rust/tantivy
1
0
Fork 0
mirror of https://github.com/quickwit-oss/tantivy.git synced 2026-01-03 07:42:54 +00:00
Code Issues Packages Projects Releases Wiki Activity

Compare commits

base: rust:path_collector
Branches Tags
rust:main rust:paul.masurel/skip-first-doc-on-intersection rust:paul.masurel/search_into_danger_zone_fastfield 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: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:experimental-do-not-merge
Branches Tags
rust:paul.masurel/skip-first-doc-on-intersection rust:main rust:paul.masurel/search_into_danger_zone_fastfield 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: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

2 Commits
path_colle ... experiment

Author SHA1 Message Date
Paul Masurel
4e64801ec9 Experimental do not merge 2024-04-19 18:32:03 +09:00
Paul Masurel
34cb0c48bf Owned value relying on Vec instead of BTreeMap 2024-04-19 15:08:50 +09:00
46 changed files with 996 additions and 902 deletions
Expand all files Collapse all files

62
CHANGELOG.md
View File

@@ -1,65 +1,3 @@
Tantivy 0.22
================================
Tantivy 0.22 will be able to read indices created with Tantivy 0.21.
#### Bugfixes
- Fix null byte handling in JSON paths (null bytes in json keys caused panic during indexing) [#2345](https://github.com/quickwit-oss/tantivy/pull/2345)(@PSeitz)
- Fix bug that can cause `get_docids_for_value_range` to panic. [#2295](https://github.com/quickwit-oss/tantivy/pull/2295)(@fulmicoton)
- Avoid 1 document indices by increase min memory to 15MB for indexing [#2176](https://github.com/quickwit-oss/tantivy/pull/2176)(@PSeitz)
- Fix merge panic for JSON fields [#2284](https://github.com/quickwit-oss/tantivy/pull/2284)(@PSeitz)
- Fix bug occuring when merging JSON object indexed with positions. [#2253](https://github.com/quickwit-oss/tantivy/pull/2253)(@fulmicoton)
- Fix empty DateHistogram gap bug [#2183](https://github.com/quickwit-oss/tantivy/pull/2183)(@PSeitz)
- Fix range query end check (fields with less than 1 value per doc are affected) [#2226](https://github.com/quickwit-oss/tantivy/pull/2226)(@PSeitz)
- Handle exclusive out of bounds ranges on fastfield range queries [#2174](https://github.com/quickwit-oss/tantivy/pull/2174)(@PSeitz)
#### Breaking API Changes
- rename ReloadPolicy onCommit to onCommitWithDelay [#2235](https://github.com/quickwit-oss/tantivy/pull/2235)(@giovannicuccu)
- Move exports from the root into modules [#2220](https://github.com/quickwit-oss/tantivy/pull/2220)(@PSeitz)
- Accept field name instead of `Field` in FilterCollector [#2196](https://github.com/quickwit-oss/tantivy/pull/2196)(@PSeitz)
- remove deprecated IntOptions and DateTime [#2353](https://github.com/quickwit-oss/tantivy/pull/2353)(@PSeitz)
#### Features/Improvements
- Tantivy documents as a trait: Index data directly without converting to tantivy types first [#2071](https://github.com/quickwit-oss/tantivy/pull/2071)(@ChillFish8)
- encode some part of posting list as -1 instead of direct values (smaller inverted indices) [#2185](https://github.com/quickwit-oss/tantivy/pull/2185)(@trinity-1686a)
- **Aggregation**
- Support to deserialize f64 from string [#2311](https://github.com/quickwit-oss/tantivy/pull/2311)(@PSeitz)
- Add a top_hits aggregator [#2198](https://github.com/quickwit-oss/tantivy/pull/2198)(@ditsuke)
- Support bool type in term aggregation [#2318](https://github.com/quickwit-oss/tantivy/pull/2318)(@PSeitz)
- Support ip adresses in term aggregation [#2319](https://github.com/quickwit-oss/tantivy/pull/2319)(@PSeitz)
- Support date type in term aggregation [#2172](https://github.com/quickwit-oss/tantivy/pull/2172)(@PSeitz)
- Support escaped dot when addressing field [#2250](https://github.com/quickwit-oss/tantivy/pull/2250)(@PSeitz)
- Add ExistsQuery to check documents that have a value [#2160](https://github.com/quickwit-oss/tantivy/pull/2160)(@imotov)
- Expose TopDocs::order_by_u64_field again [#2282](https://github.com/quickwit-oss/tantivy/pull/2282)(@ditsuke)
- **Memory/Performance**
- Faster TopN: replace BinaryHeap with TopNComputer [#2186](https://github.com/quickwit-oss/tantivy/pull/2186)(@PSeitz)
- reduce number of allocations during indexing [#2257](https://github.com/quickwit-oss/tantivy/pull/2257)(@PSeitz)
- Less Memory while indexing: docid deltas while indexing [#2249](https://github.com/quickwit-oss/tantivy/pull/2249)(@PSeitz)
- Faster indexing: use term hashmap in fastfield [#2243](https://github.com/quickwit-oss/tantivy/pull/2243)(@PSeitz)
- term hashmap remove copy in is_empty, unused unordered_id [#2229](https://github.com/quickwit-oss/tantivy/pull/2229)(@PSeitz)
- add method to fetch block of first values in columnar [#2330](https://github.com/quickwit-oss/tantivy/pull/2330)(@PSeitz)
- Faster aggregations: add fast path for full columns in fetch_block [#2328](https://github.com/quickwit-oss/tantivy/pull/2328)(@PSeitz)
- Faster sstable loading: use fst for sstable index [#2268](https://github.com/quickwit-oss/tantivy/pull/2268)(@trinity-1686a)
- **QueryParser**
- allow newline where we allow space in query parser [#2302](https://github.com/quickwit-oss/tantivy/pull/2302)(@trinity-1686a)
- allow some mixing of occur and bool in strict query parser [#2323](https://github.com/quickwit-oss/tantivy/pull/2323)(@trinity-1686a)
- handle * inside term in lenient query parser [#2228](https://github.com/quickwit-oss/tantivy/pull/2228)(@trinity-1686a)
- add support for exists query syntax in query parser [#2170](https://github.com/quickwit-oss/tantivy/pull/2170)(@trinity-1686a)
- Add shared search executor [#2312](https://github.com/quickwit-oss/tantivy/pull/2312)(@MochiXu)
- Truncate keys to u16::MAX in term hashmap [#2299](https://github.com/quickwit-oss/tantivy/pull/2299)(@PSeitz)
- report if a term matched when warming up posting list [#2309](https://github.com/quickwit-oss/tantivy/pull/2309)(@trinity-1686a)
- Support json fields in FuzzyTermQuery [#2173](https://github.com/quickwit-oss/tantivy/pull/2173)(@PingXia-at)
- Read list of fields encoded in term dictionary for JSON fields [#2184](https://github.com/quickwit-oss/tantivy/pull/2184)(@PSeitz)
- add collect_block to BoxableSegmentCollector [#2331](https://github.com/quickwit-oss/tantivy/pull/2331)(@PSeitz)
- expose collect_block buffer size [#2326](https://github.com/quickwit-oss/tantivy/pull/2326)(@PSeitz)
- Forward regex parser errors [#2288](https://github.com/quickwit-oss/tantivy/pull/2288)(@adamreichold)
- Make FacetCounts defaultable and cloneable. [#2322](https://github.com/quickwit-oss/tantivy/pull/2322)(@adamreichold)
- Derive Debug for SchemaBuilder [#2254](https://github.com/quickwit-oss/tantivy/pull/2254)(@GodTamIt)
- add missing inlines to tantivy options [#2245](https://github.com/quickwit-oss/tantivy/pull/2245)(@PSeitz)
Tantivy 0.21.1
================================
#### Bugfixes

16
Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy"
version = "0.22.0"
version = "0.22.0-dev"
authors = ["Paul Masurel <paul.masurel@gmail.com>"]
license = "MIT"
categories = ["database-implementations", "data-structures"]
@@ -52,13 +52,13 @@ itertools = "0.12.0"
measure_time = "0.8.2"
arc-swap = "1.5.0"
columnar = { version= "0.3", path="./columnar", package ="tantivy-columnar" }
sstable = { version= "0.3", path="./sstable", package ="tantivy-sstable", optional = true }
stacker = { version= "0.3", path="./stacker", package ="tantivy-stacker" }
query-grammar = { version= "0.22.0", path="./query-grammar", package = "tantivy-query-grammar" }
tantivy-bitpacker = { version= "0.6", path="./bitpacker" }
common = { version= "0.7", path = "./common/", package = "tantivy-common" }
tokenizer-api = { version= "0.3", path="./tokenizer-api", package="tantivy-tokenizer-api" }
columnar = { version= "0.2", path="./columnar", package ="tantivy-columnar" }
sstable = { version= "0.2", path="./sstable", package ="tantivy-sstable", optional = true }
stacker = { version= "0.2", path="./stacker", package ="tantivy-stacker" }
query-grammar = { version= "0.21.0", path="./query-grammar", package = "tantivy-query-grammar" }
tantivy-bitpacker = { version= "0.5", path="./bitpacker" }
common = { version= "0.6", path = "./common/", package = "tantivy-common" }
tokenizer-api = { version= "0.2", path="./tokenizer-api", package="tantivy-tokenizer-api" }
sketches-ddsketch = { version = "0.2.1", features = ["use_serde"] }
futures-util = { version = "0.3.28", optional = true }
fnv = "1.0.7"

2
bitpacker/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy-bitpacker"
version = "0.6.0"
version = "0.5.0"
edition = "2021"
authors = ["Paul Masurel <paul.masurel@gmail.com>"]
license = "MIT"

83
cliff.toml
View File

@@ -1,10 +1,6 @@
# configuration file for git-cliff{ pattern = "foo", replace = "bar"}
# see https://github.com/orhun/git-cliff#configuration-file
[remote.github]
owner = "quickwit-oss"
repo = "tantivy"
[changelog]
# changelog header
header = """
@@ -12,43 +8,15 @@ header = """
# template for the changelog body
# https://tera.netlify.app/docs/#introduction
body = """
## What's Changed
{%- if version %} in {{ version }}{%- endif -%}
{% if version %}\
{{ version | trim_start_matches(pat="v") }} ({{ timestamp | date(format="%Y-%m-%d") }})
==================
{% else %}\
## [unreleased]
{% endif %}\
{% for commit in commits %}
{% if commit.github.pr_title -%}
{%- set commit_message = commit.github.pr_title -%}
{%- else -%}
{%- set commit_message = commit.message -%}
{%- endif -%}
- {{ commit_message | split(pat="\n") | first | trim }}\
{% if commit.github.pr_number %} \
[#{{ commit.github.pr_number }}]({{ self::remote_url() }}/pull/{{ commit.github.pr_number }}){% if commit.github.username %}(@{{ commit.github.username }}){%- endif -%} \
{%- endif %}
{%- endfor -%}
{% if github.contributors | filter(attribute="is_first_time", value=true) | length != 0 %}
{% raw %}\n{% endraw -%}
## New Contributors
{%- endif %}\
{% for contributor in github.contributors | filter(attribute="is_first_time", value=true) %}
* @{{ contributor.username }} made their first contribution
{%- if contributor.pr_number %} in \
[#{{ contributor.pr_number }}]({{ self::remote_url() }}/pull/{{ contributor.pr_number }}) \
{%- endif %}
{%- endfor -%}
{% if version %}
{% if previous.version %}
**Full Changelog**: {{ self::remote_url() }}/compare/{{ previous.version }}...{{ version }}
{% endif %}
{% else -%}
{% raw %}\n{% endraw %}
{% endif %}
{%- macro remote_url() -%}
https://github.com/{{ remote.github.owner }}/{{ remote.github.repo }}
{%- endmacro -%}
- {% if commit.breaking %}[**breaking**] {% endif %}{{ commit.message | split(pat="\n") | first | trim | upper_first }}(@{{ commit.author.name }})\
{% endfor %}
"""
# remove the leading and trailing whitespace from the template
trim = true
@@ -57,24 +25,53 @@ footer = """
"""
postprocessors = [
{ pattern = 'Paul Masurel', replace = "fulmicoton"}, # replace with github user
{ pattern = 'PSeitz', replace = "PSeitz"}, # replace with github user
{ pattern = 'Adam Reichold', replace = "adamreichold"}, # replace with github user
{ pattern = 'trinity-1686a', replace = "trinity-1686a"}, # replace with github user
{ pattern = 'Michael Kleen', replace = "mkleen"}, # replace with github user
{ pattern = 'Adrien Guillo', replace = "guilload"}, # replace with github user
{ pattern = 'François Massot', replace = "fmassot"}, # replace with github user
{ pattern = 'Naveen Aiathurai', replace = "naveenann"}, # replace with github user
{ pattern = '', replace = ""}, # replace with github user
]
[git]
# parse the commits based on https://www.conventionalcommits.org
# This is required or commit.message contains the whole commit message and not just the title
conventional_commits = false
conventional_commits = true
# filter out the commits that are not conventional
filter_unconventional = true
filter_unconventional = false
# process each line of a commit as an individual commit
split_commits = false
# regex for preprocessing the commit messages
commit_preprocessors = [
{ pattern = '\((\w+\s)?#([0-9]+)\)', replace = ""},
{ pattern = '\((\w+\s)?#([0-9]+)\)', replace = "[#${2}](https://github.com/quickwit-oss/tantivy/issues/${2})"}, # replace issue numbers
]
#link_parsers = [
#{ pattern = "#(\\d+)", href = "https://github.com/quickwit-oss/tantivy/pulls/$1"},
#]
# regex for parsing and grouping commits
commit_parsers = [
{ message = "^feat", group = "Features"},
{ message = "^fix", group = "Bug Fixes"},
{ message = "^doc", group = "Documentation"},
{ message = "^perf", group = "Performance"},
{ message = "^refactor", group = "Refactor"},
{ message = "^style", group = "Styling"},
{ message = "^test", group = "Testing"},
{ message = "^chore\\(release\\): prepare for", skip = true},
{ message = "(?i)clippy", skip = true},
{ message = "(?i)dependabot", skip = true},
{ message = "(?i)fmt", skip = true},
{ message = "(?i)bump", skip = true},
{ message = "(?i)readme", skip = true},
{ message = "(?i)comment", skip = true},
{ message = "(?i)spelling", skip = true},
{ message = "^chore", group = "Miscellaneous Tasks"},
{ body = ".*security", group = "Security"},
{ message = ".*", group = "Other", default_scope = "other"},
]
# protect breaking changes from being skipped due to matching a skipping commit_parser
protect_breaking_commits = false
# filter out the commits that are not matched by commit parsers

10
columnar/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy-columnar"
version = "0.3.0"
version = "0.2.0"
edition = "2021"
license = "MIT"
homepage = "https://github.com/quickwit-oss/tantivy"
@@ -12,10 +12,10 @@ categories = ["database-implementations", "data-structures", "compression"]
itertools = "0.12.0"
fastdivide = "0.4.0"
stacker = { version= "0.3", path = "../stacker", package="tantivy-stacker"}
sstable = { version= "0.3", path = "../sstable", package = "tantivy-sstable" }
common = { version= "0.7", path = "../common", package = "tantivy-common" }
tantivy-bitpacker = { version= "0.6", path = "../bitpacker/" }
stacker = { version= "0.2", path = "../stacker", package="tantivy-stacker"}
sstable = { version= "0.2", path = "../sstable", package = "tantivy-sstable" }
common = { version= "0.6", path = "../common", package = "tantivy-common" }
tantivy-bitpacker = { version= "0.5", path = "../bitpacker/" }
serde = "1.0.152"
downcast-rs = "1.2.0"

4
common/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy-common"
version = "0.7.0"
version = "0.6.0"
authors = ["Paul Masurel <paul@quickwit.io>", "Pascal Seitz <pascal@quickwit.io>"]
license = "MIT"
edition = "2021"
@@ -14,7 +14,7 @@ repository = "https://github.com/quickwit-oss/tantivy"
[dependencies]
byteorder = "1.4.3"
ownedbytes = { version= "0.7", path="../ownedbytes" }
ownedbytes = { version= "0.6", path="../ownedbytes" }
async-trait = "0.1"
time = { version = "0.3.10", features = ["serde-well-known"] }
serde = { version = "1.0.136", features = ["derive"] }

2
common/src/bitset.rs
View File

@@ -1,5 +1,5 @@
use std::io::Write;
use std::{fmt, io};
use std::{fmt, io, u64};
use ownedbytes::OwnedBytes;

5
common/src/datetime.rs
View File

@@ -1,3 +1,5 @@
#![allow(deprecated)]
use std::fmt;
use std::io::{Read, Write};
@@ -25,6 +27,9 @@ pub enum DateTimePrecision {
Nanoseconds,
}
#[deprecated(since = "0.20.0", note = "Use `DateTimePrecision` instead")]
pub type DatePrecision = DateTimePrecision;
/// A date/time value with nanoseconds precision.
///
/// This timestamp does not carry any explicit time zone information.

36
common/src/json_path_writer.rs
View File

@@ -5,12 +5,6 @@ pub const JSON_PATH_SEGMENT_SEP: u8 = 1u8;
pub const JSON_PATH_SEGMENT_SEP_STR: &str =
unsafe { std::str::from_utf8_unchecked(&[JSON_PATH_SEGMENT_SEP]) };
/// Separates the json path and the value in
/// a JSON term binary representation.
pub const JSON_END_OF_PATH: u8 = 0u8;
pub const JSON_END_OF_PATH_STR: &str =
unsafe { std::str::from_utf8_unchecked(&[JSON_END_OF_PATH]) };
/// Create a new JsonPathWriter, that creates flattened json paths for tantivy.
#[derive(Clone, Debug, Default)]
pub struct JsonPathWriter {
@@ -20,14 +14,6 @@ pub struct JsonPathWriter {
}
impl JsonPathWriter {
pub fn with_expand_dots(expand_dots: bool) -> Self {
JsonPathWriter {
path: String::new(),
indices: Vec::new(),
expand_dots,
}
}
pub fn new() -> Self {
JsonPathWriter {
path: String::new(),
@@ -53,8 +39,8 @@ impl JsonPathWriter {
pub fn push(&mut self, segment: &str) {
let len_path = self.path.len();
self.indices.push(len_path);
if self.indices.len() > 1 {
self.path.push(JSON_PATH_SEGMENT_SEP as char);
if !self.path.is_empty() {
self.path.push_str(JSON_PATH_SEGMENT_SEP_STR);
}
self.path.push_str(segment);
if self.expand_dots {
@@ -69,12 +55,6 @@ impl JsonPathWriter {
}
}
/// Set the end of JSON path marker.
#[inline]
pub fn set_end(&mut self) {
self.path.push_str(JSON_END_OF_PATH_STR);
}
/// Remove the last segment. Does nothing if the path is empty.
#[inline]
pub fn pop(&mut self) {
@@ -111,7 +91,6 @@ mod tests {
#[test]
fn json_path_writer_test() {
let mut writer = JsonPathWriter::new();
writer.set_expand_dots(false);
writer.push("root");
assert_eq!(writer.as_str(), "root");
@@ -130,15 +109,4 @@ mod tests {
writer.push("k8s.node.id");
assert_eq!(writer.as_str(), "root\u{1}k8s\u{1}node\u{1}id");
}
#[test]
fn test_json_path_expand_dots_enabled_pop_segment() {
let mut json_writer = JsonPathWriter::with_expand_dots(true);
json_writer.push("hello");
assert_eq!(json_writer.as_str(), "hello");
json_writer.push("color.hue");
assert_eq!(json_writer.as_str(), "hello\x01color\x01hue");
json_writer.pop();
assert_eq!(json_writer.as_str(), "hello");
}
}

4
common/src/lib.rs
View File

@@ -9,12 +9,14 @@ mod byte_count;
mod datetime;
pub mod file_slice;
mod group_by;
pub mod json_path_writer;
mod json_path_writer;
mod serialize;
mod vint;
mod writer;
pub use bitset::*;
pub use byte_count::ByteCount;
#[allow(deprecated)]
pub use datetime::DatePrecision;
pub use datetime::{DateTime, DateTimePrecision};
pub use group_by::GroupByIteratorExtended;
pub use json_path_writer::JsonPathWriter;

2
ownedbytes/Cargo.toml
View File

@@ -1,7 +1,7 @@
[package]
authors = ["Paul Masurel <paul@quickwit.io>", "Pascal Seitz <pascal@quickwit.io>"]
name = "ownedbytes"
version = "0.7.0"
version = "0.6.0"
edition = "2021"
description = "Expose data as static slice"
license = "MIT"

2
query-grammar/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy-query-grammar"
version = "0.22.0"
version = "0.21.0"
authors = ["Paul Masurel <paul.masurel@gmail.com>"]
license = "MIT"
categories = ["database-implementations", "data-structures"]

81
query-grammar/src/query_grammar.rs
View File

@@ -218,14 +218,27 @@ fn term_or_phrase_infallible(inp: &str) -> JResult<&str, Option<UserInputLeaf>>
}
fn term_group(inp: &str) -> IResult<&str, UserInputAst> {
let occur_symbol = alt((
value(Occur::MustNot, char('-')),
value(Occur::Must, char('+')),
));
map(
tuple((
terminated(field_name, multispace0),
delimited(tuple((char('('), multispace0)), ast, char(')')),
delimited(
tuple((char('('), multispace0)),
separated_list0(multispace1, tuple((opt(occur_symbol), term_or_phrase))),
char(')'),
),
)),
|(field_name, mut ast)| {
ast.set_default_field(field_name);
ast
|(field_name, terms)| {
UserInputAst::Clause(
terms
.into_iter()
.map(|(occur, leaf)| (occur, leaf.set_field(Some(field_name.clone())).into()))
.collect(),
)
},
)(inp)
}
@@ -245,18 +258,46 @@ fn term_group_precond(inp: &str) -> IResult<&str, (), ()> {
}
fn term_group_infallible(inp: &str) -> JResult<&str, UserInputAst> {
let (inp, (field_name, _, _, _)) =
let (mut inp, (field_name, _, _, _)) =
tuple((field_name, multispace0, char('('), multispace0))(inp).expect("precondition failed");
let res = delimited_infallible(
nothing,
map(ast_infallible, |(mut ast, errors)| {
ast.set_default_field(field_name.to_string());
(ast, errors)
}),
opt_i_err(char(')'), "expected ')'"),
)(inp);
res
let mut terms = Vec::new();
let mut errs = Vec::new();
let mut first_round = true;
loop {
let mut space_error = if first_round {
first_round = false;
Vec::new()
} else {
let (rest, (_, err)) = space1_infallible(inp)?;
inp = rest;
err
};
if inp.is_empty() {
errs.push(LenientErrorInternal {
pos: inp.len(),
message: "missing )".to_string(),
});
break Ok((inp, (UserInputAst::Clause(terms), errs)));
}
if let Some(inp) = inp.strip_prefix(')') {
break Ok((inp, (UserInputAst::Clause(terms), errs)));
}
// only append missing space error if we did not reach the end of group
errs.append(&mut space_error);
// here we do the assumption term_or_phrase_infallible always consume something if the
// first byte is not `)` or ' '. If it did not, we would end up looping.
let (rest, ((occur, leaf), mut err)) =
tuple_infallible((occur_symbol, term_or_phrase_infallible))(inp)?;
errs.append(&mut err);
if let Some(leaf) = leaf {
terms.push((occur, leaf.set_field(Some(field_name.clone())).into()));
}
inp = rest;
}
}
fn exists(inp: &str) -> IResult<&str, UserInputLeaf> {
@@ -1427,18 +1468,8 @@ mod test {
#[test]
fn test_parse_query_term_group() {
test_parse_query_to_ast_helper(r#"field:(abc)"#, r#""field":abc"#);
test_parse_query_to_ast_helper(r#"field:(abc)"#, r#"(*"field":abc)"#);
test_parse_query_to_ast_helper(r#"field:(+a -"b c")"#, r#"(+"field":a -"field":"b c")"#);
test_parse_query_to_ast_helper(r#"field:(a AND "b c")"#, r#"(+"field":a +"field":"b c")"#);
test_parse_query_to_ast_helper(r#"field:(a OR "b c")"#, r#"(?"field":a ?"field":"b c")"#);
test_parse_query_to_ast_helper(
r#"field:(a OR (b AND c))"#,
r#"(?"field":a ?(+"field":b +"field":c))"#,
);
test_parse_query_to_ast_helper(
r#"field:(a [b TO c])"#,
r#"(*"field":a *"field":["b" TO "c"])"#,
);
test_is_parse_err(r#"field:(+a -"b c""#, r#"(+"field":a -"field":"b c")"#);
}

30
query-grammar/src/user_input_ast.rs
View File

@@ -44,26 +44,6 @@ impl UserInputLeaf {
},
}
}
pub(crate) fn set_default_field(&mut self, default_field: String) {
match self {
UserInputLeaf::Literal(ref mut literal) if literal.field_name.is_none() => {
literal.field_name = Some(default_field)
}
UserInputLeaf::All => {
*self = UserInputLeaf::Exists {
field: default_field,
}
}
UserInputLeaf::Range { ref mut field, .. } if field.is_none() => {
*field = Some(default_field)
}
UserInputLeaf::Set { ref mut field, .. } if field.is_none() => {
*field = Some(default_field)
}
_ => (), // field was already set, do nothing
}
}
}
impl Debug for UserInputLeaf {
@@ -225,16 +205,6 @@ impl UserInputAst {
pub fn or(asts: Vec<UserInputAst>) -> UserInputAst {
UserInputAst::compose(Occur::Should, asts)
}
pub(crate) fn set_default_field(&mut self, field: String) {
match self {
UserInputAst::Clause(clauses) => clauses
.iter_mut()
.for_each(|(_, ast)| ast.set_default_field(field.clone())),
UserInputAst::Leaf(leaf) => leaf.set_default_field(field),
UserInputAst::Boost(ref mut ast, _) => ast.set_default_field(field),
}
}
}
impl From<UserInputLiteral> for UserInputLeaf {

11
src/aggregation/bucket/mod.rs
View File

@@ -28,7 +28,6 @@ mod term_agg;
mod term_missing_agg;
use std::collections::HashMap;
use std::fmt;
pub use histogram::*;
pub use range::*;
@@ -73,12 +72,12 @@ impl From<&str> for OrderTarget {
}
}
impl fmt::Display for OrderTarget {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
impl ToString for OrderTarget {
fn to_string(&self) -> String {
match self {
OrderTarget::Key => f.write_str("_key"),
OrderTarget::Count => f.write_str("_count"),
OrderTarget::SubAggregation(agg) => agg.fmt(f),
OrderTarget::Key => "_key".to_string(),
OrderTarget::Count => "_count".to_string(),
OrderTarget::SubAggregation(agg) => agg.to_string(),
}
}
}

48
src/aggregation/metric/top_hits.rs
View File

@@ -2,7 +2,6 @@ use std::collections::HashMap;
use std::net::Ipv6Addr;
use columnar::{ColumnarReader, DynamicColumn};
use common::json_path_writer::JSON_PATH_SEGMENT_SEP_STR;
use common::DateTime;
use regex::Regex;
use serde::ser::SerializeMap;
@@ -14,8 +13,8 @@ use crate::aggregation::intermediate_agg_result::{
IntermediateAggregationResult, IntermediateMetricResult,
};
use crate::aggregation::segment_agg_result::SegmentAggregationCollector;
use crate::aggregation::AggregationError;
use crate::collector::TopNComputer;
use crate::schema::term::JSON_PATH_SEGMENT_SEP_STR;
use crate::schema::OwnedValue;
use crate::{DocAddress, DocId, SegmentOrdinal};
@@ -97,14 +96,6 @@ pub struct TopHitsAggregation {
#[serde(rename = "docvalue_fields")]
#[serde(default)]
doc_value_fields: Vec<String>,
// Not supported
_source: Option<serde_json::Value>,
fields: Option<serde_json::Value>,
script_fields: Option<serde_json::Value>,
highlight: Option<serde_json::Value>,
explain: Option<serde_json::Value>,
version: Option<serde_json::Value>,
}
#[derive(Debug, Clone, PartialEq, Default)]
@@ -151,49 +142,12 @@ fn globbed_string_to_regex(glob: &str) -> Result<Regex, crate::TantivyError> {
})
}
fn use_doc_value_fields_err(parameter: &str) -> crate::Result<()> {
Err(crate::TantivyError::AggregationError(
AggregationError::InvalidRequest(format!(
"The `{}` parameter is not supported, only `docvalue_fields` is supported in \
`top_hits` aggregation",
parameter
)),
))
}
fn unsupported_err(parameter: &str) -> crate::Result<()> {
Err(crate::TantivyError::AggregationError(
AggregationError::InvalidRequest(format!(
"The `{}` parameter is not supported in the `top_hits` aggregation",
parameter
)),
))
}
impl TopHitsAggregation {
/// Validate and resolve field retrieval parameters
pub fn validate_and_resolve_field_names(
&mut self,
reader: &ColumnarReader,
) -> crate::Result<()> {
if self._source.is_some() {
use_doc_value_fields_err("_source")?;
}
if self.fields.is_some() {
use_doc_value_fields_err("fields")?;
}
if self.script_fields.is_some() {
use_doc_value_fields_err("script_fields")?;
}
if self.explain.is_some() {
unsupported_err("explain")?;
}
if self.highlight.is_some() {
unsupported_err("highlight")?;
}
if self.version.is_some() {
unsupported_err("version")?;
}
self.doc_value_fields = self
.doc_value_fields
.iter()

2
src/collector/facet_collector.rs
View File

@@ -1,7 +1,7 @@
use std::cmp::Ordering;
use std::collections::{btree_map, BTreeMap, BTreeSet, BinaryHeap};
use std::io;
use std::ops::Bound;
use std::{io, u64, usize};
use crate::collector::{Collector, SegmentCollector};
use crate::fastfield::FacetReader;

373
src/core/json_utils.rs
View File

@@ -1,11 +1,12 @@
use common::json_path_writer::JSON_PATH_SEGMENT_SEP;
use columnar::MonotonicallyMappableToU64;
use common::{replace_in_place, JsonPathWriter};
use rustc_hash::FxHashMap;
use crate::fastfield::FastValue;
use crate::postings::{IndexingContext, IndexingPosition, PostingsWriter};
use crate::schema::document::{ReferenceValue, ReferenceValueLeaf, Value};
use crate::schema::indexing_term::IndexingTerm;
use crate::schema::{Field, Type};
use crate::schema::term::JSON_PATH_SEGMENT_SEP;
use crate::schema::{Field, Type, DATE_TIME_PRECISION_INDEXED};
use crate::time::format_description::well_known::Rfc3339;
use crate::time::{OffsetDateTime, UtcOffset};
use crate::tokenizer::TextAnalyzer;
@@ -75,7 +76,7 @@ pub(crate) fn index_json_values<'a, V: Value<'a>>(
json_visitors: impl Iterator<Item = crate::Result<V::ObjectIter>>,
text_analyzer: &mut TextAnalyzer,
expand_dots_enabled: bool,
term_buffer: &mut IndexingTerm,
term_buffer: &mut Term,
postings_writer: &mut dyn PostingsWriter,
json_path_writer: &mut JsonPathWriter,
ctx: &mut IndexingContext,
@@ -104,7 +105,7 @@ fn index_json_object<'a, V: Value<'a>>(
doc: DocId,
json_visitor: V::ObjectIter,
text_analyzer: &mut TextAnalyzer,
term_buffer: &mut IndexingTerm,
term_buffer: &mut Term,
json_path_writer: &mut JsonPathWriter,
postings_writer: &mut dyn PostingsWriter,
ctx: &mut IndexingContext,
@@ -131,16 +132,19 @@ fn index_json_value<'a, V: Value<'a>>(
doc: DocId,
json_value: V,
text_analyzer: &mut TextAnalyzer,
term_buffer: &mut IndexingTerm,
term_buffer: &mut Term,
json_path_writer: &mut JsonPathWriter,
postings_writer: &mut dyn PostingsWriter,
ctx: &mut IndexingContext,
positions_per_path: &mut IndexingPositionsPerPath,
) {
let set_path_id = |term_buffer: &mut IndexingTerm, unordered_id: u32| {
let set_path_id = |term_buffer: &mut Term, unordered_id: u32| {
term_buffer.truncate_value_bytes(0);
term_buffer.append_bytes(&unordered_id.to_be_bytes());
};
let set_type = |term_buffer: &mut Term, typ: Type| {
term_buffer.append_bytes(&[typ.to_code()]);
};
match json_value.as_value() {
ReferenceValue::Leaf(leaf) => match leaf {
@@ -153,7 +157,7 @@ fn index_json_value<'a, V: Value<'a>>(
// TODO: make sure the chain position works out.
set_path_id(term_buffer, unordered_id);
term_buffer.append_bytes(&[Type::Str.to_code()]);
set_type(term_buffer, Type::Str);
let indexing_position = positions_per_path.get_position_from_id(unordered_id);
postings_writer.index_text(
doc,
@@ -209,16 +213,18 @@ fn index_json_value<'a, V: Value<'a>>(
postings_writer.subscribe(doc, 0u32, term_buffer, ctx);
}
ReferenceValueLeaf::PreTokStr(_) => {
unimplemented!("Pre-tokenized string support in JSON fields is not yet implemented")
unimplemented!(
"Pre-tokenized string support in dynamic fields is not yet implemented"
)
}
ReferenceValueLeaf::Bytes(_) => {
unimplemented!("Bytes support in JSON fields is not yet implemented")
unimplemented!("Bytes support in dynamic fields is not yet implemented")
}
ReferenceValueLeaf::Facet(_) => {
unimplemented!("Facet support in JSON fields is not yet implemented")
unimplemented!("Facet support in dynamic fields is not yet implemented")
}
ReferenceValueLeaf::IpAddr(_) => {
unimplemented!("IP address support in JSON fields is not yet implemented")
unimplemented!("IP address support in dynamic fields is not yet implemented")
}
},
ReferenceValue::Array(elements) => {
@@ -250,43 +256,71 @@ fn index_json_value<'a, V: Value<'a>>(
}
}
/// Tries to infer a JSON type from a string and append it to the term.
///
/// The term must be json + JSON path.
pub fn convert_to_fast_value_and_append_to_json_term(mut term: Term, phrase: &str) -> Option<Term> {
assert_eq!(
term.value()
.as_json()
.expect("expecting a Term with a json type and json path")
.1
.as_serialized()
.len(),
0,
"JSON value bytes should be empty"
);
// Tries to infer a JSON type from a string.
pub fn convert_to_fast_value_and_get_term(
json_term_writer: &mut JsonTermWriter,
phrase: &str,
) -> Option<Term> {
if let Ok(dt) = OffsetDateTime::parse(phrase, &Rfc3339) {
let dt_utc = dt.to_offset(UtcOffset::UTC);
term.append_type_and_fast_value(DateTime::from_utc(dt_utc));
return Some(term);
return Some(set_fastvalue_and_get_term(
json_term_writer,
DateTime::from_utc(dt_utc),
));
}
if let Ok(i64_val) = str::parse::<i64>(phrase) {
term.append_type_and_fast_value(i64_val);
return Some(term);
return Some(set_fastvalue_and_get_term(json_term_writer, i64_val));
}
if let Ok(u64_val) = str::parse::<u64>(phrase) {
term.append_type_and_fast_value(u64_val);
return Some(term);
return Some(set_fastvalue_and_get_term(json_term_writer, u64_val));
}
if let Ok(f64_val) = str::parse::<f64>(phrase) {
term.append_type_and_fast_value(f64_val);
return Some(term);
return Some(set_fastvalue_and_get_term(json_term_writer, f64_val));
}
if let Ok(bool_val) = str::parse::<bool>(phrase) {
term.append_type_and_fast_value(bool_val);
return Some(term);
return Some(set_fastvalue_and_get_term(json_term_writer, bool_val));
}
None
}
// helper function to generate a Term from a json fastvalue
pub(crate) fn set_fastvalue_and_get_term<T: FastValue>(
json_term_writer: &mut JsonTermWriter,
value: T,
) -> Term {
json_term_writer.set_fast_value(value);
json_term_writer.term().clone()
}
// helper function to generate a list of terms with their positions from a textual json value
pub(crate) fn set_string_and_get_terms(
json_term_writer: &mut JsonTermWriter,
value: &str,
text_analyzer: &mut TextAnalyzer,
) -> Vec<(usize, Term)> {
let mut positions_and_terms = Vec::<(usize, Term)>::new();
json_term_writer.close_path_and_set_type(Type::Str);
let term_num_bytes = json_term_writer.term_buffer.len_bytes();
let mut token_stream = text_analyzer.token_stream(value);
token_stream.process(&mut |token| {
json_term_writer
.term_buffer
.truncate_value_bytes(term_num_bytes);
json_term_writer
.term_buffer
.append_bytes(token.text.as_bytes());
positions_and_terms.push((token.position, json_term_writer.term().clone()));
});
positions_and_terms
}
/// Writes a value of a JSON field to a `Term`.
/// The Term format is as follows:
/// `[JSON_TYPE][JSON_PATH][JSON_END_OF_PATH][VALUE_BYTES]`
pub struct JsonTermWriter<'a> {
term_buffer: &'a mut Term,
path_stack: Vec<usize>,
expand_dots_enabled: bool,
}
/// Splits a json path supplied to the query parser in such a way that
/// `.` can be escaped.
@@ -343,106 +377,275 @@ pub(crate) fn encode_column_name(
path.into()
}
pub fn term_from_json_paths<'a>(
json_field: Field,
paths: impl Iterator<Item = &'a str>,
expand_dots_enabled: bool,
) -> Term {
let mut json_path = JsonPathWriter::with_expand_dots(expand_dots_enabled);
for path in paths {
json_path.push(path);
impl<'a> JsonTermWriter<'a> {
pub fn from_field_and_json_path(
field: Field,
json_path: &str,
expand_dots_enabled: bool,
term_buffer: &'a mut Term,
) -> Self {
term_buffer.set_field_and_type(field, Type::Json);
let mut json_term_writer = Self::wrap(term_buffer, expand_dots_enabled);
for segment in split_json_path(json_path) {
json_term_writer.push_path_segment(&segment);
}
json_term_writer
}
json_path.set_end();
let mut term = Term::with_type_and_field(Type::Json, json_field);
term.append_bytes(json_path.as_str().as_bytes());
term
pub fn wrap(term_buffer: &'a mut Term, expand_dots_enabled: bool) -> Self {
term_buffer.clear_with_type(Type::Json);
let mut path_stack = Vec::with_capacity(10);
path_stack.push(0);
Self {
term_buffer,
path_stack,
expand_dots_enabled,
}
}
fn trim_to_end_of_path(&mut self) {
let end_of_path = *self.path_stack.last().unwrap();
self.term_buffer.truncate_value_bytes(end_of_path);
}
pub fn close_path_and_set_type(&mut self, typ: Type) {
self.trim_to_end_of_path();
self.term_buffer.set_json_path_end();
self.term_buffer.append_bytes(&[typ.to_code()]);
}
// TODO: Remove this function and use JsonPathWriter instead.
pub fn push_path_segment(&mut self, segment: &str) {
// the path stack should never be empty.
self.trim_to_end_of_path();
if self.path_stack.len() > 1 {
self.term_buffer.set_json_path_separator();
}
let appended_segment = self.term_buffer.append_bytes(segment.as_bytes());
if self.expand_dots_enabled {
// We need to replace `.` by JSON_PATH_SEGMENT_SEP.
replace_in_place(b'.', JSON_PATH_SEGMENT_SEP, appended_segment);
}
self.term_buffer.add_json_path_separator();
self.path_stack.push(self.term_buffer.len_bytes());
}
pub fn pop_path_segment(&mut self) {
self.path_stack.pop();
assert!(!self.path_stack.is_empty());
self.trim_to_end_of_path();
}
/// Returns the json path of the term being currently built.
#[cfg(test)]
pub(crate) fn path(&self) -> &[u8] {
let end_of_path = self.path_stack.last().cloned().unwrap_or(1);
&self.term().serialized_value_bytes()[..end_of_path - 1]
}
pub(crate) fn set_fast_value<T: FastValue>(&mut self, val: T) {
self.close_path_and_set_type(T::to_type());
let value = if T::to_type() == Type::Date {
DateTime::from_u64(val.to_u64())
.truncate(DATE_TIME_PRECISION_INDEXED)
.to_u64()
} else {
val.to_u64()
};
self.term_buffer
.append_bytes(value.to_be_bytes().as_slice());
}
pub fn set_str(&mut self, text: &str) {
self.close_path_and_set_type(Type::Str);
self.term_buffer.append_bytes(text.as_bytes());
}
pub fn term(&self) -> &Term {
self.term_buffer
}
}
#[cfg(test)]
mod tests {
use super::split_json_path;
use crate::json_utils::term_from_json_paths;
use crate::schema::Field;
use super::{split_json_path, JsonTermWriter};
use crate::schema::{Field, Type};
use crate::Term;
#[test]
fn test_json_writer() {
let field = Field::from_field_id(1);
let mut term = term_from_json_paths(field, ["attributes", "color"].into_iter(), false);
term.append_type_and_str("red");
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_writer = JsonTermWriter::wrap(&mut term, false);
json_writer.push_path_segment("attributes");
json_writer.push_path_segment("color");
json_writer.set_str("red");
assert_eq!(
format!("{:?}", term),
format!("{:?}", json_writer.term()),
"Term(field=1, type=Json, path=attributes.color, type=Str, \"red\")"
);
let mut term = term_from_json_paths(
field,
["attributes", "dimensions", "width"].into_iter(),
false,
);
term.append_type_and_fast_value(400i64);
json_writer.set_str("blue");
assert_eq!(
format!("{:?}", term),
format!("{:?}", json_writer.term()),
"Term(field=1, type=Json, path=attributes.color, type=Str, \"blue\")"
);
json_writer.pop_path_segment();
json_writer.push_path_segment("dimensions");
json_writer.push_path_segment("width");
json_writer.set_fast_value(400i64);
assert_eq!(
format!("{:?}", json_writer.term()),
"Term(field=1, type=Json, path=attributes.dimensions.width, type=I64, 400)"
);
json_writer.pop_path_segment();
json_writer.push_path_segment("height");
json_writer.set_fast_value(300i64);
assert_eq!(
format!("{:?}", json_writer.term()),
"Term(field=1, type=Json, path=attributes.dimensions.height, type=I64, 300)"
);
}
#[test]
fn test_string_term() {
let field = Field::from_field_id(1);
let mut term = term_from_json_paths(field, ["color"].into_iter(), false);
term.append_type_and_str("red");
assert_eq!(term.serialized_term(), b"\x00\x00\x00\x01jcolor\x00sred")
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_writer = JsonTermWriter::wrap(&mut term, false);
json_writer.push_path_segment("color");
json_writer.set_str("red");
assert_eq!(
json_writer.term().serialized_term(),
b"\x00\x00\x00\x01jcolor\x00sred"
)
}
#[test]
fn test_i64_term() {
let field = Field::from_field_id(1);
let mut term = term_from_json_paths(field, ["color"].into_iter(), false);
term.append_type_and_fast_value(-4i64);
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_writer = JsonTermWriter::wrap(&mut term, false);
json_writer.push_path_segment("color");
json_writer.set_fast_value(-4i64);
assert_eq!(
term.value().as_serialized(),
b"jcolor\x00i\x7f\xff\xff\xff\xff\xff\xff\xfc"
json_writer.term().serialized_term(),
b"\x00\x00\x00\x01jcolor\x00i\x7f\xff\xff\xff\xff\xff\xff\xfc"
)
}
#[test]
fn test_u64_term() {
let field = Field::from_field_id(1);
let mut term = term_from_json_paths(field, ["color"].into_iter(), false);
term.append_type_and_fast_value(4u64);
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_writer = JsonTermWriter::wrap(&mut term, false);
json_writer.push_path_segment("color");
json_writer.set_fast_value(4u64);
assert_eq!(
term.value().as_serialized(),
b"jcolor\x00u\x00\x00\x00\x00\x00\x00\x00\x04"
json_writer.term().serialized_term(),
b"\x00\x00\x00\x01jcolor\x00u\x00\x00\x00\x00\x00\x00\x00\x04"
)
}
#[test]
fn test_f64_term() {
let field = Field::from_field_id(1);
let mut term = term_from_json_paths(field, ["color"].into_iter(), false);
term.append_type_and_fast_value(4.0f64);
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_writer = JsonTermWriter::wrap(&mut term, false);
json_writer.push_path_segment("color");
json_writer.set_fast_value(4.0f64);
assert_eq!(
term.value().as_serialized(),
b"jcolor\x00f\xc0\x10\x00\x00\x00\x00\x00\x00"
json_writer.term().serialized_term(),
b"\x00\x00\x00\x01jcolor\x00f\xc0\x10\x00\x00\x00\x00\x00\x00"
)
}
#[test]
fn test_bool_term() {
let field = Field::from_field_id(1);
let mut term = term_from_json_paths(field, ["color"].into_iter(), false);
term.append_type_and_fast_value(true);
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_writer = JsonTermWriter::wrap(&mut term, false);
json_writer.push_path_segment("color");
json_writer.set_fast_value(true);
assert_eq!(
term.value().as_serialized(),
b"jcolor\x00o\x00\x00\x00\x00\x00\x00\x00\x01"
json_writer.term().serialized_term(),
b"\x00\x00\x00\x01jcolor\x00o\x00\x00\x00\x00\x00\x00\x00\x01"
)
}
#[test]
fn test_push_after_set_path_segment() {
let field = Field::from_field_id(1);
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_writer = JsonTermWriter::wrap(&mut term, false);
json_writer.push_path_segment("attribute");
json_writer.set_str("something");
json_writer.push_path_segment("color");
json_writer.set_str("red");
assert_eq!(
json_writer.term().serialized_term(),
b"\x00\x00\x00\x01jattribute\x01color\x00sred"
)
}
#[test]
fn test_pop_segment() {
let field = Field::from_field_id(1);
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_writer = JsonTermWriter::wrap(&mut term, false);
json_writer.push_path_segment("color");
json_writer.push_path_segment("hue");
json_writer.pop_path_segment();
json_writer.set_str("red");
assert_eq!(
json_writer.term().serialized_term(),
b"\x00\x00\x00\x01jcolor\x00sred"
)
}
#[test]
fn test_json_writer_path() {
let field = Field::from_field_id(1);
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_writer = JsonTermWriter::wrap(&mut term, false);
json_writer.push_path_segment("color");
assert_eq!(json_writer.path(), b"color");
json_writer.push_path_segment("hue");
assert_eq!(json_writer.path(), b"color\x01hue");
json_writer.set_str("pink");
assert_eq!(json_writer.path(), b"color\x01hue");
}
#[test]
fn test_json_path_expand_dots_disabled() {
let field = Field::from_field_id(1);
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_writer = JsonTermWriter::wrap(&mut term, false);
json_writer.push_path_segment("color.hue");
assert_eq!(json_writer.path(), b"color.hue");
}
#[test]
fn test_json_path_expand_dots_enabled() {
let field = Field::from_field_id(1);
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_writer = JsonTermWriter::wrap(&mut term, true);
json_writer.push_path_segment("color.hue");
assert_eq!(json_writer.path(), b"color\x01hue");
}
#[test]
fn test_json_path_expand_dots_enabled_pop_segment() {
let field = Field::from_field_id(1);
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_writer = JsonTermWriter::wrap(&mut term, true);
json_writer.push_path_segment("hello");
assert_eq!(json_writer.path(), b"hello");
json_writer.push_path_segment("color.hue");
assert_eq!(json_writer.path(), b"hello\x01color\x01hue");
json_writer.pop_path_segment();
assert_eq!(json_writer.path(), b"hello");
}
#[test]
fn test_split_json_path_simple() {
let json_path = split_json_path("titi.toto");

32
src/core/tests.rs
View File

@@ -1,9 +1,9 @@
use crate::collector::Count;
use crate::directory::{RamDirectory, WatchCallback};
use crate::indexer::{LogMergePolicy, NoMergePolicy};
use crate::json_utils::term_from_json_paths;
use crate::json_utils::JsonTermWriter;
use crate::query::TermQuery;
use crate::schema::{Field, IndexRecordOption, Schema, INDEXED, STRING, TEXT};
use crate::schema::{Field, IndexRecordOption, Schema, Type, INDEXED, STRING, TEXT};
use crate::tokenizer::TokenizerManager;
use crate::{
Directory, DocSet, Index, IndexBuilder, IndexReader, IndexSettings, IndexWriter, Postings,
@@ -416,12 +416,16 @@ fn test_non_text_json_term_freq() {
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(0u32);
let inv_idx = segment_reader.inverted_index(field).unwrap();
let mut term = term_from_json_paths(field, ["tenant_id"].iter().cloned(), false);
term.append_type_and_fast_value(75u64);
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_term_writer = JsonTermWriter::wrap(&mut term, false);
json_term_writer.push_path_segment("tenant_id");
json_term_writer.close_path_and_set_type(Type::U64);
json_term_writer.set_fast_value(75u64);
let postings = inv_idx
.read_postings(&term, IndexRecordOption::WithFreqsAndPositions)
.read_postings(
json_term_writer.term(),
IndexRecordOption::WithFreqsAndPositions,
)
.unwrap()
.unwrap();
assert_eq!(postings.doc(), 0);
@@ -450,12 +454,16 @@ fn test_non_text_json_term_freq_bitpacked() {
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(0u32);
let inv_idx = segment_reader.inverted_index(field).unwrap();
let mut term = term_from_json_paths(field, ["tenant_id"].iter().cloned(), false);
term.append_type_and_fast_value(75u64);
let mut term = Term::with_type_and_field(Type::Json, field);
let mut json_term_writer = JsonTermWriter::wrap(&mut term, false);
json_term_writer.push_path_segment("tenant_id");
json_term_writer.close_path_and_set_type(Type::U64);
json_term_writer.set_fast_value(75u64);
let mut postings = inv_idx
.read_postings(&term, IndexRecordOption::WithFreqsAndPositions)
.read_postings(
json_term_writer.term(),
IndexRecordOption::WithFreqsAndPositions,
)
.unwrap()
.unwrap();
assert_eq!(postings.doc(), 0);

2
src/functional_test.rs
View File

@@ -1,5 +1,3 @@
#![allow(deprecated)] // Remove with index sorting
use std::collections::HashSet;
use rand::{thread_rng, Rng};

18
src/index/index.rs
View File

@@ -20,7 +20,7 @@ use crate::indexer::segment_updater::save_metas;
use crate::indexer::{IndexWriter, SingleSegmentIndexWriter};
use crate::reader::{IndexReader, IndexReaderBuilder};
use crate::schema::document::Document;
use crate::schema::{Field, FieldType, Schema, Type};
use crate::schema::{Field, FieldType, Schema};
use crate::tokenizer::{TextAnalyzer, TokenizerManager};
use crate::SegmentReader;
@@ -83,7 +83,7 @@ fn save_new_metas(
///
/// ```
/// use tantivy::schema::*;
/// use tantivy::{Index, IndexSettings};
/// use tantivy::{Index, IndexSettings, IndexSortByField, Order};
///
/// let mut schema_builder = Schema::builder();
/// let id_field = schema_builder.add_text_field("id", STRING);
@@ -96,7 +96,10 @@ fn save_new_metas(
///
/// let schema = schema_builder.build();
/// let settings = IndexSettings{
/// docstore_blocksize: 100_000,
/// sort_by_field: Some(IndexSortByField{
/// field: "number".to_string(),
/// order: Order::Asc
/// }),
/// ..Default::default()
/// };
/// let index = Index::builder().schema(schema).settings(settings).create_in_ram();
@@ -248,15 +251,6 @@ impl IndexBuilder {
sort_by_field.field
)));
}
let supported_field_types = [Type::I64, Type::U64, Type::F64, Type::Date];
let field_type = entry.field_type().value_type();
if !supported_field_types.contains(&field_type) {
return Err(TantivyError::InvalidArgument(format!(
"Unsupported field type in sort_by_field: {:?}. Supported field types: \
{:?} ",
field_type, supported_field_types,
)));
}
}
Ok(())
} else {

4
src/index/index_meta.rs
View File

@@ -288,10 +288,6 @@ impl Default for IndexSettings {
/// Presorting documents can greatly improve performance
/// in some scenarios, by applying top n
/// optimizations.
#[deprecated(
since = "0.22.0",
note = "We plan to remove index sorting in `0.23`. If you need index sorting, please comment on the related issue https://github.com/quickwit-oss/tantivy/issues/2352 and explain your use case."
)]
#[derive(Clone, Debug, Serialize, Deserialize, Eq, PartialEq)]
pub struct IndexSortByField {
/// The field to sort the documents by

3
src/index/inverted_index_reader.rs
View File

@@ -1,13 +1,12 @@
use std::io;
use common::json_path_writer::JSON_END_OF_PATH;
use common::BinarySerializable;
use fnv::FnvHashSet;
use crate::directory::FileSlice;
use crate::positions::PositionReader;
use crate::postings::{BlockSegmentPostings, SegmentPostings, TermInfo};
use crate::schema::{IndexRecordOption, Term, Type};
use crate::schema::{IndexRecordOption, Term, Type, JSON_END_OF_PATH};
use crate::termdict::TermDictionary;
/// The inverted index reader is in charge of accessing

2
src/index/segment_reader.rs
View File

@@ -406,7 +406,7 @@ impl SegmentReader {
}
/// Returns an iterator that will iterate over the alive document ids
pub fn doc_ids_alive(&self) -> Box<dyn Iterator<Item = DocId> + Send + '_> {
pub fn doc_ids_alive(&self) -> Box<dyn Iterator<Item = DocId> + '_> {
if let Some(alive_bitset) = &self.alive_bitset_opt {
Box::new(alive_bitset.iter_alive())
} else {

26
src/indexer/doc_id_mapping.rs
View File

@@ -159,7 +159,7 @@ mod tests_indexsorting {
use crate::indexer::NoMergePolicy;
use crate::query::QueryParser;
use crate::schema::*;
use crate::{DocAddress, Index, IndexBuilder, IndexSettings, IndexSortByField, Order};
use crate::{DocAddress, Index, IndexSettings, IndexSortByField, Order};
fn create_test_index(
index_settings: Option<IndexSettings>,
@@ -557,28 +557,4 @@ mod tests_indexsorting {
&[2000, 8000, 3000]
);
}
#[test]
fn test_text_sort() -> crate::Result<()> {
let mut schema_builder = SchemaBuilder::new();
schema_builder.add_text_field("id", STRING | FAST | STORED);
schema_builder.add_text_field("name", TEXT | STORED);
let resp = IndexBuilder::new()
.schema(schema_builder.build())
.settings(IndexSettings {
sort_by_field: Some(IndexSortByField {
field: "id".to_string(),
order: Order::Asc,
}),
..Default::default()
})
.create_in_ram();
assert!(resp
.unwrap_err()
.to_string()
.contains("Unsupported field type"));
Ok(())
}
}

18
src/indexer/mod.rs
View File

@@ -156,24 +156,16 @@ mod tests_mmap {
}
#[test]
fn test_json_field_1byte() {
// Test when field name contains a '1' byte, which has special meaning in tantivy.
// The 1 byte can be addressed as '1' byte or '.'.
// Test when field name contains a 1 byte, which has special meaning in tantivy.
let field_name_in = "\u{0001}";
let field_name_out = "\u{0001}";
test_json_field_name(field_name_in, field_name_out);
// Test when field name contains a '1' byte, which has special meaning in tantivy.
// Test when field name contains a 1 byte, which has special meaning in tantivy.
let field_name_in = "\u{0001}";
let field_name_out = ".";
test_json_field_name(field_name_in, field_name_out);
}
#[test]
fn test_json_field_dot() {
// Test when field name contains a '.'
let field_name_in = ".";
let field_name_out = ".";
test_json_field_name(field_name_in, field_name_out);
}
fn test_json_field_name(field_name_in: &str, field_name_out: &str) {
let mut schema_builder = Schema::builder();
@@ -213,10 +205,10 @@ mod tests_mmap {
let reader = index.reader().unwrap();
let searcher = reader.searcher();
let parse_query = QueryParser::for_index(&index, Vec::new());
let test_query = |query_str: &str| {
let query = parse_query.parse_query(query_str).unwrap();
let test_query = |field_name: &str| {
let query = parse_query.parse_query(field_name).unwrap();
let num_docs = searcher.search(&query, &Count).unwrap();
assert_eq!(num_docs, 1, "{}", query_str);
assert_eq!(num_docs, 1);
};
test_query(format!("json.{field_name_out}:test1").as_str());
test_query(format!("json.a{field_name_out}:test2").as_str());

179
src/indexer/segment_writer.rs
View File

@@ -1,3 +1,4 @@
use columnar::MonotonicallyMappableToU64;
use common::JsonPathWriter;
use itertools::Itertools;
use tokenizer_api::BoxTokenStream;
@@ -14,8 +15,7 @@ use crate::postings::{
PerFieldPostingsWriter, PostingsWriter,
};
use crate::schema::document::{Document, ReferenceValue, Value};
use crate::schema::indexing_term::IndexingTerm;
use crate::schema::{FieldEntry, FieldType, Schema};
use crate::schema::{FieldEntry, FieldType, Schema, Term, DATE_TIME_PRECISION_INDEXED};
use crate::store::{StoreReader, StoreWriter};
use crate::tokenizer::{FacetTokenizer, PreTokenizedStream, TextAnalyzer, Tokenizer};
use crate::{DocId, Opstamp, SegmentComponent, TantivyError};
@@ -70,7 +70,7 @@ pub struct SegmentWriter {
pub(crate) json_path_writer: JsonPathWriter,
pub(crate) doc_opstamps: Vec<Opstamp>,
per_field_text_analyzers: Vec<TextAnalyzer>,
term_buffer: IndexingTerm,
term_buffer: Term,
schema: Schema,
}
@@ -126,7 +126,7 @@ impl SegmentWriter {
)?,
doc_opstamps: Vec::with_capacity(1_000),
per_field_text_analyzers,
term_buffer: IndexingTerm::new(),
term_buffer: Term::with_capacity(16),
schema,
})
}
@@ -195,7 +195,7 @@ impl SegmentWriter {
let (term_buffer, ctx) = (&mut self.term_buffer, &mut self.ctx);
let postings_writer: &mut dyn PostingsWriter =
self.per_field_postings_writers.get_for_field_mut(field);
term_buffer.clear_with_field(field);
term_buffer.clear_with_field_and_type(field_entry.field_type().value_type(), field);
match field_entry.field_type() {
FieldType::Facet(_) => {
@@ -271,7 +271,8 @@ impl SegmentWriter {
num_vals += 1;
let date_val = value.as_datetime().ok_or_else(make_schema_error)?;
term_buffer.set_date(date_val);
term_buffer
.set_u64(date_val.truncate(DATE_TIME_PRECISION_INDEXED).to_u64());
postings_writer.subscribe(doc_id, 0u32, term_buffer, ctx);
}
if field_entry.has_fieldnorms() {
@@ -331,7 +332,7 @@ impl SegmentWriter {
num_vals += 1;
let bytes = value.as_bytes().ok_or_else(make_schema_error)?;
term_buffer.set_value_bytes(bytes);
term_buffer.set_bytes(bytes);
postings_writer.subscribe(doc_id, 0u32, term_buffer, ctx);
}
if field_entry.has_fieldnorms() {
@@ -495,14 +496,14 @@ mod tests {
use tempfile::TempDir;
use crate::collector::{Count, TopDocs};
use crate::core::json_utils::JsonTermWriter;
use crate::directory::RamDirectory;
use crate::fastfield::FastValue;
use crate::json_utils::term_from_json_paths;
use crate::postings::TermInfo;
use crate::query::{PhraseQuery, QueryParser};
use crate::schema::document::Value;
use crate::schema::{
Document, IndexRecordOption, Schema, TextFieldIndexing, TextOptions, STORED, STRING, TEXT,
Document, IndexRecordOption, Schema, TextFieldIndexing, TextOptions, Type, STORED, STRING,
TEXT,
};
use crate::store::{Compressor, StoreReader, StoreWriter};
use crate::time::format_description::well_known::Rfc3339;
@@ -644,117 +645,115 @@ mod tests {
let inv_idx = segment_reader.inverted_index(json_field).unwrap();
let term_dict = inv_idx.terms();
let mut term = Term::with_type_and_field(Type::Json, json_field);
let mut term_stream = term_dict.stream().unwrap();
let term_from_path = |paths: &[&str]| -> Term {
term_from_json_paths(json_field, paths.iter().cloned(), false)
};
let mut json_term_writer = JsonTermWriter::wrap(&mut term, false);
fn set_fast_val<T: FastValue>(val: T, mut term: Term) -> Term {
term.append_type_and_fast_value(val);
term
}
fn set_str(val: &str, mut term: Term) -> Term {
term.append_type_and_str(val);
term
}
let term = term_from_path(&["bool"]);
json_term_writer.push_path_segment("bool");
json_term_writer.set_fast_value(true);
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
set_fast_val(true, term).serialized_value_bytes()
json_term_writer.term().serialized_value_bytes()
);
let term = term_from_path(&["complexobject", "field.with.dot"]);
json_term_writer.pop_path_segment();
json_term_writer.push_path_segment("complexobject");
json_term_writer.push_path_segment("field.with.dot");
json_term_writer.set_fast_value(1i64);
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
set_fast_val(1i64, term).serialized_value_bytes()
json_term_writer.term().serialized_value_bytes()
);
// Date
let term = term_from_path(&["date"]);
json_term_writer.pop_path_segment();
json_term_writer.pop_path_segment();
json_term_writer.push_path_segment("date");
json_term_writer.set_fast_value(DateTime::from_utc(
OffsetDateTime::parse("1985-04-12T23:20:50.52Z", &Rfc3339).unwrap(),
));
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
set_fast_val(
DateTime::from_utc(
OffsetDateTime::parse("1985-04-12T23:20:50.52Z", &Rfc3339).unwrap(),
),
term
)
.serialized_value_bytes()
json_term_writer.term().serialized_value_bytes()
);
// Float
let term = term_from_path(&["float"]);
json_term_writer.pop_path_segment();
json_term_writer.push_path_segment("float");
json_term_writer.set_fast_value(-0.2f64);
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
set_fast_val(-0.2f64, term).serialized_value_bytes()
json_term_writer.term().serialized_value_bytes()
);
// Number In Array
let term = term_from_path(&["my_arr"]);
json_term_writer.pop_path_segment();
json_term_writer.push_path_segment("my_arr");
json_term_writer.set_fast_value(2i64);
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
set_fast_val(2i64, term).serialized_value_bytes()
json_term_writer.term().serialized_value_bytes()
);
let term = term_from_path(&["my_arr"]);
json_term_writer.set_fast_value(3i64);
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
set_fast_val(3i64, term).serialized_value_bytes()
json_term_writer.term().serialized_value_bytes()
);
let term = term_from_path(&["my_arr"]);
json_term_writer.set_fast_value(4i64);
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
set_fast_val(4i64, term).serialized_value_bytes()
json_term_writer.term().serialized_value_bytes()
);
// El in Array
let term = term_from_path(&["my_arr", "my_key"]);
json_term_writer.push_path_segment("my_key");
json_term_writer.set_str("tokens");
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
set_str("tokens", term).serialized_value_bytes()
);
let term = term_from_path(&["my_arr", "my_key"]);
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
set_str("two", term).serialized_value_bytes()
json_term_writer.term().serialized_value_bytes()
);
// Signed
let term = term_from_path(&["signed"]);
json_term_writer.set_str("two");
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
set_fast_val(-2i64, term).serialized_value_bytes()
json_term_writer.term().serialized_value_bytes()
);
let term = term_from_path(&["toto"]);
json_term_writer.pop_path_segment();
json_term_writer.pop_path_segment();
json_term_writer.push_path_segment("signed");
json_term_writer.set_fast_value(-2i64);
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
set_str("titi", term).serialized_value_bytes()
);
// Unsigned
let term = term_from_path(&["unsigned"]);
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
set_fast_val(1i64, term).serialized_value_bytes()
json_term_writer.term().serialized_value_bytes()
);
json_term_writer.pop_path_segment();
json_term_writer.push_path_segment("toto");
json_term_writer.set_str("titi");
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
json_term_writer.term().serialized_value_bytes()
);
json_term_writer.pop_path_segment();
json_term_writer.push_path_segment("unsigned");
json_term_writer.set_fast_value(1i64);
assert!(term_stream.advance());
assert_eq!(
term_stream.key(),
json_term_writer.term().serialized_value_bytes()
);
assert!(!term_stream.advance());
}
@@ -775,9 +774,14 @@ mod tests {
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(0u32);
let inv_index = segment_reader.inverted_index(json_field).unwrap();
let mut term = term_from_json_paths(json_field, ["mykey"].into_iter(), false);
term.append_type_and_str("token");
let term_info = inv_index.get_term_info(&term).unwrap().unwrap();
let mut term = Term::with_type_and_field(Type::Json, json_field);
let mut json_term_writer = JsonTermWriter::wrap(&mut term, false);
json_term_writer.push_path_segment("mykey");
json_term_writer.set_str("token");
let term_info = inv_index
.get_term_info(json_term_writer.term())
.unwrap()
.unwrap();
assert_eq!(
term_info,
TermInfo {
@@ -814,9 +818,14 @@ mod tests {
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(0u32);
let inv_index = segment_reader.inverted_index(json_field).unwrap();
let mut term = term_from_json_paths(json_field, ["mykey"].into_iter(), false);
term.append_type_and_str("two tokens");
let term_info = inv_index.get_term_info(&term).unwrap().unwrap();
let mut term = Term::with_type_and_field(Type::Json, json_field);
let mut json_term_writer = JsonTermWriter::wrap(&mut term, false);
json_term_writer.push_path_segment("mykey");
json_term_writer.set_str("two tokens");
let term_info = inv_index
.get_term_info(json_term_writer.term())
.unwrap()
.unwrap();
assert_eq!(
term_info,
TermInfo {
@@ -854,18 +863,16 @@ mod tests {
writer.commit().unwrap();
let reader = index.reader().unwrap();
let searcher = reader.searcher();
let term = term_from_json_paths(json_field, ["mykey", "field"].into_iter(), false);
let mut hello_term = term.clone();
hello_term.append_type_and_str("hello");
let mut nothello_term = term.clone();
nothello_term.append_type_and_str("nothello");
let mut happy_term = term.clone();
happy_term.append_type_and_str("happy");
let mut term = Term::with_type_and_field(Type::Json, json_field);
let mut json_term_writer = JsonTermWriter::wrap(&mut term, false);
json_term_writer.push_path_segment("mykey");
json_term_writer.push_path_segment("field");
json_term_writer.set_str("hello");
let hello_term = json_term_writer.term().clone();
json_term_writer.set_str("nothello");
let nothello_term = json_term_writer.term().clone();
json_term_writer.set_str("happy");
let happy_term = json_term_writer.term().clone();
let phrase_query = PhraseQuery::new(vec![hello_term, happy_term.clone()]);
assert_eq!(searcher.search(&phrase_query, &Count).unwrap(), 1);
let phrase_query = PhraseQuery::new(vec![nothello_term, happy_term]);

14
src/lib.rs
View File

@@ -178,7 +178,6 @@ pub use crate::future_result::FutureResult;
pub type Result<T> = std::result::Result<T, TantivyError>;
mod core;
#[allow(deprecated)] // Remove with index sorting
pub mod indexer;
#[allow(unused_doc_comments)]
@@ -190,7 +189,6 @@ pub mod collector;
pub mod directory;
pub mod fastfield;
pub mod fieldnorm;
#[allow(deprecated)] // Remove with index sorting
pub mod index;
pub mod positions;
pub mod postings;
@@ -225,7 +223,6 @@ pub use self::snippet::{Snippet, SnippetGenerator};
pub use crate::core::json_utils;
pub use crate::core::{Executor, Searcher, SearcherGeneration};
pub use crate::directory::Directory;
#[allow(deprecated)] // Remove with index sorting
pub use crate::index::{
Index, IndexBuilder, IndexMeta, IndexSettings, IndexSortByField, InvertedIndexReader, Order,
Segment, SegmentComponent, SegmentId, SegmentMeta, SegmentReader,
@@ -237,6 +234,8 @@ pub use crate::index::{
pub use crate::indexer::PreparedCommit;
pub use crate::indexer::{IndexWriter, SingleSegmentIndexWriter};
pub use crate::postings::Postings;
#[allow(deprecated)]
pub use crate::schema::DatePrecision;
pub use crate::schema::{DateOptions, DateTimePrecision, Document, TantivyDocument, Term};
/// Index format version.
@@ -255,7 +254,7 @@ pub struct Version {
impl fmt::Debug for Version {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
fmt::Display::fmt(self, f)
write!(f, "{}", self.to_string())
}
}
@@ -266,10 +265,9 @@ static VERSION: Lazy<Version> = Lazy::new(|| Version {
index_format_version: INDEX_FORMAT_VERSION,
});
impl fmt::Display for Version {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
write!(
f,
impl ToString for Version {
fn to_string(&self) -> String {
format!(
"tantivy v{}.{}.{}, index_format v{}",
self.major, self.minor, self.patch, self.index_format_version
)

97
src/postings/json_postings_writer.rs
View File

@@ -1,6 +1,5 @@
use std::io;
use common::json_path_writer::JSON_END_OF_PATH;
use stacker::Addr;
use crate::indexer::doc_id_mapping::DocIdMapping;
@@ -8,10 +7,9 @@ use crate::indexer::path_to_unordered_id::OrderedPathId;
use crate::postings::postings_writer::SpecializedPostingsWriter;
use crate::postings::recorder::{BufferLender, DocIdRecorder, Recorder};
use crate::postings::{FieldSerializer, IndexingContext, IndexingPosition, PostingsWriter};
use crate::schema::indexing_term::IndexingTerm;
use crate::schema::{Field, Type, ValueBytes};
use crate::schema::{Field, Type, JSON_END_OF_PATH};
use crate::tokenizer::TokenStream;
use crate::DocId;
use crate::{DocId, Term};
/// The `JsonPostingsWriter` is odd in that it relies on a hidden contract:
///
@@ -35,7 +33,7 @@ impl<Rec: Recorder> PostingsWriter for JsonPostingsWriter<Rec> {
&mut self,
doc: crate::DocId,
pos: u32,
term: &IndexingTerm,
term: &crate::Term,
ctx: &mut IndexingContext,
) {
self.non_str_posting_writer.subscribe(doc, pos, term, ctx);
@@ -45,7 +43,7 @@ impl<Rec: Recorder> PostingsWriter for JsonPostingsWriter<Rec> {
&mut self,
doc_id: DocId,
token_stream: &mut dyn TokenStream,
term_buffer: &mut IndexingTerm,
term_buffer: &mut Term,
ctx: &mut IndexingContext,
indexing_position: &mut IndexingPosition,
) {
@@ -67,40 +65,42 @@ impl<Rec: Recorder> PostingsWriter for JsonPostingsWriter<Rec> {
ctx: &IndexingContext,
serializer: &mut FieldSerializer,
) -> io::Result<()> {
let mut term_buffer = JsonTermSerializer(Vec::with_capacity(48));
let mut term_buffer = Term::with_capacity(48);
let mut buffer_lender = BufferLender::default();
term_buffer.clear_with_field_and_type(Type::Json, Field::from_field_id(0));
let mut prev_term_id = u32::MAX;
let mut term_path_len = 0; // this will be set in the first iteration
for (_field, path_id, term, addr) in term_addrs {
if prev_term_id != path_id.path_id() {
term_buffer.clear();
term_buffer.truncate_value_bytes(0);
term_buffer.append_path(ordered_id_to_path[path_id.path_id() as usize].as_bytes());
term_buffer.append_bytes(&[JSON_END_OF_PATH]);
term_path_len = term_buffer.len();
term_path_len = term_buffer.len_bytes();
prev_term_id = path_id.path_id();
}
term_buffer.truncate(term_path_len);
term_buffer.truncate_value_bytes(term_path_len);
term_buffer.append_bytes(term);
let json_value = ValueBytes::wrap(term);
let typ = json_value.typ();
if typ == Type::Str {
SpecializedPostingsWriter::<Rec>::serialize_one_term(
term_buffer.as_bytes(),
*addr,
doc_id_map,
&mut buffer_lender,
ctx,
serializer,
)?;
} else {
SpecializedPostingsWriter::<DocIdRecorder>::serialize_one_term(
term_buffer.as_bytes(),
*addr,
doc_id_map,
&mut buffer_lender,
ctx,
serializer,
)?;
if let Some(json_value) = term_buffer.value().as_json_value_bytes() {
let typ = json_value.typ();
if typ == Type::Str {
SpecializedPostingsWriter::<Rec>::serialize_one_term(
term_buffer.serialized_value_bytes(),
*addr,
doc_id_map,
&mut buffer_lender,
ctx,
serializer,
)?;
} else {
SpecializedPostingsWriter::<DocIdRecorder>::serialize_one_term(
term_buffer.serialized_value_bytes(),
*addr,
doc_id_map,
&mut buffer_lender,
ctx,
serializer,
)?;
}
}
}
Ok(())
@@ -110,40 +110,3 @@ impl<Rec: Recorder> PostingsWriter for JsonPostingsWriter<Rec> {
self.str_posting_writer.total_num_tokens() + self.non_str_posting_writer.total_num_tokens()
}
}
struct JsonTermSerializer(Vec<u8>);
impl JsonTermSerializer {
#[inline]
pub fn append_path(&mut self, bytes: &[u8]) {
if bytes.contains(&0u8) {
self.0
.extend(bytes.iter().map(|&b| if b == 0 { b'0' } else { b }));
} else {
self.0.extend_from_slice(bytes);
}
}
/// Appends value bytes to the Term.
///
/// This function returns the segment that has just been added.
#[inline]
pub fn append_bytes(&mut self, bytes: &[u8]) -> &mut [u8] {
let len_before = self.0.len();
self.0.extend_from_slice(bytes);
&mut self.0[len_before..]
}
fn clear(&mut self) {
self.0.clear();
}
fn truncate(&mut self, len: usize) {
self.0.truncate(len);
}
fn len(&self) -> usize {
self.0.len()
}
fn as_bytes(&self) -> &[u8] {
&self.0
}
}

54
src/postings/postings_writer.rs
View File

@@ -11,8 +11,7 @@ use crate::postings::recorder::{BufferLender, Recorder};
use crate::postings::{
FieldSerializer, IndexingContext, InvertedIndexSerializer, PerFieldPostingsWriter,
};
use crate::schema::indexing_term::{get_field_from_indexing_term, IndexingTerm};
use crate::schema::{Field, Schema, Type};
use crate::schema::{Field, Schema, Term, Type};
use crate::tokenizer::{Token, TokenStream, MAX_TOKEN_LEN};
use crate::DocId;
@@ -61,14 +60,14 @@ pub(crate) fn serialize_postings(
let mut term_offsets: Vec<(Field, OrderedPathId, &[u8], Addr)> =
Vec::with_capacity(ctx.term_index.len());
term_offsets.extend(ctx.term_index.iter().map(|(key, addr)| {
let field = get_field_from_indexing_term(key);
let field = Term::wrap(key).field();
if schema.get_field_entry(field).field_type().value_type() == Type::Json {
let byte_range_path = 4..4 + 4;
let byte_range_path = 5..5 + 4;
let unordered_id = u32::from_be_bytes(key[byte_range_path.clone()].try_into().unwrap());
let path_id = unordered_id_to_ordered_id[unordered_id as usize];
(field, path_id, &key[byte_range_path.end..], addr)
} else {
(field, 0.into(), &key[4..], addr)
(field, 0.into(), &key[5..], addr)
}
}));
// Sort by field, path, and term
@@ -115,7 +114,7 @@ pub(crate) trait PostingsWriter: Send + Sync {
/// * term - the term
/// * ctx - Contains a term hashmap and a memory arena to store all necessary posting list
/// information.
fn subscribe(&mut self, doc: DocId, pos: u32, term: &IndexingTerm, ctx: &mut IndexingContext);
fn subscribe(&mut self, doc: DocId, pos: u32, term: &Term, ctx: &mut IndexingContext);
/// Serializes the postings on disk.
/// The actual serialization format is handled by the `PostingsSerializer`.
@@ -133,7 +132,7 @@ pub(crate) trait PostingsWriter: Send + Sync {
&mut self,
doc_id: DocId,
token_stream: &mut dyn TokenStream,
term_buffer: &mut IndexingTerm,
term_buffer: &mut Term,
ctx: &mut IndexingContext,
indexing_position: &mut IndexingPosition,
) {
@@ -204,35 +203,26 @@ impl<Rec: Recorder> SpecializedPostingsWriter<Rec> {
impl<Rec: Recorder> PostingsWriter for SpecializedPostingsWriter<Rec> {
#[inline]
fn subscribe(
&mut self,
doc: DocId,
position: u32,
term: &IndexingTerm,
ctx: &mut IndexingContext,
) {
debug_assert!(term.serialized_for_hashmap().len() >= 4);
fn subscribe(&mut self, doc: DocId, position: u32, term: &Term, ctx: &mut IndexingContext) {
debug_assert!(term.serialized_term().len() >= 4);
self.total_num_tokens += 1;
let (term_index, arena) = (&mut ctx.term_index, &mut ctx.arena);
term_index.mutate_or_create(
term.serialized_for_hashmap(),
|opt_recorder: Option<Rec>| {
if let Some(mut recorder) = opt_recorder {
let current_doc = recorder.current_doc();
if current_doc != doc {
recorder.close_doc(arena);
recorder.new_doc(doc, arena);
}
recorder.record_position(position, arena);
recorder
} else {
let mut recorder = Rec::default();
term_index.mutate_or_create(term.serialized_term(), |opt_recorder: Option<Rec>| {
if let Some(mut recorder) = opt_recorder {
let current_doc = recorder.current_doc();
if current_doc != doc {
recorder.close_doc(arena);
recorder.new_doc(doc, arena);
recorder.record_position(position, arena);
recorder
}
},
);
recorder.record_position(position, arena);
recorder
} else {
let mut recorder = Rec::default();
recorder.new_doc(doc, arena);
recorder.record_position(position, arena);
recorder
}
});
}
fn serialize(

45
src/query/fuzzy_query.rs
View File

@@ -3,7 +3,7 @@ use once_cell::sync::OnceCell;
use tantivy_fst::Automaton;
use crate::query::{AutomatonWeight, EnableScoring, Query, Weight};
use crate::schema::Term;
use crate::schema::{Term, Type};
use crate::TantivyError::InvalidArgument;
pub(crate) struct DfaWrapper(pub DFA);
@@ -133,33 +133,40 @@ impl FuzzyTermQuery {
let term_value = self.term.value();
let get_automaton = |term_text: &str| {
if self.prefix {
automaton_builder.build_prefix_dfa(term_text)
} else {
automaton_builder.build_dfa(term_text)
let term_text = if term_value.typ() == Type::Json {
if let Some(json_path_type) = term_value.json_path_type() {
if json_path_type != Type::Str {
return Err(InvalidArgument(format!(
"The fuzzy term query requires a string path type for a json term. Found \
{:?}",
json_path_type
)));
}
}
std::str::from_utf8(self.term.serialized_value_bytes()).map_err(|_| {
InvalidArgument(
"Failed to convert json term value bytes to utf8 string.".to_string(),
)
})?
} else {
term_value.as_str().ok_or_else(|| {
InvalidArgument("The fuzzy term query requires a string term.".to_string())
})?
};
let automaton = if self.prefix {
automaton_builder.build_prefix_dfa(term_text)
} else {
automaton_builder.build_dfa(term_text)
};
if let Some((json_path_bytes, _term_value)) = term_value.as_json() {
let term_text =
std::str::from_utf8(self.term.serialized_value_bytes()).map_err(|_| {
InvalidArgument(
"Failed to convert json term value bytes to utf8 string.".to_string(),
)
})?;
let automaton = get_automaton(term_text);
if let Some((json_path_bytes, _)) = term_value.as_json() {
Ok(AutomatonWeight::new_for_json_path(
self.term.field(),
DfaWrapper(automaton),
json_path_bytes,
))
} else {
let term_text = term_value.as_str().ok_or_else(|| {
InvalidArgument("The fuzzy term query requires a string term.".to_string())
})?;
let automaton = get_automaton(term_text);
Ok(AutomatonWeight::new(
self.term.field(),
DfaWrapper(automaton),

2
src/query/phrase_prefix_query/phrase_prefix_query.rs
View File

@@ -137,7 +137,7 @@ impl Query for PhrasePrefixQuery {
// There are no prefix. Let's just match the suffix.
let end_term =
if let Some(end_value) = prefix_end(self.prefix.1.serialized_value_bytes()) {
let mut end_term = Term::new();
let mut end_term = Term::with_capacity(end_value.len());
end_term.set_field_and_type(self.field, self.prefix.1.typ());
end_term.append_bytes(&end_value);
Bound::Excluded(end_term)

45
src/query/query_parser/query_parser.rs
View File

@@ -10,10 +10,10 @@ use query_grammar::{UserInputAst, UserInputBound, UserInputLeaf, UserInputLitera
use rustc_hash::FxHashMap;
use super::logical_ast::*;
use crate::index::Index;
use crate::json_utils::{
convert_to_fast_value_and_append_to_json_term, split_json_path, term_from_json_paths,
use crate::core::json_utils::{
convert_to_fast_value_and_get_term, set_string_and_get_terms, JsonTermWriter,
};
use crate::index::Index;
use crate::query::range_query::{is_type_valid_for_fastfield_range_query, RangeQuery};
use crate::query::{
AllQuery, BooleanQuery, BoostQuery, EmptyQuery, FuzzyTermQuery, Occur, PhrasePrefixQuery,
@@ -965,33 +965,20 @@ fn generate_literals_for_json_object(
})?;
let index_record_option = text_options.index_option();
let mut logical_literals = Vec::new();
let paths = split_json_path(json_path);
let get_term_with_path = || {
term_from_json_paths(
field,
paths.iter().map(|el| el.as_str()),
json_options.is_expand_dots_enabled(),
)
};
// Try to convert the phrase to a fast value
if let Some(term) = convert_to_fast_value_and_append_to_json_term(get_term_with_path(), phrase)
{
let mut term = Term::with_capacity(100);
let mut json_term_writer = JsonTermWriter::from_field_and_json_path(
field,
json_path,
json_options.is_expand_dots_enabled(),
&mut term,
);
if let Some(term) = convert_to_fast_value_and_get_term(&mut json_term_writer, phrase) {
logical_literals.push(LogicalLiteral::Term(term));
}
// Try to tokenize the phrase and create Terms.
let mut positions_and_terms = Vec::<(usize, Term)>::new();
let mut token_stream = text_analyzer.token_stream(phrase);
token_stream.process(&mut |token| {
let mut term = get_term_with_path();
term.append_type_and_str(&token.text);
positions_and_terms.push((token.position, term.clone()));
});
if positions_and_terms.len() <= 1 {
for (_, term) in positions_and_terms {
let terms = set_string_and_get_terms(&mut json_term_writer, phrase, &mut text_analyzer);
drop(json_term_writer);
if terms.len() <= 1 {
for (_, term) in terms {
logical_literals.push(LogicalLiteral::Term(term));
}
return Ok(logical_literals);
@@ -1002,7 +989,7 @@ fn generate_literals_for_json_object(
));
}
logical_literals.push(LogicalLiteral::Phrase {
terms: positions_and_terms,
terms,
slop: 0,
prefix: false,
});

2
src/schema/date_time_options.rs
View File

@@ -1,5 +1,7 @@
use std::ops::BitOr;
#[allow(deprecated)]
pub use common::DatePrecision;
pub use common::DateTimePrecision;
use serde::{Deserialize, Serialize};

2
src/schema/document/de.rs
View File

@@ -160,7 +160,7 @@ pub enum ValueType {
/// A dynamic object value.
Object,
/// A JSON object value. Deprecated.
#[deprecated(note = "We keep this for backwards compatibility, use Object instead")]
#[deprecated]
JSONObject,
}

37
src/schema/document/default_document.rs
View File

@@ -1,15 +1,16 @@
use std::collections::{BTreeMap, HashMap, HashSet};
use std::net::Ipv6Addr;
use common::DateTime;
use common::{BinarySerializable, DateTime, VInt};
use serde_json::Map;
use crate::schema::document::se::BinaryValueSerializer;
use crate::schema::document::{
DeserializeError, Document, DocumentDeserialize, DocumentDeserializer,
BinaryDocumentDeserializer, BinaryDocumentSerializer, DeserializeError, Document, DocumentDeserialize, DocumentDeserializer, ReferenceValue, ReferenceValueLeaf
};
use crate::schema::field_type::ValueParsingError;
use crate::schema::field_value::FieldValueIter;
use crate::schema::{Facet, Field, FieldValue, NamedFieldDocument, OwnedValue, Schema};
use crate::schema::{Facet, Field, FieldValue, NamedFieldDocument, OwnedValue, Schema, Value};
use crate::tokenizer::PreTokenizedString;
/// TantivyDocument provides a default implementation of the `Document` trait.
@@ -85,6 +86,36 @@ impl IntoIterator for TantivyDocument {
}
impl TantivyDocument {
pub fn to_bytes(&self, buffer: &mut Vec<u8>) -> std::io::Result<()> {
buffer.clear();
let num_field_values = self.field_values.len();
VInt(num_field_values as u64).serialize(buffer)?;
for (field, value_access) in self.iter_fields_and_values() {
field.serialize(buffer)?;
let mut serializer = BinaryValueSerializer::new(buffer);
match value_access.as_value() {
ReferenceValue::Leaf(ReferenceValueLeaf::PreTokStr(pre_tokenized_text)) => {
serializer.serialize_value(ReferenceValue::Leaf::<&'_ OwnedValue>(
ReferenceValueLeaf::Str(&pre_tokenized_text.text),
))?;
}
_ => {
serializer.serialize_value(value_access.as_value())?;
}
}
}
Ok(())
}
pub fn from_bytes(mut payload: &[u8]) -> Self {
let deserializer = BinaryDocumentDeserializer::from_reader(&mut payload).unwrap();
Self::deserialize(deserializer).unwrap()
}
/// Creates a new, empty document object
pub fn new() -> TantivyDocument {
TantivyDocument::default()

31
src/schema/document/owned_value.rs
View File

@@ -1,4 +1,4 @@
use std::collections::{btree_map, BTreeMap};
use std::collections::BTreeMap;
use std::fmt;
use std::net::Ipv6Addr;
@@ -45,7 +45,7 @@ pub enum OwnedValue {
/// A set of values.
Array(Vec<Self>),
/// Dynamic object value.
Object(BTreeMap<String, Self>),
Object(Vec<(String, Self)>),
/// IpV6 Address. Internally there is no IpV4, it needs to be converted to `Ipv6Addr`.
IpAddr(Ipv6Addr),
}
@@ -148,10 +148,10 @@ impl ValueDeserialize for OwnedValue {
fn visit_object<'de, A>(&self, mut access: A) -> Result<Self::Value, DeserializeError>
where A: ObjectAccess<'de> {
let mut elements = BTreeMap::new();
let mut elements = Vec::new();
while let Some((key, value)) = access.next_entry()? {
elements.insert(key, value);
elements.push((key, value));
}
Ok(OwnedValue::Object(elements))
@@ -248,12 +248,13 @@ impl<'de> serde::Deserialize<'de> for OwnedValue {
fn visit_map<A>(self, mut map: A) -> Result<Self::Value, A::Error>
where A: MapAccess<'de> {
let mut object = BTreeMap::new();
let mut object =
map.size_hint()
.map(Vec::with_capacity)
.unwrap_or_default();
while let Some((key, value)) = map.next_entry()? {
object.insert(key, value);
object.push((key, value));
}
Ok(OwnedValue::Object(object))
}
}
@@ -363,7 +364,8 @@ impl From<PreTokenizedString> for OwnedValue {
impl From<BTreeMap<String, OwnedValue>> for OwnedValue {
fn from(object: BTreeMap<String, OwnedValue>) -> OwnedValue {
OwnedValue::Object(object)
let key_values = object.into_iter().collect();
OwnedValue::Object(key_values)
}
}
@@ -417,18 +419,15 @@ impl From<serde_json::Value> for OwnedValue {
impl From<serde_json::Map<String, serde_json::Value>> for OwnedValue {
fn from(map: serde_json::Map<String, serde_json::Value>) -> Self {
let mut object = BTreeMap::new();
for (key, value) in map {
object.insert(key, OwnedValue::from(value));
}
let object: Vec<(String, OwnedValue)> = map.into_iter()
.map(|(key, value)| (key, OwnedValue::from(value)))
.collect();
OwnedValue::Object(object)
}
}
/// A wrapper type for iterating over a serde_json object producing reference values.
pub struct ObjectMapIter<'a>(btree_map::Iter<'a, String, OwnedValue>);
pub struct ObjectMapIter<'a>(std::slice::Iter<'a, (String, OwnedValue)>);
impl<'a> Iterator for ObjectMapIter<'a> {
type Item = (&'a str, &'a OwnedValue);

147
src/schema/indexing_term.rs
View File

@@ -1,147 +0,0 @@
use std::net::Ipv6Addr;
use columnar::{MonotonicallyMappableToU128, MonotonicallyMappableToU64};
use super::date_time_options::DATE_TIME_PRECISION_INDEXED;
use super::Field;
use crate::fastfield::FastValue;
use crate::schema::Type;
use crate::DateTime;
/// IndexingTerm represents is the serialized information of a term during indexing.
/// It's a serialized representation over different types.
///
/// It actually wraps a `Vec<u8>`.
///
/// The format is as follow:
/// `[field id: u32][serialized value]`
///
/// For JSON it equals to:
/// `[field id: u32][path id: u32][type code: u8][serialized value]`
///
/// The format is chosen to easily partition the terms by field during serialization, as all terms
/// are stored in one hashmap.
#[derive(Clone)]
pub(crate) struct IndexingTerm(Vec<u8>);
/// The number of bytes used as for the field id by `Term`.
const FIELD_ID_LENGTH: usize = 4;
impl IndexingTerm {
/// Create a new IndexingTerm.
pub fn new() -> IndexingTerm {
let mut data = Vec::with_capacity(FIELD_ID_LENGTH + 32);
data.resize(FIELD_ID_LENGTH, 0u8);
IndexingTerm(data)
}
/// Is empty if there are no value bytes.
pub fn is_empty(&self) -> bool {
self.0.len() == FIELD_ID_LENGTH
}
/// Removes the value_bytes and set the field
pub(crate) fn clear_with_field(&mut self, field: Field) {
self.truncate_value_bytes(0);
self.0[0..4].clone_from_slice(field.field_id().to_be_bytes().as_ref());
}
/// Sets a u64 value in the term.
///
/// U64 are serialized using (8-byte) BigEndian
/// representation.
/// The use of BigEndian has the benefit of preserving
/// the natural order of the values.
pub fn set_u64(&mut self, val: u64) {
self.set_fast_value(val);
}
/// Sets a `DateTime` value in the term.
pub fn set_date(&mut self, val: DateTime) {
self.set_fast_value(val);
}
/// Sets a `i64` value in the term.
pub fn set_i64(&mut self, val: i64) {
self.set_fast_value(val);
}
/// Sets a `f64` value in the term.
pub fn set_f64(&mut self, val: f64) {
self.set_fast_value(val);
}
/// Sets a `bool` value in the term.
pub fn set_bool(&mut self, val: bool) {
self.set_fast_value(val);
}
fn set_fast_value<T: FastValue>(&mut self, val: T) {
self.truncate_value_bytes(0);
self.append_fast_value(val);
}
/// Sets a `Ipv6Addr` value in the term.
pub fn set_ip_addr(&mut self, val: Ipv6Addr) {
self.set_value_bytes(val.to_u128().to_be_bytes().as_ref());
}
/// Sets the value bytes of the term.
pub fn set_value_bytes(&mut self, bytes: &[u8]) {
self.truncate_value_bytes(0);
self.0.extend(bytes);
}
/// Append a type marker + fast value to a term.
/// This is used in JSON type to append a fast value after the path.
///
/// It will not clear existing bytes.
pub(crate) fn append_type_and_fast_value<T: FastValue>(&mut self, val: T) {
self.0.push(T::to_type().to_code());
self.append_fast_value(val)
}
/// Append a fast value to a term.
///
/// It will not clear existing bytes.
pub fn append_fast_value<T: FastValue>(&mut self, val: T) {
let value = if T::to_type() == Type::Date {
DateTime::from_u64(val.to_u64())
.truncate(DATE_TIME_PRECISION_INDEXED)
.to_u64()
} else {
val.to_u64()
};
self.0.extend(value.to_be_bytes().as_ref());
}
/// Truncates the value bytes of the term. Value and field type stays the same.
pub fn truncate_value_bytes(&mut self, len: usize) {
self.0.truncate(len + FIELD_ID_LENGTH);
}
/// The length of the bytes.
pub fn len_bytes(&self) -> usize {
self.0.len() - FIELD_ID_LENGTH
}
/// Appends bytes to the Term.
///
/// This function returns the segment that has just been added.
#[inline]
pub fn append_bytes(&mut self, bytes: &[u8]) {
self.0.extend_from_slice(bytes);
}
/// Returns the serialized representation of Term.
/// This includes field_id, value bytes
#[inline]
pub fn serialized_for_hashmap(&self) -> &[u8] {
self.0.as_ref()
}
}
pub fn get_field_from_indexing_term(bytes: &[u8]) -> Field {
let field_id_bytes: [u8; 4] = bytes[..4].try_into().unwrap();
Field::from_field_id(u32::from_be_bytes(field_id_bytes))
}

7
src/schema/mod.rs
View File

@@ -109,7 +109,6 @@
pub mod document;
mod facet;
mod facet_options;
pub(crate) mod indexing_term;
mod schema;
pub(crate) mod term;
@@ -131,6 +130,8 @@ mod text_options;
use columnar::ColumnType;
pub use self::bytes_options::BytesOptions;
#[allow(deprecated)]
pub use self::date_time_options::DatePrecision;
pub use self::date_time_options::{DateOptions, DateTimePrecision, DATE_TIME_PRECISION_INDEXED};
pub use self::document::{DocParsingError, Document, OwnedValue, TantivyDocument, Value};
pub(crate) use self::facet::FACET_SEP_BYTE;
@@ -145,9 +146,11 @@ pub use self::index_record_option::IndexRecordOption;
pub use self::ip_options::{IntoIpv6Addr, IpAddrOptions};
pub use self::json_object_options::JsonObjectOptions;
pub use self::named_field_document::NamedFieldDocument;
#[allow(deprecated)]
pub use self::numeric_options::IntOptions;
pub use self::numeric_options::NumericOptions;
pub use self::schema::{Schema, SchemaBuilder};
pub use self::term::{Term, ValueBytes};
pub use self::term::{Term, ValueBytes, JSON_END_OF_PATH};
pub use self::text_options::{TextFieldIndexing, TextOptions, STRING, TEXT};
/// Validator for a potential `field_name`.

4
src/schema/numeric_options.rs
View File

@@ -5,6 +5,10 @@ use serde::{Deserialize, Serialize};
use super::flags::CoerceFlag;
use crate::schema::flags::{FastFlag, IndexedFlag, SchemaFlagList, StoredFlag};
#[deprecated(since = "0.17.0", note = "Use NumericOptions instead.")]
/// Deprecated use [`NumericOptions`] instead.
pub type IntOptions = NumericOptions;
/// Define how an `u64`, `i64`, or `f64` field should be handled by tantivy.
#[derive(Clone, Debug, PartialEq, Eq, Serialize, Deserialize, Default)]
#[serde(from = "NumericOptionsDeser")]

337
src/schema/term.rs
View File

@@ -1,9 +1,8 @@
use std::hash::Hash;
use std::hash::{Hash, Hasher};
use std::net::Ipv6Addr;
use std::{fmt, str};
use columnar::{MonotonicallyMappableToU128, MonotonicallyMappableToU64};
use common::json_path_writer::{JSON_END_OF_PATH, JSON_PATH_SEGMENT_SEP_STR};
use super::date_time_options::DATE_TIME_PRECISION_INDEXED;
use super::Field;
@@ -11,6 +10,15 @@ use crate::fastfield::FastValue;
use crate::schema::{Facet, Type};
use crate::DateTime;
/// Separates the different segments of a json path.
pub const JSON_PATH_SEGMENT_SEP: u8 = 1u8;
pub const JSON_PATH_SEGMENT_SEP_STR: &str =
unsafe { std::str::from_utf8_unchecked(&[JSON_PATH_SEGMENT_SEP]) };
/// Separates the json path and the value in
/// a JSON term binary representation.
pub const JSON_END_OF_PATH: u8 = 0u8;
/// Term represents the value that the token can take.
/// It's a serialized representation over different types.
///
@@ -18,45 +26,37 @@ use crate::DateTime;
/// 4 bytes are the field id, and the last byte is the type.
///
/// The serialized value `ValueBytes` is considered everything after the 4 first bytes (term id).
#[derive(Clone, Hash, PartialEq, Ord, PartialOrd, Eq)]
pub struct Term(Vec<u8>);
impl Default for Term {
fn default() -> Self {
Self::new()
}
}
#[derive(Clone)]
pub struct Term<B = Vec<u8>>(B)
where B: AsRef<[u8]>;
/// The number of bytes used as metadata by `Term`.
const TERM_METADATA_LENGTH: usize = 5;
impl Term {
/// Create a new Term
pub fn new() -> Term {
let mut data = Vec::with_capacity(TERM_METADATA_LENGTH + 32);
/// Create a new Term with a buffer with a given capacity.
pub fn with_capacity(capacity: usize) -> Term {
let mut data = Vec::with_capacity(TERM_METADATA_LENGTH + capacity);
data.resize(TERM_METADATA_LENGTH, 0u8);
Term(data)
}
pub(crate) fn with_type_and_field(typ: Type, field: Field) -> Term {
Self::with_bytes_and_field_and_payload(typ, field, &[])
let mut term = Self::with_capacity(8);
term.set_field_and_type(field, typ);
term
}
fn with_bytes_and_field_and_payload(typ: Type, field: Field, bytes: &[u8]) -> Term {
let mut term = Self::new();
let mut term = Self::with_capacity(bytes.len());
term.set_field_and_type(field, typ);
term.0.extend_from_slice(bytes);
term
}
/// Sets a fast value in the term.
///
/// fast values are converted to u64 and then serialized using (8-byte) BigEndian
/// representation.
/// The use of BigEndian has the benefit of preserving
/// the natural order of the values.
fn from_fast_value<T: FastValue>(field: Field, val: &T) -> Term {
let mut term = Self::with_type_and_field(T::to_type(), field);
term.set_bytes(val.to_u64().to_be_bytes().as_ref());
term.set_u64(val.to_u64());
term
}
@@ -78,7 +78,7 @@ impl Term {
/// Builds a term given a field, and a `Ipv6Addr`-value
pub fn from_field_ip_addr(field: Field, ip_addr: Ipv6Addr) -> Term {
let mut term = Self::with_type_and_field(Type::IpAddr, field);
term.set_bytes(ip_addr.to_u128().to_be_bytes().as_ref());
term.set_ip_addr(ip_addr);
term
}
@@ -123,16 +123,52 @@ impl Term {
Term::with_bytes_and_field_and_payload(Type::Bytes, field, bytes)
}
/// Removes the value_bytes and set the field and type code.
pub(crate) fn clear_with_field_and_type(&mut self, typ: Type, field: Field) {
self.truncate_value_bytes(0);
self.set_field_and_type(field, typ);
}
/// Removes the value_bytes and set the type code.
pub fn clear_with_type(&mut self, typ: Type) {
self.truncate_value_bytes(0);
self.0[4] = typ.to_code();
}
/// Append a type marker + fast value to a term.
/// This is used in JSON type to append a fast value after the path.
/// Sets a u64 value in the term.
///
/// It will not clear existing bytes.
/// U64 are serialized using (8-byte) BigEndian
/// representation.
/// The use of BigEndian has the benefit of preserving
/// the natural order of the values.
pub fn set_u64(&mut self, val: u64) {
self.set_fast_value(val);
}
/// Sets a `i64` value in the term.
pub fn set_i64(&mut self, val: i64) {
self.set_fast_value(val);
}
/// Sets a `DateTime` value in the term.
pub fn set_date(&mut self, date: DateTime) {
self.set_fast_value(date);
}
/// Sets a `f64` value in the term.
pub fn set_f64(&mut self, val: f64) {
self.set_fast_value(val);
}
/// Sets a `bool` value in the term.
pub fn set_bool(&mut self, val: bool) {
self.set_fast_value(val);
}
fn set_fast_value<T: FastValue>(&mut self, val: T) {
self.set_bytes(val.to_u64().to_be_bytes().as_ref());
}
pub(crate) fn append_type_and_fast_value<T: FastValue>(&mut self, val: T) {
self.0.push(T::to_type().to_code());
let value = if T::to_type() == Type::Date {
@@ -145,26 +181,32 @@ impl Term {
self.0.extend(value.to_be_bytes().as_ref());
}
/// Append a string type marker + string to a term.
/// This is used in JSON type to append a str after the path.
///
/// It will not clear existing bytes.
pub(crate) fn append_type_and_str(&mut self, val: &str) {
self.0.push(Type::Str.to_code());
self.0.extend(val.as_bytes().as_ref());
/// Sets a `Ipv6Addr` value in the term.
pub fn set_ip_addr(&mut self, val: Ipv6Addr) {
self.set_bytes(val.to_u128().to_be_bytes().as_ref());
}
/// Sets the value of a `Bytes` field.
fn set_bytes(&mut self, bytes: &[u8]) {
pub fn set_bytes(&mut self, bytes: &[u8]) {
self.truncate_value_bytes(0);
self.0.extend(bytes);
}
/// Set the texts only, keeping the field untouched.
pub fn set_text(&mut self, text: &str) {
self.set_bytes(text.as_bytes());
}
/// Truncates the value bytes of the term. Value and field type stays the same.
pub fn truncate_value_bytes(&mut self, len: usize) {
self.0.truncate(len + TERM_METADATA_LENGTH);
}
/// The length of the bytes.
pub fn len_bytes(&self) -> usize {
self.0.len() - TERM_METADATA_LENGTH
}
/// Appends value bytes to the Term.
///
/// This function returns the segment that has just been added.
@@ -175,19 +217,78 @@ impl Term {
&mut self.0[len_before..]
}
/// Appends json path bytes to the Term.
/// If the path contains 0 bytes, they are replaced by a "0" string.
/// The 0 byte is used to mark the end of the path.
///
/// This function returns the segment that has just been added.
#[inline]
pub fn append_path(&mut self, bytes: &[u8]) -> &mut [u8] {
let len_before = self.0.len();
if bytes.contains(&0u8) {
self.0
.extend(bytes.iter().map(|&b| if b == 0 { b'0' } else { b }));
} else {
self.0.extend_from_slice(bytes);
}
&mut self.0[len_before..]
}
/// Appends a JSON_PATH_SEGMENT_SEP to the term.
/// Only used for JSON type.
#[inline]
pub fn add_json_path_separator(&mut self) {
self.0.push(JSON_PATH_SEGMENT_SEP);
}
/// Sets the current end to JSON_END_OF_PATH.
/// Only used for JSON type.
#[inline]
pub fn set_json_path_end(&mut self) {
let buffer_len = self.0.len();
self.0[buffer_len - 1] = JSON_END_OF_PATH;
}
/// Sets the current end to JSON_PATH_SEGMENT_SEP.
/// Only used for JSON type.
#[inline]
pub fn set_json_path_separator(&mut self) {
let buffer_len = self.0.len();
self.0[buffer_len - 1] = JSON_PATH_SEGMENT_SEP;
}
}
impl<B> Term<B>
where B: AsRef<[u8]>
{
/// Wraps a object holding bytes
pub fn wrap(data: B) -> Term<B> {
Term(data)
}
/// Return the type of the term.
pub fn typ(&self) -> Type {
self.value().typ()
}
/// Returns the field.
pub fn field(&self) -> Field {
let field_id_bytes: [u8; 4] = (&self.0.as_ref()[..4]).try_into().unwrap();
Field::from_field_id(u32::from_be_bytes(field_id_bytes))
}
/// Returns the serialized representation of the value.
/// (this does neither include the field id nor the value type.)
///
/// If the term is a string, its value is utf-8 encoded.
/// If the term is a u64, its value is encoded according
/// to `byteorder::BigEndian`.
pub(crate) fn serialized_value_bytes(&self) -> &[u8] {
&self.0[TERM_METADATA_LENGTH..]
pub fn serialized_value_bytes(&self) -> &[u8] {
&self.0.as_ref()[TERM_METADATA_LENGTH..]
}
/// Returns the value of the term.
/// address or JSON path + value. (this does not include the field.)
pub fn value(&self) -> ValueBytes<&[u8]> {
ValueBytes::wrap(&self.0.as_ref()[4..])
}
/// Returns the serialized representation of Term.
@@ -196,22 +297,9 @@ impl Term {
/// Do NOT rely on this byte representation in the index.
/// This value is likely to change in the future.
#[inline]
#[cfg(test)]
pub fn serialized_term(&self) -> &[u8] {
self.0.as_ref()
}
/// Returns the field.
pub fn field(&self) -> Field {
let field_id_bytes: [u8; 4] = (&self.0[..4]).try_into().unwrap();
Field::from_field_id(u32::from_be_bytes(field_id_bytes))
}
/// Returns the value of the term.
/// address or JSON path + value. (this does not include the field.)
pub fn value(&self) -> ValueBytes<&[u8]> {
ValueBytes::wrap(&self.0[4..])
}
}
/// ValueBytes represents a serialized value.
@@ -242,10 +330,18 @@ where B: AsRef<[u8]>
}
/// Return the type of the term.
pub(crate) fn typ(&self) -> Type {
pub fn typ(&self) -> Type {
Type::from_code(self.typ_code()).expect("The term has an invalid type code")
}
/// Returns the `u64` value stored in a term.
///
/// Returns `None` if the term is not of the u64 type, or if the term byte representation
/// is invalid.
pub fn as_u64(&self) -> Option<u64> {
self.get_fast_type::<u64>()
}
fn get_fast_type<T: FastValue>(&self) -> Option<T> {
if self.typ() != T::to_type() {
return None;
@@ -255,6 +351,38 @@ where B: AsRef<[u8]>
Some(T::from_u64(value_u64))
}
/// Returns the `i64` value stored in a term.
///
/// Returns `None` if the term is not of the i64 type, or if the term byte representation
/// is invalid.
pub fn as_i64(&self) -> Option<i64> {
self.get_fast_type::<i64>()
}
/// Returns the `f64` value stored in a term.
///
/// Returns `None` if the term is not of the f64 type, or if the term byte representation
/// is invalid.
pub fn as_f64(&self) -> Option<f64> {
self.get_fast_type::<f64>()
}
/// Returns the `bool` value stored in a term.
///
/// Returns `None` if the term is not of the bool type, or if the term byte representation
/// is invalid.
pub fn as_bool(&self) -> Option<bool> {
self.get_fast_type::<bool>()
}
/// Returns the `Date` value stored in a term.
///
/// Returns `None` if the term is not of the Date type, or if the term byte representation
/// is invalid.
pub fn as_date(&self) -> Option<DateTime> {
self.get_fast_type::<DateTime>()
}
/// Returns the text associated with the term.
///
/// Returns `None` if the field is not of string type
@@ -270,7 +398,7 @@ where B: AsRef<[u8]>
///
/// Returns `None` if the field is not of facet type
/// or if the bytes are not valid utf-8.
pub(crate) fn as_facet(&self) -> Option<Facet> {
pub fn as_facet(&self) -> Option<Facet> {
if self.typ() != Type::Facet {
return None;
}
@@ -281,7 +409,7 @@ where B: AsRef<[u8]>
/// Returns the bytes associated with the term.
///
/// Returns `None` if the field is not of bytes type.
pub(crate) fn as_bytes(&self) -> Option<&[u8]> {
pub fn as_bytes(&self) -> Option<&[u8]> {
if self.typ() != Type::Bytes {
return None;
}
@@ -289,7 +417,7 @@ where B: AsRef<[u8]>
}
/// Returns a `Ipv6Addr` value from the term.
pub(crate) fn as_ip_addr(&self) -> Option<Ipv6Addr> {
pub fn as_ip_addr(&self) -> Option<Ipv6Addr> {
if self.typ() != Type::IpAddr {
return None;
}
@@ -297,6 +425,15 @@ where B: AsRef<[u8]>
Some(Ipv6Addr::from_u128(ip_u128))
}
/// Returns the json path type.
///
/// Returns `None` if the value is not JSON.
pub fn json_path_type(&self) -> Option<Type> {
let json_value_bytes = self.as_json_value_bytes()?;
Some(json_value_bytes.typ())
}
/// Returns the json path bytes (including the JSON_END_OF_PATH byte),
/// and the encoded ValueBytes after the json path.
///
@@ -313,6 +450,18 @@ where B: AsRef<[u8]>
Some((json_path_bytes, ValueBytes::wrap(term)))
}
/// Returns the encoded ValueBytes after the json path.
///
/// Returns `None` if the value is not JSON.
pub(crate) fn as_json_value_bytes(&self) -> Option<ValueBytes<&[u8]>> {
if self.typ() != Type::Json {
return None;
}
let bytes = self.value_bytes();
let pos = bytes.iter().cloned().position(|b| b == JSON_END_OF_PATH)?;
Some(ValueBytes::wrap(&bytes[pos + 1..]))
}
/// Returns the serialized value of ValueBytes without the type.
fn value_bytes(&self) -> &[u8] {
&self.0.as_ref()[1..]
@@ -335,20 +484,20 @@ where B: AsRef<[u8]>
write_opt(f, s)?;
}
Type::U64 => {
write_opt(f, self.get_fast_type::<u64>())?;
write_opt(f, self.as_u64())?;
}
Type::I64 => {
write_opt(f, self.get_fast_type::<i64>())?;
write_opt(f, self.as_i64())?;
}
Type::F64 => {
write_opt(f, self.get_fast_type::<f64>())?;
write_opt(f, self.as_f64())?;
}
Type::Bool => {
write_opt(f, self.get_fast_type::<bool>())?;
write_opt(f, self.as_bool())?;
}
// TODO pretty print these types too.
Type::Date => {
write_opt(f, self.get_fast_type::<DateTime>())?;
write_opt(f, self.as_date())?;
}
Type::Facet => {
write_opt(f, self.as_facet())?;
@@ -374,6 +523,40 @@ where B: AsRef<[u8]>
}
}
impl<B> Ord for Term<B>
where B: AsRef<[u8]>
{
fn cmp(&self, other: &Self) -> std::cmp::Ordering {
self.serialized_term().cmp(other.serialized_term())
}
}
impl<B> PartialOrd for Term<B>
where B: AsRef<[u8]>
{
fn partial_cmp(&self, other: &Self) -> Option<std::cmp::Ordering> {
Some(self.cmp(other))
}
}
impl<B> PartialEq for Term<B>
where B: AsRef<[u8]>
{
fn eq(&self, other: &Self) -> bool {
self.serialized_term() == other.serialized_term()
}
}
impl<B> Eq for Term<B> where B: AsRef<[u8]> {}
impl<B> Hash for Term<B>
where B: AsRef<[u8]>
{
fn hash<H: Hasher>(&self, state: &mut H) {
self.0.as_ref().hash(state)
}
}
fn write_opt<T: std::fmt::Debug>(f: &mut fmt::Formatter, val_opt: Option<T>) -> fmt::Result {
if let Some(val) = val_opt {
write!(f, "{val:?}")?;
@@ -381,11 +564,13 @@ fn write_opt<T: std::fmt::Debug>(f: &mut fmt::Formatter, val_opt: Option<T>) ->
Ok(())
}
impl fmt::Debug for Term {
impl<B> fmt::Debug for Term<B>
where B: AsRef<[u8]>
{
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
let field_id = self.field().field_id();
write!(f, "Term(field={field_id}, ")?;
let value_bytes = ValueBytes::wrap(&self.0[4..]);
let value_bytes = ValueBytes::wrap(&self.0.as_ref()[4..]);
value_bytes.debug_value_bytes(f)?;
write!(f, ")",)?;
Ok(())
@@ -407,4 +592,38 @@ mod tests {
assert_eq!(term.typ(), Type::Str);
assert_eq!(term.value().as_str(), Some("test"))
}
/// Size (in bytes) of the buffer of a fast value (u64, i64, f64, or date) term.
/// <field> + <type byte> + <value len>
///
/// - <field> is a big endian encoded u32 field id
/// - <type_byte>'s most significant bit expresses whether the term is a json term or not
/// The remaining 7 bits are used to encode the type of the value.
/// If this is a JSON term, the type is the type of the leaf of the json.
///
/// - <value> is, if this is not the json term, a binary representation specific to the type.
/// If it is a JSON Term, then it is prepended with the path that leads to this leaf value.
const FAST_VALUE_TERM_LEN: usize = 4 + 1 + 8;
#[test]
pub fn test_term_u64() {
let mut schema_builder = Schema::builder();
let count_field = schema_builder.add_u64_field("count", INDEXED);
let term = Term::from_field_u64(count_field, 983u64);
assert_eq!(term.field(), count_field);
assert_eq!(term.typ(), Type::U64);
assert_eq!(term.serialized_term().len(), FAST_VALUE_TERM_LEN);
assert_eq!(term.value().as_u64(), Some(983u64))
}
#[test]
pub fn test_term_bool() {
let mut schema_builder = Schema::builder();
let bool_field = schema_builder.add_bool_field("bool", INDEXED);
let term = Term::from_field_bool(bool_field, true);
assert_eq!(term.field(), bool_field);
assert_eq!(term.typ(), Type::Bool);
assert_eq!(term.serialized_term().len(), FAST_VALUE_TERM_LEN);
assert_eq!(term.value().as_bool(), Some(true))
}
}

4
src/store/reader.rs
View File

@@ -14,7 +14,7 @@ use super::Decompressor;
use crate::directory::FileSlice;
use crate::error::DataCorruption;
use crate::fastfield::AliveBitSet;
use crate::schema::document::{BinaryDocumentDeserializer, DocumentDeserialize};
use crate::schema::document::{BinaryDocumentDeserializer, Document, DocumentDeserialize};
use crate::space_usage::StoreSpaceUsage;
use crate::store::index::Checkpoint;
use crate::DocId;
@@ -235,7 +235,7 @@ impl StoreReader {
/// Iterator over all Documents in their order as they are stored in the doc store.
/// Use this, if you want to extract all Documents from the doc store.
/// The `alive_bitset` has to be forwarded from the `SegmentReader` or the results may be wrong.
pub fn iter<'a: 'b, 'b, D: DocumentDeserialize>(
pub fn iter<'a: 'b, 'b, D: Document + DocumentDeserialize>(
&'b self,
alive_bitset: Option<&'a AliveBitSet>,
) -> impl Iterator<Item = crate::Result<D>> + 'b {

6
sstable/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy-sstable"
version = "0.3.0"
version = "0.2.0"
edition = "2021"
license = "MIT"
homepage = "https://github.com/quickwit-oss/tantivy"
@@ -10,8 +10,8 @@ categories = ["database-implementations", "data-structures", "compression"]
description = "sstables for tantivy"
[dependencies]
common = {version= "0.7", path="../common", package="tantivy-common"}
tantivy-bitpacker = { version= "0.6", path="../bitpacker" }
common = {version= "0.6", path="../common", package="tantivy-common"}
tantivy-bitpacker = { version= "0.5", path="../bitpacker" }
tantivy-fst = "0.5"
# experimental gives us access to Decompressor::upper_bound
zstd = { version = "0.13", features = ["experimental"] }

1
sstable/src/lib.rs
View File

@@ -1,5 +1,6 @@
use std::io::{self, Write};
use std::ops::Range;
use std::usize;
use merge::ValueMerger;

4
stacker/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy-stacker"
version = "0.3.0"
version = "0.2.0"
edition = "2021"
license = "MIT"
homepage = "https://github.com/quickwit-oss/tantivy"
@@ -9,7 +9,7 @@ description = "term hashmap used for indexing"
[dependencies]
murmurhash32 = "0.3"
common = { version = "0.7", path = "../common/", package = "tantivy-common" }
common = { version = "0.6", path = "../common/", package = "tantivy-common" }
ahash = { version = "0.8.11", default-features = false, optional = true }
rand_distr = "0.4.3"

2
tokenizer-api/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy-tokenizer-api"
version = "0.3.0"
version = "0.2.0"
license = "MIT"
edition = "2021"
description = "Tokenizer API of tantivy"