}
}
/// Reads a range of values from the fast field.
///
/// The range of values read is from
/// `[start..start + output.len()[`
pub fn get_range(&self, start: u32, output: &mut [u64]) {
if self.num_bits == 0 {
for val in output.iter_mut() {
*val = 0u64;
}
} else {
let data: &[u8] = &*self.data;
let num_bits = self.num_bits;
let mask = self.mask;
let mut addr_in_bits = (start as usize) * num_bits;
for output_val in output.iter_mut() {
let addr = addr_in_bits >> 3;
let bit_shift = addr_in_bits & 7;
let val_unshifted_unmasked: u64 = unsafe { ptr::read_unaligned(data[addr..].as_ptr() as *const u64) };
let val_shifted = (val_unshifted_unmasked >> bit_shift) as u64;
*output_val = val_shifted & mask;
addr_in_bits += num_bits;
}
}
debug_assert!(
addr + 8 <= data.len() as u64,
"The fast field field should have been padded with 7 bytes."
);
let val_unshifted_unmasked: u64 = LittleEndian::read_u64(&data[(addr as usize)..]);
let val_shifted = (val_unshifted_unmasked >> bit_shift) as u64;
val_shifted & mask
}
}
@@ -173,7 +129,7 @@ mod test {
fn test_bitpacker_util(len: usize, num_bits: u8) {
let (bitunpacker, vals) = create_fastfield_bitpacker(len, num_bits);
for (i, val) in vals.iter().enumerate() {
assert_eq!(bitunpacker.get(i), *val);
assert_eq!(bitunpacker.get(i as u64), *val);
}
}
@@ -185,17 +141,4 @@ mod test {
test_bitpacker_util(6, 14);
test_bitpacker_util(1000, 14);
}
#[test]
fn test_bitpacker_range() {
let (bitunpacker, vals) = create_fastfield_bitpacker(100_000, 12);
let buffer_len = 100;
let mut buffer = vec![0u64; buffer_len];
for start in vec![0, 10, 20, 100, 1_000] {
bitunpacker.get_range(start as u32, &mut buffer[..]);
for i in 0..buffer_len {
assert_eq!(buffer[i], vals[start + i]);
}
}
}
}

38
src/common/bitset.rs
View File

@@ -34,17 +34,17 @@ impl TinySet {
}
/// Returns the complement of the set in `[0, 64[`.
fn complement(&self) -> TinySet {
fn complement(self) -> TinySet {
TinySet(!self.0)
}
/// Returns true iff the `TinySet` contains the element `el`.
pub fn contains(&self, el: u32) -> bool {
pub fn contains(self, el: u32) -> bool {
!self.intersect(TinySet::singleton(el)).is_empty()
}
/// Returns the intersection of `self` and `other`
pub fn intersect(&self, other: TinySet) -> TinySet {
pub fn intersect(self, other: TinySet) -> TinySet {
TinySet(self.0 & other.0)
}
@@ -77,7 +77,7 @@ impl TinySet {
/// Returns true iff the `TinySet` is empty.
#[inline(always)]
pub fn is_empty(&self) -> bool {
pub fn is_empty(self) -> bool {
self.0 == 0u64
}
@@ -114,7 +114,7 @@ impl TinySet {
self.0 = 0u64;
}
pub fn len(&self) -> u32 {
pub fn len(self) -> u32 {
self.0.count_ones()
}
}
@@ -202,15 +202,14 @@ impl BitSet {
#[cfg(test)]
mod tests {
extern crate test;
use tests;
use std::collections::HashSet;
use super::BitSet;
use super::TinySet;
use tests::generate_nonunique_unsorted;
use std::collections::BTreeSet;
use query::BitSetDocSet;
use docset::DocSet;
use query::BitSetDocSet;
use std::collections::BTreeSet;
use std::collections::HashSet;
use tests;
use tests::generate_nonunique_unsorted;
#[test]
fn test_tiny_set() {
@@ -267,14 +266,14 @@ mod tests {
#[test]
fn test_bitset_large() {
let arr = generate_nonunique_unsorted(1_000_000, 50_000);
let arr = generate_nonunique_unsorted(100_000, 5_000);
let mut btreeset: BTreeSet<u32> = BTreeSet::new();
let mut bitset = BitSet::with_max_value(1_000_000);
let mut bitset = BitSet::with_max_value(100_000);
for el in arr {
btreeset.insert(el);
bitset.insert(el);
}
for i in 0..1_000_000 {
for i in 0..100_000 {
assert_eq!(btreeset.contains(&i), bitset.contains(i));
}
assert_eq!(btreeset.len(), bitset.len());
@@ -343,7 +342,7 @@ mod tests {
#[test]
fn test_bitset_clear() {
let mut bitset = BitSet::with_max_value(1_000);
let els = tests::sample(1_000, 0.01f32);
let els = tests::sample(1_000, 0.01f64);
for &el in &els {
bitset.insert(el);
}
@@ -353,6 +352,14 @@ mod tests {
assert!(!bitset.contains(el));
}
}
}
#[cfg(all(test, feature = "unstable"))]
mod bench {
use super::BitSet;
use super::TinySet;
use test;
#[bench]
fn bench_tinyset_pop(b: &mut test::Bencher) {
@@ -385,5 +392,4 @@ mod tests {
fn bench_bitset_initialize(b: &mut test::Bencher) {
b.iter(|| BitSet::with_max_value(1_000_000));
}
}

50
src/common/composite_file.rs
View File

@@ -1,12 +1,14 @@
use std::io::Write;
use common::CountingWriter;
use std::collections::HashMap;
use schema::Field;
use common::VInt;
use directory::WritePtr;
use std::io::{self, Read};
use directory::ReadOnlySource;
use common::BinarySerializable;
use common::CountingWriter;
use common::VInt;
use directory::ReadOnlySource;
use directory::WritePtr;
use schema::Field;
use space_usage::FieldUsage;
use space_usage::PerFieldSpaceUsage;
use std::collections::HashMap;
use std::io::Write;
use std::io::{self, Read};
#[derive(Eq, PartialEq, Hash, Copy, Ord, PartialOrd, Clone, Debug)]
pub struct FileAddr {
@@ -30,17 +32,14 @@ impl BinarySerializable for FileAddr {
fn deserialize<R: Read>(reader: &mut R) -> io::Result<Self> {
let field = Field::deserialize(reader)?;
let idx = VInt::deserialize(reader)?.0 as usize;
Ok(FileAddr {
field,
idx,
})
Ok(FileAddr { field, idx })
}
}
/// A `CompositeWrite` is used to write a `CompositeFile`.
pub struct CompositeWrite<W = WritePtr> {
write: CountingWriter<W>,
offsets: HashMap<FileAddr, usize>,
offsets: HashMap<FileAddr, u64>,
}
impl<W: Write> CompositeWrite<W> {
@@ -67,7 +66,7 @@ impl<W: Write> CompositeWrite<W> {
&mut self.write
}
/// Close the composite file.
/// Close the composite file
///
/// An index of the different field offsets
/// will be written as a footer.
@@ -75,7 +74,8 @@ impl<W: Write> CompositeWrite<W> {
let footer_offset = self.write.written_bytes();
VInt(self.offsets.len() as u64).serialize(&mut self.write)?;
let mut offset_fields: Vec<_> = self.offsets
let mut offset_fields: Vec<_> = self
.offsets
.iter()
.map(|(file_addr, offset)| (*offset, *file_addr))
.collect();
@@ -115,7 +115,6 @@ impl CompositeFile {
let end = data.len();
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);
let mut footer_buffer = footer_data.as_slice();
@@ -166,20 +165,31 @@ impl CompositeFile {
/// to a given `Field` and stored in a `CompositeFile`.
pub fn open_read_with_idx(&self, field: Field, idx: usize) -> Option<ReadOnlySource> {
self.offsets_index
.get(&FileAddr { field, idx, })
.get(&FileAddr { field, idx })
.map(|&(from, to)| self.data.slice(from, to))
}
pub fn space_usage(&self) -> PerFieldSpaceUsage {
let mut fields = HashMap::new();
for (&field_addr, &(start, end)) in self.offsets_index.iter() {
fields
.entry(field_addr.field)
.or_insert_with(|| FieldUsage::empty(field_addr.field))
.add_field_idx(field_addr.idx, end - start);
}
PerFieldSpaceUsage::new(fields)
}
}
#[cfg(test)]
mod test {
use std::io::Write;
use super::{CompositeFile, CompositeWrite};
use common::BinarySerializable;
use common::VInt;
use directory::{Directory, RAMDirectory};
use schema::Field;
use common::VInt;
use common::BinarySerializable;
use std::io::Write;
use std::path::Path;
#[test]

20
src/common/counting_writer.rs
View File

@@ -1,9 +1,9 @@
use std::io::Write;
use std::io;
use std::io::Write;
pub struct CountingWriter<W> {
underlying: W,
written_bytes: usize,
written_bytes: u64,
}
impl<W: Write> CountingWriter<W> {
@@ -14,11 +14,11 @@ impl<W: Write> CountingWriter<W> {
}
}
pub fn written_bytes(&self) -> usize {
pub fn written_bytes(&self) -> u64 {
self.written_bytes
}
pub fn finish(mut self) -> io::Result<(W, usize)> {
pub fn finish(mut self) -> io::Result<(W, u64)> {
self.flush()?;
Ok((self.underlying, self.written_bytes))
}
@@ -27,10 +27,16 @@ impl<W: Write> CountingWriter<W> {
impl<W: Write> Write for CountingWriter<W> {
fn write(&mut self, buf: &[u8]) -> io::Result<usize> {
let written_size = self.underlying.write(buf)?;
self.written_bytes += written_size;
self.written_bytes += written_size as u64;
Ok(written_size)
}
fn write_all(&mut self, buf: &[u8]) -> io::Result<()> {
self.underlying.write_all(buf)?;
self.written_bytes += buf.len() as u64;
Ok(())
}
fn flush(&mut self) -> io::Result<()> {
self.underlying.flush()
}
@@ -48,8 +54,8 @@ mod test {
let mut counting_writer = CountingWriter::wrap(buffer);
let bytes = (0u8..10u8).collect::<Vec<u8>>();
counting_writer.write_all(&bytes).unwrap();
let (w, len): (Vec<u8>, usize) = counting_writer.finish().unwrap();
assert_eq!(len, 10);
let (w, len): (Vec<u8>, u64) = counting_writer.finish().unwrap();
assert_eq!(len, 10u64);
assert_eq!(w.len(), 10);
}
}

35
src/common/mod.rs
View File

@@ -1,19 +1,22 @@
mod serialize;
mod vint;
mod counting_writer;
mod composite_file;
pub mod bitpacker;
mod bitset;
mod composite_file;
mod counting_writer;
mod serialize;
mod vint;
pub(crate) use self::composite_file::{CompositeFile, CompositeWrite};
pub use self::serialize::{BinarySerializable, FixedSize};
pub use self::vint::VInt;
pub use self::counting_writer::CountingWriter;
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 use self::serialize::{BinarySerializable, FixedSize};
pub use self::vint::{read_u32_vint, serialize_vint_u32, write_u32_vint, VInt};
pub use byteorder::LittleEndian as Endianness;
use std::io;
/// Segment's max doc must be `< MAX_DOC_LIMIT`.
///
/// We do not allow segments with more than
pub const MAX_DOC_LIMIT: u32 = 1 << 31;
/// Computes the number of bits that will be used for bitpacking.
///
@@ -52,11 +55,6 @@ pub(crate) fn is_power_of_2(n: usize) -> bool {
(n > 0) && (n & (n - 1) == 0)
}
/// Create a default io error given a string.
pub(crate) fn make_io_err(msg: String) -> io::Error {
io::Error::new(io::ErrorKind::Other, msg)
}
/// Has length trait
pub trait HasLen {
/// Return length
@@ -104,8 +102,8 @@ pub fn u64_to_i64(val: u64) -> i64 {
#[cfg(test)]
pub(crate) mod test {
use super::{compute_num_bits, i64_to_u64, u64_to_i64};
pub use super::serialize::test::fixed_size_test;
use super::{compute_num_bits, i64_to_u64, u64_to_i64};
fn test_i64_converter_helper(val: i64) {
assert_eq!(u64_to_i64(i64_to_u64(val)), val);
@@ -134,4 +132,11 @@ pub(crate) mod test {
assert_eq!(compute_num_bits(256), 9u8);
assert_eq!(compute_num_bits(5_000_000_000), 33u8);
}
#[test]
fn test_max_doc() {
// this is the first time I write a unit test for a constant.
assert!(((super::MAX_DOC_LIMIT - 1) as i32) >= 0);
assert!((super::MAX_DOC_LIMIT as i32) < 0);
}
}

10
src/common/serialize.rs
View File

@@ -1,10 +1,10 @@
use byteorder::{ReadBytesExt, WriteBytesExt};
use common::Endianness;
use std::fmt;
use std::io::Write;
use std::io::Read;
use std::io;
use common::VInt;
use std::fmt;
use std::io;
use std::io::Read;
use std::io::Write;
/// Trait for a simple binary serialization.
pub trait BinarySerializable: fmt::Debug + Sized {
@@ -135,8 +135,8 @@ impl BinarySerializable for String {
#[cfg(test)]
pub mod test {
use common::VInt;
use super::*;
use common::VInt;
pub fn fixed_size_test<O: BinarySerializable + FixedSize + Default>() {
let mut buffer = Vec::new();

211
src/common/vint.rs
View File

@@ -1,12 +1,109 @@
use super::BinarySerializable;
use byteorder::{ByteOrder, LittleEndian};
use std::io;
use std::io::Write;
use std::io::Read;
use std::io::Write;
/// Wrapper over a `u64` that serializes as a variable int.
#[derive(Debug, Eq, PartialEq)]
pub struct VInt(pub u64);
const STOP_BIT: u8 = 128;
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;
const START_5: u64 = 1 << 28;
const STOP_1: u64 = START_2 - 1;
const STOP_2: u64 = START_3 - 1;
const STOP_3: u64 = START_4 - 1;
const STOP_4: u64 = START_5 - 1;
const MASK_1: u64 = 127;
const MASK_2: u64 = MASK_1 << 7;
const MASK_3: u64 = MASK_2 << 7;
const MASK_4: u64 = MASK_3 << 7;
const MASK_5: u64 = MASK_4 << 7;
let val = u64::from(val);
const STOP_BIT: u64 = 128u64;
match val {
0...STOP_1 => (val | STOP_BIT, 1),
START_2...STOP_2 => (
(val & MASK_1) | ((val & MASK_2) << 1) | (STOP_BIT << (8)),
2,
),
START_3...STOP_3 => (
(val & MASK_1) | ((val & MASK_2) << 1) | ((val & MASK_3) << 2) | (STOP_BIT << (8 * 2)),
3,
),
START_4...STOP_4 => (
(val & MASK_1)
| ((val & MASK_2) << 1)
| ((val & MASK_3) << 2)
| ((val & MASK_4) << 3)
| (STOP_BIT << (8 * 3)),
4,
),
_ => (
(val & MASK_1)
| ((val & MASK_2) << 1)
| ((val & MASK_3) << 2)
| ((val & MASK_4) << 3)
| ((val & MASK_5) << 4)
| (STOP_BIT << (8 * 4)),
5,
),
}
}
/// Returns the number of bytes covered by a
/// serialized vint `u32`.
///
/// Expects a buffer data that starts
/// by the serialized `vint`, scans at most 5 bytes ahead until
/// it finds the vint final byte.
///
/// # May Panic
/// If the payload does not start by a valid `vint`
fn vint_len(data: &[u8]) -> usize {
for (i, &val) in data.iter().enumerate().take(5) {
if val >= STOP_BIT {
return i + 1;
}
}
panic!("Corrupted data. Invalid VInt 32");
}
/// Reads a vint `u32` from a buffer, and
/// consumes its payload data.
///
/// # Panics
///
/// If the buffer does not start by a valid
/// vint payload
pub fn read_u32_vint(data: &mut &[u8]) -> u32 {
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;
}
*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 (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 {
pub fn val(&self) -> u64 {
self.0
@@ -15,24 +112,34 @@ impl VInt {
pub fn deserialize_u64<R: Read>(reader: &mut R) -> io::Result<u64> {
VInt::deserialize(reader).map(|vint| vint.0)
}
pub fn serialize_into_vec(&self, output: &mut Vec<u8>) {
let mut buffer = [0u8; 10];
let num_bytes = self.serialize_into(&mut buffer);
output.extend(&buffer[0..num_bytes]);
}
pub fn serialize_into(&self, buffer: &mut [u8; 10]) -> usize {
let mut remaining = self.0;
for (i, b) in buffer.iter_mut().enumerate() {
let next_byte: u8 = (remaining % 128u64) as u8;
remaining /= 128u64;
if remaining == 0u64 {
*b = next_byte | STOP_BIT;
return i + 1;
} else {
*b = next_byte;
}
}
unreachable!();
}
}
impl BinarySerializable for VInt {
fn serialize<W: Write>(&self, writer: &mut W) -> io::Result<()> {
let mut remaining = self.0;
let mut buffer = [0u8; 10];
let mut i = 0;
loop {
let next_byte: u8 = (remaining % 128u64) as u8;
remaining /= 128u64;
if remaining == 0u64 {
buffer[i] = next_byte | 128u8;
return writer.write_all(&buffer[0..i + 1]);
} else {
buffer[i] = next_byte;
}
i += 1;
}
let num_bytes = self.serialize_into(&mut buffer);
writer.write_all(&buffer[0..num_bytes])
}
fn deserialize<R: Read>(reader: &mut R) -> io::Result<Self> {
@@ -42,20 +149,84 @@ impl BinarySerializable for VInt {
loop {
match bytes.next() {
Some(Ok(b)) => {
result += u64::from(b % 128u8) << shift;
if b & 128u8 != 0u8 {
break;
result |= u64::from(b % 128u8) << shift;
if b >= STOP_BIT {
return Ok(VInt(result));
}
shift += 7;
}
_ => {
return Err(io::Error::new(
io::ErrorKind::InvalidData,
"Reach end of buffer",
))
"Reach end of buffer while reading VInt",
));
}
}
}
Ok(VInt(result))
}
}
#[cfg(test)]
mod tests {
use super::serialize_vint_u32;
use super::VInt;
use byteorder::{ByteOrder, LittleEndian};
use common::BinarySerializable;
fn aux_test_vint(val: u64) {
let mut v = [14u8; 10];
let num_bytes = VInt(val).serialize_into(&mut v);
for i in num_bytes..10 {
assert_eq!(v[i], 14u8);
}
assert!(num_bytes > 0);
if num_bytes < 10 {
assert!(1u64 << (7 * num_bytes) > val);
}
if num_bytes > 1 {
assert!(1u64 << (7 * (num_bytes - 1)) <= val);
}
let serdeser_val = VInt::deserialize(&mut &v[..]).unwrap();
assert_eq!(val, serdeser_val.0);
}
#[test]
fn test_vint() {
aux_test_vint(0);
aux_test_vint(1);
aux_test_vint(5);
aux_test_vint(u64::max_value());
for i in 1..9 {
let power_of_128 = 1u64 << (7 * i);
aux_test_vint(power_of_128 - 1u64);
aux_test_vint(power_of_128);
aux_test_vint(power_of_128 + 1u64);
}
aux_test_vint(10);
}
fn aux_test_serialize_vint_u32(val: u32) {
let mut buffer = [0u8; 10];
let mut buffer2 = [0u8; 10];
let len_vint = VInt(val as u64).serialize_into(&mut buffer);
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]
fn test_vint_u32() {
aux_test_serialize_vint_u32(0);
aux_test_serialize_vint_u32(1);
aux_test_serialize_vint_u32(5);
for i in 1..3 {
let power_of_128 = 1u32 << (7 * i);
aux_test_serialize_vint_u32(power_of_128 - 1u32);
aux_test_serialize_vint_u32(power_of_128);
aux_test_serialize_vint_u32(power_of_128 + 1u32);
}
aux_test_serialize_vint_u32(u32::max_value());
}
}

158
src/compression/stream.rs
View File

@@ -1,158 +0,0 @@
use compression::BlockDecoder;
use compression::COMPRESSION_BLOCK_SIZE;
use compression::compressed_block_size;
use directory::{ReadOnlySource, SourceRead};
/// Reads a stream of compressed ints.
///
/// Tantivy uses `CompressedIntStream` to read
/// the position file.
/// The `.skip(...)` makes it possible to avoid
/// decompressing blocks that are not required.
pub struct CompressedIntStream {
buffer: SourceRead,
block_decoder: BlockDecoder,
cached_addr: usize, // address of the currently decoded block
cached_next_addr: usize, // address following the currently decoded block
addr: usize, // address of the block associated to the current position
inner_offset: usize,
}
impl CompressedIntStream {
/// Opens a compressed int stream.
pub(crate) fn wrap(source: ReadOnlySource) -> CompressedIntStream {
CompressedIntStream {
buffer: SourceRead::from(source),
block_decoder: BlockDecoder::new(),
cached_addr: usize::max_value(),
cached_next_addr: usize::max_value(),
addr: 0,
inner_offset: 0,
}
}
/// Loads the block at the given address and return the address of the
/// following block
pub fn read_block(&mut self, addr: usize) -> usize {
if self.cached_addr == addr {
// we are already on this block.
// no need to read.
self.cached_next_addr
} else {
let next_addr = addr + self.block_decoder.uncompress_block_unsorted(self.buffer.slice_from(addr));
self.cached_addr = addr;
self.cached_next_addr = next_addr;
next_addr
}
}
/// 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 mut cursor = self.addr;
let mut inner_offset = self.inner_offset;
let mut num_els: usize = output.len();
let mut start = 0;
loop {
cursor = self.read_block(cursor);
let block = &self.block_decoder.output_array()[inner_offset..];
let block_len = block.len();
if num_els >= block_len {
output[start..start + block_len].clone_from_slice(&block);
start += block_len;
num_els -= block_len;
inner_offset = 0;
} else {
output[start..].clone_from_slice(&block[..num_els]);
break;
}
}
}
/// 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, mut skip_len: usize) {
loop {
let available = COMPRESSION_BLOCK_SIZE - self.inner_offset;
if available >= skip_len {
self.inner_offset += skip_len;
break;
} else {
skip_len -= available;
// entirely skip decompressing some blocks.
let num_bits: u8 = self.buffer.get(self.addr);
let block_len = compressed_block_size(num_bits);
self.addr += block_len;
self.inner_offset = 0;
}
}
}
}
#[cfg(test)]
pub mod tests {
use super::CompressedIntStream;
use compression::compressed_block_size;
use compression::COMPRESSION_BLOCK_SIZE;
use compression::BlockEncoder;
use directory::ReadOnlySource;
fn create_stream_buffer() -> ReadOnlySource {
let mut buffer: Vec<u8> = vec![];
let mut encoder = BlockEncoder::new();
let vals: Vec<u32> = (0u32..1152u32).collect();
for chunk in vals.chunks(COMPRESSION_BLOCK_SIZE) {
let compressed_block = encoder.compress_block_unsorted(chunk);
let num_bits = compressed_block[0];
assert_eq!(compressed_block_size(num_bits), compressed_block.len());
buffer.extend_from_slice(compressed_block);
}
if cfg!(simd) {
buffer.extend_from_slice(&[0u8; 7]);
}
ReadOnlySource::from(buffer)
}
#[test]
fn test_compressed_int_stream() {
let buffer = create_stream_buffer();
let mut stream = CompressedIntStream::wrap(buffer);
let mut block: [u32; COMPRESSION_BLOCK_SIZE] = [0u32; COMPRESSION_BLOCK_SIZE];
stream.read(&mut block[0..2]);
assert_eq!(block[0], 0);
assert_eq!(block[1], 1);
// reading does not consume the stream
stream.read(&mut block[0..2]);
assert_eq!(block[0], 0);
assert_eq!(block[1], 1);
stream.skip(2);
stream.skip(5);
stream.read(&mut block[0..3]);
stream.skip(3);
assert_eq!(block[0], 7);
assert_eq!(block[1], 8);
assert_eq!(block[2], 9);
stream.skip(500);
stream.read(&mut block[0..3]);
stream.skip(3);
assert_eq!(block[0], 510);
assert_eq!(block[1], 511);
assert_eq!(block[2], 512);
stream.skip(511);
stream.read(&mut block[..1]);
assert_eq!(block[0], 1024);
}
}

136
src/core/executor.rs Normal file
View File

@@ -0,0 +1,136 @@
use crossbeam::channel;
use scoped_pool::{Pool, ThreadConfig};
use Result;
/// Search executor whether search request are single thread or multithread.
///
/// We don't expose Rayon thread pool directly here for several reasons.
///
/// First dependency hell. It is not a good idea to expose the
/// API of a dependency, knowing it might conflict with a different version
/// used by the client. Second, we may stop using rayon in the future.
pub enum Executor {
SingleThread,
ThreadPool(Pool),
}
impl Executor {
/// Creates an Executor that performs all task in the caller thread.
pub fn single_thread() -> Executor {
Executor::SingleThread
}
// Creates an Executor that dispatches the tasks in a thread pool.
pub fn multi_thread(num_threads: usize, prefix: &'static str) -> Executor {
let thread_config = ThreadConfig::new().prefix(prefix);
let pool = Pool::with_thread_config(num_threads, thread_config);
Executor::ThreadPool(pool)
}
// Perform a map in the thread pool.
//
// Regardless of the executor (`SingleThread` or `ThreadPool`), panics in the task
// will propagate to the caller.
pub fn map<
A: Send,
R: Send,
AIterator: Iterator<Item = A>,
F: Sized + Sync + Fn(A) -> Result<R>,
>(
&self,
f: F,
args: AIterator,
) -> Result<Vec<R>> {
match self {
Executor::SingleThread => args.map(f).collect::<Result<_>>(),
Executor::ThreadPool(pool) => {
let args_with_indices: Vec<(usize, A)> = args.enumerate().collect();
let num_fruits = args_with_indices.len();
let fruit_receiver = {
let (fruit_sender, fruit_receiver) = channel::unbounded();
pool.scoped(|scope| {
for arg_with_idx in args_with_indices {
scope.execute(|| {
let (idx, arg) = arg_with_idx;
let fruit = f(arg);
if let Err(err) = fruit_sender.send((idx, fruit)) {
error!("Failed to send search task. It probably means all search threads have panicked. {:?}", err);
}
});
}
});
fruit_receiver
// This ends the scope of fruit_sender.
// This is important as it makes it possible for the fruit_receiver iteration to
// terminate.
};
// This is lame, but safe.
let mut results_with_position = Vec::with_capacity(num_fruits);
for (pos, fruit_res) in fruit_receiver {
let fruit = fruit_res?;
results_with_position.push((pos, fruit));
}
results_with_position.sort_by_key(|(pos, _)| *pos);
assert_eq!(results_with_position.len(), num_fruits);
Ok(results_with_position
.into_iter()
.map(|(_, fruit)| fruit)
.collect::<Vec<_>>())
}
}
}
}
#[cfg(test)]
mod tests {
use super::Executor;
#[test]
#[should_panic(expected = "panic should propagate")]
fn test_panic_propagates_single_thread() {
let _result: Vec<usize> = Executor::single_thread()
.map(
|_| {
panic!("panic should propagate");
},
vec![0].into_iter(),
)
.unwrap();
}
#[test]
#[should_panic] //< unfortunately the panic message is not propagated
fn test_panic_propagates_multi_thread() {
let _result: Vec<usize> = Executor::multi_thread(1, "search-test")
.map(
|_| {
panic!("panic should propagate");
},
vec![0].into_iter(),
)
.unwrap();
}
#[test]
fn test_map_singlethread() {
let result: Vec<usize> = Executor::single_thread()
.map(|i| Ok(i * 2), 0..1_000)
.unwrap();
assert_eq!(result.len(), 1_000);
for i in 0..1_000 {
assert_eq!(result[i], i * 2);
}
}
#[test]
fn test_map_multithread() {
let result: Vec<usize> = Executor::multi_thread(3, "search-test")
.map(|i| Ok(i * 2), 0..10)
.unwrap();
assert_eq!(result.len(), 10);
for i in 0..10 {
assert_eq!(result[i], i * 2);
}
}
}

501
src/core/index.rs
View File

@@ -1,80 +1,125 @@
use Result;
use error::{ErrorKind, ResultExt};
use serde_json;
use super::segment::create_segment;
use super::segment::Segment;
use core::Executor;
use core::IndexMeta;
use core::SegmentId;
use core::SegmentMeta;
use core::META_FILEPATH;
use directory::ManagedDirectory;
#[cfg(feature = "mmap")]
use directory::MmapDirectory;
use directory::INDEX_WRITER_LOCK;
use directory::{Directory, RAMDirectory};
use error::DataCorruption;
use error::TantivyError;
use indexer::index_writer::open_index_writer;
use indexer::index_writer::HEAP_SIZE_MIN;
use indexer::segment_updater::save_new_metas;
use num_cpus;
use reader::IndexReader;
use reader::IndexReaderBuilder;
use schema::Field;
use schema::FieldType;
use schema::Schema;
use std::sync::Arc;
use serde_json;
use std::borrow::BorrowMut;
use std::fmt;
use core::SegmentId;
#[cfg(feature="mmap")]
use directory::MmapDirectory;
use directory::{Directory, RAMDirectory};
use indexer::index_writer::open_index_writer;
use core::searcher::Searcher;
use std::convert::From;
use num_cpus;
use super::segment::Segment;
use core::SegmentReader;
use super::pool::Pool;
use core::SegmentMeta;
use super::pool::LeasedItem;
#[cfg(feature = "mmap")]
use std::path::Path;
use core::IndexMeta;
use indexer::DirectoryLock;
use IndexWriter;
use directory::ManagedDirectory;
use core::META_FILEPATH;
use super::segment::create_segment;
use indexer::segment_updater::save_new_metas;
use std::sync::Arc;
use tokenizer::BoxedTokenizer;
use tokenizer::TokenizerManager;
const NUM_SEARCHERS: usize = 12;
use IndexWriter;
use Result;
fn load_metas(directory: &Directory) -> Result<IndexMeta> {
let meta_data = directory.atomic_read(&META_FILEPATH)?;
let meta_string = String::from_utf8_lossy(&meta_data);
serde_json::from_str(&meta_string).chain_err(|| ErrorKind::CorruptedFile(META_FILEPATH.clone()))
serde_json::from_str(&meta_string)
.map_err(|e| {
DataCorruption::new(
META_FILEPATH.clone(),
format!("Meta file cannot be deserialized. {:?}.", e),
)
})
.map_err(From::from)
}
/// Search Index
#[derive(Clone)]
pub struct Index {
directory: ManagedDirectory,
schema: Schema,
searcher_pool: Arc<Pool<Searcher>>,
executor: Arc<Executor>,
tokenizers: TokenizerManager,
}
impl Index {
/// Examines the director to see if it contains an index
pub fn exists<Dir: Directory>(dir: &Dir) -> bool {
dir.exists(&META_FILEPATH)
}
/// Accessor to the search executor.
///
/// This pool is used by default when calling `searcher.search(...)`
/// to perform search on the individual segments.
///
/// By default the executor is single thread, and simply runs in the calling thread.
pub fn search_executor(&self) -> &Executor {
self.executor.as_ref()
}
/// Replace the default single thread search executor pool
/// by a thread pool with a given number of threads.
pub fn set_multithread_executor(&mut self, num_threads: usize) {
self.executor = Arc::new(Executor::multi_thread(num_threads, "thrd-tantivy-search-"));
}
/// Replace the default single thread search executor pool
/// by a thread pool with a given number of threads.
pub fn set_default_multithread_executor(&mut self) {
let default_num_threads = num_cpus::get();
self.set_multithread_executor(default_num_threads);
}
/// Creates a new index using the `RAMDirectory`.
///
/// The index will be allocated in anonymous memory.
/// This should only be used for unit tests.
pub fn create_in_ram(schema: Schema) -> Index {
let ram_directory = RAMDirectory::create();
// unwrap is ok here
let directory = ManagedDirectory::new(ram_directory).expect(
"Creating a managed directory from a brand new RAM directory \
should never fail.",
);
Index::from_directory(directory, schema).expect("Creating a RAMDirectory should never fail")
Index::create(ram_directory, schema).expect("Creating a RAMDirectory should never fail")
}
/// Creates a new index in a given filepath.
/// The index will use the `MMapDirectory`.
///
/// If a previous index was in this directory, then its meta file will be destroyed.
#[cfg(feature="mmap")]
pub fn create<P: AsRef<Path>>(directory_path: P, schema: Schema) -> Result<Index> {
#[cfg(feature = "mmap")]
pub fn create_in_dir<P: AsRef<Path>>(directory_path: P, schema: Schema) -> Result<Index> {
let mmap_directory = MmapDirectory::open(directory_path)?;
let directory = ManagedDirectory::new(mmap_directory)?;
Index::from_directory(directory, schema)
if Index::exists(&mmap_directory) {
return Err(TantivyError::IndexAlreadyExists);
}
Index::create(mmap_directory, schema)
}
/// Accessor for the tokenizer manager.
pub fn tokenizers(&self) -> &TokenizerManager {
&self.tokenizers
/// Opens or creates a new index in the provided directory
pub fn open_or_create<Dir: Directory>(dir: Dir, schema: Schema) -> Result<Index> {
if Index::exists(&dir) {
let index = Index::open(dir)?;
if index.schema() == schema {
Ok(index)
} else {
Err(TantivyError::SchemaError(
"An index exists but the schema does not match.".to_string(),
))
}
} else {
Index::create(dir, schema)
}
}
/// Creates a new index in a temp directory.
@@ -85,50 +130,95 @@ impl Index {
///
/// The temp directory is only used for testing the `MmapDirectory`.
/// For other unit tests, prefer the `RAMDirectory`, see: `create_in_ram`.
#[cfg(feature="mmap")]
#[cfg(test)]
#[cfg(feature = "mmap")]
pub fn create_from_tempdir(schema: Schema) -> Result<Index> {
let mmap_directory = MmapDirectory::create_from_tempdir()?;
let directory = ManagedDirectory::new(mmap_directory)?;
Index::create(mmap_directory, schema)
}
/// Creates a new index given an implementation of the trait `Directory`
pub fn create<Dir: Directory>(dir: Dir, schema: Schema) -> Result<Index> {
let directory = ManagedDirectory::wrap(dir)?;
Index::from_directory(directory, schema)
}
/// Create a new index from a directory.
///
/// This will overwrite existing meta.json
fn from_directory(mut directory: ManagedDirectory, schema: Schema) -> Result<Index> {
save_new_metas(schema.clone(), directory.borrow_mut())?;
let metas = IndexMeta::with_schema(schema);
Index::create_from_metas(directory, &metas)
}
/// Creates a new index given a directory and an `IndexMeta`.
fn create_from_metas(directory: ManagedDirectory, metas: &IndexMeta) -> Result<Index> {
let schema = metas.schema.clone();
let index = Index {
directory,
schema,
searcher_pool: Arc::new(Pool::new()),
tokenizers: TokenizerManager::default(),
executor: Arc::new(Executor::single_thread()),
};
index.load_searchers()?;
Ok(index)
}
/// Create a new index from a directory.
pub fn from_directory(mut directory: ManagedDirectory, schema: Schema) -> Result<Index> {
save_new_metas(schema.clone(), 0, directory.borrow_mut())?;
let metas = IndexMeta::with_schema(schema);
Index::create_from_metas(directory, &metas)
/// Accessor for the tokenizer manager.
pub fn tokenizers(&self) -> &TokenizerManager {
&self.tokenizers
}
/// Helper to access the tokenizer associated to a specific field.
pub fn tokenizer_for_field(&self, field: Field) -> Result<Box<BoxedTokenizer>> {
let field_entry = self.schema.get_field_entry(field);
let field_type = field_entry.field_type();
let tokenizer_manager: &TokenizerManager = self.tokenizers();
let tokenizer_name_opt: Option<Box<BoxedTokenizer>> = match field_type {
FieldType::Str(text_options) => text_options
.get_indexing_options()
.map(|text_indexing_options| text_indexing_options.tokenizer().to_string())
.and_then(|tokenizer_name| tokenizer_manager.get(&tokenizer_name)),
_ => None,
};
match tokenizer_name_opt {
Some(tokenizer) => Ok(tokenizer),
None => Err(TantivyError::SchemaError(format!(
"{:?} is not a text field.",
field_entry.name()
))),
}
}
/// Create a default `IndexReader` for the given index.
///
/// See [`Index.reader_builder()`](#method.reader_builder).
pub fn reader(&self) -> Result<IndexReader> {
self.reader_builder().try_into()
}
/// Create a `IndexReader` for the given index.
///
/// Most project should create at most one reader for a given index.
/// This method is typically called only once per `Index` instance,
/// over the lifetime of most problem.
pub fn reader_builder(&self) -> IndexReaderBuilder {
IndexReaderBuilder::new(self.clone())
}
/// Opens a new directory from an index path.
#[cfg(feature="mmap")]
pub fn open<P: AsRef<Path>>(directory_path: P) -> Result<Index> {
#[cfg(feature = "mmap")]
pub fn open_in_dir<P: AsRef<Path>>(directory_path: P) -> Result<Index> {
let mmap_directory = MmapDirectory::open(directory_path)?;
let directory = ManagedDirectory::new(mmap_directory)?;
Index::open(mmap_directory)
}
/// Open the index using the provided directory
pub fn open<D: Directory>(directory: D) -> Result<Index> {
let directory = ManagedDirectory::wrap(directory)?;
let metas = load_metas(&directory)?;
Index::create_from_metas(directory, &metas)
}
pub fn open_directory<TDirectory: Directory>(directory: TDirectory) -> Result<Index> {
let directory = ManagedDirectory::new(directory)?;
let metas = load_metas(&directory)?;
Index::create_from_metas(directory, &metas)
}
/// Reads the index meta file from the directory.
pub fn load_metas(&self) -> Result<IndexMeta> {
load_metas(self.directory())
@@ -142,31 +232,64 @@ impl Index {
/// `IndexWriter` on the system is accessing the index directory,
/// it is safe to manually delete the lockfile.
///
/// num_threads specifies the number of indexing workers that
/// - `num_threads` defines the number of indexing workers that
/// should work at the same time.
///
/// - `overall_heap_size_in_bytes` sets the amount of memory
/// allocated for all indexing thread.
/// Each thread will receive a budget of `overall_heap_size_in_bytes / num_threads`.
///
/// # Errors
/// If the lockfile already exists, returns `Error::FileAlreadyExists`.
/// If the lockfile already exists, returns `Error::DirectoryLockBusy` or an `Error::IOError`.
///
/// # Panics
/// If the heap size per thread is too small, panics.
pub fn writer_with_num_threads(
&self,
num_threads: usize,
heap_size_in_bytes: usize,
overall_heap_size_in_bytes: usize,
) -> Result<IndexWriter> {
let directory_lock = DirectoryLock::lock(self.directory().box_clone())?;
open_index_writer(self, num_threads, heap_size_in_bytes, directory_lock)
let directory_lock = self
.directory
.acquire_lock(&INDEX_WRITER_LOCK)
.map_err(|err| {
TantivyError::LockFailure(
err,
Some(
"Failed to acquire index lock. If you are using\
a regular directory, this means there is already an \
`IndexWriter` working on this `Directory`, in this process \
or in a different process."
.to_string(),
),
)
})?;
let heap_size_in_bytes_per_thread = overall_heap_size_in_bytes / num_threads;
open_index_writer(
self,
num_threads,
heap_size_in_bytes_per_thread,
directory_lock,
)
}
/// Creates a multithreaded writer
/// It just calls `writer_with_num_threads` with the number of cores as `num_threads`
///
/// 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, heap_size_in_bytes: usize) -> Result<IndexWriter> {
self.writer_with_num_threads(num_cpus::get(), heap_size_in_bytes)
pub fn writer(&self, overall_heap_size_in_bytes: usize) -> 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)
}
/// Accessor to the index schema
@@ -178,7 +301,8 @@ impl Index {
/// Returns the list of segments that are searchable
pub fn searchable_segments(&self) -> Result<Vec<Segment>> {
Ok(self.searchable_segment_metas()?
Ok(self
.searchable_segment_metas()?
.into_iter()
.map(|segment_meta| self.segment(segment_meta))
.collect())
@@ -191,8 +315,8 @@ impl Index {
/// Creates a new segment.
pub fn new_segment(&self) -> Segment {
let segment_meta = SegmentMeta::new(SegmentId::generate_random());
create_segment(self.clone(), segment_meta)
let segment_meta = SegmentMeta::new(SegmentId::generate_random(), 0);
self.segment(segment_meta)
}
/// Return a reference to the index directory.
@@ -213,44 +337,12 @@ impl Index {
/// Returns the list of segment ids that are searchable.
pub fn searchable_segment_ids(&self) -> Result<Vec<SegmentId>> {
Ok(self.searchable_segment_metas()?
Ok(self
.searchable_segment_metas()?
.iter()
.map(|segment_meta| segment_meta.id())
.map(SegmentMeta::id)
.collect())
}
/// Creates a new generation of searchers after
/// a change of the set of searchable indexes.
///
/// This needs to be called when a new segment has been
/// published or after a merge.
pub fn load_searchers(&self) -> Result<()> {
let searchable_segments = self.searchable_segments()?;
let segment_readers: Vec<SegmentReader> = searchable_segments
.iter()
.map(SegmentReader::open)
.collect::<Result<_>>()?;
let searchers = (0..NUM_SEARCHERS)
.map(|_| Searcher::from(segment_readers.clone()))
.collect();
self.searcher_pool.publish_new_generation(searchers);
Ok(())
}
/// Returns a searcher
///
/// This method should be called every single time a search
/// query is performed.
/// The searchers are taken from a pool of `NUM_SEARCHERS` searchers.
/// If no searcher is available
/// this may block.
///
/// The same searcher must be used for a given query, as it ensures
/// the use of a consistent segment set.
pub fn searcher(&self) -> LeasedItem<Searcher> {
self.searcher_pool.acquire()
}
}
impl fmt::Debug for Index {
@@ -259,13 +351,190 @@ impl fmt::Debug for Index {
}
}
impl Clone for Index {
fn clone(&self) -> Index {
Index {
directory: self.directory.clone(),
schema: self.schema.clone(),
searcher_pool: Arc::clone(&self.searcher_pool),
tokenizers: self.tokenizers.clone(),
#[cfg(test)]
mod tests {
use directory::RAMDirectory;
use schema::Field;
use schema::{Schema, INDEXED, TEXT};
use std::thread;
use std::time::Duration;
use Index;
use IndexReader;
use IndexWriter;
use ReloadPolicy;
#[test]
fn test_indexer_for_field() {
let mut schema_builder = Schema::builder();
let num_likes_field = schema_builder.add_u64_field("num_likes", INDEXED);
let body_field = schema_builder.add_text_field("body", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
assert!(index.tokenizer_for_field(body_field).is_ok());
assert_eq!(
format!("{:?}", index.tokenizer_for_field(num_likes_field).err()),
"Some(SchemaError(\"\\\"num_likes\\\" is not a text field.\"))"
);
}
#[test]
fn test_index_exists() {
let directory = RAMDirectory::create();
assert!(!Index::exists(&directory));
assert!(Index::create(directory.clone(), throw_away_schema()).is_ok());
assert!(Index::exists(&directory));
}
#[test]
fn open_or_create_should_create() {
let directory = RAMDirectory::create();
assert!(!Index::exists(&directory));
assert!(Index::open_or_create(directory.clone(), throw_away_schema()).is_ok());
assert!(Index::exists(&directory));
}
#[test]
fn open_or_create_should_open() {
let directory = RAMDirectory::create();
assert!(Index::create(directory.clone(), throw_away_schema()).is_ok());
assert!(Index::exists(&directory));
assert!(Index::open_or_create(directory, throw_away_schema()).is_ok());
}
#[test]
fn create_should_wipeoff_existing() {
let directory = RAMDirectory::create();
assert!(Index::create(directory.clone(), throw_away_schema()).is_ok());
assert!(Index::exists(&directory));
assert!(Index::create(directory.clone(), Schema::builder().build()).is_ok());
}
#[test]
fn open_or_create_exists_but_schema_does_not_match() {
let directory = RAMDirectory::create();
assert!(Index::create(directory.clone(), throw_away_schema()).is_ok());
assert!(Index::exists(&directory));
assert!(Index::open_or_create(directory.clone(), throw_away_schema()).is_ok());
let err = Index::open_or_create(directory, Schema::builder().build());
assert_eq!(
format!("{:?}", err.unwrap_err()),
"SchemaError(\"An index exists but the schema does not match.\")"
);
}
fn throw_away_schema() -> Schema {
let mut schema_builder = Schema::builder();
let _ = schema_builder.add_u64_field("num_likes", INDEXED);
schema_builder.build()
}
#[test]
fn test_index_on_commit_reload_policy() {
let schema = throw_away_schema();
let field = schema.get_field("num_likes").unwrap();
let index = Index::create_in_ram(schema);
let reader = index
.reader_builder()
.reload_policy(ReloadPolicy::OnCommit)
.try_into()
.unwrap();
assert_eq!(reader.searcher().num_docs(), 0);
let mut writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
test_index_on_commit_reload_policy_aux(field, &mut writer, &reader);
}
#[cfg(feature = "mmap")]
mod mmap_specific {
use super::*;
use std::path::PathBuf;
use tempdir::TempDir;
#[test]
fn test_index_on_commit_reload_policy_mmap() {
let schema = throw_away_schema();
let field = schema.get_field("num_likes").unwrap();
let tempdir = TempDir::new("index").unwrap();
let tempdir_path = PathBuf::from(tempdir.path());
let index = Index::create_in_dir(&tempdir_path, schema).unwrap();
let mut writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
writer.commit().unwrap();
let reader = index
.reader_builder()
.reload_policy(ReloadPolicy::OnCommit)
.try_into()
.unwrap();
assert_eq!(reader.searcher().num_docs(), 0);
test_index_on_commit_reload_policy_aux(field, &mut writer, &reader);
}
#[test]
fn test_index_manual_policy_mmap() {
let schema = throw_away_schema();
let field = schema.get_field("num_likes").unwrap();
let index = Index::create_from_tempdir(schema).unwrap();
let mut writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
writer.commit().unwrap();
let reader = index
.reader_builder()
.reload_policy(ReloadPolicy::Manual)
.try_into()
.unwrap();
assert_eq!(reader.searcher().num_docs(), 0);
writer.add_document(doc!(field=>1u64));
writer.commit().unwrap();
thread::sleep(Duration::from_millis(500));
assert_eq!(reader.searcher().num_docs(), 0);
reader.reload().unwrap();
assert_eq!(reader.searcher().num_docs(), 1);
}
#[test]
fn test_index_on_commit_reload_policy_different_directories() {
let schema = throw_away_schema();
let field = schema.get_field("num_likes").unwrap();
let tempdir = TempDir::new("index").unwrap();
let tempdir_path = PathBuf::from(tempdir.path());
let write_index = Index::create_in_dir(&tempdir_path, schema).unwrap();
let read_index = Index::open_in_dir(&tempdir_path).unwrap();
let reader = read_index
.reader_builder()
.reload_policy(ReloadPolicy::OnCommit)
.try_into()
.unwrap();
assert_eq!(reader.searcher().num_docs(), 0);
let mut writer = write_index.writer_with_num_threads(1, 3_000_000).unwrap();
test_index_on_commit_reload_policy_aux(field, &mut writer, &reader);
}
}
fn test_index_on_commit_reload_policy_aux(
field: Field,
writer: &mut IndexWriter,
reader: &IndexReader,
) {
assert_eq!(reader.searcher().num_docs(), 0);
writer.add_document(doc!(field=>1u64));
writer.commit().unwrap();
let mut count = 0;
for _ in 0..100 {
count = reader.searcher().num_docs();
if count > 0 {
break;
}
thread::sleep(Duration::from_millis(100));
}
assert_eq!(count, 1);
writer.add_document(doc!(field=>2u64));
writer.commit().unwrap();
let mut count = 0;
for _ in 0..10 {
count = reader.searcher().num_docs();
if count > 1 {
break;
}
thread::sleep(Duration::from_millis(100));
}
assert_eq!(count, 2);
}
}

15
src/core/index_meta.rs
View File

@@ -1,7 +1,8 @@
use schema::Schema;
use core::SegmentMeta;
use std::fmt;
use schema::Schema;
use serde_json;
use std::fmt;
use Opstamp;
/// Meta information about the `Index`.
///
@@ -15,7 +16,7 @@ use serde_json;
pub struct IndexMeta {
pub segments: Vec<SegmentMeta>,
pub schema: Schema,
pub opstamp: u64,
pub opstamp: Opstamp,
#[serde(skip_serializing_if = "Option::is_none")]
pub payload: Option<String>,
}
@@ -45,20 +46,20 @@ impl fmt::Debug for IndexMeta {
#[cfg(test)]
mod tests {
use serde_json;
use super::IndexMeta;
use schema::{SchemaBuilder, TEXT};
use schema::{Schema, TEXT};
use serde_json;
#[test]
fn test_serialize_metas() {
let schema = {
let mut schema_builder = SchemaBuilder::new();
let mut schema_builder = Schema::builder();
schema_builder.add_text_field("text", TEXT);
schema_builder.build()
};
let index_metas = IndexMeta {
segments: Vec::new(),
schema: schema,
schema,
opstamp: 0u64,
payload: None,
};

94
src/core/inverted_index_reader.rs
View File

@@ -1,13 +1,13 @@
use directory::{ReadOnlySource, SourceRead};
use termdict::{TermDictionary, TermDictionaryImpl};
use postings::{BlockSegmentPostings, SegmentPostings};
use common::BinarySerializable;
use directory::ReadOnlySource;
use owned_read::OwnedRead;
use positions::PositionReader;
use postings::TermInfo;
use postings::{BlockSegmentPostings, SegmentPostings};
use schema::FieldType;
use schema::IndexRecordOption;
use schema::Term;
use compression::CompressedIntStream;
use postings::FreqReadingOption;
use common::BinarySerializable;
use schema::FieldType;
use termdict::TermDictionary;
/// The inverted index reader is in charge of accessing
/// the inverted index associated to a specific field.
@@ -23,18 +23,21 @@ use schema::FieldType;
/// `InvertedIndexReader` are created by calling
/// the `SegmentReader`'s [`.inverted_index(...)`] method
pub struct InvertedIndexReader {
termdict: TermDictionaryImpl,
termdict: TermDictionary,
postings_source: ReadOnlySource,
positions_source: ReadOnlySource,
positions_idx_source: ReadOnlySource,
record_option: IndexRecordOption,
total_num_tokens: u64
total_num_tokens: u64,
}
impl InvertedIndexReader {
#[cfg_attr(feature = "cargo-clippy", allow(clippy::needless_pass_by_value))] // for symetry
pub(crate) fn new(
termdict: TermDictionaryImpl,
termdict: TermDictionary,
postings_source: ReadOnlySource,
positions_source: ReadOnlySource,
positions_idx_source: ReadOnlySource,
record_option: IndexRecordOption,
) -> InvertedIndexReader {
let total_num_tokens_data = postings_source.slice(0, 8);
@@ -44,23 +47,25 @@ impl InvertedIndexReader {
termdict,
postings_source: postings_source.slice_from(8),
positions_source,
positions_idx_source,
record_option,
total_num_tokens
total_num_tokens,
}
}
/// Creates an empty `InvertedIndexReader` object, which
/// contains no terms at all.
pub fn empty(field_type: FieldType) -> InvertedIndexReader {
pub fn empty(field_type: &FieldType) -> InvertedIndexReader {
let record_option = field_type
.get_index_record_option()
.unwrap_or(IndexRecordOption::Basic);
InvertedIndexReader {
termdict: TermDictionaryImpl::empty(field_type),
termdict: TermDictionary::empty(&field_type),
postings_source: ReadOnlySource::empty(),
positions_source: ReadOnlySource::empty(),
positions_idx_source: ReadOnlySource::empty(),
record_option,
total_num_tokens: 0u64
total_num_tokens: 0u64,
}
}
@@ -70,7 +75,7 @@ impl InvertedIndexReader {
}
/// Return the term dictionary datastructure.
pub fn terms(&self) -> &TermDictionaryImpl {
pub fn terms(&self) -> &TermDictionary {
&self.termdict
}
@@ -92,8 +97,21 @@ impl InvertedIndexReader {
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 = SourceRead::from(postings_slice);
block_postings.reset(term_info.doc_freq as usize, postings_reader);
let postings_reader = OwnedRead::new(postings_slice);
block_postings.reset(term_info.doc_freq, postings_reader);
}
/// Returns a block postings given a `Term`.
/// This method is for an advanced usage only.
///
/// Most user should prefer using `read_postings` instead.
pub fn read_block_postings(
&self,
term: &Term,
option: IndexRecordOption,
) -> Option<BlockSegmentPostings> {
self.get_term_info(term)
.map(move |term_info| self.read_block_postings_from_terminfo(&term_info, option))
}
/// Returns a block postings given a `term_info`.
@@ -107,15 +125,11 @@ impl InvertedIndexReader {
) -> BlockSegmentPostings {
let offset = term_info.postings_offset as usize;
let postings_data = self.postings_source.slice_from(offset);
let freq_reading_option = match (self.record_option, requested_option) {
(IndexRecordOption::Basic, _) => FreqReadingOption::NoFreq,
(_, IndexRecordOption::Basic) => FreqReadingOption::SkipFreq,
(_, _) => FreqReadingOption::ReadFreq,
};
BlockSegmentPostings::from_data(
term_info.doc_freq as usize,
SourceRead::from(postings_data),
freq_reading_option,
term_info.doc_freq,
OwnedRead::new(postings_data),
self.record_option,
requested_option,
)
}
@@ -131,11 +145,11 @@ impl InvertedIndexReader {
let block_postings = self.read_block_postings_from_terminfo(term_info, option);
let position_stream = {
if option.has_positions() {
let position_offset = term_info.positions_offset;
let positions_source = self.positions_source.slice_from(position_offset as usize);
let mut stream = CompressedIntStream::wrap(positions_source);
stream.skip(term_info.positions_inner_offset as usize);
Some(stream)
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);
Some(position_reader)
} else {
None
}
@@ -149,8 +163,6 @@ impl InvertedIndexReader {
self.total_num_tokens
}
/// Returns the segment postings associated with the term, and with the given option,
/// or `None` if the term has never been encountered and indexed.
///
@@ -162,16 +174,19 @@ impl InvertedIndexReader {
/// `TextIndexingOptions` that does not index position will return a `SegmentPostings`
/// with `DocId`s and frequencies.
pub fn read_postings(&self, term: &Term, option: IndexRecordOption) -> Option<SegmentPostings> {
let term_info = get!(self.get_term_info(term));
Some(self.read_postings_from_terminfo(&term_info, option))
self.get_term_info(term)
.map(move |term_info| self.read_postings_from_terminfo(&term_info, option))
}
pub(crate) fn read_postings_no_deletes(&self, term: &Term, option: IndexRecordOption) -> Option<SegmentPostings> {
let term_info = get!(self.get_term_info(term));
Some(self.read_postings_from_terminfo(&term_info, option))
pub(crate) fn read_postings_no_deletes(
&self,
term: &Term,
option: IndexRecordOption,
) -> Option<SegmentPostings> {
self.get_term_info(term)
.map(|term_info| self.read_postings_from_terminfo(&term_info, option))
}
/// Returns the number of documents containing the term.
pub fn doc_freq(&self, term: &Term) -> u32 {
self.get_term_info(term)
@@ -179,6 +194,3 @@ impl InvertedIndexReader {
.unwrap_or(0u32)
}
}

32
src/core/mod.rs
View File

@@ -1,28 +1,30 @@
pub mod searcher;
mod executor;
pub mod index;
mod segment_reader;
mod segment_id;
mod segment_component;
mod segment;
mod index_meta;
mod pool;
mod segment_meta;
mod inverted_index_reader;
pub mod searcher;
mod segment;
mod segment_component;
mod segment_id;
mod segment_meta;
mod segment_reader;
pub use self::executor::Executor;
pub use self::index::Index;
pub use self::index_meta::IndexMeta;
pub use self::inverted_index_reader::InvertedIndexReader;
pub use self::searcher::Searcher;
pub use self::segment_component::SegmentComponent;
pub use self::segment_id::SegmentId;
pub use self::segment_reader::SegmentReader;
pub use self::segment::Segment;
pub use self::segment::SerializableSegment;
pub use self::index::Index;
pub use self::segment_component::SegmentComponent;
pub use self::segment_id::SegmentId;
pub use self::segment_meta::SegmentMeta;
pub use self::index_meta::IndexMeta;
pub use self::segment_reader::SegmentReader;
use std::path::PathBuf;
lazy_static! {
/// The meta file contains all the information about the list of segments and the schema
/// of the index.
pub static ref META_FILEPATH: PathBuf = PathBuf::from("meta.json");
@@ -33,10 +35,4 @@ lazy_static! {
/// Removing this file is safe, but will prevent the garbage collection of all of the file that
/// are currently in the directory
pub static ref MANAGED_FILEPATH: PathBuf = PathBuf::from(".managed.json");
/// Only one process should be able to write tantivy's index at a time.
/// This file, when present, is in charge of preventing other processes to open an IndexWriter.
///
/// If the process is killed and this file remains, it is safe to remove it manually.
pub static ref LOCKFILE_FILEPATH: PathBuf = PathBuf::from(".tantivy-indexer.lock");
}

134
src/core/pool.rs
View File

@@ -1,134 +0,0 @@
use std::sync::atomic::AtomicUsize;
use std::sync::atomic::Ordering;
use std::mem;
use std::ops::{Deref, DerefMut};
use crossbeam::sync::MsQueue;
use std::sync::Arc;
pub struct GenerationItem<T> {
generation: usize,
item: T,
}
pub struct Pool<T> {
queue: Arc<MsQueue<GenerationItem<T>>>,
freshest_generation: AtomicUsize,
next_generation: AtomicUsize,
}
impl<T> Pool<T> {
pub fn new() -> Pool<T> {
let queue = Arc::new(MsQueue::new());
Pool {
queue,
freshest_generation: AtomicUsize::default(),
next_generation: AtomicUsize::default(),
}
}
pub fn publish_new_generation(&self, items: Vec<T>) {
let next_generation = self.next_generation.fetch_add(1, Ordering::SeqCst) + 1;
for item in items {
let gen_item = GenerationItem {
item,
generation: next_generation,
};
self.queue.push(gen_item);
}
self.advertise_generation(next_generation);
}
/// At the exit of this method,
/// - freshest_generation has a value greater or equal than generation
/// - freshest_generation has a value that has been advertised
/// - freshest_generation has)
fn advertise_generation(&self, generation: usize) {
// not optimal at all but the easiest to read proof.
loop {
let former_generation = self.freshest_generation.load(Ordering::Acquire);
if former_generation >= generation {
break;
}
self.freshest_generation.compare_and_swap(
former_generation,
generation,
Ordering::SeqCst,
);
}
}
fn generation(&self) -> usize {
self.freshest_generation.load(Ordering::Acquire)
}
pub fn acquire(&self) -> LeasedItem<T> {
let generation = self.generation();
loop {
let gen_item = self.queue.pop();
if gen_item.generation >= generation {
return LeasedItem {
gen_item: Some(gen_item),
recycle_queue: Arc::clone(&self.queue),
};
} else {
// this searcher is obsolete,
// removing it from the pool.
}
}
}
}
pub struct LeasedItem<T> {
gen_item: Option<GenerationItem<T>>,
recycle_queue: Arc<MsQueue<GenerationItem<T>>>,
}
impl<T> Deref for LeasedItem<T> {
type Target = T;
fn deref(&self) -> &T {
&self.gen_item
.as_ref()
.expect("Unwrapping a leased item should never fail")
.item // unwrap is safe here
}
}
impl<T> DerefMut for LeasedItem<T> {
fn deref_mut(&mut self) -> &mut T {
&mut self.gen_item
.as_mut()
.expect("Unwrapping a mut leased item should never fail")
.item // unwrap is safe here
}
}
impl<T> Drop for LeasedItem<T> {
fn drop(&mut self) {
let gen_item: GenerationItem<T> = mem::replace(&mut self.gen_item, None)
.expect("Unwrapping a leased item should never fail");
self.recycle_queue.push(gen_item);
}
}
#[cfg(test)]
mod tests {
use std::iter;
use super::Pool;
#[test]
fn test_pool() {
let items10: Vec<usize> = iter::repeat(10).take(10).collect();
let pool = Pool::new();
pool.publish_new_generation(items10);
for _ in 0..20 {
assert_eq!(*pool.acquire(), 10);
}
let items11: Vec<usize> = iter::repeat(11).take(10).collect();
pool.publish_new_generation(items11);
for _ in 0..20 {
assert_eq!(*pool.acquire(), 11);
}
}
}

172
src/core/searcher.rs
View File

@@ -1,14 +1,42 @@
use Result;
use core::SegmentReader;
use schema::Document;
use collector::Collector;
use query::Query;
use DocAddress;
use schema::{Field, Term};
use termdict::{TermDictionary, TermMerger};
use std::sync::Arc;
use std::fmt;
use collector::SegmentCollector;
use core::Executor;
use core::InvertedIndexReader;
use core::SegmentReader;
use query::Query;
use query::Scorer;
use query::Weight;
use schema::Document;
use schema::Schema;
use schema::{Field, Term};
use space_usage::SearcherSpaceUsage;
use std::fmt;
use std::sync::Arc;
use store::StoreReader;
use termdict::TermMerger;
use DocAddress;
use Index;
use Result;
fn collect_segment<C: Collector>(
collector: &C,
weight: &Weight,
segment_ord: u32,
segment_reader: &SegmentReader,
) -> Result<C::Fruit> {
let mut scorer = weight.scorer(segment_reader)?;
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_deleted(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.
///
@@ -16,25 +44,56 @@ use core::InvertedIndexReader;
/// the destruction of the `Searcher`.
///
pub struct Searcher {
schema: Schema,
index: Index,
segment_readers: Vec<SegmentReader>,
store_readers: Vec<StoreReader>,
}
impl Searcher {
/// Creates a new `Searcher`
pub(crate) fn new(
schema: Schema,
index: Index,
segment_readers: Vec<SegmentReader>,
) -> Searcher {
let store_readers = segment_readers
.iter()
.map(SegmentReader::get_store_reader)
.collect();
Searcher {
schema,
index,
segment_readers,
store_readers,
}
}
/// Returns the `Index` associated to the `Searcher`
pub fn index(&self) -> &Index {
&self.index
}
/// Fetches a document from tantivy's store given a `DocAddress`.
///
/// The searcher uses the segment ordinal to route the
/// the request to the right `Segment`.
pub fn doc(&self, doc_address: &DocAddress) -> Result<Document> {
let DocAddress(segment_local_id, doc_id) = *doc_address;
let segment_reader = &self.segment_readers[segment_local_id as usize];
segment_reader.doc(doc_id)
pub fn doc(&self, doc_address: DocAddress) -> Result<Document> {
let DocAddress(segment_local_id, doc_id) = doc_address;
let store_reader = &self.store_readers[segment_local_id as usize];
store_reader.get(doc_id)
}
/// Access the schema associated to the index of this searcher.
pub fn schema(&self) -> &Schema {
&self.schema
}
/// Returns the overall number of documents in the index.
pub fn num_docs(&self) -> u64 {
self.segment_readers
.iter()
.map(|segment_reader| segment_reader.num_docs() as u64)
.map(|segment_reader| u64::from(segment_reader.num_docs()))
.sum::<u64>()
}
@@ -43,7 +102,9 @@ impl Searcher {
pub fn doc_freq(&self, term: &Term) -> u64 {
self.segment_readers
.iter()
.map(|segment_reader| segment_reader.inverted_index(term.field()).doc_freq(term) as u64)
.map(|segment_reader| {
u64::from(segment_reader.inverted_index(term.field()).doc_freq(term))
})
.sum::<u64>()
}
@@ -57,19 +118,78 @@ impl Searcher {
&self.segment_readers[segment_ord as usize]
}
/// Runs a query on the segment readers wrapped by the searcher
pub fn search<C: Collector>(&self, query: &Query, collector: &mut C) -> Result<()> {
query.search(self, collector)
/// Runs a query on the segment readers wrapped by the searcher.
///
/// Search works as follows :
///
/// First the weight object associated to the query is created.
///
/// Then, the query loops over the segments and for each segment :
/// - setup the collector and informs it that the segment being processed has changed.
/// - creates a SegmentCollector for collecting documents associated to the segment
/// - creates a `Scorer` object associated for this segment
/// - iterate through the matched documents and push them to the segment collector.
///
/// Finally, the Collector merges each of the child collectors into itself for result usability
/// by the caller.
pub fn search<C: Collector>(&self, query: &Query, collector: &C) -> Result<C::Fruit> {
let executor = self.index.search_executor();
self.search_with_executor(query, collector, executor)
}
/// Same as [`search(...)`](#method.search) but multithreaded.
///
/// The current implementation is rather naive :
/// multithreading is by splitting search into as many task
/// as there are segments.
///
/// It is powerless at making search faster if your index consists in
/// one large segment.
///
/// Also, keep in my multithreading a single query on several
/// threads will not improve your throughput. It can actually
/// hurt it. It will however, decrease the average response time.
pub fn search_with_executor<C: Collector>(
&self,
query: &Query,
collector: &C,
executor: &Executor,
) -> Result<C::Fruit> {
let scoring_enabled = collector.requires_scoring();
let weight = query.weight(self, scoring_enabled)?;
let segment_readers = self.segment_readers();
let fruits = executor.map(
|(segment_ord, segment_reader)| {
collect_segment(
collector,
weight.as_ref(),
segment_ord as u32,
segment_reader,
)
},
segment_readers.iter().enumerate(),
)?;
collector.merge_fruits(fruits)
}
/// Return the field searcher associated to a `Field`.
pub fn field(&self, field: Field) -> FieldSearcher {
let inv_index_readers = self.segment_readers
let inv_index_readers = self
.segment_readers
.iter()
.map(|segment_reader| segment_reader.inverted_index(field))
.collect::<Vec<_>>();
FieldSearcher::new(inv_index_readers)
}
/// Summarize total space usage of this searcher.
pub fn space_usage(&self) -> SearcherSpaceUsage {
let mut space_usage = SearcherSpaceUsage::new();
for segment_reader in self.segment_readers.iter() {
space_usage.add_segment(segment_reader.space_usage());
}
space_usage
}
}
pub struct FieldSearcher {
@@ -84,7 +204,8 @@ impl FieldSearcher {
/// Returns a Stream over all of the sorted unique terms of
/// for the given field.
pub fn terms(&self) -> TermMerger {
let term_streamers: Vec<_> = self.inv_index_readers
let term_streamers: Vec<_> = self
.inv_index_readers
.iter()
.map(|inverted_index| inverted_index.terms().stream())
.collect();
@@ -92,17 +213,12 @@ impl FieldSearcher {
}
}
impl From<Vec<SegmentReader>> for Searcher {
fn from(segment_readers: Vec<SegmentReader>) -> Searcher {
Searcher { segment_readers }
}
}
impl fmt::Debug for Searcher {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
let segment_ids = self.segment_readers
let segment_ids = self
.segment_readers
.iter()
.map(|segment_reader| segment_reader.segment_id())
.map(SegmentReader::segment_id)
.collect::<Vec<_>>();
write!(f, "Searcher({:?})", segment_ids)
}

69
src/core/segment.rs
View File

@@ -1,16 +1,17 @@
use Result;
use std::path::PathBuf;
use schema::Schema;
use std::fmt;
use core::SegmentId;
use directory::{FileProtection, ReadOnlySource, WritePtr};
use indexer::segment_serializer::SegmentSerializer;
use super::SegmentComponent;
use core::Index;
use std::result;
use directory::Directory;
use core::SegmentId;
use core::SegmentMeta;
use directory::error::{OpenReadError, OpenWriteError};
use directory::Directory;
use directory::{ReadOnlySource, WritePtr};
use indexer::segment_serializer::SegmentSerializer;
use schema::Schema;
use std::fmt;
use std::path::PathBuf;
use std::result;
use Opstamp;
use Result;
/// A segment is a piece of the index.
#[derive(Clone)]
@@ -28,6 +29,7 @@ impl fmt::Debug for Segment {
/// Creates a new segment given an `Index` and a `SegmentId`
///
/// The function is here to make it private outside `tantivy`.
/// #[doc(hidden)]
pub fn create_segment(index: Index, meta: SegmentMeta) -> Segment {
Segment { index, meta }
}
@@ -49,8 +51,11 @@ impl Segment {
}
#[doc(hidden)]
pub fn set_delete_meta(&mut self, num_deleted_docs: u32, opstamp: u64) {
self.meta.set_delete_meta(num_deleted_docs, opstamp);
pub fn with_delete_meta(self, num_deleted_docs: u32, opstamp: Opstamp) -> Segment {
Segment {
index: self.index,
meta: self.meta.with_delete_meta(num_deleted_docs, opstamp),
}
}
/// Returns the segment's id.
@@ -66,16 +71,6 @@ impl Segment {
self.meta.relative_path(component)
}
/// Protects a specific component file from being deleted.
///
/// Returns a FileProtection object. The file is guaranteed
/// to not be garbage collected as long as this `FileProtection` object
/// lives.
pub fn protect_from_delete(&self, component: SegmentComponent) -> FileProtection {
let path = self.relative_path(component);
self.index.directory().protect_file_from_delete(&path)
}
/// Open one of the component file for a *regular* read.
pub fn open_read(
&self,
@@ -105,35 +100,3 @@ pub trait SerializableSegment {
/// The number of documents in the segment.
fn write(&self, serializer: SegmentSerializer) -> Result<u32>;
}
#[cfg(test)]
mod tests {
use core::SegmentComponent;
use directory::Directory;
use std::collections::HashSet;
use schema::SchemaBuilder;
use Index;
#[test]
fn test_segment_protect_component() {
let mut index = Index::create_in_ram(SchemaBuilder::new().build());
let segment = index.new_segment();
let path = segment.relative_path(SegmentComponent::POSTINGS);
let directory = index.directory_mut();
directory.atomic_write(&*path, &vec![0u8]).unwrap();
let living_files = HashSet::new();
{
let _file_protection = segment.protect_from_delete(SegmentComponent::POSTINGS);
assert!(directory.exists(&*path));
directory.garbage_collect(|| living_files.clone());
assert!(directory.exists(&*path));
}
directory.garbage_collect(|| living_files);
assert!(!directory.exists(&*path));
}
}

11
src/core/segment_component.rs
View File

@@ -1,3 +1,5 @@
use std::slice;
/// Enum describing each component of a tantivy segment.
/// Each component is stored in its own file,
/// using the pattern `segment_uuid`.`component_extension`,
@@ -8,6 +10,8 @@ pub enum SegmentComponent {
POSTINGS,
/// Positions of terms in each document.
POSITIONS,
/// Index to seek within the position file
POSITIONSSKIP,
/// Column-oriented random-access storage of fields.
FASTFIELDS,
/// Stores the sum of the length (in terms) of each field for each document.
@@ -26,16 +30,17 @@ pub enum SegmentComponent {
impl SegmentComponent {
/// Iterates through the components.
pub fn iterator() -> impl Iterator<Item = &'static SegmentComponent> {
static SEGMENT_COMPONENTS: [SegmentComponent; 7] = [
pub fn iterator() -> slice::Iter<'static, SegmentComponent> {
static SEGMENT_COMPONENTS: [SegmentComponent; 8] = [
SegmentComponent::POSTINGS,
SegmentComponent::POSITIONS,
SegmentComponent::POSITIONSSKIP,
SegmentComponent::FASTFIELDS,
SegmentComponent::FIELDNORMS,
SegmentComponent::TERMS,
SegmentComponent::STORE,
SegmentComponent::DELETE,
];
SEGMENT_COMPONENTS.into_iter()
SEGMENT_COMPONENTS.iter()
}
}

12
src/core/segment_id.rs
View File

@@ -1,6 +1,6 @@
use uuid::Uuid;
use std::fmt;
use std::cmp::{Ord, Ordering};
use std::fmt;
use uuid::Uuid;
#[cfg(test)]
use std::sync::atomic;
@@ -19,7 +19,7 @@ pub struct SegmentId(Uuid);
#[cfg(test)]
lazy_static! {
static ref AUTO_INC_COUNTER: atomic::AtomicUsize = atomic::AtomicUsize::default();
static ref EMPTY_ARR: [u8; 8] = [0u8; 8];
static ref ZERO_ARRAY: [u8; 8] = [0u8; 8];
}
// During tests, we generate the segment id in a autoincrement manner
@@ -30,7 +30,7 @@ lazy_static! {
#[cfg(test)]
fn create_uuid() -> Uuid {
let new_auto_inc_id = (*AUTO_INC_COUNTER).fetch_add(1, atomic::Ordering::SeqCst);
Uuid::from_fields(new_auto_inc_id as u32, 0, 0, &*EMPTY_ARR).unwrap()
Uuid::from_fields(new_auto_inc_id as u32, 0, 0, &*ZERO_ARRAY).unwrap()
}
#[cfg(not(test))]
@@ -52,12 +52,12 @@ impl SegmentId {
/// Picking the first 8 chars is ok to identify
/// segments in a display message.
pub fn short_uuid_string(&self) -> String {
(&self.0.simple().to_string()[..8]).to_string()
(&self.0.to_simple_ref().to_string()[..8]).to_string()
}
/// Returns a segment uuid string.
pub fn uuid_string(&self) -> String {
self.0.simple().to_string()
self.0.to_simple_ref().to_string()
}
}

114
src/core/segment_meta.rs
View File

@@ -1,44 +1,92 @@
use core::SegmentId;
use super::SegmentComponent;
use std::path::PathBuf;
use census::{Inventory, TrackedObject};
use core::SegmentId;
use serde;
use std::collections::HashSet;
use std::fmt;
use std::path::PathBuf;
use Opstamp;
lazy_static! {
static ref INVENTORY: Inventory<InnerSegmentMeta> = { Inventory::new() };
}
#[derive(Clone, Debug, Serialize, Deserialize)]
struct DeleteMeta {
num_deleted_docs: u32,
opstamp: u64,
opstamp: Opstamp,
}
/// `SegmentMeta` contains simple meta information about a segment.
///
/// For instance the number of docs it contains,
/// how many are deleted, etc.
#[derive(Clone, Debug, Serialize, Deserialize)]
#[derive(Clone)]
pub struct SegmentMeta {
segment_id: SegmentId,
max_doc: u32,
deletes: Option<DeleteMeta>,
tracked: TrackedObject<InnerSegmentMeta>,
}
impl fmt::Debug for SegmentMeta {
fn fmt(&self, f: &mut fmt::Formatter) -> Result<(), fmt::Error> {
self.tracked.fmt(f)
}
}
impl serde::Serialize for SegmentMeta {
fn serialize<S>(
&self,
serializer: S,
) -> Result<<S as serde::Serializer>::Ok, <S as serde::Serializer>::Error>
where
S: serde::Serializer,
{
self.tracked.serialize(serializer)
}
}
impl<'a> serde::Deserialize<'a> for SegmentMeta {
fn deserialize<D>(deserializer: D) -> Result<Self, <D as serde::Deserializer<'a>>::Error>
where
D: serde::Deserializer<'a>,
{
let inner = InnerSegmentMeta::deserialize(deserializer)?;
let tracked = INVENTORY.track(inner);
Ok(SegmentMeta { tracked })
}
}
impl SegmentMeta {
/// Creates a new segment meta for
/// a segment with no deletes and no documents.
pub fn new(segment_id: SegmentId) -> SegmentMeta {
SegmentMeta {
/// Lists all living `SegmentMeta` object at the time of the call.
pub fn all() -> Vec<SegmentMeta> {
INVENTORY
.list()
.into_iter()
.map(|inner| SegmentMeta { tracked: inner })
.collect::<Vec<_>>()
}
/// Creates a new `SegmentMeta` object.
#[doc(hidden)]
pub fn new(segment_id: SegmentId, max_doc: u32) -> SegmentMeta {
let inner = InnerSegmentMeta {
segment_id,
max_doc: 0,
max_doc,
deletes: None,
};
SegmentMeta {
tracked: INVENTORY.track(inner),
}
}
/// Returns the segment id.
pub fn id(&self) -> SegmentId {
self.segment_id
self.tracked.segment_id
}
/// Returns the number of deleted documents.
pub fn num_deleted_docs(&self) -> u32 {
self.deletes
self.tracked
.deletes
.as_ref()
.map(|delete_meta| delete_meta.num_deleted_docs)
.unwrap_or(0u32)
@@ -63,8 +111,9 @@ impl SegmentMeta {
pub fn relative_path(&self, component: SegmentComponent) -> PathBuf {
let mut path = self.id().uuid_string();
path.push_str(&*match component {
SegmentComponent::POSITIONS => ".pos".to_string(),
SegmentComponent::POSTINGS => ".idx".to_string(),
SegmentComponent::POSITIONS => ".pos".to_string(),
SegmentComponent::POSITIONSSKIP => ".posidx".to_string(),
SegmentComponent::TERMS => ".term".to_string(),
SegmentComponent::STORE => ".store".to_string(),
SegmentComponent::FASTFIELDS => ".fast".to_string(),
@@ -80,7 +129,7 @@ impl SegmentMeta {
/// and all the doc ids contains in this segment
/// are exactly (0..max_doc).
pub fn max_doc(&self) -> u32 {
self.max_doc
self.tracked.max_doc
}
/// Return the number of documents in the segment.
@@ -88,28 +137,39 @@ impl SegmentMeta {
self.max_doc() - self.num_deleted_docs()
}
/// Returns the opstamp of the last delete operation
/// Returns the `Opstamp` of the last delete operation
/// taken in account in this segment.
pub fn delete_opstamp(&self) -> Option<u64> {
self.deletes.as_ref().map(|delete_meta| delete_meta.opstamp)
pub fn delete_opstamp(&self) -> Option<Opstamp> {
self.tracked
.deletes
.as_ref()
.map(|delete_meta| delete_meta.opstamp)
}
/// Returns true iff the segment meta contains
/// delete information.
pub fn has_deletes(&self) -> bool {
self.deletes.is_some()
self.num_deleted_docs() > 0
}
#[doc(hidden)]
pub fn set_max_doc(&mut self, max_doc: u32) {
self.max_doc = max_doc;
}
#[doc(hidden)]
pub fn set_delete_meta(&mut self, num_deleted_docs: u32, opstamp: u64) {
self.deletes = Some(DeleteMeta {
pub fn with_delete_meta(self, num_deleted_docs: u32, opstamp: Opstamp) -> SegmentMeta {
let delete_meta = DeleteMeta {
num_deleted_docs,
opstamp,
};
let tracked = self.tracked.map(move |inner_meta| InnerSegmentMeta {
segment_id: inner_meta.segment_id,
max_doc: inner_meta.max_doc,
deletes: Some(delete_meta),
});
SegmentMeta { tracked }
}
}
#[derive(Clone, Debug, Serialize, Deserialize)]
struct InnerSegmentMeta {
segment_id: SegmentId,
max_doc: u32,
deletes: Option<DeleteMeta>,
}

305
src/core/segment_reader.rs
View File

@@ -1,31 +1,26 @@
use Result;
use core::Segment;
use core::SegmentId;
use core::SegmentComponent;
use std::sync::RwLock;
use common::HasLen;
use core::SegmentMeta;
use fastfield::{self, FastFieldNotAvailableError};
use fastfield::DeleteBitSet;
use store::StoreReader;
use schema::Document;
use DocId;
use std::sync::Arc;
use std::collections::HashMap;
use common::CompositeFile;
use std::fmt;
use common::HasLen;
use core::InvertedIndexReader;
use core::Segment;
use core::SegmentComponent;
use core::SegmentId;
use directory::ReadOnlySource;
use fastfield::DeleteBitSet;
use fastfield::FacetReader;
use fastfield::FastFieldReaders;
use fieldnorm::FieldNormReader;
use schema::Field;
use schema::FieldType;
use error::ErrorKind;
use termdict::TermDictionaryImpl;
use fastfield::FacetReader;
use fastfield::FastFieldReader;
use schema::Schema;
use space_usage::SegmentSpaceUsage;
use std::collections::HashMap;
use std::fmt;
use std::sync::Arc;
use std::sync::RwLock;
use store::StoreReader;
use termdict::TermDictionary;
use fastfield::{FastValue, MultiValueIntFastFieldReader};
use schema::Cardinality;
use fieldnorm::FieldNormReader;
use DocId;
use Result;
/// Entry point to access all of the datastructures of the `Segment`
///
@@ -45,15 +40,17 @@ pub struct SegmentReader {
inv_idx_reader_cache: Arc<RwLock<HashMap<Field, Arc<InvertedIndexReader>>>>,
segment_id: SegmentId,
segment_meta: SegmentMeta,
max_doc: DocId,
num_docs: DocId,
termdict_composite: CompositeFile,
postings_composite: CompositeFile,
positions_composite: CompositeFile,
fast_fields_composite: CompositeFile,
positions_idx_composite: CompositeFile,
fast_fields_readers: Arc<FastFieldReaders>,
fieldnorms_composite: CompositeFile,
store_reader: StoreReader,
store_source: ReadOnlySource,
delete_bitset_opt: Option<DeleteBitSet>,
schema: Schema,
}
@@ -64,7 +61,7 @@ impl SegmentReader {
/// Today, `tantivy` does not handle deletes, so it happens
/// to also be the number of documents in the index.
pub fn max_doc(&self) -> DocId {
self.segment_meta.max_doc()
self.max_doc
}
/// Returns the number of documents.
@@ -73,7 +70,12 @@ impl SegmentReader {
/// Today, `tantivy` does not handle deletes so max doc and
/// num_docs are the same.
pub fn num_docs(&self) -> DocId {
self.segment_meta.num_docs()
self.num_docs
}
/// Returns the schema of the index this segment belongs to.
pub fn schema(&self) -> &Schema {
&self.schema
}
/// Return the number of documents that have been
@@ -99,91 +101,46 @@ impl SegmentReader {
///
/// # Panics
/// May panic if the index is corrupted.
pub fn fast_field_reader<Item: FastValue>(
&self,
field: Field,
) -> fastfield::Result<FastFieldReader<Item>> {
let field_entry = self.schema.get_field_entry(field);
if Item::fast_field_cardinality(field_entry.field_type()) == Some(Cardinality::SingleValue)
{
self.fast_fields_composite
.open_read(field)
.ok_or_else(|| FastFieldNotAvailableError::new(field_entry))
.map(FastFieldReader::open)
} else {
Err(FastFieldNotAvailableError::new(field_entry))
}
}
/// Accessor to the `MultiValueIntFastFieldReader` associated to a given `Field`.
/// May panick if the field is not a multivalued fastfield of the type `Item`.
pub fn multi_fast_field_reader<Item: FastValue>(
&self,
field: Field,
) -> fastfield::Result<MultiValueIntFastFieldReader<Item>> {
let field_entry = self.schema.get_field_entry(field);
if Item::fast_field_cardinality(field_entry.field_type()) == Some(Cardinality::MultiValues)
{
let idx_reader = self.fast_fields_composite
.open_read_with_idx(field, 0)
.ok_or_else(|| FastFieldNotAvailableError::new(field_entry))
.map(FastFieldReader::open)?;
let vals_reader = self.fast_fields_composite
.open_read_with_idx(field, 1)
.ok_or_else(|| FastFieldNotAvailableError::new(field_entry))
.map(FastFieldReader::open)?;
Ok(MultiValueIntFastFieldReader::open(idx_reader, vals_reader))
} else {
Err(FastFieldNotAvailableError::new(field_entry))
}
pub fn fast_fields(&self) -> &FastFieldReaders {
&self.fast_fields_readers
}
/// Accessor to the `FacetReader` associated to a given `Field`.
pub fn facet_reader(&self, field: Field) -> 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(ErrorKind::InvalidArgument(format!(
"The field {:?} is not a \
hierarchical facet.",
field_entry
)).into());
return None;
}
let term_ords_reader = self.multi_fast_field_reader(field)?;
let termdict_source = self.termdict_composite.open_read(field).ok_or_else(|| {
ErrorKind::InvalidArgument(format!(
"The field \"{}\" is a hierarchical \
but this segment does not seem to have the field term \
dictionary.",
field_entry.name()
))
})?;
let termdict = TermDictionaryImpl::from_source(termdict_source);
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);
Ok(facet_reader)
Some(facet_reader)
}
/// Accessor to the segment's `Field norms`'s reader.
///
/// Field norms are the length (in tokens) of the fields.
/// It is used in the computation of the [TfIdf]
/// (https://fulmicoton.gitbooks.io/tantivy-doc/content/tfidf.html).
/// It is used in the computation of the [TfIdf](https://fulmicoton.gitbooks.io/tantivy-doc/content/tfidf.html).
///
/// They are simply stored as a fast field, serialized in
/// the `.fieldnorm` file of the segment.
pub fn get_fieldnorms_reader(&self, field: Field) -> FieldNormReader {
if let Some(fieldnorm_source) = self.fieldnorms_composite
.open_read(field) {
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 market as indexed during indexing.", field_name);
let err_msg = format!(
"Field norm not found for field {:?}. Was it market as indexed during indexing.",
field_name
);
panic!(err_msg);
}
}
/// Accessor to the segment's `StoreReader`.
pub fn get_store_reader(&self) -> &StoreReader {
&self.store_reader
pub fn get_store_reader(&self) -> StoreReader {
StoreReader::from_source(self.store_source.clone())
}
/// Open a new segment for reading.
@@ -192,7 +149,8 @@ impl SegmentReader {
let termdict_composite = CompositeFile::open(&termdict_source)?;
let store_source = segment.open_read(SegmentComponent::STORE)?;
let store_reader = StoreReader::from_source(store_source);
fail_point!("SegmentReader::open#middle");
let postings_source = segment.open_read(SegmentComponent::POSTINGS)?;
let postings_composite = CompositeFile::open(&postings_source)?;
@@ -205,32 +163,44 @@ impl SegmentReader {
}
};
let positions_idx_composite = {
if let Ok(source) = segment.open_read(SegmentComponent::POSITIONSSKIP) {
CompositeFile::open(&source)?
} else {
CompositeFile::empty()
}
};
let schema = segment.schema();
let fast_fields_data = segment.open_read(SegmentComponent::FASTFIELDS)?;
let fast_fields_composite = CompositeFile::open(&fast_fields_data)?;
let fast_field_readers =
Arc::new(FastFieldReaders::load_all(&schema, &fast_fields_composite)?);
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)?;
Some(DeleteBitSet::open(delete_data))
} else {
None
};
let delete_bitset_opt = if segment.meta().has_deletes() {
let delete_data = segment.open_read(SegmentComponent::DELETE)?;
Some(DeleteBitSet::open(delete_data))
} else {
None
};
let schema = segment.schema();
Ok(SegmentReader {
inv_idx_reader_cache: Arc::new(RwLock::new(HashMap::new())),
segment_meta: segment.meta().clone(),
max_doc: segment.meta().max_doc(),
num_docs: segment.meta().num_docs(),
termdict_composite,
postings_composite,
fast_fields_composite,
fast_fields_readers: fast_field_readers,
fieldnorms_composite,
segment_id: segment.id(),
store_reader,
store_source,
delete_bitset_opt,
positions_composite,
positions_idx_composite,
schema,
})
}
@@ -243,7 +213,8 @@ impl SegmentReader {
/// term dictionary associated to a specific field,
/// and opening the posting list associated to any term.
pub fn inverted_index(&self, field: Field) -> Arc<InvertedIndexReader> {
if let Some(inv_idx_reader) = self.inv_idx_reader_cache
if let Some(inv_idx_reader) = self
.inv_idx_reader_cache
.read()
.expect("Lock poisoned. This should never happen")
.get(&field)
@@ -267,23 +238,31 @@ impl SegmentReader {
// As a result, no data is associated to the inverted index.
//
// Returns an empty inverted index.
return Arc::new(InvertedIndexReader::empty(field_type.clone()));
return Arc::new(InvertedIndexReader::empty(field_type));
}
let postings_source = postings_source_opt.unwrap();
let termdict_source = self.termdict_composite
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_source = self.positions_composite
let positions_source = self
.positions_composite
.open_read(field)
.expect("Index corrupted. Failed to open field positions in composite file.");
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(
TermDictionaryImpl::from_source(termdict_source),
TermDictionary::from_source(&termdict_source),
postings_source,
positions_source,
positions_idx_source,
record_option,
));
@@ -297,14 +276,6 @@ impl SegmentReader {
inv_idx_reader
}
/// Returns the document (or to be accurate, its stored field)
/// bearing the given doc id.
/// This method is slow and should seldom be called from
/// within a collector.
pub fn doc(&self, doc_id: DocId) -> Result<Document> {
self.store_reader.get(doc_id)
}
/// Returns the segment id
pub fn segment_id(&self) -> SegmentId {
self.segment_id
@@ -323,6 +294,29 @@ impl SegmentReader {
.map(|delete_set| delete_set.is_deleted(doc))
.unwrap_or(false)
}
/// Returns an iterator that will iterate over the alive document ids
pub fn doc_ids_alive(&self) -> SegmentReaderAliveDocsIterator {
SegmentReaderAliveDocsIterator::new(&self)
}
/// Summarize total space usage of this segment.
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.fieldnorms_composite.space_usage(),
self.get_store_reader().space_usage(),
self.delete_bitset_opt
.as_ref()
.map(DeleteBitSet::space_usage)
.unwrap_or(0),
)
}
}
impl fmt::Debug for SegmentReader {
@@ -330,3 +324,88 @@ impl fmt::Debug for SegmentReader {
write!(f, "SegmentReader({:?})", self.segment_id)
}
}
/// 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 core::Index;
use schema::{Schema, Term, STORED, TEXT};
use DocId;
#[test]
fn test_alive_docs_iterator() {
let mut schema_builder = Schema::builder();
schema_builder.add_text_field("name", TEXT | STORED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema.clone());
let name = schema.get_field("name").unwrap();
{
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().unwrap();
}
{
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().unwrap();
}
let searcher = index.reader().unwrap().searcher();
let docs: Vec<DocId> = searcher.segment_reader(0).doc_ids_alive().collect();
assert_eq!(vec![0u32, 2u32], docs);
}
}

4
src/datastruct/mod.rs
View File

@@ -1,4 +0,0 @@
mod skip;
pub mod stacker;
pub use self::skip::{SkipList, SkipListBuilder};

161
src/datastruct/stacker/expull.rs
View File

@@ -1,161 +0,0 @@
use std::mem;
use super::heap::{Heap, HeapAllocable};
#[inline]
pub fn is_power_of_2(val: u32) -> bool {
val & (val - 1) == 0
}
#[inline]
pub fn jump_needed(val: u32) -> bool {
val > 3 && is_power_of_2(val)
}
#[derive(Debug, Clone)]
pub struct ExpUnrolledLinkedList {
len: u32,
end: u32,
val0: u32,
val1: u32,
val2: u32,
next: u32, // inline of the first block
}
impl ExpUnrolledLinkedList {
pub fn iter<'a>(&self, addr: u32, heap: &'a Heap) -> ExpUnrolledLinkedListIterator<'a> {
ExpUnrolledLinkedListIterator {
heap,
addr: addr + 2u32 * (mem::size_of::<u32>() as u32),
len: self.len,
consumed: 0,
}
}
pub fn push(&mut self, val: u32, heap: &Heap) {
self.len += 1;
if jump_needed(self.len) {
// we need to allocate another block.
// ... As we want to grow block exponentially
// the next block as a size of (length so far),
// and we need to add 1u32 to store the pointer
// to the next element.
let new_block_size: usize = (self.len as usize + 1) * mem::size_of::<u32>();
let new_block_addr: u32 = heap.allocate_space(new_block_size);
heap.set(self.end, &new_block_addr);
self.end = new_block_addr;
}
heap.set(self.end, &val);
self.end += mem::size_of::<u32>() as u32;
}
}
impl HeapAllocable for u32 {
fn with_addr(_addr: u32) -> u32 {
0u32
}
}
impl HeapAllocable for ExpUnrolledLinkedList {
fn with_addr(addr: u32) -> ExpUnrolledLinkedList {
let last_addr = addr + mem::size_of::<u32>() as u32 * 2u32;
ExpUnrolledLinkedList {
len: 0u32,
end: last_addr,
val0: 0u32,
val1: 0u32,
val2: 0u32,
next: 0u32,
}
}
}
pub struct ExpUnrolledLinkedListIterator<'a> {
heap: &'a Heap,
addr: u32,
len: u32,
consumed: u32,
}
impl<'a> Iterator for ExpUnrolledLinkedListIterator<'a> {
type Item = u32;
fn next(&mut self) -> Option<u32> {
if self.consumed == self.len {
None
} else {
let addr: u32;
self.consumed += 1;
if jump_needed(self.consumed) {
addr = *self.heap.get_mut_ref(self.addr);
} else {
addr = self.addr;
}
self.addr = addr + mem::size_of::<u32>() as u32;
Some(*self.heap.get_mut_ref(addr))
}
}
}
#[cfg(test)]
mod tests {
use super::*;
use super::super::heap::Heap;
use test::Bencher;
const NUM_STACK: usize = 10_000;
const STACK_SIZE: u32 = 1000;
#[test]
fn test_stack() {
let heap = Heap::with_capacity(1_000_000);
let (addr, stack) = heap.allocate_object::<ExpUnrolledLinkedList>();
stack.push(1u32, &heap);
stack.push(2u32, &heap);
stack.push(4u32, &heap);
stack.push(8u32, &heap);
{
let mut it = stack.iter(addr, &heap);
assert_eq!(it.next().unwrap(), 1u32);
assert_eq!(it.next().unwrap(), 2u32);
assert_eq!(it.next().unwrap(), 4u32);
assert_eq!(it.next().unwrap(), 8u32);
assert!(it.next().is_none());
}
}
#[bench]
fn bench_push_vec(bench: &mut Bencher) {
bench.iter(|| {
let mut vecs = Vec::with_capacity(100);
for _ in 0..NUM_STACK {
vecs.push(Vec::new());
}
for s in 0..NUM_STACK {
for i in 0u32..STACK_SIZE {
let t = s * 392017 % NUM_STACK;
vecs[t].push(i);
}
}
});
}
#[bench]
fn bench_push_stack(bench: &mut Bencher) {
let heap = Heap::with_capacity(64_000_000);
bench.iter(|| {
let mut stacks = Vec::with_capacity(100);
for _ in 0..NUM_STACK {
let (_, stack) = heap.allocate_object::<ExpUnrolledLinkedList>();
stacks.push(stack);
}
for s in 0..NUM_STACK {
for i in 0u32..STACK_SIZE {
let t = s * 392017 % NUM_STACK;
stacks[t].push(i, &heap);
}
}
heap.clear();
});
}
}

309
src/datastruct/stacker/hashmap.rs
View File

@@ -1,309 +0,0 @@
use std::iter;
use std::mem;
use postings::UnorderedTermId;
use super::heap::{BytesRef, Heap, HeapAllocable};
mod murmurhash2 {
const SEED: u32 = 3_242_157_231u32;
#[inline(always)]
pub fn murmurhash2(key: &[u8]) -> u32 {
let mut key_ptr: *const u32 = key.as_ptr() as *const u32;
let m: u32 = 0x5bd1_e995;
let r = 24;
let len = key.len() as u32;
let mut h: u32 = SEED ^ len;
let num_blocks = len >> 2;
for _ in 0..num_blocks {
let mut k: u32 = unsafe { *key_ptr };
k = k.wrapping_mul(m);
k ^= k >> r;
k = k.wrapping_mul(m);
k = k.wrapping_mul(m);
h ^= k;
key_ptr = key_ptr.wrapping_offset(1);
}
// Handle the last few bytes of the input array
let remaining = len & 3;
let key_ptr_u8: *const u8 = key_ptr as *const u8;
match remaining {
3 => {
h ^= unsafe { u32::from(*key_ptr_u8.wrapping_offset(2)) } << 16;
h ^= unsafe { u32::from(*key_ptr_u8.wrapping_offset(1)) } << 8;
h ^= unsafe { u32::from(*key_ptr_u8) };
h = h.wrapping_mul(m);
}
2 => {
h ^= unsafe { u32::from(*key_ptr_u8.wrapping_offset(1)) } << 8;
h ^= unsafe { u32::from(*key_ptr_u8) };
h = h.wrapping_mul(m);
}
1 => {
h ^= unsafe { u32::from(*key_ptr_u8) };
h = h.wrapping_mul(m);
}
_ => {}
}
h ^= h >> 13;
h = h.wrapping_mul(m);
h ^ (h >> 15)
}
}
/// Split the thread memory budget into
/// - the heap size
/// - the hash table "table" itself.
///
/// Returns (the heap size in bytes, the hash table size in number of bits)
pub(crate) fn split_memory(per_thread_memory_budget: usize) -> (usize, usize) {
let table_size_limit: usize = per_thread_memory_budget / 3;
let compute_table_size = |num_bits: usize| (1 << num_bits) * mem::size_of::<KeyValue>();
let table_num_bits: usize = (1..)
.into_iter()
.take_while(|num_bits: &usize| compute_table_size(*num_bits) < table_size_limit)
.last()
.expect(&format!(
"Per thread memory is too small: {}",
per_thread_memory_budget
));
let table_size = compute_table_size(table_num_bits);
let heap_size = per_thread_memory_budget - table_size;
(heap_size, table_num_bits)
}
/// `KeyValue` is the item stored in the hash table.
/// The key is actually a `BytesRef` object stored in an external heap.
/// The `value_addr` also points to an address in the heap.
///
/// The key and the value are actually stored contiguously.
/// For this reason, the (start, stop) information is actually redundant
/// and can be simplified in the future
#[derive(Copy, Clone, Default)]
struct KeyValue {
key_value_addr: BytesRef,
hash: u32,
}
impl KeyValue {
fn is_empty(&self) -> bool {
self.key_value_addr.is_null()
}
}
/// Customized `HashMap` with string keys
///
/// This `HashMap` takes String as keys. Keys are
/// stored in a user defined heap.
///
/// The quirky API has the benefit of avoiding
/// the computation of the hash of the key twice,
/// or copying the key as long as there is no insert.
///
pub struct TermHashMap<'a> {
table: Box<[KeyValue]>,
heap: &'a Heap,
mask: usize,
occupied: Vec<usize>,
}
struct QuadraticProbing {
hash: usize,
i: usize,
mask: usize,
}
impl QuadraticProbing {
fn compute(hash: usize, mask: usize) -> QuadraticProbing {
QuadraticProbing {
hash,
i: 0,
mask,
}
}
#[inline]
fn next_probe(&mut self) -> usize {
self.i += 1;
(self.hash + self.i * self.i) & self.mask
}
}
impl<'a> TermHashMap<'a> {
pub fn new(num_bucket_power_of_2: usize, heap: &'a Heap) -> TermHashMap<'a> {
let table_size = 1 << num_bucket_power_of_2;
let table: Vec<KeyValue> = iter::repeat(KeyValue::default()).take(table_size).collect();
TermHashMap {
table: table.into_boxed_slice(),
heap,
mask: table_size - 1,
occupied: Vec::with_capacity(table_size / 2),
}
}
fn probe(&self, hash: u32) -> QuadraticProbing {
QuadraticProbing::compute(hash as usize, self.mask)
}
pub fn is_saturated(&self) -> bool {
self.table.len() < self.occupied.len() * 3
}
#[inline(never)]
fn get_key_value(&self, bytes_ref: BytesRef) -> (&[u8], u32) {
let key_bytes: &[u8] = self.heap.get_slice(bytes_ref);
let expull_addr: u32 = bytes_ref.addr() + 2 + key_bytes.len() as u32;
(key_bytes, expull_addr)
}
pub fn set_bucket(&mut self, hash: u32, key_value_addr: BytesRef, bucket: usize) {
self.occupied.push(bucket);
self.table[bucket] = KeyValue {
key_value_addr, hash
};
}
pub fn iter<'b: 'a>(&'b self) -> impl Iterator<Item = (&'a [u8], u32, UnorderedTermId)> + 'b {
self.occupied.iter().cloned().map(move |bucket: usize| {
let kv = self.table[bucket];
let (key, offset) = self.get_key_value(kv.key_value_addr);
(key, offset, bucket as UnorderedTermId)
})
}
pub fn get_or_create<S: AsRef<[u8]>, V: HeapAllocable>(
&mut self,
key: S,
) -> (UnorderedTermId, &mut V) {
let key_bytes: &[u8] = key.as_ref();
let hash = murmurhash2::murmurhash2(key.as_ref());
let mut probe = self.probe(hash);
loop {
let bucket = probe.next_probe();
let kv: KeyValue = self.table[bucket];
if kv.is_empty() {
let key_bytes_ref = self.heap.allocate_and_set(key_bytes);
let (addr, val): (u32, &mut V) = self.heap.allocate_object();
assert_eq!(addr, key_bytes_ref.addr() + 2 + key_bytes.len() as u32);
self.set_bucket(hash, key_bytes_ref, bucket);
return (bucket as UnorderedTermId, val);
} else if kv.hash == hash {
let (stored_key, expull_addr): (&[u8], u32) = self.get_key_value(kv.key_value_addr);
if stored_key == key_bytes {
return (
bucket as UnorderedTermId,
self.heap.get_mut_ref(expull_addr),
);
}
}
}
}
}
#[cfg(test)]
mod tests {
use super::*;
use super::super::heap::{Heap, HeapAllocable};
use super::murmurhash2::murmurhash2;
use test::Bencher;
use std::collections::HashSet;
use super::split_memory;
struct TestValue {
val: u32,
_addr: u32,
}
impl HeapAllocable for TestValue {
fn with_addr(addr: u32) -> TestValue {
TestValue {
val: 0u32,
_addr: addr,
}
}
}
#[test]
fn test_hashmap_size() {
assert_eq!(split_memory(100_000), (67232, 12));
assert_eq!(split_memory(1_000_000), (737856, 15));
assert_eq!(split_memory(10_000_000), (7902848, 18));
}
#[test]
fn test_hash_map() {
let heap = Heap::with_capacity(2_000_000);
let mut hash_map: TermHashMap = TermHashMap::new(18, &heap);
{
let v: &mut TestValue = hash_map.get_or_create("abc").1;
assert_eq!(v.val, 0u32);
v.val = 3u32;
}
{
let v: &mut TestValue = hash_map.get_or_create("abcd").1;
assert_eq!(v.val, 0u32);
v.val = 4u32;
}
{
let v: &mut TestValue = hash_map.get_or_create("abc").1;
assert_eq!(v.val, 3u32);
}
{
let v: &mut TestValue = hash_map.get_or_create("abcd").1;
assert_eq!(v.val, 4u32);
}
let mut iter_values = hash_map.iter();
{
let (_, addr, _) = iter_values.next().unwrap();
let val: &TestValue = heap.get_ref(addr);
assert_eq!(val.val, 3u32);
}
{
let (_, addr, _) = iter_values.next().unwrap();
let val: &TestValue = heap.get_ref(addr);
assert_eq!(val.val, 4u32);
}
assert!(iter_values.next().is_none());
}
#[test]
fn test_murmur() {
let s1 = "abcdef";
let s2 = "abcdeg";
for i in 0..5 {
assert_eq!(
murmurhash2(&s1[i..5].as_bytes()),
murmurhash2(&s2[i..5].as_bytes())
);
}
}
#[test]
fn test_murmur_collisions() {
let mut set: HashSet<u32> = HashSet::default();
for i in 0..10_000 {
let s = format!("hash{}", i);
let hash = murmurhash2(s.as_bytes());
set.insert(hash);
}
assert_eq!(set.len(), 10_000);
}
#[bench]
fn bench_murmurhash_2(b: &mut Bencher) {
let keys: Vec<&'static str> =
vec!["wer qwe qwe qwe ", "werbq weqweqwe2 ", "weraq weqweqwe3 "];
b.iter(|| {
keys.iter()
.map(|&s| s.as_bytes())
.map(murmurhash2::murmurhash2)
.map(|h| h as u64)
.last()
.unwrap()
});
}
}

233
src/datastruct/stacker/heap.rs
View File

@@ -1,233 +0,0 @@
use std::cell::UnsafeCell;
use std::mem;
use std::ptr;
use byteorder::{ByteOrder, NativeEndian};
/// `BytesRef` refers to a slice in tantivy's custom `Heap`.
///
/// The slice will encode the length of the `&[u8]` slice
/// on 16-bits, and then the data is encoded.
#[derive(Copy, Clone)]
pub struct BytesRef(u32);
impl BytesRef {
pub fn is_null(&self) -> bool {
self.0 == u32::max_value()
}
pub fn addr(&self) -> u32 {
self.0
}
}
impl Default for BytesRef {
fn default() -> BytesRef {
BytesRef(u32::max_value())
}
}
/// Object that can be allocated in tantivy's custom `Heap`.
pub trait HeapAllocable {
fn with_addr(addr: u32) -> Self;
}
/// Tantivy's custom `Heap`.
pub struct Heap {
inner: UnsafeCell<InnerHeap>,
}
#[cfg_attr(feature = "cargo-clippy", allow(mut_from_ref))]
impl Heap {
/// Creates a new heap with a given capacity
pub fn with_capacity(num_bytes: usize) -> Heap {
Heap {
inner: UnsafeCell::new(InnerHeap::with_capacity(num_bytes)),
}
}
fn inner(&self) -> &mut InnerHeap {
unsafe { &mut *self.inner.get() }
}
/// Clears the heap. All the underlying data is lost.
///
/// This heap does not support deallocation.
/// This method is the only way to free memory.
pub fn clear(&self) {
self.inner().clear();
}
/// Return amount of free space, in bytes.
pub fn num_free_bytes(&self) -> u32 {
self.inner().num_free_bytes()
}
/// Allocate a given amount of space and returns an address
/// in the Heap.
pub fn allocate_space(&self, num_bytes: usize) -> u32 {
self.inner().allocate_space(num_bytes)
}
/// Allocate an object in the heap
pub fn allocate_object<V: HeapAllocable>(&self) -> (u32, &mut V) {
let addr = self.inner().allocate_space(mem::size_of::<V>());
let v: V = V::with_addr(addr);
self.inner().set(addr, &v);
(addr, self.inner().get_mut_ref(addr))
}
/// Stores a `&[u8]` in the heap and returns the destination BytesRef.
pub fn allocate_and_set(&self, data: &[u8]) -> BytesRef {
self.inner().allocate_and_set(data)
}
/// Fetches the `&[u8]` stored on the slice defined by the `BytesRef`
/// given as argumetn
pub fn get_slice(&self, bytes_ref: BytesRef) -> &[u8] {
self.inner().get_slice(bytes_ref)
}
/// Stores an item's data in the heap, at the given `address`.
pub fn set<Item>(&self, addr: u32, val: &Item) {
self.inner().set(addr, val);
}
/// Returns a mutable reference for an object at a given Item.
pub fn get_mut_ref<Item>(&self, addr: u32) -> &mut Item {
self.inner().get_mut_ref(addr)
}
/// Returns a mutable reference to an `Item` at a given `addr`.
#[cfg(test)]
pub fn get_ref<Item>(&self, addr: u32) -> &mut Item {
self.get_mut_ref(addr)
}
}
struct InnerHeap {
buffer: Vec<u8>,
buffer_len: u32,
used: u32,
next_heap: Option<Box<InnerHeap>>,
}
impl InnerHeap {
pub fn with_capacity(num_bytes: usize) -> InnerHeap {
let buffer: Vec<u8> = vec![0u8; num_bytes];
InnerHeap {
buffer,
buffer_len: num_bytes as u32,
next_heap: None,
used: 0u32,
}
}
pub fn clear(&mut self) {
self.used = 0u32;
self.next_heap = None;
}
// Returns the number of free bytes. If the buffer
// has reached it's capacity and overflowed to another buffer, return 0.
pub fn num_free_bytes(&self) -> u32 {
if self.next_heap.is_some() {
0u32
} else {
self.buffer_len - self.used
}
}
pub fn allocate_space(&mut self, num_bytes: usize) -> u32 {
let addr = self.used;
self.used += num_bytes as u32;
if self.used <= self.buffer_len {
addr
} else {
if self.next_heap.is_none() {
info!(
r#"Exceeded heap size. The segment will be committed right
after indexing this document."#,
);
self.next_heap = Some(Box::new(InnerHeap::with_capacity(self.buffer_len as usize)));
}
self.next_heap.as_mut().unwrap().allocate_space(num_bytes) + self.buffer_len
}
}
fn get_slice(&self, bytes_ref: BytesRef) -> &[u8] {
let start = bytes_ref.0;
if start >= self.buffer_len {
self.next_heap
.as_ref()
.unwrap()
.get_slice(BytesRef(start - self.buffer_len))
} else {
let start = start as usize;
let len = NativeEndian::read_u16(&self.buffer[start..start + 2]) as usize;
&self.buffer[start + 2..start + 2 + len]
}
}
fn get_mut_slice(&mut self, start: u32, stop: u32) -> &mut [u8] {
if start >= self.buffer_len {
self.next_heap
.as_mut()
.unwrap()
.get_mut_slice(start - self.buffer_len, stop - self.buffer_len)
} else {
&mut self.buffer[start as usize..stop as usize]
}
}
fn allocate_and_set(&mut self, data: &[u8]) -> BytesRef {
assert!(data.len() < u16::max_value() as usize);
let total_len = 2 + data.len();
let start = self.allocate_space(total_len);
let total_buff = self.get_mut_slice(start, start + total_len as u32);
NativeEndian::write_u16(&mut total_buff[0..2], data.len() as u16);
total_buff[2..].clone_from_slice(data);
BytesRef(start)
}
fn get_mut(&mut self, addr: u32) -> *mut u8 {
if addr >= self.buffer_len {
self.next_heap
.as_mut()
.unwrap()
.get_mut(addr - self.buffer_len)
} else {
let addr_isize = addr as isize;
unsafe { self.buffer.as_mut_ptr().offset(addr_isize) }
}
}
fn get_mut_ref<Item>(&mut self, addr: u32) -> &mut Item {
if addr >= self.buffer_len {
self.next_heap
.as_mut()
.unwrap()
.get_mut_ref(addr - self.buffer_len)
} else {
let v_ptr_u8 = self.get_mut(addr) as *mut u8;
let v_ptr = v_ptr_u8 as *mut Item;
unsafe { &mut *v_ptr }
}
}
pub fn set<Item>(&mut self, addr: u32, val: &Item) {
if addr >= self.buffer_len {
self.next_heap
.as_mut()
.unwrap()
.set(addr - self.buffer_len, val);
} else {
let v_ptr: *const Item = val as *const Item;
let v_ptr_u8: *const u8 = v_ptr as *const u8;
debug_assert!(addr + mem::size_of::<Item>() as u32 <= self.used);
unsafe {
let dest_ptr: *mut u8 = self.get_mut(addr);
ptr::copy(v_ptr_u8, dest_ptr, mem::size_of::<Item>());
}
}
}
}

43
src/datastruct/stacker/mod.rs
View File

@@ -1,43 +0,0 @@
pub(crate) mod hashmap;
mod heap;
mod expull;
pub use self::heap::{Heap, HeapAllocable};
pub use self::expull::ExpUnrolledLinkedList;
pub use self::hashmap::TermHashMap;
#[test]
fn test_unrolled_linked_list() {
use std::collections;
let heap = Heap::with_capacity(30_000_000);
{
heap.clear();
let mut ks: Vec<usize> = (1..5).map(|k| k * 100).collect();
ks.push(2);
ks.push(3);
for k in (1..5).map(|k| k * 100) {
let mut hashmap: TermHashMap = TermHashMap::new(10, &heap);
for j in 0..k {
for i in 0..500 {
let v: &mut ExpUnrolledLinkedList = hashmap.get_or_create(i.to_string()).1;
v.push(i * j, &heap);
}
}
let mut map_addr: collections::HashMap<Vec<u8>, u32> = collections::HashMap::new();
for (key, addr, _) in hashmap.iter() {
map_addr.insert(Vec::from(key), addr);
}
for i in 0..500 {
let key: String = i.to_string();
let addr: u32 = *map_addr.get(key.as_bytes()).unwrap();
let exp_pull: &ExpUnrolledLinkedList = heap.get_ref(addr);
let mut it = exp_pull.iter(addr, &heap);
for j in 0..k {
assert_eq!(it.next().unwrap(), i * j);
}
assert!(!it.next().is_some());
}
}
}
}

154
src/directory/directory.rs
View File

@@ -1,11 +1,104 @@
use std::marker::Send;
use std::fmt;
use std::path::Path;
use directory::directory_lock::Lock;
use directory::error::LockError;
use directory::error::{DeleteError, OpenReadError, OpenWriteError};
use directory::WatchCallback;
use directory::WatchHandle;
use directory::{ReadOnlySource, WritePtr};
use std::result;
use std::fmt;
use std::io;
use std::io::Write;
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;
/// Retry the logic of acquiring locks is pretty simple.
/// We just retry `n` times after a given `duratio`, both
/// depending on the type of lock.
struct RetryPolicy {
num_retries: usize,
wait_in_ms: u64,
}
impl RetryPolicy {
fn no_retry() -> RetryPolicy {
RetryPolicy {
num_retries: 0,
wait_in_ms: 0,
}
}
fn wait_and_retry(&mut self) -> bool {
if self.num_retries == 0 {
false
} else {
self.num_retries -= 1;
let wait_duration = Duration::from_millis(self.wait_in_ms);
thread::sleep(wait_duration);
true
}
}
}
/// The `DirectoryLock` is an object that represents a file lock.
/// See [`LockType`](struct.LockType.html)
///
/// It is transparently associated to a lock file, that gets deleted
/// on `Drop.` The lock is released automatically on `Drop`.
pub struct DirectoryLock(Box<Drop + Send + Sync + 'static>);
struct DirectoryLockGuard {
directory: Box<Directory>,
path: PathBuf,
}
impl<T: Drop + Send + Sync + 'static> From<Box<T>> for DirectoryLock {
fn from(underlying: Box<T>) -> Self {
DirectoryLock(underlying)
}
}
impl Drop for DirectoryLockGuard {
fn drop(&mut self) {
if let Err(e) = self.directory.delete(&*self.path) {
error!("Failed to remove the lock file. {:?}", e);
}
}
}
enum TryAcquireLockError {
FileExists,
IOError(io::Error),
}
fn try_acquire_lock(
filepath: &Path,
directory: &mut Directory,
) -> 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.into()),
})?;
write.flush().map_err(TryAcquireLockError::IOError)?;
Ok(DirectoryLock::from(Box::new(DirectoryLockGuard {
directory: directory.box_clone(),
path: filepath.to_owned(),
})))
}
fn retry_policy(is_blocking: bool) -> RetryPolicy {
if is_blocking {
RetryPolicy {
num_retries: 100,
wait_in_ms: 100,
}
} else {
RetryPolicy::no_retry()
}
}
/// Write-once read many (WORM) abstraction for where
/// tantivy's data should be stored.
@@ -17,7 +110,7 @@ use std::marker::Sync;
/// - The [`RAMDirectory`](struct.RAMDirectory.html), which
/// should be used mostly for tests.
///
pub trait Directory: fmt::Debug + Send + Sync + 'static {
pub trait Directory: DirectoryClone + fmt::Debug + Send + Sync + 'static {
/// Opens a virtual file for read.
///
/// Once a virtual file is open, its data may not
@@ -74,6 +167,57 @@ pub trait Directory: fmt::Debug + Send + Sync + 'static {
/// The file may or may not previously exist.
fn atomic_write(&mut self, path: &Path, data: &[u8]) -> io::Result<()>;
/// Acquire a lock in the given directory.
///
/// The method is blocking or not depending on the `Lock` object.
fn acquire_lock(&self, lock: &Lock) -> Result<DirectoryLock, LockError> {
let mut box_directory = self.box_clone();
let mut retry_policy = retry_policy(lock.is_blocking);
loop {
match try_acquire_lock(&lock.filepath, &mut *box_directory) {
Ok(result) => {
return Ok(result);
}
Err(TryAcquireLockError::FileExists) => {
if !retry_policy.wait_and_retry() {
return Err(LockError::LockBusy);
}
}
Err(TryAcquireLockError::IOError(io_error)) => {
return Err(LockError::IOError(io_error));
}
}
}
}
/// Registers a callback that will be called whenever a change on the `meta.json`
/// using the `atomic_write` API is detected.
///
/// The behavior when using `.watch()` on a file using `.open_write(...)` is, on the other
/// hand, undefined.
///
/// The file will be watched for the lifetime of the returned `WatchHandle`. The caller is
/// required to keep it.
/// It does not override previous callbacks. When the file is modified, all callback that are
/// registered (and whose `WatchHandle` is still alive) are triggered.
///
/// Internally, tantivy only uses this API to detect new commits to implement the
/// `OnCommit` `ReloadPolicy`. Not implementing watch in a `Directory` only prevents the
/// `OnCommit` `ReloadPolicy` to work properly.
fn watch(&self, watch_callback: WatchCallback) -> WatchHandle;
}
/// DirectoryClone
pub trait DirectoryClone {
/// Clones the directory and boxes the clone
fn box_clone(&self) -> Box<Directory>;
}
impl<T> DirectoryClone for T
where
T: 'static + Directory + Clone,
{
fn box_clone(&self) -> Box<Directory> {
Box::new(self.clone())
}
}

56
src/directory/directory_lock.rs Normal file
View File

@@ -0,0 +1,56 @@
use std::path::PathBuf;
/// A directory lock.
///
/// A lock is associated to a specific path and some
/// [`LockParams`](./enum.LockParams.html).
/// Tantivy itself uses only two locks but client application
/// can use the directory facility to define their own locks.
/// - [INDEX_WRITER_LOCK](./struct.INDEX_WRITER_LOCK.html)
/// - [META_LOCK](./struct.META_LOCK.html)
///
/// Check out these locks documentation for more information.
///
#[derive(Debug)]
pub struct Lock {
/// The lock needs to be associated with its own file `path`.
/// Depending on the platform, the lock might rely on the creation
/// and deletion of this filepath.
pub filepath: PathBuf,
/// `lock_params` describes whether acquiring the lock is meant
/// to be a blocking operation or a non-blocking.
///
/// Acquiring a blocking lock blocks until the lock is
/// available.
/// Acquiring a blocking lock returns rapidly, either successfully
/// or with an error signifying that someone is already holding
/// the lock.
pub is_blocking: bool,
}
lazy_static! {
/// Only one process should be able to write tantivy's index at a time.
/// This lock file, when present, is in charge of preventing other processes to open an IndexWriter.
///
/// If the process is killed and this file remains, it is safe to remove it manually.
///
/// Failing to acquire this lock usually means a misuse of tantivy's API,
/// (creating more than one instance of the `IndexWriter`), are a spurious
/// lock file remaining after a crash. In the latter case, removing the file after
/// checking no process running tantivy is running is safe.
pub static ref INDEX_WRITER_LOCK: Lock = Lock {
filepath: PathBuf::from(".tantivy-writer.lock"),
is_blocking: false
};
/// The meta lock file is here to protect the segment files being opened by
/// `IndexReader::reload()` from being garbage collected.
/// It makes it possible for another process to safely consume
/// our index in-writing. Ideally, we may have prefered `RWLock` semantics
/// here, but it is difficult to achieve on Windows.
///
/// Opening segment readers is a very fast process.
pub static ref META_LOCK: Lock = Lock {
filepath: PathBuf::from(".tantivy-meta.lock"),
is_blocking: true
};
}

47
src/directory/error.rs
View File

@@ -1,7 +1,23 @@
use std::error::Error as StdError;
use std::path::PathBuf;
use std::io;
use std::fmt;
use std::io;
use std::path::PathBuf;
/// Error while trying to acquire a directory lock.
#[derive(Debug, Fail)]
pub enum LockError {
/// Failed to acquired a lock as it is already hold 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.
#[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`
#[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)]
@@ -10,6 +26,12 @@ pub struct IOError {
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 {
@@ -51,6 +73,14 @@ pub enum OpenDirectoryError {
DoesNotExist(PathBuf),
/// The path exists but is not a directory.
NotADirectory(PathBuf),
/// IoError
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 {
@@ -62,6 +92,11 @@ impl fmt::Display for OpenDirectoryError {
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
),
}
}
}
@@ -173,9 +208,6 @@ pub enum DeleteError {
/// Any kind of IO error that happens when
/// interacting with the underlying IO device.
IOError(IOError),
/// The file may not be deleted because it is
/// protected.
FileProtected(PathBuf),
}
impl From<IOError> for DeleteError {
@@ -190,9 +222,6 @@ impl fmt::Display for DeleteError {
DeleteError::FileDoesNotExist(ref path) => {
write!(f, "the file '{:?}' does not exist", path)
}
DeleteError::FileProtected(ref path) => {
write!(f, "the file '{:?}' is protected and can't be deleted", path)
}
DeleteError::IOError(ref err) => {
write!(f, "an io error occurred while deleting a file: '{}'", err)
}
@@ -207,7 +236,7 @@ impl StdError for DeleteError {
fn cause(&self) -> Option<&StdError> {
match *self {
DeleteError::FileDoesNotExist(_) | DeleteError::FileProtected(_) => None,
DeleteError::FileDoesNotExist(_) => None,
DeleteError::IOError(ref err) => Some(err),
}
}

368
src/directory/managed_directory.rs
View File

@@ -1,18 +1,32 @@
use std::path::{Path, PathBuf};
use serde_json;
use directory::error::{DeleteError, IOError, OpenReadError, OpenWriteError};
use directory::{ReadOnlySource, WritePtr};
use std::result;
use std::io;
use Directory;
use std::sync::{Arc, RwLock};
use std::collections::HashSet;
use std::sync::RwLockWriteGuard;
use std::io::Write;
use core::MANAGED_FILEPATH;
use std::collections::HashMap;
use std::fmt;
use error::{ErrorKind, Result, ResultExt};
use directory::error::{DeleteError, IOError, LockError, OpenReadError, OpenWriteError};
use directory::DirectoryLock;
use directory::Lock;
use directory::META_LOCK;
use directory::{ReadOnlySource, WritePtr};
use directory::{WatchCallback, WatchHandle};
use error::DataCorruption;
use serde_json;
use std::collections::HashSet;
use std::io;
use std::io::Write;
use std::path::{Path, PathBuf};
use std::result;
use std::sync::RwLockWriteGuard;
use std::sync::{Arc, RwLock};
use Directory;
use Result;
/// Returns true iff the file is "managed".
/// Non-managed file are not subject to garbage collection.
///
/// Filenames that starts by a "." -typically locks-
/// are not managed.
fn is_managed(path: &Path) -> bool {
path.to_str()
.map(|p_str| !p_str.starts_with('.'))
.unwrap_or(true)
}
/// Wrapper of directories that keeps track of files created by Tantivy.
///
@@ -32,37 +46,6 @@ pub struct ManagedDirectory {
#[derive(Debug, Default)]
struct MetaInformation {
managed_paths: HashSet<PathBuf>,
protected_files: HashMap<PathBuf, usize>,
}
/// A `FileProtection` prevents the garbage collection of a file.
///
/// See `ManagedDirectory.protect_file_from_delete`.
pub struct FileProtection {
directory: ManagedDirectory,
path: PathBuf,
}
fn unprotect_file_from_delete(directory: &ManagedDirectory, path: &Path) {
let mut meta_informations_wlock = directory
.meta_informations
.write()
.expect("Managed file lock poisoned");
if let Some(counter_ref_mut) = meta_informations_wlock.protected_files.get_mut(path) {
(*counter_ref_mut) -= 1;
}
}
impl fmt::Debug for FileProtection {
fn fmt(&self, formatter: &mut fmt::Formatter) -> result::Result<(), fmt::Error> {
write!(formatter, "FileProtectionFor({:?})", self.path)
}
}
impl Drop for FileProtection {
fn drop(&mut self) {
unprotect_file_from_delete(&self.directory, &*self.path);
}
}
/// Saves the file containing the list of existing files
@@ -72,29 +55,33 @@ fn save_managed_paths(
wlock: &RwLockWriteGuard<MetaInformation>,
) -> io::Result<()> {
let mut w = serde_json::to_vec(&wlock.managed_paths)?;
write!(&mut w, "\n")?;
writeln!(&mut w)?;
directory.atomic_write(&MANAGED_FILEPATH, &w[..])?;
Ok(())
}
impl ManagedDirectory {
/// Wraps a directory as managed directory.
pub fn new<Dir: Directory>(directory: Dir) -> Result<ManagedDirectory> {
pub fn wrap<Dir: Directory>(directory: Dir) -> Result<ManagedDirectory> {
match directory.atomic_read(&MANAGED_FILEPATH) {
Ok(data) => {
let managed_files_json = String::from_utf8_lossy(&data);
let managed_files: HashSet<PathBuf> = serde_json::from_str(&managed_files_json)
.chain_err(|| ErrorKind::CorruptedFile(MANAGED_FILEPATH.clone()))?;
.map_err(|e| {
DataCorruption::new(
MANAGED_FILEPATH.clone(),
format!("Managed file cannot be deserialized: {:?}. ", e),
)
})?;
Ok(ManagedDirectory {
directory: box directory,
directory: Box::new(directory),
meta_informations: Arc::new(RwLock::new(MetaInformation {
managed_paths: managed_files,
protected_files: HashMap::default(),
})),
})
}
Err(OpenReadError::FileDoesNotExist(_)) => Ok(ManagedDirectory {
directory: box directory,
directory: Box::new(directory),
meta_informations: Arc::default(),
}),
Err(OpenReadError::IOError(e)) => Err(From::from(e)),
@@ -108,6 +95,9 @@ impl ManagedDirectory {
///
/// * `living_files` - List of files that are still used by the index.
///
/// The use a callback ensures that the list of living_files is computed
/// while we hold the lock on meta.
///
/// This method does not panick nor returns errors.
/// If a file cannot be deleted (for permission reasons for instance)
/// an error is simply logged, and the file remains in the list of managed
@@ -115,25 +105,35 @@ impl ManagedDirectory {
pub fn garbage_collect<L: FnOnce() -> HashSet<PathBuf>>(&mut self, get_living_files: L) {
info!("Garbage collect");
let mut files_to_delete = vec![];
// It is crucial to get the living files after acquiring the
// read lock of meta informations. That way, we
// avoid the following scenario.
//
// 1) we get the list of living files.
// 2) someone creates a new file.
// 3) we start garbage collection and remove this file
// even though it is a living file.
//
// releasing the lock as .delete() will use it too.
{
// releasing the lock as .delete() will use it too.
let meta_informations_rlock = self.meta_informations
let meta_informations_rlock = self
.meta_informations
.read()
.expect("Managed directory rlock poisoned in garbage collect.");
// It is crucial to get the living files after acquiring the
// read lock of meta informations. That way, we
// avoid the following scenario.
//
// 1) we get the list of living files.
// 2) someone creates a new file.
// 3) we start garbage collection and remove this file
// even though it is a living file.
let living_files = get_living_files();
for managed_path in &meta_informations_rlock.managed_paths {
if !living_files.contains(managed_path) {
files_to_delete.push(managed_path.clone());
// The point of this second "file" lock is to enforce the following scenario
// 1) process B tries to load a new set of searcher.
// The list of segments is loaded
// 2) writer change meta.json (for instance after a merge or a commit)
// 3) gc kicks in.
// 4) gc removes a file that was useful for process B, before process B opened it.
if let Ok(_meta_lock) = self.acquire_lock(&META_LOCK) {
let living_files = get_living_files();
for managed_path in &meta_informations_rlock.managed_paths {
if !living_files.contains(managed_path) {
files_to_delete.push(managed_path.clone());
}
}
}
}
@@ -158,9 +158,6 @@ impl ManagedDirectory {
error!("Failed to delete {:?}", file_to_delete);
}
}
DeleteError::FileProtected(_) => {
// this is expected.
}
}
}
}
@@ -170,7 +167,8 @@ impl ManagedDirectory {
if !deleted_files.is_empty() {
// update the list of managed files by removing
// the file that were removed.
let mut meta_informations_wlock = self.meta_informations
let mut meta_informations_wlock = self
.meta_informations
.write()
.expect("Managed directory wlock poisoned (2).");
{
@@ -185,28 +183,6 @@ impl ManagedDirectory {
}
}
/// Protects a file from being garbage collected.
///
/// The method returns a `FileProtection` object.
/// The file will not be garbage collected as long as the
/// `FileProtection` object is kept alive.
pub fn protect_file_from_delete(&self, path: &Path) -> FileProtection {
let pathbuf = path.to_owned();
{
let mut meta_informations_wlock = self.meta_informations
.write()
.expect("Managed file lock poisoned on protect");
*meta_informations_wlock
.protected_files
.entry(pathbuf.clone())
.or_insert(0) += 1;
}
FileProtection {
directory: self.clone(),
path: pathbuf.clone(),
}
}
/// Registers a file as managed
///
/// This method must be called before the file is
@@ -214,8 +190,17 @@ impl ManagedDirectory {
/// registering the filepath and creating the file
/// will not lead to garbage files that will
/// never get removed.
///
/// File starting by "." are reserved to locks.
/// They are not managed and cannot be subjected
/// to garbage collection.
fn register_file_as_managed(&mut self, filepath: &Path) -> io::Result<()> {
let mut meta_wlock = self.meta_informations
// Files starting by "." (e.g. lock files) are not managed.
if !is_managed(filepath) {
return Ok(());
}
let mut meta_wlock = self
.meta_informations
.write()
.expect("Managed file lock poisoned");
let has_changed = meta_wlock.managed_paths.insert(filepath.to_owned());
@@ -247,16 +232,6 @@ impl Directory for ManagedDirectory {
}
fn delete(&self, path: &Path) -> result::Result<(), DeleteError> {
{
let metas_rlock = self.meta_informations
.read()
.expect("poisoned lock in managed directory meta");
if let Some(counter) = metas_rlock.protected_files.get(path) {
if *counter > 0 {
return Err(DeleteError::FileProtected(path.to_owned()));
}
}
}
self.directory.delete(path)
}
@@ -264,8 +239,12 @@ impl Directory for ManagedDirectory {
self.directory.exists(path)
}
fn box_clone(&self) -> Box<Directory> {
box self.clone()
fn acquire_lock(&self, lock: &Lock) -> result::Result<DirectoryLock, LockError> {
self.directory.acquire_lock(lock)
}
fn watch(&self, watch_callback: WatchCallback) -> WatchHandle {
self.directory.watch(watch_callback)
}
}
@@ -281,119 +260,98 @@ impl Clone for ManagedDirectory {
#[cfg(test)]
mod tests {
use super::*;
#[cfg(feature="mmap")]
use directory::MmapDirectory;
use std::path::Path;
use std::io::Write;
use tempdir::TempDir;
#[cfg(feature = "mmap")]
mod mmap_specific {
lazy_static! {
static ref TEST_PATH1: &'static Path = Path::new("some_path_for_test");
static ref TEST_PATH2: &'static Path = Path::new("some_path_for_test2");
}
use super::super::*;
use std::path::Path;
use tempdir::TempDir;
#[test]
#[cfg(feature="mmap")]
fn test_managed_directory() {
let tempdir = TempDir::new("index").unwrap();
let tempdir_path = PathBuf::from(tempdir.path());
{
let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
let mut managed_directory = ManagedDirectory::new(mmap_directory).unwrap();
lazy_static! {
static ref TEST_PATH1: &'static Path = Path::new("some_path_for_test");
static ref TEST_PATH2: &'static Path = Path::new("some_path_for_test2");
}
use directory::MmapDirectory;
use std::io::Write;
#[test]
fn test_managed_directory() {
let tempdir = TempDir::new("index").unwrap();
let tempdir_path = PathBuf::from(tempdir.path());
{
let mut write_file = managed_directory.open_write(*TEST_PATH1).unwrap();
write_file.flush().unwrap();
let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
let mut managed_directory = ManagedDirectory::wrap(mmap_directory).unwrap();
{
let mut write_file = managed_directory.open_write(*TEST_PATH1).unwrap();
write_file.flush().unwrap();
}
{
managed_directory
.atomic_write(*TEST_PATH2, &vec![0u8, 1u8])
.unwrap();
}
{
assert!(managed_directory.exists(*TEST_PATH1));
assert!(managed_directory.exists(*TEST_PATH2));
}
{
let living_files: HashSet<PathBuf> =
[TEST_PATH1.to_owned()].into_iter().cloned().collect();
managed_directory.garbage_collect(|| living_files);
}
{
assert!(managed_directory.exists(*TEST_PATH1));
assert!(!managed_directory.exists(*TEST_PATH2));
}
}
{
managed_directory
.atomic_write(*TEST_PATH2, &vec![0u8, 1u8])
.unwrap();
}
{
assert!(managed_directory.exists(*TEST_PATH1));
assert!(managed_directory.exists(*TEST_PATH2));
}
{
let living_files: HashSet<PathBuf> =
[TEST_PATH1.to_owned()].into_iter().cloned().collect();
managed_directory.garbage_collect(|| living_files);
}
{
assert!(managed_directory.exists(*TEST_PATH1));
assert!(!managed_directory.exists(*TEST_PATH2));
let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
let mut managed_directory = ManagedDirectory::wrap(mmap_directory).unwrap();
{
assert!(managed_directory.exists(*TEST_PATH1));
assert!(!managed_directory.exists(*TEST_PATH2));
}
{
let living_files: HashSet<PathBuf> = HashSet::new();
managed_directory.garbage_collect(|| living_files);
}
{
assert!(!managed_directory.exists(*TEST_PATH1));
assert!(!managed_directory.exists(*TEST_PATH2));
}
}
}
{
#[test]
fn test_managed_directory_gc_while_mmapped() {
let tempdir = TempDir::new("index").unwrap();
let tempdir_path = PathBuf::from(tempdir.path());
let living_files = HashSet::new();
let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
let mut managed_directory = ManagedDirectory::new(mmap_directory).unwrap();
{
assert!(managed_directory.exists(*TEST_PATH1));
assert!(!managed_directory.exists(*TEST_PATH2));
}
{
let living_files: HashSet<PathBuf> = HashSet::new();
managed_directory.garbage_collect(|| living_files);
}
{
assert!(!managed_directory.exists(*TEST_PATH1));
assert!(!managed_directory.exists(*TEST_PATH2));
}
}
}
#[test]
#[cfg(feature="mmap ")]
fn test_managed_directory_gc_while_mmapped() {
let tempdir = TempDir::new("index").unwrap();
let tempdir_path = PathBuf::from(tempdir.path());
let living_files = HashSet::new();
let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
let mut managed_directory = ManagedDirectory::new(mmap_directory).unwrap();
managed_directory
.atomic_write(*TEST_PATH1, &vec![0u8, 1u8])
.unwrap();
assert!(managed_directory.exists(*TEST_PATH1));
let _mmap_read = managed_directory.open_read(*TEST_PATH1).unwrap();
managed_directory.garbage_collect(|| living_files.clone());
if cfg!(target_os = "windows") {
// On Windows, gc should try and fail the file as it is mmapped.
let mut managed_directory = ManagedDirectory::wrap(mmap_directory).unwrap();
managed_directory
.atomic_write(*TEST_PATH1, &vec![0u8, 1u8])
.unwrap();
assert!(managed_directory.exists(*TEST_PATH1));
// unmap should happen here.
drop(_mmap_read);
// The file should still be in the list of managed file and
// eventually be deleted once mmap is released.
managed_directory.garbage_collect(|| living_files);
assert!(!managed_directory.exists(*TEST_PATH1));
} else {
assert!(!managed_directory.exists(*TEST_PATH1));
}
}
#[test]
#[cfg(feature="mmap")]
fn test_managed_directory_protect() {
let tempdir = TempDir::new("index").unwrap();
let tempdir_path = PathBuf::from(tempdir.path());
let living_files = HashSet::new();
let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
let mut managed_directory = ManagedDirectory::new(mmap_directory).unwrap();
managed_directory
.atomic_write(*TEST_PATH1, &vec![0u8, 1u8])
.unwrap();
assert!(managed_directory.exists(*TEST_PATH1));
{
let _file_protection = managed_directory.protect_file_from_delete(*TEST_PATH1);
let _mmap_read = managed_directory.open_read(*TEST_PATH1).unwrap();
managed_directory.garbage_collect(|| living_files.clone());
assert!(managed_directory.exists(*TEST_PATH1));
if cfg!(target_os = "windows") {
// On Windows, gc should try and fail the file as it is mmapped.
assert!(managed_directory.exists(*TEST_PATH1));
// unmap should happen here.
drop(_mmap_read);
// The file should still be in the list of managed file and
// eventually be deleted once mmap is released.
managed_directory.garbage_collect(|| living_files);
assert!(!managed_directory.exists(*TEST_PATH1));
} else {
assert!(!managed_directory.exists(*TEST_PATH1));
}
}
managed_directory.garbage_collect(|| living_files.clone());
assert!(!managed_directory.exists(*TEST_PATH1));
}
}

480
src/directory/mmap_directory.rs
View File

@@ -1,29 +1,49 @@
extern crate fs2;
extern crate notify;
use self::fs2::FileExt;
use self::notify::RawEvent;
use self::notify::RecursiveMode;
use self::notify::Watcher;
use atomicwrites;
use common::make_io_err;
use directory::Directory;
use core::META_FILEPATH;
use directory::error::LockError;
use directory::error::{DeleteError, IOError, OpenDirectoryError, OpenReadError, OpenWriteError};
use directory::read_only_source::BoxedData;
use directory::Directory;
use directory::DirectoryLock;
use directory::Lock;
use directory::ReadOnlySource;
use directory::shared_vec_slice::SharedVecSlice;
use directory::WatchCallback;
use directory::WatchCallbackList;
use directory::WatchHandle;
use directory::WritePtr;
use fst::raw::MmapReadOnly;
use std::collections::hash_map::Entry as HashMapEntry;
use memmap::Mmap;
use std::collections::HashMap;
use std::convert::From;
use std::fmt;
use std::fs::{self, File};
use std::fs::OpenOptions;
use std::fs::{self, File};
use std::io::{self, Seek, SeekFrom};
use std::io::{BufWriter, Read, Write};
use std::path::{Path, PathBuf};
use std::result;
use std::sync::mpsc::{channel, Receiver, Sender};
use std::sync::Arc;
use std::sync::Mutex;
use std::sync::RwLock;
use std::sync::Weak;
use std::thread;
use tempdir::TempDir;
/// Create a default io error given a string.
pub(crate) fn make_io_err(msg: String) -> io::Error {
io::Error::new(io::ErrorKind::Other, msg)
}
/// 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<MmapReadOnly>, OpenReadError> {
/// cannot be mmapped)
fn open_mmap(full_path: &Path) -> result::Result<Option<Mmap>, OpenReadError> {
let file = File::open(full_path).map_err(|e| {
if e.kind() == io::ErrorKind::NotFound {
OpenReadError::FileDoesNotExist(full_path.to_owned())
@@ -32,7 +52,8 @@ fn open_mmap(full_path: &Path) -> result::Result<Option<MmapReadOnly>, OpenReadE
}
})?;
let meta_data = file.metadata()
let meta_data = file
.metadata()
.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
@@ -40,9 +61,11 @@ fn open_mmap(full_path: &Path) -> result::Result<Option<MmapReadOnly>, OpenReadE
// instead.
return Ok(None);
}
MmapReadOnly::open(&file)
.map(Some)
.map_err(|e| From::from(IOError::with_path(full_path.to_owned(), e)))
unsafe {
memmap::Mmap::map(&file)
.map(Some)
.map_err(|e| From::from(IOError::with_path(full_path.to_owned(), e)))
}
}
#[derive(Default, Clone, Debug, Serialize, Deserialize)]
@@ -62,7 +85,7 @@ pub struct CacheInfo {
struct MmapCache {
counters: CacheCounters,
cache: HashMap<PathBuf, MmapReadOnly>,
cache: HashMap<PathBuf, Weak<BoxedData>>,
}
impl Default for MmapCache {
@@ -75,12 +98,7 @@ impl Default for MmapCache {
}
impl MmapCache {
/// Removes a `MmapReadOnly` entry from the mmap cache.
fn discard_from_cache(&mut self, full_path: &Path) -> bool {
self.cache.remove(full_path).is_some()
}
fn get_info(&mut self) -> CacheInfo {
fn get_info(&self) -> CacheInfo {
let paths: Vec<PathBuf> = self.cache.keys().cloned().collect();
CacheInfo {
counters: self.counters.clone(),
@@ -88,57 +106,178 @@ impl MmapCache {
}
}
fn get_mmap(&mut self, full_path: &Path) -> Result<Option<MmapReadOnly>, OpenReadError> {
Ok(match self.cache.entry(full_path.to_owned()) {
HashMapEntry::Occupied(occupied_entry) => {
let mmap = occupied_entry.get();
fn remove_weak_ref(&mut self) {
let keys_to_remove: Vec<PathBuf> = self
.cache
.iter()
.filter(|(_, mmap_weakref)| mmap_weakref.upgrade().is_none())
.map(|(key, _)| key.clone())
.collect();
for key in keys_to_remove {
self.cache.remove(&key);
}
}
// Returns None if the file exists but as a len of 0 (and hence is not mmappable).
fn get_mmap(&mut self, full_path: &Path) -> Result<Option<Arc<BoxedData>>, OpenReadError> {
if let Some(mmap_weak) = self.cache.get(full_path) {
if let Some(mmap_arc) = mmap_weak.upgrade() {
self.counters.hit += 1;
Some(mmap.clone())
}
HashMapEntry::Vacant(vacant_entry) => {
self.counters.miss += 1;
if let Some(mmap) = open_mmap(full_path)? {
vacant_entry.insert(mmap.clone());
Some(mmap)
} else {
None
}
return Ok(Some(mmap_arc));
}
}
self.cache.remove(full_path);
self.counters.miss += 1;
Ok(if let Some(mmap) = open_mmap(full_path)? {
let mmap_arc: Arc<BoxedData> = Arc::new(Box::new(mmap));
let mmap_weak = Arc::downgrade(&mmap_arc);
self.cache.insert(full_path.to_owned(), mmap_weak);
Some(mmap_arc)
} else {
None
})
}
}
struct InnerWatcherWrapper {
_watcher: Mutex<notify::RecommendedWatcher>,
watcher_router: WatchCallbackList,
}
impl InnerWatcherWrapper {
pub fn new(path: &Path) -> Result<(Self, Receiver<notify::RawEvent>), notify::Error> {
let (tx, watcher_recv): (Sender<RawEvent>, Receiver<RawEvent>) = channel();
// We need to initialize the
let mut watcher = notify::raw_watcher(tx)?;
watcher.watch(path, RecursiveMode::Recursive)?;
let inner = InnerWatcherWrapper {
_watcher: Mutex::new(watcher),
watcher_router: Default::default(),
};
Ok((inner, watcher_recv))
}
}
#[derive(Clone)]
pub(crate) struct WatcherWrapper {
inner: Arc<InnerWatcherWrapper>,
}
impl WatcherWrapper {
pub fn new(path: &Path) -> Result<Self, OpenDirectoryError> {
let (inner, watcher_recv) = InnerWatcherWrapper::new(path).map_err(|err| match err {
notify::Error::PathNotFound => OpenDirectoryError::DoesNotExist(path.to_owned()),
_ => {
panic!("Unknown error while starting watching directory {:?}", path);
}
})?;
let watcher_wrapper = WatcherWrapper {
inner: Arc::new(inner),
};
let watcher_wrapper_clone = watcher_wrapper.clone();
thread::Builder::new()
.name("meta-file-watch-thread".to_string())
.spawn(move || {
loop {
match watcher_recv.recv().map(|evt| evt.path) {
Ok(Some(changed_path)) => {
// ... Actually subject to false positive.
// We might want to be more accurate than this at one point.
if let Some(filename) = changed_path.file_name() {
if filename == *META_FILEPATH {
watcher_wrapper_clone.inner.watcher_router.broadcast();
}
}
}
Ok(None) => {
// not an event we are interested in.
}
Err(_e) => {
// the watch send channel was dropped
break;
}
}
}
})
.expect("Failed to spawn thread to watch meta.json");
Ok(watcher_wrapper)
}
pub fn watch(&mut self, watch_callback: WatchCallback) -> WatchHandle {
self.inner.watcher_router.subscribe(watch_callback)
}
}
/// Directory storing data in files, read via mmap.
///
/// The Mmap object are cached to limit the
/// system calls.
///
/// In the `MmapDirectory`, locks are implemented using the `fs2` crate definition of locks.
///
/// On MacOS & linux, it relies on `flock` (aka `BSD Lock`). These locks solve most of the
/// problems related to POSIX Locks, but may their contract may not be respected on `NFS`
/// depending on the implementation.
///
/// On Windows the semantics are again different.
#[derive(Clone)]
pub struct MmapDirectory {
inner: Arc<MmapDirectoryInner>,
}
struct MmapDirectoryInner {
root_path: PathBuf,
mmap_cache: Arc<RwLock<MmapCache>>,
_temp_directory: Arc<Option<TempDir>>,
mmap_cache: RwLock<MmapCache>,
_temp_directory: Option<TempDir>,
watcher: RwLock<WatcherWrapper>,
}
impl MmapDirectoryInner {
fn new(
root_path: PathBuf,
temp_directory: Option<TempDir>,
) -> Result<MmapDirectoryInner, OpenDirectoryError> {
let watch_wrapper = WatcherWrapper::new(&root_path)?;
let mmap_directory_inner = MmapDirectoryInner {
root_path,
mmap_cache: Default::default(),
_temp_directory: temp_directory,
watcher: RwLock::new(watch_wrapper),
};
Ok(mmap_directory_inner)
}
fn watch(&self, watch_callback: WatchCallback) -> WatchHandle {
let mut wlock = self.watcher.write().unwrap();
wlock.watch(watch_callback)
}
}
impl fmt::Debug for MmapDirectory {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
write!(f, "MmapDirectory({:?})", self.root_path)
write!(f, "MmapDirectory({:?})", self.inner.root_path)
}
}
impl 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.
///
/// This is mostly useful to test the MmapDirectory itself.
/// For your unit tests, prefer the RAMDirectory.
pub fn create_from_tempdir() -> io::Result<MmapDirectory> {
let tempdir = TempDir::new("index")?;
pub fn create_from_tempdir() -> Result<MmapDirectory, OpenDirectoryError> {
let tempdir = TempDir::new("index").map_err(OpenDirectoryError::IoError)?;
let tempdir_path = PathBuf::from(tempdir.path());
let directory = MmapDirectory {
root_path: tempdir_path,
mmap_cache: Arc::new(RwLock::new(MmapCache::default())),
_temp_directory: Arc::new(Some(tempdir)),
};
Ok(directory)
MmapDirectory::new(tempdir_path, Some(tempdir))
}
/// Opens a MmapDirectory in a directory.
@@ -156,18 +295,14 @@ impl MmapDirectory {
directory_path,
)))
} else {
Ok(MmapDirectory {
root_path: PathBuf::from(directory_path),
mmap_cache: Arc::new(RwLock::new(MmapCache::default())),
_temp_directory: Arc::new(None),
})
Ok(MmapDirectory::new(PathBuf::from(directory_path), None)?)
}
}
/// Joins a relative_path to the directory `root_path`
/// to create a proper complete `filepath`.
fn resolve_path(&self, relative_path: &Path) -> PathBuf {
self.root_path.join(relative_path)
self.inner.root_path.join(relative_path)
}
/// Sync the root directory.
@@ -192,7 +327,7 @@ impl MmapDirectory {
.custom_flags(winbase::FILE_FLAG_BACKUP_SEMANTICS);
}
let fd = open_opts.open(&self.root_path)?;
let fd = open_opts.open(&self.inner.root_path)?;
fd.sync_all()?;
Ok(())
}
@@ -202,14 +337,35 @@ impl MmapDirectory {
///
/// The `MmapDirectory` embeds a `MmapDirectory`
/// to avoid multiplying the `mmap` system calls.
pub fn get_cache_info(&mut self) -> CacheInfo {
self.mmap_cache
pub fn get_cache_info(&self) -> CacheInfo {
self.inner
.mmap_cache
.write()
.expect("mmap cache lock is poisoned")
.remove_weak_ref();
self.inner
.mmap_cache
.read()
.expect("Mmap cache lock is poisoned.")
.get_info()
}
}
/// We rely on fs2 for file locking. On Windows & MacOS this
/// uses BSD locks (`flock`). The lock is actually released when
/// the `File` object is dropped and its associated file descriptor
/// is closed.
struct ReleaseLockFile {
_file: File,
path: PathBuf,
}
impl Drop for ReleaseLockFile {
fn drop(&mut self) {
debug!("Releasing lock {:?}", self.path);
}
}
/// This Write wraps a File, but has the specificity of
/// call `sync_all` on flush.
struct SafeFileWriter(File);
@@ -242,7 +398,7 @@ impl Directory for MmapDirectory {
debug!("Open Read {:?}", path);
let full_path = self.resolve_path(path);
let mut mmap_cache = self.mmap_cache.write().map_err(|_| {
let mut mmap_cache = self.inner.mmap_cache.write().map_err(|_| {
let msg = format!(
"Failed to acquired write lock \
on mmap cache while reading {:?}",
@@ -250,11 +406,34 @@ impl Directory for MmapDirectory {
);
IOError::with_path(path.to_owned(), make_io_err(msg))
})?;
Ok(mmap_cache
.get_mmap(&full_path)?
.map(ReadOnlySource::Mmap)
.unwrap_or_else(|| ReadOnlySource::Anonymous(SharedVecSlice::empty())))
.map(ReadOnlySource::from)
.unwrap_or_else(ReadOnlySource::empty))
}
/// Any entry associated to the path in the mmap will be
/// removed before the file is deleted.
fn delete(&self, path: &Path) -> result::Result<(), DeleteError> {
debug!("Deleting file {:?}", path);
let full_path = self.resolve_path(path);
match fs::remove_file(&full_path) {
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(IOError::with_path(path.to_owned(), e).into())
}
}
}
}
fn exists(&self, path: &Path) -> bool {
let full_path = self.resolve_path(path);
full_path.exists()
}
fn open_write(&mut self, path: &Path) -> Result<WritePtr, OpenWriteError> {
@@ -287,43 +466,6 @@ impl Directory for MmapDirectory {
Ok(BufWriter::new(Box::new(writer)))
}
/// Any entry associated to the path in the mmap will be
/// removed before the file is deleted.
fn delete(&self, path: &Path) -> result::Result<(), DeleteError> {
debug!("Deleting file {:?}", path);
let full_path = self.resolve_path(path);
let mut mmap_cache = self.mmap_cache.write().map_err(|_| {
let msg = format!(
"Failed to acquired write lock \
on mmap cache while deleting {:?}",
path
);
IOError::with_path(path.to_owned(), make_io_err(msg))
})?;
mmap_cache.discard_from_cache(path);
// Removing the entry in the MMap cache.
// The munmap will appear on Drop,
// when the last reference is gone.
mmap_cache.cache.remove(&full_path);
match fs::remove_file(&full_path) {
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(IOError::with_path(path.to_owned(), e).into())
}
}
}
}
fn exists(&self, path: &Path) -> bool {
let full_path = self.resolve_path(path);
full_path.exists()
}
fn atomic_read(&self, path: &Path) -> Result<Vec<u8>, OpenReadError> {
let full_path = self.resolve_path(path);
let mut buffer = Vec::new();
@@ -351,8 +493,28 @@ impl Directory for MmapDirectory {
Ok(())
}
fn box_clone(&self) -> Box<Directory> {
Box::new(self.clone())
fn acquire_lock(&self, lock: &Lock) -> Result<DirectoryLock, LockError> {
let full_path = self.resolve_path(&lock.filepath);
// We make sure that the file exists.
let file: File = OpenOptions::new()
.write(true)
.create(true) //< if the file does not exist yet, create it.
.open(&full_path)
.map_err(LockError::IOError)?;
if lock.is_blocking {
file.lock_exclusive().map_err(LockError::IOError)?;
} else {
file.try_lock_exclusive().map_err(|_| LockError::LockBusy)?
}
// dropping the file handle will release the lock.
Ok(DirectoryLock::from(Box::new(ReleaseLockFile {
path: lock.filepath.clone(),
_file: file,
})))
}
fn watch(&self, watch_callback: WatchCallback) -> WatchHandle {
self.inner.watch(watch_callback)
}
}
@@ -363,6 +525,18 @@ mod tests {
// The following tests are specific to the MmapDirectory
use super::*;
use schema::{Schema, SchemaBuilder, TEXT};
use std::fs;
use std::sync::atomic::{AtomicUsize, Ordering};
use std::thread;
use std::time::Duration;
use Index;
use ReloadPolicy;
#[test]
fn test_open_non_existant_path() {
assert!(MmapDirectory::open(PathBuf::from("./nowhere")).is_err());
}
#[test]
fn test_open_empty() {
@@ -382,7 +556,7 @@ mod tests {
#[test]
fn test_cache() {
let content = "abc".as_bytes();
let content = b"abc";
// here we test if the cache releases
// mmaps correctly.
@@ -398,26 +572,104 @@ mod tests {
w.flush().unwrap();
}
}
{
for (i, path) in paths.iter().enumerate() {
let _r = mmap_directory.open_read(path).unwrap();
assert_eq!(mmap_directory.get_cache_info().mmapped.len(), i + 1);
}
for path in paths.iter() {
let _r = mmap_directory.open_read(path).unwrap();
assert_eq!(mmap_directory.get_cache_info().mmapped.len(), num_paths);
}
for (i, path) in paths.iter().enumerate() {
mmap_directory.delete(path).unwrap();
assert_eq!(
mmap_directory.get_cache_info().mmapped.len(),
num_paths - i - 1
);
}
let mut keep = vec![];
for (i, path) in paths.iter().enumerate() {
keep.push(mmap_directory.open_read(path).unwrap());
assert_eq!(mmap_directory.get_cache_info().mmapped.len(), i + 1);
}
assert_eq!(mmap_directory.get_cache_info().counters.hit, 0);
assert_eq!(mmap_directory.get_cache_info().counters.miss, 10);
assert_eq!(mmap_directory.get_cache_info().mmapped.len(), 10);
for path in paths.iter() {
let _r = mmap_directory.open_read(path).unwrap();
assert_eq!(mmap_directory.get_cache_info().mmapped.len(), num_paths);
}
assert_eq!(mmap_directory.get_cache_info().counters.hit, 10);
assert_eq!(mmap_directory.get_cache_info().counters.miss, 10);
assert_eq!(mmap_directory.get_cache_info().mmapped.len(), 10);
for path in paths.iter() {
let _r = mmap_directory.open_read(path).unwrap();
assert_eq!(mmap_directory.get_cache_info().mmapped.len(), 10);
}
assert_eq!(mmap_directory.get_cache_info().counters.hit, 20);
assert_eq!(mmap_directory.get_cache_info().counters.miss, 10);
assert_eq!(mmap_directory.get_cache_info().mmapped.len(), 10);
drop(keep);
for path in paths.iter() {
let _r = mmap_directory.open_read(path).unwrap();
assert_eq!(mmap_directory.get_cache_info().mmapped.len(), 1);
}
assert_eq!(mmap_directory.get_cache_info().counters.hit, 20);
assert_eq!(mmap_directory.get_cache_info().counters.miss, 20);
assert_eq!(mmap_directory.get_cache_info().mmapped.len(), 0);
for path in &paths {
mmap_directory.delete(path).unwrap();
}
assert_eq!(mmap_directory.get_cache_info().counters.hit, 20);
assert_eq!(mmap_directory.get_cache_info().counters.miss, 20);
assert_eq!(mmap_directory.get_cache_info().mmapped.len(), 0);
for path in paths.iter() {
assert!(mmap_directory.open_read(path).is_err());
}
assert_eq!(mmap_directory.get_cache_info().counters.hit, 20);
assert_eq!(mmap_directory.get_cache_info().counters.miss, 30);
assert_eq!(mmap_directory.get_cache_info().mmapped.len(), 0);
}
#[test]
fn test_watch_wrapper() {
let counter: Arc<AtomicUsize> = Default::default();
let counter_clone = counter.clone();
let tmp_dir: TempDir = tempdir::TempDir::new("test_watch_wrapper").unwrap();
let tmp_dirpath = tmp_dir.path().to_owned();
let mut watch_wrapper = WatcherWrapper::new(&tmp_dirpath).unwrap();
let tmp_file = tmp_dirpath.join("coucou");
let _handle = watch_wrapper.watch(Box::new(move || {
counter_clone.fetch_add(1, Ordering::SeqCst);
}));
assert_eq!(counter.load(Ordering::SeqCst), 0);
fs::write(&tmp_file, b"whateverwilldo").unwrap();
thread::sleep(Duration::new(0, 1_000u32));
}
#[test]
fn test_mmap_released() {
let mmap_directory = MmapDirectory::create_from_tempdir().unwrap();
let mut schema_builder: SchemaBuilder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
{
let index = Index::create(mmap_directory.clone(), schema).unwrap();
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
for _num_commits in 0..16 {
for _ in 0..10 {
index_writer.add_document(doc!(text_field=>"abc"));
}
index_writer.commit().unwrap();
}
let reader = index
.reader_builder()
.reload_policy(ReloadPolicy::Manual)
.try_into()
.unwrap();
for _ in 0..30 {
index_writer.add_document(doc!(text_field=>"abc"));
index_writer.commit().unwrap();
reader.reload().unwrap();
}
index_writer.wait_merging_threads().unwrap();
reader.reload().unwrap();
let num_segments = reader.searcher().segment_readers().len();
assert_eq!(num_segments, 4);
assert_eq!(
num_segments * 7,
mmap_directory.get_cache_info().mmapped.len()
);
}
assert_eq!(mmap_directory.get_cache_info().mmapped.len(), 0);
}
}

154
src/directory/mod.rs
View File

@@ -4,32 +4,32 @@ WORM directory abstraction.
*/
#[cfg(feature="mmap")]
#[cfg(feature = "mmap")]
mod mmap_directory;
mod ram_directory;
mod directory;
mod read_only_source;
mod shared_vec_slice;
mod directory_lock;
mod managed_directory;
mod static_directory;
mod ram_directory;
mod read_only_source;
mod watch_event_router;
/// Errors specific to the directory module.
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 use self::ram_directory::RAMDirectory;
pub use self::read_only_source::ReadOnlySource;
pub(crate) use self::watch_event_router::WatchCallbackList;
pub use self::watch_event_router::{WatchCallback, WatchHandle};
use std::io::{BufWriter, Seek, Write};
pub use self::static_directory::StaticDirectory;
pub use self::static_directory::write_static_from_directory;
pub use self::read_only_source::ReadOnlySource;
pub use self::directory::Directory;
pub use self::ram_directory::RAMDirectory;
#[cfg(feature="mmap")]
#[cfg(feature = "mmap")]
pub use self::mmap_directory::MmapDirectory;
pub(crate) use self::read_only_source::SourceRead;
pub(crate) use self::managed_directory::{FileProtection, ManagedDirectory};
pub(crate) use self::managed_directory::ManagedDirectory;
/// Synonym of Seek + Write
pub trait SeekableWrite: Seek + Write {}
@@ -42,128 +42,4 @@ impl<T: Seek + Write> SeekableWrite for T {}
pub type WritePtr = BufWriter<Box<SeekableWrite>>;
#[cfg(test)]
mod tests {
use super::*;
use std::path::Path;
use std::io::{Seek, SeekFrom, Write};
lazy_static! {
static ref TEST_PATH: &'static Path = Path::new("some_path_for_test");
}
#[test]
fn test_ram_directory() {
let mut ram_directory = RAMDirectory::create();
test_directory(&mut ram_directory);
}
#[test]
#[cfg(feature="mmap")]
fn test_mmap_directory() {
let mut mmap_directory = MmapDirectory::create_from_tempdir().unwrap();
test_directory(&mut mmap_directory);
}
#[test]
#[should_panic]
fn ram_directory_panics_if_flush_forgotten() {
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: &mut Directory) {
{
{
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));
}
fn test_seek(directory: &mut Directory) {
{
{
let mut write_file = directory.open_write(*TEST_PATH).unwrap();
write_file.write_all(&[4, 3, 7, 3, 5]).unwrap();
write_file.seek(SeekFrom::Start(0)).unwrap();
write_file.write_all(&[3, 1]).unwrap();
write_file.flush().unwrap();
}
let read_file = directory.open_read(*TEST_PATH).unwrap();
let data: &[u8] = &*read_file;
assert_eq!(data, &[3u8, 1u8, 7u8, 3u8, 5u8]);
}
assert!(directory.delete(*TEST_PATH).is_ok());
}
fn test_rewrite_forbidden(directory: &mut Directory) {
{
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());
}
fn test_write_create_the_file(directory: &mut Directory) {
{
assert!(directory.open_read(*TEST_PATH).is_err());
let _w = directory.open_write(*TEST_PATH).unwrap();
assert!(directory.exists(*TEST_PATH));
assert!(directory.open_read(*TEST_PATH).is_ok());
assert!(directory.delete(*TEST_PATH).is_ok());
}
}
fn test_directory_delete(directory: &mut Directory) {
assert!(directory.open_read(*TEST_PATH).is_err());
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).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, &[1u8, 2u8, 3u8, 4u8]);
}
assert!(directory.delete(Path::new("SomeOtherPath")).is_err());
}
}
if cfg!(windows) {
assert!(directory.delete(*TEST_PATH).is_ok());
}
assert!(directory.open_read(*TEST_PATH).is_err());
assert!(directory.delete(*TEST_PATH).is_err());
}
fn test_directory(directory: &mut Directory) {
test_simple(directory);
test_seek(directory);
test_rewrite_forbidden(directory);
test_write_create_the_file(directory);
test_directory_delete(directory);
}
}
mod tests;

149
src/directory/ram_directory.rs
View File

@@ -1,14 +1,14 @@
use core::META_FILEPATH;
use directory::error::{DeleteError, OpenReadError, OpenWriteError};
use directory::WatchCallbackList;
use directory::WritePtr;
use directory::{Directory, ReadOnlySource, WatchCallback, WatchHandle};
use std::collections::HashMap;
use std::fmt;
use std::io::{self, BufWriter, Cursor, Seek, SeekFrom, Write};
use std::path::{Path, PathBuf};
use std::result;
use std::sync::{Arc, RwLock};
use common::make_io_err;
use directory::{Directory, ReadOnlySource};
use directory::error::{DeleteError, IOError, OpenReadError, OpenWriteError};
use directory::WritePtr;
use super::shared_vec_slice::SharedVecSlice;
/// Writer associated with the `RAMDirectory`
///
@@ -22,13 +22,13 @@ use super::shared_vec_slice::SharedVecSlice;
///
struct VecWriter {
path: PathBuf,
shared_directory: InnerDirectory,
shared_directory: RAMDirectory,
data: Cursor<Vec<u8>>,
is_flushed: bool,
}
impl VecWriter {
fn new(path_buf: PathBuf, shared_directory: InnerDirectory) -> VecWriter {
fn new(path_buf: PathBuf, shared_directory: RAMDirectory) -> VecWriter {
VecWriter {
path: path_buf,
data: Cursor::new(Vec::new()),
@@ -64,75 +64,44 @@ impl Write for VecWriter {
fn flush(&mut self) -> io::Result<()> {
self.is_flushed = true;
self.shared_directory
.write(self.path.clone(), self.data.get_ref())?;
let mut fs = self.shared_directory.fs.write().unwrap();
fs.write(self.path.clone(), self.data.get_ref());
Ok(())
}
}
#[derive(Clone)]
struct InnerDirectory(Arc<RwLock<HashMap<PathBuf, Arc<Vec<u8>>>>>);
#[derive(Default)]
struct InnerDirectory {
fs: HashMap<PathBuf, ReadOnlySource>,
watch_router: WatchCallbackList,
}
impl InnerDirectory {
fn new() -> InnerDirectory {
InnerDirectory(Arc::new(RwLock::new(HashMap::new())))
}
fn write(&self, path: PathBuf, data: &[u8]) -> io::Result<bool> {
let mut map = self.0.write().map_err(|_| {
make_io_err(format!(
"Failed to lock the directory, when trying to write {:?}",
path
))
})?;
let prev_value = map.insert(path, Arc::new(Vec::from(data)));
Ok(prev_value.is_some())
fn write(&mut self, path: PathBuf, data: &[u8]) -> bool {
let data = ReadOnlySource::new(Vec::from(data));
self.fs.insert(path, data).is_some()
}
fn open_read(&self, path: &Path) -> Result<ReadOnlySource, OpenReadError> {
self.0
.read()
.map_err(|_| {
let msg = format!(
"Failed to acquire read lock for the \
directory when trying to read {:?}",
path
);
let io_err = make_io_err(msg);
OpenReadError::IOError(IOError::with_path(path.to_owned(), io_err))
})
.and_then(|readable_map| {
readable_map
.get(path)
.ok_or_else(|| OpenReadError::FileDoesNotExist(PathBuf::from(path)))
.map(Arc::clone)
.map(|data| ReadOnlySource::Anonymous(SharedVecSlice::new(data)))
})
self.fs
.get(path)
.ok_or_else(|| OpenReadError::FileDoesNotExist(PathBuf::from(path)))
.map(Clone::clone)
}
fn delete(&self, path: &Path) -> result::Result<(), DeleteError> {
self.0
.write()
.map_err(|_| {
let msg = format!(
"Failed to acquire write lock for the \
directory when trying to delete {:?}",
path
);
let io_err = make_io_err(msg);
DeleteError::IOError(IOError::with_path(path.to_owned(), io_err))
})
.and_then(|mut writable_map| match writable_map.remove(path) {
Some(_) => Ok(()),
None => Err(DeleteError::FileDoesNotExist(PathBuf::from(path))),
})
fn delete(&mut self, path: &Path) -> result::Result<(), DeleteError> {
match self.fs.remove(path) {
Some(_) => Ok(()),
None => Err(DeleteError::FileDoesNotExist(PathBuf::from(path))),
}
}
fn exists(&self, path: &Path) -> bool {
self.0
.read()
.expect("Failed to get read lock directory.")
.contains_key(path)
self.fs.contains_key(path)
}
fn watch(&mut self, watch_handle: WatchCallback) -> WatchHandle {
self.watch_router.subscribe(watch_handle)
}
}
@@ -147,33 +116,36 @@ impl fmt::Debug for RAMDirectory {
/// It is mainly meant for unit testing.
/// Writes are only made visible upon flushing.
///
#[derive(Clone)]
#[derive(Clone, Default)]
pub struct RAMDirectory {
fs: InnerDirectory,
fs: Arc<RwLock<InnerDirectory>>,
}
impl RAMDirectory {
/// Constructor
pub fn create() -> RAMDirectory {
RAMDirectory {
fs: InnerDirectory::new(),
}
Self::default()
}
}
impl Directory for RAMDirectory {
fn open_read(&self, path: &Path) -> result::Result<ReadOnlySource, OpenReadError> {
self.fs.open_read(path)
self.fs.read().unwrap().open_read(path)
}
fn delete(&self, path: &Path) -> result::Result<(), DeleteError> {
self.fs.write().unwrap().delete(path)
}
fn exists(&self, path: &Path) -> bool {
self.fs.read().unwrap().exists(path)
}
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.fs.clone());
let exists = self.fs
.write(path_buf.clone(), &Vec::new())
.map_err(|err| IOError::with_path(path.to_owned(), err))?;
let vec_writer = VecWriter::new(path_buf.clone(), self.clone());
let exists = fs.write(path_buf.clone(), &[]);
// force the creation of the file to mimic the MMap directory.
if exists {
Err(OpenWriteError::FileAlreadyExists(path_buf))
@@ -182,29 +154,30 @@ impl Directory for RAMDirectory {
}
}
fn delete(&self, path: &Path) -> result::Result<(), DeleteError> {
self.fs.delete(path)
}
fn exists(&self, path: &Path) -> bool {
self.fs.exists(path)
}
fn atomic_read(&self, path: &Path) -> Result<Vec<u8>, OpenReadError> {
let read = self.open_read(path)?;
Ok(read.as_slice().to_owned())
Ok(self.open_read(path)?.as_slice().to_owned())
}
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("Undefined".to_string())
)));
let path_buf = PathBuf::from(path);
let mut vec_writer = VecWriter::new(path_buf.clone(), self.fs.clone());
self.fs.write(path_buf, &Vec::new())?;
// Reserve the path to prevent calls to .write() to succeed.
self.fs.write().unwrap().write(path_buf.clone(), &[]);
let mut vec_writer = VecWriter::new(path_buf.clone(), self.clone());
vec_writer.write_all(data)?;
vec_writer.flush()?;
if path == Path::new(&*META_FILEPATH) {
self.fs.write().unwrap().watch_router.broadcast();
}
Ok(())
}
fn box_clone(&self) -> Box<Directory> {
Box::new(self.clone())
fn watch(&self, watch_callback: WatchCallback) -> WatchHandle {
self.fs.write().unwrap().watch(watch_callback)
}
}

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