rust/tantivy
1
0
Fork 0
mirror of https://github.com/quickwit-oss/tantivy.git synced 2026-01-04 16:22:55 +00:00
Code Issues Packages Projects Releases Wiki Activity

Compare commits

base: rust:0.16.1
Branches Tags
rust:main rust:paul.masurel/seek_into_the_danger_zone_result 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:revert-1222-issue/1195
Branches Tags
rust:paul.masurel/seek_into_the_danger_zone_result 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

73 Commits
0.16.1 ... revert-122

Author SHA1 Message Date
Paul Masurel
bd083159ac Revert "Closes #1195 (#1222)" 
This reverts commit dde49ac8e2.
 2021-12-02 14:37:59 +09:00
Paul Masurel
dde49ac8e2 Closes #1195 (#1222) 
Removes the indexed option for facets.
Facets are now always indexed.

Closes #1195
 2021-12-02 14:37:19 +09:00
Paul Masurel
c3cc93406d Bugfix: adds missing fdatasync on atomic_write. 
In addition this PR:
- removes unnecessary flushes and fsyncs on files.
- replace all fsync by fdatasync. The latter triggers
a meta sync if a metadata required to read the file
has changed. It is therefore sufficient for us.

Closes #1224
 2021-12-02 13:42:44 +09:00
Kanji Yomoda
bd0f9211da Remove unused sort for segmenta meta list (#1218) 
* Remove unused sort for segment meta list
* Fix segment meta order dependent test
 2021-12-01 11:18:17 +09:00
PSeitz
c503c6e4fa Switch to non-strict schema (#1216) 
Fixes #1211
 2021-11-29 10:38:59 +09:00
PSeitz
02174d26af Merge pull request #1209 from quickwit-inc/lz4_flex_version 
fix lz4_flex version
 2021-11-16 14:12:45 +08:00
PSeitz
cf92be3bd6 fix lz4_flex version 2021-11-16 06:03:04 +00:00
Shikhar Bhushan
72cef12db1 Add none compression (#1208) 2021-11-16 10:50:42 +09:00
Paul Masurel
bbc0a2e233 Fixing the build 2021-11-16 09:37:25 +09:00
François Massot
4fd1a6c84b Merge pull request #1207 from quickwit-inc/fix-chat-links 
Remove patron link and changer gitter links to discord links.
 2021-11-15 19:23:21 +01:00
François Massot
c83d99c414 Remove patron link and changer gitter links to discord links. 2021-11-15 19:17:35 +01:00
Paul Masurel
eacf510175 Exchange gitter link for discord 2021-11-15 16:44:13 +09:00
Paul Masurel
8802d125f8 Prepare commit is public again (#1202) 
- Simplified some of the prepare commit & segment updater code using
async.
- Made PrepareCommit public again.
 2021-11-12 23:25:39 +09:00
dependabot[bot]
33301a3eb4 Update fail requirement from 0.4 to 0.5 (#1197) 
Updates the requirements on [fail](https://github.com/tikv/fail-rs) to permit the latest version.
- [Release notes](https://github.com/tikv/fail-rs/releases)
- [Changelog](https://github.com/tikv/fail-rs/blob/master/CHANGELOG.md)
- [Commits](https://github.com/tikv/fail-rs/compare/v0.4.0...v0.5.0)

---
updated-dependencies:
- dependency-name: fail
  dependency-type: direct:production
...

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

Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2021-11-12 23:21:16 +09:00
Paul Masurel
7234bef0eb Issue/1198 (#1201) 
* Unit test reproducing #1198
* Fixing unit test to handle the error from add_document.
* Bump project version
 2021-11-11 16:42:19 +09:00
azerowall
fcff91559b Fix the deserialization error of FieldEntry when the 'options' field appears before the 'type' field (#1199) 
Co-authored-by: quel <azerowall>
 2021-11-10 18:39:58 +09:00
Paul Masurel
b75d4e59d1 Remove the broken panic on drop unit test. (#1200) 2021-11-10 18:39:37 +09:00
Paul Masurel
c6b5ab1dbe Replacing the panic check in the RAM Directory on lack of flush. 2021-11-09 11:04:31 +09:00
PSeitz
c12e07f0ce Merge pull request #1196 from quickwit-inc/dependabot/cargo/measure_time-0.8.0 
Update measure_time requirement from 0.7.0 to 0.8.0
 2021-11-05 08:47:51 +08:00
dependabot[bot]
8b877a4c26 Update measure_time requirement from 0.7.0 to 0.8.0 
Updates the requirements on [measure_time](https://github.com/PSeitz/rust_measure_time) to permit the latest version.
- [Release notes](https://github.com/PSeitz/rust_measure_time/releases)
- [Commits](https://github.com/PSeitz/rust_measure_time/commits)

---
updated-dependencies:
- dependency-name: measure_time
  dependency-type: direct:production
...

Signed-off-by: dependabot[bot] <support@github.com>
 2021-11-04 20:27:16 +00:00
PSeitz
7dc0dc1c9b extend proptests with adding case (#1191) 
This extends the proptest to cover a case where up to a 100 documents are added to an index.
 2021-11-01 09:27:10 +09:00
François Massot
0462754673 Optimize block wand for one and several TermScorer. (#1190) 
* Added optimisation using block wand for single TermScorer.

A proptest was also added.

* Fix block wand algorithm by taking the last doc id of scores until the pivot scorer (included).
* In block wand, when block max score is lower than the threshold, advance the scorer with best score.
* Fix wrong condition in block_wand_single_scorer and add debug_assert to have an equality check on doc to break the loop.
 2021-11-01 09:18:05 +09:00
PSeitz
5916ceda73 Merge pull request #1188 from PSeitz/sort_issue 
fix incorrect padding in bitset for multiple of 64
 2021-10-29 17:06:38 +08:00
Pascal Seitz
70283dc6c8 fix incorrect padding in bitset for multiple of 64 2021-10-29 16:49:22 +08:00
PSeitz
dbaf4f3623 Merge pull request #1187 from PSeitz/sort_issue 
check searcher num docs in proptest
 2021-10-29 16:19:24 +08:00
Pascal Seitz
4808648322 check searcher num docs in proptest 2021-10-29 14:38:30 +08:00
Paul Masurel
54afb9b34a Made PrepareCommit private 2021-10-29 14:13:14 +09:00
Paul Masurel
d336c8b938 Fixed logo 2021-10-27 08:54:16 +09:00
Paul Masurel
980d1b2796 Removing Patreon link 2021-10-27 08:53:45 +09:00
Dan Cecile
6317982876 Make indexer::prepared_commit public (#1184) 
* Make indexer::prepared_commit public

* Add PreparedCommit to lib
 2021-10-26 12:21:24 +09:00
PSeitz
e2fbbc08ca Merge pull request #1182 from PSeitz/remove_directory_generic 
use Box<dyn Directory> as parameter to open/create an Index
 2021-10-25 12:49:55 +08:00
Pascal Seitz
99cd25beae use <T: Into<Box<dyn Directory>>> as parameter to open/create an Index 
This is done in order to support Box<dyn Directory> additionally to generic implementations of the trait Directory.
Remove boxing in ManagedDirectory.
 2021-10-25 12:34:40 +08:00
Kanji Yomoda
737ecc7015 Fix outdated comment for IndexWriter::new (#1183) 2021-10-25 10:59:18 +09:00
Kanji Yomoda
09668459c8 Update codecov-action to v2 and make it possible to keep it up-to-date with dependabot (#1181) 
* Update codecov-action to v2

* Add github-actions to dependabot
 2021-10-25 10:58:16 +09:00
Evance Soumaoro
e5fd30f438 Fixed links (#1177) 2021-10-25 10:56:04 +09:00
Tom Parker-Shemilt
c412a46105 Remove travis config (#1180) 2021-10-24 15:40:43 +09:00
PSeitz
3a78402496 update links (#1176) 2021-10-18 20:45:40 +09:00
Paul Masurel
d18ac136c0 Search simplified (#1175) 2021-10-18 12:52:43 +09:00
Paul Masurel
b5b1244857 More functionality in the ownedbytes crate (#1172) 2021-10-07 18:14:49 +09:00
Paul Masurel
27acfa4dea Removing dead file (#1170) 2021-10-07 14:15:21 +09:00
Paul Masurel
02cffa4dea Code simplification. (#1169) 
Code simplification and Clippy
 2021-10-07 14:11:44 +09:00
Paul Masurel
b52abbc771 Bugfix transposition_cost_one in FuzzyQuery (#1167) 2021-10-07 09:38:39 +09:00
Paul Masurel
894c61867f Fix test compilation (#1168) 2021-10-06 17:50:10 +09:00
PSeitz
352e0cc58d Adde demux operation (#1150) 
* add merge for DeleteBitSet, allow custom DeleteBitSet on merge
* forward delete bitsets on merge, add tests
* add demux operation and tests
 2021-10-06 16:05:16 +09:00
Paul Masurel
ffe4446d90 Minor lint comments (#1166) 2021-10-06 11:27:48 +09:00
dependabot[bot]
4d05b26e7a Update lru requirement from 0.6.5 to 0.7.0 (#1165) 
Updates the requirements on [lru](https://github.com/jeromefroe/lru-rs) to permit the latest version.
- [Release notes](https://github.com/jeromefroe/lru-rs/releases)
- [Changelog](https://github.com/jeromefroe/lru-rs/blob/master/CHANGELOG.md)
- [Commits](https://github.com/jeromefroe/lru-rs/compare/0.6.5...0.7.0)

---
updated-dependencies:
- dependency-name: lru
  dependency-type: direct:production
...

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

Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2021-10-06 05:50:24 +09:00
Paul Masurel
0855649986 Leaning more on the alive (vs delete) semantics. (#1164) 2021-10-05 18:53:29 +09:00
PSeitz
d828e58903 Merge pull request #1163 from PSeitz/reduce_mem_usage 
reduce mem usage
 2021-10-01 08:03:41 +02:00
Pascal Seitz
aa0396fe27 fix variable names 2021-10-01 13:48:51 +08:00
Pascal Seitz
8d8315f8d0 prealloc vec in postinglist 2021-09-29 09:02:38 +08:00
Pascal Seitz
078c0a2e2e reserve vec 2021-09-29 08:45:04 +08:00
Pascal Seitz
f21e8dd875 use only segment ordinal in docidmapping 2021-09-29 08:44:56 +08:00
Tomoko Uchida
74e36c7e97 Add unit tests for tokenizers and filters (#1156) 
* add unit test for SimpleTokenizer
* add unit tests for tokenizers and filters.
 2021-09-27 10:22:01 +09:00
PSeitz
f27ae04282 fix slope calculation in multilinear interpol (#1161) 
add test to check for compression
 2021-09-27 10:14:03 +09:00
PSeitz
0ce49c9dd4 use lz4_flex 0.9.0 (#1160) 2021-09-27 10:12:20 +09:00
PSeitz
fe8e58e078 Merge pull request #1154 from PSeitz/delete_bitset 
add DeleteBitSet iterator
 2021-09-24 09:37:39 +02:00
Pascal Seitz
efc0d8341b fix comment 2021-09-24 15:09:21 +08:00
Pascal Seitz
22bcc83d10 fix padding in initialization 2021-09-24 14:43:04 +08:00
Pascal Seitz
5ee5037934 create and use ReadSerializedBitSet 2021-09-24 12:53:33 +08:00
Pascal Seitz
c217bfed1e cargo fmt 2021-09-23 21:02:19 +08:00
Pascal Seitz
c27ccd3e24 improve naming 2021-09-23 21:02:09 +08:00
Paul Masurel
367f5da782 Fixed comment to the index accessor 2021-09-23 21:53:48 +09:00
Mestery
b256df6599 add index accessor for index writer (#1159) 
* add index accessor for index writer

* Update src/indexer/index_writer.rs

Co-authored-by: Paul Masurel <paul@quickwit.io>
 2021-09-23 21:49:20 +09:00
Pascal Seitz
d7a6a409a1 renames 2021-09-23 20:33:11 +08:00
Pascal Seitz
a1f5cead96 AliveBitSet instead of DeleteBitSet 2021-09-23 20:03:57 +08:00
dependabot[bot]
37c5fe3c86 Update memmap2 requirement from 0.4 to 0.5 (#1157) 
Updates the requirements on [memmap2](https://github.com/RazrFalcon/memmap2-rs) to permit the latest version.
- [Release notes](https://github.com/RazrFalcon/memmap2-rs/releases)
- [Changelog](https://github.com/RazrFalcon/memmap2-rs/blob/master/CHANGELOG.md)
- [Commits](https://github.com/RazrFalcon/memmap2-rs/compare/v0.4.0...v0.5.0)

---
updated-dependencies:
- dependency-name: memmap2
  dependency-type: direct:production
...

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

Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2021-09-23 20:18:27 +09:00
Pascal Seitz
4583fa270b fixes 2021-09-23 10:39:53 +08:00
Pascal Seitz
beb3a5bd73 fix len 2021-09-18 17:58:15 +08:00
Pascal Seitz
93cbd52bf0 move code to biset, add inline, add benchmark 2021-09-18 17:35:22 +08:00
Pascal Seitz
c22177a005 add iterator 2021-09-17 15:29:27 +08:00
Pascal Seitz
4da71273e1 add de/serialization for bitset 
remove len footgun
 2021-09-17 10:28:12 +08:00
dependabot[bot]
2c78b31aab Update memmap2 requirement from 0.3 to 0.4 (#1155) 
Updates the requirements on [memmap2](https://github.com/RazrFalcon/memmap2-rs) to permit the latest version.
- [Release notes](https://github.com/RazrFalcon/memmap2-rs/releases)
- [Changelog](https://github.com/RazrFalcon/memmap2-rs/blob/master/CHANGELOG.md)
- [Commits](https://github.com/RazrFalcon/memmap2-rs/compare/v.0.3.0...v0.4.0)
 2021-09-17 08:52:52 +09:00
Pascal Seitz
4ae1d87632 add DeleteBitSet iterator 2021-09-15 23:10:04 +08:00
118 changed files with 3688 additions and 2283 deletions
Expand all files Collapse all files

7
.github/dependabot.yml vendored
View File

@@ -6,3 +6,10 @@ updates:
interval: daily
time: "20:00"
open-pull-requests-limit: 10
- package-ecosystem: "github-actions"
directory: "/"
schedule:
interval: daily
time: "20:00"
open-pull-requests-limit: 10

2
.github/workflows/coverage.yml vendored
View File

@@ -18,7 +18,7 @@ jobs:
- name: Generate code coverage
run: cargo llvm-cov --all-features --workspace --lcov --output-path lcov.info
- name: Upload coverage to Codecov
uses: codecov/codecov-action@v1
uses: codecov/codecov-action@v2
with:
token: ${{ secrets.CODECOV_TOKEN }} # not required for public repos
files: lcov.info

1
.gitignore vendored
View File

@@ -1,4 +1,5 @@
tantivy.iml
.cargo
proptest-regressions
*.swp
target

92
.travis.yml
View File

@@ -1,92 +0,0 @@
# Based on the "trust" template v0.1.2
# https://github.com/japaric/trust/tree/v0.1.2
dist: trusty
language: rust
services: docker
sudo: required
env:
global:
- CRATE_NAME=tantivy
- TRAVIS_CARGO_NIGHTLY_FEATURE=""
# - secure: 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
addons:
apt:
sources:
- ubuntu-toolchain-r-test
- kalakris-cmake
packages:
- gcc-4.8
- g++-4.8
- libcurl4-openssl-dev
- libelf-dev
- libdw-dev
- binutils-dev
- cmake
matrix:
include:
# Android
- env: TARGET=aarch64-linux-android DISABLE_TESTS=1
#- env: TARGET=arm-linux-androideabi DISABLE_TESTS=1
#- env: TARGET=armv7-linux-androideabi DISABLE_TESTS=1
#- env: TARGET=i686-linux-android DISABLE_TESTS=1
#- env: TARGET=x86_64-linux-android DISABLE_TESTS=1
# Linux
#- env: TARGET=aarch64-unknown-linux-gnu
#- env: TARGET=i686-unknown-linux-gnu
- env: TARGET=x86_64-unknown-linux-gnu CODECOV=1 #UPLOAD_DOCS=1
# - env: TARGET=x86_64-unknown-linux-musl CODECOV=1
# OSX
#- env: TARGET=x86_64-apple-darwin
# os: osx
before_install:
- set -e
- rustup self update
- rustup component add rustfmt
install:
- sh ci/install.sh
- source ~/.cargo/env || true
- env | grep "TRAVIS"
before_script:
- export PATH=$HOME/.cargo/bin:$PATH
- cargo install cargo-update || echo "cargo-update already installed"
- cargo install cargo-travis || echo "cargo-travis already installed"
script:
- bash ci/script.sh
- cargo fmt --all -- --check
before_deploy:
- sh ci/before_deploy.sh
after_success:
# Needs GH_TOKEN env var to be set in travis settings
- if [[ -v GH_TOKEN ]]; then echo "GH TOKEN IS SET"; else echo "GH TOKEN NOT SET"; fi
- if [[ -v UPLOAD_DOCS ]]; then cargo doc; cargo doc-upload; else echo "doc upload disabled."; fi
#cache: cargo
#before_cache:
# # Travis can't cache files that are not readable by "others"
# - chmod -R a+r $HOME/.cargo
# - find ./target/debug -type f -maxdepth 1 -delete
# - rm -f ./target/.rustc_info.json
# - rm -fr ./target/debug/{deps,.fingerprint}/tantivy*
# - rm -r target/debug/examples/
# - ls -1 examples/ | sed -e 's/\.rs$//' | xargs -I "{}" find target/* -name "*{}*" -type f -delete
#branches:
# only:
# # release tags
# - /^v\d+\.\d+\.\d+.*$/
# - master
notifications:
email:
on_success: never

11
CHANGELOG.md
View File

@@ -1,6 +1,15 @@
Tantivy 0.17
================================
- Change to non-strict schema. Ignore fields in data which are not defined in schema. Previously this returned an error. #1211
Tantivy 0.16.2
================================
- Bugfix in FuzzyTermQuery. (tranposition_cost_one was not doing anything)
Tantivy 0.16.1
========================
- Major Bugfix on multivalued fastfield. #1151
- Demux operation (@PSeitz)
Tantivy 0.16.0
=========================
@@ -113,7 +122,7 @@ Tantivy 0.12.0
## How to update?
Crates relying on custom tokenizer, or registering tokenizer in the manager will require some
minor changes. Check https://github.com/tantivy-search/tantivy/blob/main/examples/custom_tokenizer.rs
minor changes. Check https://github.com/quickwit-inc/tantivy/blob/main/examples/custom_tokenizer.rs
to check for some code sample.
Tantivy 0.11.3

20
Cargo.toml
View File

@@ -1,13 +1,13 @@
[package]
name = "tantivy"
version = "0.16.1"
version = "0.17.0-dev"
authors = ["Paul Masurel <paul.masurel@gmail.com>"]
license = "MIT"
categories = ["database-implementations", "data-structures"]
description = """Search engine library"""
documentation = "https://docs.rs/tantivy/"
homepage = "https://github.com/tantivy-search/tantivy"
repository = "https://github.com/tantivy-search/tantivy"
homepage = "https://github.com/quickwit-inc/tantivy"
repository = "https://github.com/quickwit-inc/tantivy"
readme = "README.md"
keywords = ["search", "information", "retrieval"]
edition = "2018"
@@ -19,8 +19,8 @@ crc32fast = "1.2.1"
once_cell = "1.7.2"
regex ={ version = "1.5.4", default-features = false, features = ["std"] }
tantivy-fst = "0.3"
memmap2 = {version = "0.3", optional=true}
lz4_flex = { version = "0.8.0", default-features = false, features = ["checked-decode"], optional = true }
memmap2 = {version = "0.5", optional=true}
lz4_flex = { version = "0.9", default-features = false, features = ["checked-decode"], optional = true }
brotli = { version = "3.3", optional = true }
snap = { version = "1.0.5", optional = true }
tempfile = { version = "3.2", optional = true }
@@ -37,7 +37,7 @@ tantivy-query-grammar = { version="0.15.0", path="./query-grammar" }
tantivy-bitpacker = { version="0.1", path="./bitpacker" }
common = { version = "0.1", path = "./common/", package = "tantivy-common" }
fastfield_codecs = { version="0.1", path="./fastfield_codecs", default-features = false }
ownedbytes = { version="0.1", path="./ownedbytes" }
ownedbytes = { version="0.2", path="./ownedbytes" }
stable_deref_trait = "1.2"
rust-stemmers = "1.2"
downcast-rs = "1.2"
@@ -46,15 +46,15 @@ census = "0.4"
fnv = "1.0.7"
thiserror = "1.0.24"
htmlescape = "0.3.1"
fail = "0.4"
fail = "0.5"
murmurhash32 = "0.2"
chrono = "0.4.19"
smallvec = "1.6.1"
rayon = "1.5"
lru = "0.6.5"
lru = "0.7.0"
fastdivide = "0.3"
itertools = "0.10.0"
measure_time = "0.7.0"
measure_time = "0.8.0"
[target.'cfg(windows)'.dependencies]
winapi = "0.3.9"
@@ -69,7 +69,7 @@ test-env-log = "0.2.7"
env_logger = "0.9.0"
[dev-dependencies.fail]
version = "0.4"
version = "0.5"
features = ["failpoints"]
[profile.release]

16
README.md
View File

@@ -1,8 +1,8 @@
[![Docs](https://docs.rs/tantivy/badge.svg)](https://docs.rs/crate/tantivy/)
[![Build Status](https://github.com/tantivy-search/tantivy/actions/workflows/test.yml/badge.svg)](https://github.com/tantivy-search/tantivy/actions/workflows/test.yml)
[![codecov](https://codecov.io/gh/tantivy-search/tantivy/branch/main/graph/badge.svg)](https://codecov.io/gh/tantivy-search/tantivy)
[![Join the chat at https://gitter.im/tantivy-search/tantivy](https://badges.gitter.im/tantivy-search/tantivy.svg)](https://gitter.im/tantivy-search/tantivy?utm_source=badge&utm_medium=badge&utm_campaign=pr-badge&utm_content=badge)
[![Build Status](https://github.com/quickwit-inc/tantivy/actions/workflows/test.yml/badge.svg)](https://github.com/quickwit-inc/tantivy/actions/workflows/test.yml)
[![codecov](https://codecov.io/gh/quickwit-inc/tantivy/branch/main/graph/badge.svg)](https://codecov.io/gh/quickwit-inc/tantivy)
[![Join the chat at https://discord.gg/MT27AG5EVE](https://shields.io/discord/908281611840282624?label=chat%20on%20discord)](https://discord.gg/MT27AG5EVE)
[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
[![Crates.io](https://img.shields.io/crates/v/tantivy.svg)](https://crates.io/crates/tantivy)
@@ -17,9 +17,6 @@
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/6)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/6)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/7)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/7)
[![Become a patron](https://c5.patreon.com/external/logo/become_a_patron_button.png)](https://www.patreon.com/fulmicoton)
**Tantivy** is a **full text search engine library** written in Rust.
It is closer to [Apache Lucene](https://lucene.apache.org/) than to [Elasticsearch](https://www.elastic.co/products/elasticsearch) or [Apache Solr](https://lucene.apache.org/solr/) in the sense it is not
@@ -78,13 +75,12 @@ It walks you through getting a wikipedia search engine up and running in a few m
There are many ways to support this project.
- Use Tantivy and tell us about your experience on [Gitter](https://gitter.im/tantivy-search/tantivy) or by email (paul.masurel@gmail.com)
- Use Tantivy and tell us about your experience on [Discord](https://discord.gg/MT27AG5EVE) or by email (paul.masurel@gmail.com)
- Report bugs
- Write a blog post
- Help with documentation by asking questions or submitting PRs
- Contribute code (you can join [our Gitter](https://gitter.im/tantivy-search/tantivy))
- Contribute code (you can join [our Discord server](https://discord.gg/MT27AG5EVE))
- Talk about Tantivy around you
- [![Become a patron](https://c5.patreon.com/external/logo/become_a_patron_button.png)](https://www.patreon.com/fulmicoton)
# Contributing code
@@ -96,7 +92,7 @@ Tantivy compiles on stable Rust but requires `Rust >= 1.27`.
To check out and run tests, you can simply run:
```bash
git clone https://github.com/tantivy-search/tantivy.git
git clone https://github.com/quickwit-inc/tantivy.git
cd tantivy
cargo build
```

2
bitpacker/Cargo.toml
View File

@@ -6,7 +6,7 @@ authors = ["Paul Masurel <paul.masurel@gmail.com>"]
license = "MIT"
categories = []
description = """Tantivy-sub crate: bitpacking"""
repository = "https://github.com/tantivy-search/tantivy"
repository = "https://github.com/quickwit-inc/tantivy"
keywords = []

1
common/Cargo.toml
View File

@@ -10,6 +10,7 @@ description = "common traits and utility functions used by multiple tantivy subc
[dependencies]
byteorder = "1.4.3"
ownedbytes = { version="0.2", path="../ownedbytes" }
[dev-dependencies]
proptest = "1.0.0"

376
common/src/bitset.rs
View File

@@ -1,5 +1,8 @@
use std::fmt;
use ownedbytes::OwnedBytes;
use std::convert::TryInto;
use std::io::Write;
use std::u64;
use std::{fmt, io};
#[derive(Clone, Copy, Eq, PartialEq)]
pub struct TinySet(u64);
@@ -14,6 +17,7 @@ pub struct TinySetIterator(TinySet);
impl Iterator for TinySetIterator {
type Item = u32;
#[inline]
fn next(&mut self) -> Option<Self::Item> {
self.0.pop_lowest()
}
@@ -28,30 +32,58 @@ impl IntoIterator for TinySet {
}
impl TinySet {
pub fn serialize<T: Write>(&self, writer: &mut T) -> io::Result<()> {
writer.write_all(self.0.to_le_bytes().as_ref())
}
pub fn into_bytes(self) -> [u8; 8] {
self.0.to_le_bytes()
}
#[inline]
pub fn deserialize(data: [u8; 8]) -> Self {
let val: u64 = u64::from_le_bytes(data);
TinySet(val)
}
/// Returns an empty `TinySet`.
#[inline]
pub fn empty() -> TinySet {
TinySet(0u64)
}
/// Returns a full `TinySet`.
#[inline]
pub fn full() -> TinySet {
TinySet::empty().complement()
}
pub fn clear(&mut self) {
self.0 = 0u64;
}
#[inline]
/// Returns the complement of the set in `[0, 64[`.
///
/// Careful on making this function public, as it will break the padding handling in the last
/// bucket.
fn complement(self) -> TinySet {
TinySet(!self.0)
}
#[inline]
/// Returns true iff the `TinySet` contains the element `el`.
pub fn contains(self, el: u32) -> bool {
!self.intersect(TinySet::singleton(el)).is_empty()
}
#[inline]
/// Returns the number of elements in the TinySet.
pub fn len(self) -> u32 {
self.0.count_ones()
}
#[inline]
/// Returns the intersection of `self` and `other`
pub fn intersect(self, other: TinySet) -> TinySet {
TinySet(self.0 & other.0)
@@ -64,13 +96,21 @@ impl TinySet {
TinySet(1u64 << u64::from(el))
}
/// Insert a new element within [0..64[
/// Insert a new element within [0..64)
#[inline]
pub fn insert(self, el: u32) -> TinySet {
self.union(TinySet::singleton(el))
}
/// Insert a new element within [0..64[
/// Removes an element within [0..64)
#[inline]
pub fn remove(self, el: u32) -> TinySet {
self.intersect(TinySet::singleton(el).complement())
}
/// Insert a new element within [0..64)
///
/// returns true if the set changed
#[inline]
pub fn insert_mut(&mut self, el: u32) -> bool {
let old = *self;
@@ -78,6 +118,16 @@ impl TinySet {
old != *self
}
/// Remove a element within [0..64)
///
/// returns true if the set changed
#[inline]
pub fn remove_mut(&mut self, el: u32) -> bool {
let old = *self;
*self = old.remove(el);
old != *self
}
/// Returns the union of two tinysets
#[inline]
pub fn union(self, other: TinySet) -> TinySet {
@@ -123,7 +173,7 @@ impl TinySet {
#[derive(Clone)]
pub struct BitSet {
tinysets: Box<[TinySet]>,
len: usize,
len: u64,
max_value: u32,
}
@@ -132,18 +182,47 @@ fn num_buckets(max_val: u32) -> u32 {
}
impl BitSet {
/// serialize a `BitSet`.
///
pub fn serialize<T: Write>(&self, writer: &mut T) -> io::Result<()> {
writer.write_all(self.max_value.to_le_bytes().as_ref())?;
for tinyset in self.tinysets.iter().cloned() {
writer.write_all(&tinyset.into_bytes())?;
}
writer.flush()?;
Ok(())
}
/// Create a new `BitSet` that may contain elements
/// within `[0, max_val[`.
/// within `[0, max_val)`.
pub fn with_max_value(max_value: u32) -> BitSet {
let num_buckets = num_buckets(max_value);
let tinybisets = vec![TinySet::empty(); num_buckets as usize].into_boxed_slice();
let tinybitsets = vec![TinySet::empty(); num_buckets as usize].into_boxed_slice();
BitSet {
tinysets: tinybisets,
tinysets: tinybitsets,
len: 0,
max_value,
}
}
/// Create a new `BitSet` that may contain elements. Initially all values will be set.
/// within `[0, max_val)`.
pub fn with_max_value_and_full(max_value: u32) -> BitSet {
let num_buckets = num_buckets(max_value);
let mut tinybitsets = vec![TinySet::full(); num_buckets as usize].into_boxed_slice();
// Fix padding
let lower = max_value % 64u32;
if lower != 0 {
tinybitsets[tinybitsets.len() - 1] = TinySet::range_lower(lower);
}
BitSet {
tinysets: tinybitsets,
len: max_value as u64,
max_value,
}
}
/// Removes all elements from the `BitSet`.
pub fn clear(&mut self) {
for tinyset in self.tinysets.iter_mut() {
@@ -151,12 +230,28 @@ impl BitSet {
}
}
/// Intersect with serialized bitset
pub fn intersect_update(&mut self, other: &ReadOnlyBitSet) {
self.intersect_update_with_iter(other.iter_tinysets());
}
/// Intersect with tinysets
fn intersect_update_with_iter(&mut self, other: impl Iterator<Item = TinySet>) {
self.len = 0;
for (left, right) in self.tinysets.iter_mut().zip(other) {
*left = left.intersect(right);
self.len += left.len() as u64;
}
}
/// Returns the number of elements in the `BitSet`.
#[inline]
pub fn len(&self) -> usize {
self.len
self.len as usize
}
/// Inserts an element in the `BitSet`
#[inline]
pub fn insert(&mut self, el: u32) {
// we do not check saturated els.
let higher = el / 64u32;
@@ -168,7 +263,21 @@ impl BitSet {
};
}
/// Inserts an element in the `BitSet`
#[inline]
pub fn remove(&mut self, el: u32) {
// we do not check saturated els.
let higher = el / 64u32;
let lower = el % 64u32;
self.len -= if self.tinysets[higher as usize].remove_mut(lower) {
1
} else {
0
};
}
/// Returns true iff the elements is in the `BitSet`.
#[inline]
pub fn contains(&self, el: u32) -> bool {
self.tinyset(el / 64u32).contains(el % 64)
}
@@ -186,6 +295,7 @@ impl BitSet {
.map(|delta_bucket| bucket + delta_bucket as u32)
}
#[inline]
pub fn max_value(&self) -> u32 {
self.max_value
}
@@ -198,16 +308,237 @@ impl BitSet {
}
}
/// Serialized BitSet.
#[derive(Clone)]
pub struct ReadOnlyBitSet {
data: OwnedBytes,
max_value: u32,
}
pub fn intersect_bitsets(left: &ReadOnlyBitSet, other: &ReadOnlyBitSet) -> ReadOnlyBitSet {
assert_eq!(left.max_value(), other.max_value());
assert_eq!(left.data.len(), other.data.len());
let union_tinyset_it = left
.iter_tinysets()
.zip(other.iter_tinysets())
.map(|(left_tinyset, right_tinyset)| left_tinyset.intersect(right_tinyset));
let mut output_dataset: Vec<u8> = Vec::with_capacity(left.data.len());
for tinyset in union_tinyset_it {
output_dataset.extend_from_slice(&tinyset.into_bytes());
}
ReadOnlyBitSet {
data: OwnedBytes::new(output_dataset),
max_value: left.max_value(),
}
}
impl ReadOnlyBitSet {
pub fn open(data: OwnedBytes) -> Self {
let (max_value_data, data) = data.split(4);
assert_eq!(data.len() % 8, 0);
let max_value: u32 = u32::from_le_bytes(max_value_data.as_ref().try_into().unwrap());
ReadOnlyBitSet { data, max_value }
}
/// Number of elements in the bitset.
#[inline]
pub fn len(&self) -> usize {
self.iter_tinysets()
.map(|tinyset| tinyset.len() as usize)
.sum()
}
/// Iterate the tinyset on the fly from serialized data.
///
#[inline]
fn iter_tinysets(&self) -> impl Iterator<Item = TinySet> + '_ {
self.data.chunks_exact(8).map(move |chunk| {
let tinyset: TinySet = TinySet::deserialize(chunk.try_into().unwrap());
tinyset
})
}
/// Iterate over the positions of the elements.
///
#[inline]
pub fn iter(&self) -> impl Iterator<Item = u32> + '_ {
self.iter_tinysets()
.enumerate()
.flat_map(move |(chunk_num, tinyset)| {
let chunk_base_val = chunk_num as u32 * 64;
tinyset
.into_iter()
.map(move |val| val + chunk_base_val)
.take_while(move |doc| *doc < self.max_value)
})
}
/// Returns true iff the elements is in the `BitSet`.
#[inline]
pub fn contains(&self, el: u32) -> bool {
let byte_offset = el / 8u32;
let b: u8 = self.data[byte_offset as usize];
let shift = (el % 8) as u8;
b & (1u8 << shift) != 0
}
/// Maximum value the bitset may contain.
/// (Note this is not the maximum value contained in the set.)
///
/// A bitset has an intrinsic capacity.
/// It only stores elements within [0..max_value).
#[inline]
pub fn max_value(&self) -> u32 {
self.max_value
}
/// Number of bytes used in the bitset representation.
pub fn num_bytes(&self) -> usize {
self.data.len()
}
}
impl<'a> From<&'a BitSet> for ReadOnlyBitSet {
fn from(bitset: &'a BitSet) -> ReadOnlyBitSet {
let mut buffer = Vec::with_capacity(bitset.tinysets.len() * 8 + 4);
bitset
.serialize(&mut buffer)
.expect("serializing into a buffer should never fail");
ReadOnlyBitSet::open(OwnedBytes::new(buffer))
}
}
#[cfg(test)]
mod tests {
use super::BitSet;
use super::ReadOnlyBitSet;
use super::TinySet;
use ownedbytes::OwnedBytes;
use rand::distributions::Bernoulli;
use rand::rngs::StdRng;
use rand::{Rng, SeedableRng};
use std::collections::HashSet;
#[test]
fn test_read_serialized_bitset_full_multi() {
for i in 0..1000 {
let bitset = BitSet::with_max_value_and_full(i);
let mut out = vec![];
bitset.serialize(&mut out).unwrap();
let bitset = ReadOnlyBitSet::open(OwnedBytes::new(out));
assert_eq!(bitset.len() as usize, i as usize);
}
}
#[test]
fn test_read_serialized_bitset_full_block() {
let bitset = BitSet::with_max_value_and_full(64);
let mut out = vec![];
bitset.serialize(&mut out).unwrap();
let bitset = ReadOnlyBitSet::open(OwnedBytes::new(out));
assert_eq!(bitset.len() as usize, 64 as usize);
}
#[test]
fn test_read_serialized_bitset_full() {
let mut bitset = BitSet::with_max_value_and_full(5);
bitset.remove(3);
let mut out = vec![];
bitset.serialize(&mut out).unwrap();
let bitset = ReadOnlyBitSet::open(OwnedBytes::new(out));
assert_eq!(bitset.len(), 4);
}
#[test]
fn test_bitset_intersect() {
let bitset_serialized = {
let mut bitset = BitSet::with_max_value_and_full(5);
bitset.remove(1);
bitset.remove(3);
let mut out = vec![];
bitset.serialize(&mut out).unwrap();
ReadOnlyBitSet::open(OwnedBytes::new(out))
};
let mut bitset = BitSet::with_max_value_and_full(5);
bitset.remove(1);
bitset.intersect_update(&bitset_serialized);
assert!(bitset.contains(0));
assert!(!bitset.contains(1));
assert!(bitset.contains(2));
assert!(!bitset.contains(3));
assert!(bitset.contains(4));
bitset.intersect_update_with_iter(vec![TinySet::singleton(0)].into_iter());
assert!(bitset.contains(0));
assert!(!bitset.contains(1));
assert!(!bitset.contains(2));
assert!(!bitset.contains(3));
assert!(!bitset.contains(4));
assert_eq!(bitset.len(), 1);
bitset.intersect_update_with_iter(vec![TinySet::singleton(1)].into_iter());
assert!(!bitset.contains(0));
assert!(!bitset.contains(1));
assert!(!bitset.contains(2));
assert!(!bitset.contains(3));
assert!(!bitset.contains(4));
assert_eq!(bitset.len(), 0);
}
#[test]
fn test_read_serialized_bitset_empty() {
let mut bitset = BitSet::with_max_value(5);
bitset.insert(3);
let mut out = vec![];
bitset.serialize(&mut out).unwrap();
let bitset = ReadOnlyBitSet::open(OwnedBytes::new(out));
assert_eq!(bitset.len(), 1);
{
let bitset = BitSet::with_max_value(5);
let mut out = vec![];
bitset.serialize(&mut out).unwrap();
let bitset = ReadOnlyBitSet::open(OwnedBytes::new(out));
assert_eq!(bitset.len(), 0);
}
}
#[test]
fn test_tiny_set_remove() {
{
let mut u = TinySet::empty().insert(63u32).insert(5).remove(63u32);
assert_eq!(u.pop_lowest(), Some(5u32));
assert!(u.pop_lowest().is_none());
}
{
let mut u = TinySet::empty()
.insert(63u32)
.insert(1)
.insert(5)
.remove(63u32);
assert_eq!(u.pop_lowest(), Some(1u32));
assert_eq!(u.pop_lowest(), Some(5u32));
assert!(u.pop_lowest().is_none());
}
{
let mut u = TinySet::empty().insert(1).remove(63u32);
assert_eq!(u.pop_lowest(), Some(1u32));
assert!(u.pop_lowest().is_none());
}
{
let mut u = TinySet::empty().insert(1).remove(1u32);
assert!(u.pop_lowest().is_none());
}
}
#[test]
fn test_tiny_set() {
assert!(TinySet::empty().is_empty());
@@ -233,6 +564,17 @@ mod tests {
assert_eq!(u.pop_lowest(), Some(63u32));
assert!(u.pop_lowest().is_none());
}
{
let mut u = TinySet::empty().insert(63u32).insert(5);
assert_eq!(u.pop_lowest(), Some(5u32));
assert_eq!(u.pop_lowest(), Some(63u32));
assert!(u.pop_lowest().is_none());
}
{
let original = TinySet::empty().insert(63u32).insert(5);
let after_serialize_deserialize = TinySet::deserialize(original.into_bytes());
assert_eq!(original, after_serialize_deserialize);
}
}
#[test]
@@ -249,6 +591,16 @@ mod tests {
assert_eq!(hashset.contains(&el), bitset.contains(el));
}
assert_eq!(bitset.max_value(), max_value);
// test deser
let mut data = vec![];
bitset.serialize(&mut data).unwrap();
let ro_bitset = ReadOnlyBitSet::open(OwnedBytes::new(data));
for el in 0..max_value {
assert_eq!(hashset.contains(&el), ro_bitset.contains(el));
}
assert_eq!(ro_bitset.max_value(), max_value);
assert_eq!(ro_bitset.len(), els.len());
};
test_against_hashset(&[], 0);
@@ -313,6 +665,14 @@ mod tests {
assert_eq!(bitset.len(), 2);
bitset.insert(104u32);
assert_eq!(bitset.len(), 3);
bitset.remove(105u32);
assert_eq!(bitset.len(), 3);
bitset.remove(104u32);
assert_eq!(bitset.len(), 2);
bitset.remove(3u32);
assert_eq!(bitset.len(), 1);
bitset.remove(103u32);
assert_eq!(bitset.len(), 0);
}
pub fn sample_with_seed(n: u32, ratio: f64, seed_val: u8) -> Vec<u32> {

2
common/src/lib.rs
View File

@@ -1,3 +1,5 @@
#![allow(clippy::len_without_is_empty)]
use std::ops::Deref;
pub use byteorder::LittleEndian as Endianness;

4
doc/src/index_sorting.md
View File

@@ -38,7 +38,7 @@ Note: Tantivy 0.16 does not do this optimization yet.
In principle there are many algorithms possible that exploit the monotonically increasing nature. (aggregations maybe?)
## Usage
The index sorting can be configured setting [`sort_by_field`](https://github.com/tantivy-search/tantivy/blob/000d76b11a139a84b16b9b95060a1c93e8b9851c/src/core/index_meta.rs#L238) on `IndexSettings` and passing it to a `IndexBuilder`. As of tantvy 0.16 only fast fields are allowed to be used.
The index sorting can be configured setting [`sort_by_field`](https://github.com/quickwit-inc/tantivy/blob/000d76b11a139a84b16b9b95060a1c93e8b9851c/src/core/index_meta.rs#L238) on `IndexSettings` and passing it to a `IndexBuilder`. As of tantvy 0.16 only fast fields are allowed to be used.
```
let settings = IndexSettings {
@@ -55,7 +55,7 @@ let index = index_builder.create_in_ram().unwrap();
## Implementation details
Sorting an index is applied in the serialization step. In general there are two serialization steps: [Finishing a single segment](https://github.com/tantivy-search/tantivy/blob/000d76b11a139a84b16b9b95060a1c93e8b9851c/src/indexer/segment_writer.rs#L338) and [merging multiple segments](https://github.com/tantivy-search/tantivy/blob/000d76b11a139a84b16b9b95060a1c93e8b9851c/src/indexer/merger.rs#L1073).
Sorting an index is applied in the serialization step. In general there are two serialization steps: [Finishing a single segment](https://github.com/quickwit-inc/tantivy/blob/000d76b11a139a84b16b9b95060a1c93e8b9851c/src/indexer/segment_writer.rs#L338) and [merging multiple segments](https://github.com/quickwit-inc/tantivy/blob/000d76b11a139a84b16b9b95060a1c93e8b9851c/src/indexer/merger.rs#L1073).
In both cases we generate a docid mapping reflecting the sort. This mapping is used when serializing the different components (doc store, fastfields, posting list, normfield, facets).

6
examples/basic_search.rs
View File

@@ -96,7 +96,7 @@ fn main() -> tantivy::Result<()> {
);
// ... and add it to the `IndexWriter`.
index_writer.add_document(old_man_doc);
index_writer.add_document(old_man_doc)?;
// For convenience, tantivy also comes with a macro to
// reduce the boilerplate above.
@@ -110,7 +110,7 @@ fn main() -> tantivy::Result<()> {
fresh and green with every spring, carrying in their lower leaf junctures the \
debris of the winter’s flooding; and sycamores with mottled, white, recumbent \
limbs and branches that arch over the pool"
));
))?;
// Multivalued field just need to be repeated.
index_writer.add_document(doc!(
@@ -120,7 +120,7 @@ fn main() -> tantivy::Result<()> {
enterprise which you have regarded with such evil forebodings. I arrived here \
yesterday, and my first task is to assure my dear sister of my welfare and \
increasing confidence in the success of my undertaking."
));
))?;
// This is an example, so we will only index 3 documents
// here. You can check out tantivy's tutorial to index

8
examples/custom_collector.rs
View File

@@ -145,23 +145,23 @@ fn main() -> tantivy::Result<()> {
product_description => "While it is ok for short distance travel, this broom \
was designed quiditch. It will up your game.",
price => 30_200u64
));
))?;
index_writer.add_document(doc!(
product_name => "Turbulobroom",
product_description => "You might have heard of this broom before : it is the sponsor of the Wales team.\
You'll enjoy its sharp turns, and rapid acceleration",
price => 29_240u64
));
))?;
index_writer.add_document(doc!(
product_name => "Broomio",
product_description => "Great value for the price. This broom is a market favorite",
price => 21_240u64
));
))?;
index_writer.add_document(doc!(
product_name => "Whack a Mole",
product_description => "Prime quality bat.",
price => 5_200u64
));
))?;
index_writer.commit()?;
let reader = index.reader()?;

6
examples/custom_tokenizer.rs
View File

@@ -68,7 +68,7 @@ fn main() -> tantivy::Result<()> {
title => "The Old Man and the Sea",
body => "He was an old man who fished alone in a skiff in the Gulf Stream and \
he had gone eighty-four days now without taking a fish."
));
))?;
index_writer.add_document(doc!(
title => "Of Mice and Men",
body => r#"A few miles south of Soledad, the Salinas River drops in close to the hillside
@@ -79,14 +79,14 @@ fn main() -> tantivy::Result<()> {
fresh and green with every spring, carrying in their lower leaf junctures the
debris of the winter’s flooding; and sycamores with mottled, white, recumbent
limbs and branches that arch over the pool"#
));
))?;
index_writer.add_document(doc!(
title => "Frankenstein",
body => r#"You will rejoice to hear that no disaster has accompanied the commencement of an
enterprise which you have regarded with such evil forebodings. I arrived here
yesterday, and my first task is to assure my dear sister of my welfare and
increasing confidence in the success of my undertaking."#
));
))?;
index_writer.commit()?;
let reader = index.reader()?;

8
examples/deleting_updating_documents.rs
View File

@@ -76,15 +76,15 @@ fn main() -> tantivy::Result<()> {
index_writer.add_document(doc!(
isbn => "978-0099908401",
title => "The old Man and the see"
));
))?;
index_writer.add_document(doc!(
isbn => "978-0140177398",
title => "Of Mice and Men",
));
))?;
index_writer.add_document(doc!(
title => "Frankentein", //< Oops there is a typo here.
isbn => "978-9176370711",
));
))?;
index_writer.commit()?;
let reader = index.reader()?;
@@ -122,7 +122,7 @@ fn main() -> tantivy::Result<()> {
index_writer.add_document(doc!(
title => "Frankenstein",
isbn => "978-9176370711",
));
))?;
// You are guaranteed that your clients will only observe your index in
// the state it was in after a commit.

16
examples/faceted_search.rs
View File

@@ -35,35 +35,35 @@ fn main() -> tantivy::Result<()> {
index_writer.add_document(doc!(
name => "Cat",
classification => Facet::from("/Felidae/Felinae/Felis")
));
))?;
index_writer.add_document(doc!(
name => "Canada lynx",
classification => Facet::from("/Felidae/Felinae/Lynx")
));
))?;
index_writer.add_document(doc!(
name => "Cheetah",
classification => Facet::from("/Felidae/Felinae/Acinonyx")
));
))?;
index_writer.add_document(doc!(
name => "Tiger",
classification => Facet::from("/Felidae/Pantherinae/Panthera")
));
))?;
index_writer.add_document(doc!(
name => "Lion",
classification => Facet::from("/Felidae/Pantherinae/Panthera")
));
))?;
index_writer.add_document(doc!(
name => "Jaguar",
classification => Facet::from("/Felidae/Pantherinae/Panthera")
));
))?;
index_writer.add_document(doc!(
name => "Sunda clouded leopard",
classification => Facet::from("/Felidae/Pantherinae/Neofelis")
));
))?;
index_writer.add_document(doc!(
name => "Fossa",
classification => Facet::from("/Eupleridae/Cryptoprocta")
));
))?;
index_writer.commit()?;
let reader = index.reader()?;

6
examples/faceted_search_with_tweaked_score.rs
View File

@@ -20,14 +20,14 @@ fn main() -> tantivy::Result<()> {
title => "Fried egg",
ingredient => Facet::from("/ingredient/egg"),
ingredient => Facet::from("/ingredient/oil"),
));
))?;
index_writer.add_document(doc!(
title => "Scrambled egg",
ingredient => Facet::from("/ingredient/egg"),
ingredient => Facet::from("/ingredient/butter"),
ingredient => Facet::from("/ingredient/milk"),
ingredient => Facet::from("/ingredient/salt"),
));
))?;
index_writer.add_document(doc!(
title => "Egg rolls",
ingredient => Facet::from("/ingredient/egg"),
@@ -36,7 +36,7 @@ fn main() -> tantivy::Result<()> {
ingredient => Facet::from("/ingredient/oil"),
ingredient => Facet::from("/ingredient/tortilla-wrap"),
ingredient => Facet::from("/ingredient/mushroom"),
));
))?;
index_writer.commit()?;
let reader = index.reader()?;

8
examples/integer_range_search.rs
View File

@@ -7,7 +7,7 @@ use tantivy::query::RangeQuery;
use tantivy::schema::{Schema, INDEXED};
use tantivy::{doc, Index, Result};
fn run() -> Result<()> {
fn main() -> Result<()> {
// For the sake of simplicity, this schema will only have 1 field
let mut schema_builder = Schema::builder();
@@ -19,7 +19,7 @@ fn run() -> Result<()> {
{
let mut index_writer = index.writer_with_num_threads(1, 6_000_000)?;
for year in 1950u64..2019u64 {
index_writer.add_document(doc!(year_field => year));
index_writer.add_document(doc!(year_field => year))?;
}
index_writer.commit()?;
// The index will be a range of years
@@ -33,7 +33,3 @@ fn run() -> Result<()> {
assert_eq!(num_60s_books, 10);
Ok(())
}
fn main() {
run().unwrap()
}

6
examples/iterating_docs_and_positions.rs
View File

@@ -25,9 +25,9 @@ fn main() -> tantivy::Result<()> {
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_with_num_threads(1, 50_000_000)?;
index_writer.add_document(doc!(title => "The Old Man and the Sea"));
index_writer.add_document(doc!(title => "Of Mice and Men"));
index_writer.add_document(doc!(title => "The modern Promotheus"));
index_writer.add_document(doc!(title => "The Old Man and the Sea"))?;
index_writer.add_document(doc!(title => "Of Mice and Men"))?;
index_writer.add_document(doc!(title => "The modern Promotheus"))?;
index_writer.commit()?;
let reader = index.reader()?;

10
examples/multiple_producer.rs
View File

@@ -29,7 +29,7 @@ use std::sync::{Arc, RwLock};
use std::thread;
use std::time::Duration;
use tantivy::schema::{Schema, STORED, TEXT};
use tantivy::{doc, Index, IndexWriter, Opstamp};
use tantivy::{doc, Index, IndexWriter, Opstamp, TantivyError};
fn main() -> tantivy::Result<()> {
// # Defining the schema
@@ -59,10 +59,11 @@ fn main() -> tantivy::Result<()> {
fresh and green with every spring, carrying in their lower leaf junctures the \
debris of the winter’s flooding; and sycamores with mottled, white, recumbent \
limbs and branches that arch over the pool"
));
))?;
println!("add doc {} from thread 1 - opstamp {}", i, opstamp);
thread::sleep(Duration::from_millis(20));
}
Result::<(), TantivyError>::Ok(())
});
// # Second indexing thread.
@@ -78,11 +79,12 @@ fn main() -> tantivy::Result<()> {
index_writer_rlock.add_document(doc!(
title => "Manufacturing consent",
body => "Some great book description..."
))
))?
};
println!("add doc {} from thread 2 - opstamp {}", i, opstamp);
thread::sleep(Duration::from_millis(10));
}
Result::<(), TantivyError>::Ok(())
});
// # In the main thread, we commit 10 times, once every 500ms.
@@ -90,7 +92,7 @@ fn main() -> tantivy::Result<()> {
let opstamp: Opstamp = {
// Committing or rollbacking on the other hand requires write lock. This will block other threads.
let mut index_writer_wlock = index_writer.write().unwrap();
index_writer_wlock.commit().unwrap()
index_writer_wlock.commit()?
};
println!("committed with opstamp {}", opstamp);
thread::sleep(Duration::from_millis(500));

12
examples/pre_tokenized_text.rs
View File

@@ -68,7 +68,7 @@ fn main() -> tantivy::Result<()> {
let old_man_doc = doc!(title => title_tok, body => body_tok);
// ... now let's just add it to the IndexWriter
index_writer.add_document(old_man_doc);
index_writer.add_document(old_man_doc)?;
// Pretokenized text can also be fed as JSON
let short_man_json = r#"{
@@ -84,7 +84,7 @@ fn main() -> tantivy::Result<()> {
let short_man_doc = schema.parse_document(short_man_json)?;
index_writer.add_document(short_man_doc);
index_writer.add_document(short_man_doc)?;
// Let's commit changes
index_writer.commit()?;
@@ -106,9 +106,7 @@ fn main() -> tantivy::Result<()> {
IndexRecordOption::Basic,
);
let (top_docs, count) = searcher
.search(&query, &(TopDocs::with_limit(2), Count))
.unwrap();
let (top_docs, count) = searcher.search(&query, &(TopDocs::with_limit(2), Count))?;
assert_eq!(count, 2);
@@ -129,9 +127,7 @@ fn main() -> tantivy::Result<()> {
IndexRecordOption::Basic,
);
let (_top_docs, count) = searcher
.search(&query, &(TopDocs::with_limit(2), Count))
.unwrap();
let (_top_docs, count) = searcher.search(&query, &(TopDocs::with_limit(2), Count))?;
assert_eq!(count, 0);

2
examples/snippet.rs
View File

@@ -40,7 +40,7 @@ fn main() -> tantivy::Result<()> {
fresh and green with every spring, carrying in their lower leaf junctures the \
debris of the winter’s flooding; and sycamores with mottled, white, recumbent \
limbs and branches that arch over the pool"
));
))?;
// ...
index_writer.commit()?;

6
examples/stop_words.rs
View File

@@ -68,7 +68,7 @@ fn main() -> tantivy::Result<()> {
title => "The Old Man and the Sea",
body => "He was an old man who fished alone in a skiff in the Gulf Stream and \
he had gone eighty-four days now without taking a fish."
));
))?;
index_writer.add_document(doc!(
title => "Of Mice and Men",
@@ -80,7 +80,7 @@ fn main() -> tantivy::Result<()> {
fresh and green with every spring, carrying in their lower leaf junctures the \
debris of the winter’s flooding; and sycamores with mottled, white, recumbent \
limbs and branches that arch over the pool"
));
))?;
index_writer.add_document(doc!(
title => "Frankenstein",
@@ -88,7 +88,7 @@ fn main() -> tantivy::Result<()> {
enterprise which you have regarded with such evil forebodings. I arrived here \
yesterday, and my first task is to assure my dear sister of my welfare and \
increasing confidence in the success of my undertaking."
));
))?;
index_writer.commit()?;

2
fastfield_codecs/src/lib.rs
View File

@@ -118,7 +118,7 @@ mod tests {
);
}
}
let actual_compression = data.len() as f32 / out.len() as f32;
let actual_compression = out.len() as f32 / (data.len() as f32 * 8.0);
(estimation, actual_compression)
}
pub fn get_codec_test_data_sets() -> Vec<(Vec<u64>, &'static str)> {

14
fastfield_codecs/src/linearinterpol.rs
View File

@@ -239,11 +239,21 @@ mod tests {
use super::*;
use crate::tests::get_codec_test_data_sets;
fn create_and_validate(data: &[u64], name: &str) {
fn create_and_validate(data: &[u64], name: &str) -> (f32, f32) {
crate::tests::create_and_validate::<
LinearInterpolFastFieldSerializer,
LinearInterpolFastFieldReader,
>(data, name);
>(data, name)
}
#[test]
fn test_compression() {
let data = (10..=6_000_u64).collect::<Vec<_>>();
let (estimate, actual_compression) =
create_and_validate(&data, "simple monotonically large");
assert!(actual_compression < 0.01);
assert!(estimate < 0.01);
}
#[test]

23
fastfield_codecs/src/multilinearinterpol.rs
View File

@@ -57,7 +57,7 @@ struct Function {
impl Function {
fn calc_slope(&mut self) {
let num_vals = self.end_pos - self.start_pos;
get_slope(self.value_start_pos, self.value_end_pos, num_vals);
self.slope = get_slope(self.value_start_pos, self.value_end_pos, num_vals);
}
// split the interpolation into two function, change self and return the second split
fn split(&mut self, split_pos: u64, split_pos_value: u64) -> Function {
@@ -378,11 +378,22 @@ mod tests {
use super::*;
use crate::tests::get_codec_test_data_sets;
fn create_and_validate(data: &[u64], name: &str) {
fn create_and_validate(data: &[u64], name: &str) -> (f32, f32) {
crate::tests::create_and_validate::<
MultiLinearInterpolFastFieldSerializer,
MultiLinearInterpolFastFieldReader,
>(data, name);
>(data, name)
}
#[test]
fn test_compression() {
let data = (10..=6_000_u64).collect::<Vec<_>>();
let (estimate, actual_compression) =
create_and_validate(&data, "simple monotonically large");
assert!(actual_compression < 0.2);
assert!(estimate < 0.20);
assert!(estimate > 0.15);
assert!(actual_compression > 0.01);
}
#[test]
@@ -414,9 +425,11 @@ mod tests {
fn rand() {
for _ in 0..10 {
let mut data = (5_000..20_000)
.map(|_| rand::random::<u64>() as u64)
.map(|_| rand::random::<u32>() as u64)
.collect::<Vec<_>>();
create_and_validate(&data, "random");
let (estimate, actual_compression) = create_and_validate(&data, "random");
dbg!(estimate);
dbg!(actual_compression);
data.reverse();
create_and_validate(&data, "random");

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 = "0.1.0"
version = "0.2.0"
edition = "2018"
description = "Expose data as static slice"
license = "MIT"

52
ownedbytes/src/lib.rs
View File

@@ -76,6 +76,19 @@ impl OwnedBytes {
(left, right)
}
/// Splits the right part of the `OwnedBytes` at the given offset.
///
/// `self` is truncated to `split_len`, left with the remaining bytes.
pub fn split_off(&mut self, split_len: usize) -> OwnedBytes {
let right_box_stable_deref = self.box_stable_deref.clone();
let right_piece = OwnedBytes {
data: &self.data[split_len..],
box_stable_deref: right_box_stable_deref,
};
self.data = &self.data[..split_len];
right_piece
}
/// Returns true iff this `OwnedBytes` is empty.
#[inline]
pub fn is_empty(&self) -> bool {
@@ -124,6 +137,35 @@ impl fmt::Debug for OwnedBytes {
}
}
impl PartialEq for OwnedBytes {
fn eq(&self, other: &OwnedBytes) -> bool {
self.as_slice() == other.as_slice()
}
}
impl Eq for OwnedBytes {}
impl PartialEq<[u8]> for OwnedBytes {
fn eq(&self, other: &[u8]) -> bool {
self.as_slice() == other
}
}
impl PartialEq<str> for OwnedBytes {
fn eq(&self, other: &str) -> bool {
self.as_slice() == other.as_bytes()
}
}
impl<'a, T: ?Sized> PartialEq<&'a T> for OwnedBytes
where
OwnedBytes: PartialEq<T>,
{
fn eq(&self, other: &&'a T) -> bool {
*self == **other
}
}
impl Deref for OwnedBytes {
type Target = [u8];
@@ -287,4 +329,14 @@ mod tests {
assert_eq!(right.as_slice(), b"");
}
}
#[test]
fn test_split_off() {
let mut data = OwnedBytes::new(b"abcdef".as_ref());
assert_eq!(data, "abcdef");
assert_eq!(data.split_off(2), "cdef");
assert_eq!(data, "ab");
assert_eq!(data.split_off(1), "b");
assert_eq!(data, "a");
}
}

6
query-grammar/Cargo.toml
View File

@@ -5,9 +5,9 @@ authors = ["Paul Masurel <paul.masurel@gmail.com>"]
license = "MIT"
categories = ["database-implementations", "data-structures"]
description = """Search engine library"""
documentation = "https://tantivy-search.github.io/tantivy/tantivy/index.html"
homepage = "https://github.com/tantivy-search/tantivy"
repository = "https://github.com/tantivy-search/tantivy"
documentation = "https://quickwit-inc.github.io/tantivy/tantivy/index.html"
homepage = "https://github.com/quickwit-inc/tantivy"
repository = "https://github.com/quickwit-inc/tantivy"
readme = "README.md"
keywords = ["search", "information", "retrieval"]
edition = "2018"

8
src/collector/count_collector.rs
View File

@@ -20,10 +20,10 @@ use crate::SegmentReader;
/// let index = Index::create_in_ram(schema);
///
/// let mut index_writer = index.writer(3_000_000).unwrap();
/// index_writer.add_document(doc!(title => "The Name of the Wind"));
/// index_writer.add_document(doc!(title => "The Diary of Muadib"));
/// index_writer.add_document(doc!(title => "A Dairy Cow"));
/// index_writer.add_document(doc!(title => "The Diary of a Young Girl"));
/// index_writer.add_document(doc!(title => "The Name of the Wind")).unwrap();
/// index_writer.add_document(doc!(title => "The Diary of Muadib")).unwrap();
/// index_writer.add_document(doc!(title => "A Dairy Cow")).unwrap();
/// index_writer.add_document(doc!(title => "The Diary of a Young Girl")).unwrap();
/// assert!(index_writer.commit().is_ok());
///
/// let reader = index.reader().unwrap();

46
src/collector/facet_collector.rs
View File

@@ -103,23 +103,23 @@ fn facet_depth(facet_bytes: &[u8]) -> usize {
/// title => "The Name of the Wind",
/// facet => Facet::from("/lang/en"),
/// facet => Facet::from("/category/fiction/fantasy")
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "Dune",
/// facet => Facet::from("/lang/en"),
/// facet => Facet::from("/category/fiction/sci-fi")
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "La Vénus d'Ille",
/// facet => Facet::from("/lang/fr"),
/// facet => Facet::from("/category/fiction/fantasy"),
/// facet => Facet::from("/category/fiction/horror")
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "The Diary of a Young Girl",
/// facet => Facet::from("/lang/en"),
/// facet => Facet::from("/category/biography")
/// ));
/// ))?;
/// index_writer.commit()?;
/// }
/// let reader = index.reader()?;
@@ -470,13 +470,13 @@ mod tests {
use std::iter;
#[test]
fn test_facet_collector_drilldown() {
fn test_facet_collector_drilldown() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let facet_field = schema_builder.add_facet_field("facet", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
let num_facets: usize = 3 * 4 * 5;
let facets: Vec<Facet> = (0..num_facets)
.map(|mut n| {
@@ -491,14 +491,14 @@ mod tests {
for i in 0..num_facets * 10 {
let mut doc = Document::new();
doc.add_facet(facet_field, facets[i % num_facets].clone());
index_writer.add_document(doc);
index_writer.add_document(doc)?;
}
index_writer.commit().unwrap();
let reader = index.reader().unwrap();
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let mut facet_collector = FacetCollector::for_field(facet_field);
facet_collector.add_facet(Facet::from("/top1"));
let counts = searcher.search(&AllQuery, &facet_collector).unwrap();
let counts = searcher.search(&AllQuery, &facet_collector)?;
{
let facets: Vec<(String, u64)> = counts
@@ -518,6 +518,7 @@ mod tests {
.collect::<Vec<_>>()
);
}
Ok(())
}
#[test]
@@ -530,27 +531,28 @@ mod tests {
}
#[test]
fn test_doc_unsorted_multifacet() {
fn test_doc_unsorted_multifacet() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let facet_field = schema_builder.add_facet_field("facets", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(
facet_field => Facet::from_text(&"/subjects/A/a").unwrap(),
facet_field => Facet::from_text(&"/subjects/B/a").unwrap(),
facet_field => Facet::from_text(&"/subjects/A/b").unwrap(),
facet_field => Facet::from_text(&"/subjects/B/b").unwrap(),
));
index_writer.commit().unwrap();
let reader = index.reader().unwrap();
))?;
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
assert_eq!(searcher.num_docs(), 1);
let mut facet_collector = FacetCollector::for_field(facet_field);
facet_collector.add_facet("/subjects");
let counts = searcher.search(&AllQuery, &facet_collector).unwrap();
let counts = searcher.search(&AllQuery, &facet_collector)?;
let facets: Vec<(&Facet, u64)> = counts.get("/subjects").collect();
assert_eq!(facets[0].1, 1);
Ok(())
}
#[test]
@@ -562,16 +564,16 @@ mod tests {
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(
facet_field => Facet::from_text(&"/A/A").unwrap(),
));
))?;
index_writer.add_document(doc!(
facet_field => Facet::from_text(&"/A/B").unwrap(),
));
))?;
index_writer.add_document(doc!(
facet_field => Facet::from_text(&"/A/C/A").unwrap(),
));
))?;
index_writer.add_document(doc!(
facet_field => Facet::from_text(&"/D/C/A").unwrap(),
));
))?;
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
@@ -637,7 +639,7 @@ mod tests {
let mut index_writer = index.writer_for_tests().unwrap();
for doc in docs {
index_writer.add_document(doc);
index_writer.add_document(doc).unwrap();
}
index_writer.commit().unwrap();
let searcher = index.reader().unwrap().searcher();
@@ -677,7 +679,7 @@ mod tests {
let mut index_writer = index.writer_for_tests()?;
for doc in docs {
index_writer.add_document(doc);
index_writer.add_document(doc)?;
}
index_writer.commit()?;

23
src/collector/filter_collector_wrapper.rs
View File

@@ -25,34 +25,37 @@ use crate::{Score, SegmentReader, TantivyError};
/// use tantivy::schema::{Schema, TEXT, INDEXED, FAST};
/// use tantivy::{doc, DocAddress, Index};
///
/// # fn main() -> tantivy::Result<()> {
/// let mut schema_builder = Schema::builder();
/// let title = schema_builder.add_text_field("title", TEXT);
/// let price = schema_builder.add_u64_field("price", INDEXED | FAST);
/// let schema = schema_builder.build();
/// let index = Index::create_in_ram(schema);
///
/// let mut index_writer = index.writer_with_num_threads(1, 10_000_000).unwrap();
/// index_writer.add_document(doc!(title => "The Name of the Wind", price => 30_200u64));
/// index_writer.add_document(doc!(title => "The Diary of Muadib", price => 29_240u64));
/// index_writer.add_document(doc!(title => "A Dairy Cow", price => 21_240u64));
/// index_writer.add_document(doc!(title => "The Diary of a Young Girl", price => 20_120u64));
/// assert!(index_writer.commit().is_ok());
/// let mut index_writer = index.writer_with_num_threads(1, 10_000_000)?;
/// index_writer.add_document(doc!(title => "The Name of the Wind", price => 30_200u64))?;
/// index_writer.add_document(doc!(title => "The Diary of Muadib", price => 29_240u64))?;
/// index_writer.add_document(doc!(title => "A Dairy Cow", price => 21_240u64))?;
/// index_writer.add_document(doc!(title => "The Diary of a Young Girl", price => 20_120u64))?;
/// index_writer.commit()?;
///
/// let reader = index.reader().unwrap();
/// let reader = index.reader()?;
/// let searcher = reader.searcher();
///
/// let query_parser = QueryParser::for_index(&index, vec![title]);
/// let query = query_parser.parse_query("diary").unwrap();
/// let query = query_parser.parse_query("diary")?;
/// let no_filter_collector = FilterCollector::new(price, &|value: u64| value > 20_120u64, TopDocs::with_limit(2));
/// let top_docs = searcher.search(&query, &no_filter_collector).unwrap();
/// let top_docs = searcher.search(&query, &no_filter_collector)?;
///
/// assert_eq!(top_docs.len(), 1);
/// assert_eq!(top_docs[0].1, DocAddress::new(0, 1));
///
/// let filter_all_collector: FilterCollector<_, _, u64> = FilterCollector::new(price, &|value| value < 5u64, TopDocs::with_limit(2));
/// let filtered_top_docs = searcher.search(&query, &filter_all_collector).unwrap();
/// let filtered_top_docs = searcher.search(&query, &filter_all_collector)?;
///
/// assert_eq!(filtered_top_docs.len(), 0);
/// # Ok(())
/// # }
/// ```
pub struct FilterCollector<TCollector, TPredicate, TPredicateValue: FastValue>
where

22
src/collector/histogram_collector.rs
View File

@@ -226,10 +226,10 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut writer = index.writer_with_num_threads(1, 4_000_000)?;
writer.add_document(doc!(val_field=>12i64));
writer.add_document(doc!(val_field=>-30i64));
writer.add_document(doc!(val_field=>-12i64));
writer.add_document(doc!(val_field=>-10i64));
writer.add_document(doc!(val_field=>12i64))?;
writer.add_document(doc!(val_field=>-30i64))?;
writer.add_document(doc!(val_field=>-12i64))?;
writer.add_document(doc!(val_field=>-10i64))?;
writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
@@ -247,13 +247,13 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut writer = index.writer_with_num_threads(1, 4_000_000)?;
writer.add_document(doc!(val_field=>12i64));
writer.add_document(doc!(val_field=>12i64))?;
writer.commit()?;
writer.add_document(doc!(val_field=>-30i64));
writer.add_document(doc!(val_field=>-30i64))?;
writer.commit()?;
writer.add_document(doc!(val_field=>-12i64));
writer.add_document(doc!(val_field=>-12i64))?;
writer.commit()?;
writer.add_document(doc!(val_field=>-10i64));
writer.add_document(doc!(val_field=>-10i64))?;
writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
@@ -271,9 +271,9 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut writer = index.writer_with_num_threads(1, 4_000_000)?;
writer.add_document(doc!(date_field=>Utc.ymd(1982, 9, 17).and_hms(0, 0,0)));
writer.add_document(doc!(date_field=>Utc.ymd(1986, 3, 9).and_hms(0, 0, 0)));
writer.add_document(doc!(date_field=>Utc.ymd(1983, 9, 27).and_hms(0, 0, 0)));
writer.add_document(doc!(date_field=>Utc.ymd(1982, 9, 17).and_hms(0, 0,0)))?;
writer.add_document(doc!(date_field=>Utc.ymd(1986, 3, 9).and_hms(0, 0, 0)))?;
writer.add_document(doc!(date_field=>Utc.ymd(1983, 9, 27).and_hms(0, 0, 0)))?;
writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();

8
src/collector/mod.rs
View File

@@ -48,10 +48,10 @@ use tantivy::collector::{Count, TopDocs};
# let mut index_writer = index.writer(3_000_000)?;
# index_writer.add_document(doc!(
# title => "The Name of the Wind",
# ));
# ))?;
# index_writer.add_document(doc!(
# title => "The Diary of Muadib",
# ));
# ))?;
# index_writer.commit()?;
# let reader = index.reader()?;
# let searcher = reader.searcher();
@@ -178,9 +178,9 @@ pub trait Collector: Sync + Send {
) -> crate::Result<<Self::Child as SegmentCollector>::Fruit> {
let mut segment_collector = self.for_segment(segment_ord as u32, reader)?;
if let Some(delete_bitset) = reader.delete_bitset() {
if let Some(alive_bitset) = reader.alive_bitset() {
weight.for_each(reader, &mut |doc, score| {
if delete_bitset.is_alive(doc) {
if alive_bitset.is_alive(doc) {
segment_collector.collect(doc, score);
}
})?;

41
src/collector/multi_collector.rs
View File

@@ -112,19 +112,19 @@ impl<TFruit: Fruit> FruitHandle<TFruit> {
/// use tantivy::schema::{Schema, TEXT};
/// use tantivy::{doc, Index};
///
/// # fn main() -> tantivy::Result<()> {
/// let mut schema_builder = Schema::builder();
/// let title = schema_builder.add_text_field("title", TEXT);
/// let schema = schema_builder.build();
/// let index = Index::create_in_ram(schema);
/// let mut index_writer = index.writer(3_000_000)?;
/// index_writer.add_document(doc!(title => "The Name of the Wind"))?;
/// index_writer.add_document(doc!(title => "The Diary of Muadib"))?;
/// index_writer.add_document(doc!(title => "A Dairy Cow"))?;
/// index_writer.add_document(doc!(title => "The Diary of a Young Girl"))?;
/// index_writer.commit()?;
///
/// let mut index_writer = index.writer(3_000_000).unwrap();
/// index_writer.add_document(doc!(title => "The Name of the Wind"));
/// index_writer.add_document(doc!(title => "The Diary of Muadib"));
/// index_writer.add_document(doc!(title => "A Dairy Cow"));
/// index_writer.add_document(doc!(title => "The Diary of a Young Girl"));
/// assert!(index_writer.commit().is_ok());
///
/// let reader = index.reader().unwrap();
/// let reader = index.reader()?;
/// let searcher = reader.searcher();
///
/// let mut collectors = MultiCollector::new();
@@ -139,6 +139,8 @@ impl<TFruit: Fruit> FruitHandle<TFruit> {
///
/// assert_eq!(count, 2);
/// assert_eq!(top_docs.len(), 2);
/// # Ok(())
/// # }
/// ```
#[allow(clippy::type_complexity)]
#[derive(Default)]
@@ -252,24 +254,24 @@ mod tests {
use crate::Term;
#[test]
fn test_multi_collector() {
fn test_multi_collector() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let text = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(text=>"abc"));
index_writer.add_document(doc!(text=>"abc abc abc"));
index_writer.add_document(doc!(text=>"abc abc"));
index_writer.commit().unwrap();
index_writer.add_document(doc!(text=>""));
index_writer.add_document(doc!(text=>"abc abc abc abc"));
index_writer.add_document(doc!(text=>"abc"));
index_writer.commit().unwrap();
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text=>"abc"))?;
index_writer.add_document(doc!(text=>"abc abc abc"))?;
index_writer.add_document(doc!(text=>"abc abc"))?;
index_writer.commit()?;
index_writer.add_document(doc!(text=>""))?;
index_writer.add_document(doc!(text=>"abc abc abc abc"))?;
index_writer.add_document(doc!(text=>"abc"))?;
index_writer.commit()?;
}
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
let term = Term::from_field_text(text, "abc");
let query = TermQuery::new(term, IndexRecordOption::Basic);
@@ -280,5 +282,6 @@ mod tests {
assert_eq!(count_handler.extract(&mut multifruits), 5);
assert_eq!(topdocs_handler.extract(&mut multifruits).len(), 2);
Ok(())
}
}

29
src/collector/tests.rs
View File

@@ -25,7 +25,7 @@ pub const TEST_COLLECTOR_WITHOUT_SCORE: TestCollector = TestCollector {
};
#[test]
pub fn test_filter_collector() {
pub fn test_filter_collector() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let title = schema_builder.add_text_field("title", TEXT);
let price = schema_builder.add_u64_field("price", FAST);
@@ -33,25 +33,25 @@ pub fn test_filter_collector() {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_with_num_threads(1, 10_000_000).unwrap();
index_writer.add_document(doc!(title => "The Name of the Wind", price => 30_200u64, date => DateTime::from_str("1898-04-09T00:00:00+00:00").unwrap()));
index_writer.add_document(doc!(title => "The Diary of Muadib", price => 29_240u64, date => DateTime::from_str("2020-04-09T00:00:00+00:00").unwrap()));
index_writer.add_document(doc!(title => "The Diary of Anne Frank", price => 18_240u64, date => DateTime::from_str("2019-04-20T00:00:00+00:00").unwrap()));
index_writer.add_document(doc!(title => "A Dairy Cow", price => 21_240u64, date => DateTime::from_str("2019-04-09T00:00:00+00:00").unwrap()));
index_writer.add_document(doc!(title => "The Diary of a Young Girl", price => 20_120u64, date => DateTime::from_str("2018-04-09T00:00:00+00:00").unwrap()));
assert!(index_writer.commit().is_ok());
let mut index_writer = index.writer_with_num_threads(1, 10_000_000)?;
index_writer.add_document(doc!(title => "The Name of the Wind", price => 30_200u64, date => DateTime::from_str("1898-04-09T00:00:00+00:00").unwrap()))?;
index_writer.add_document(doc!(title => "The Diary of Muadib", price => 29_240u64, date => DateTime::from_str("2020-04-09T00:00:00+00:00").unwrap()))?;
index_writer.add_document(doc!(title => "The Diary of Anne Frank", price => 18_240u64, date => DateTime::from_str("2019-04-20T00:00:00+00:00").unwrap()))?;
index_writer.add_document(doc!(title => "A Dairy Cow", price => 21_240u64, date => DateTime::from_str("2019-04-09T00:00:00+00:00").unwrap()))?;
index_writer.add_document(doc!(title => "The Diary of a Young Girl", price => 20_120u64, date => DateTime::from_str("2018-04-09T00:00:00+00:00").unwrap()))?;
index_writer.commit()?;
let reader = index.reader().unwrap();
let reader = index.reader()?;
let searcher = reader.searcher();
let query_parser = QueryParser::for_index(&index, vec![title]);
let query = query_parser.parse_query("diary").unwrap();
let query = query_parser.parse_query("diary")?;
let filter_some_collector = FilterCollector::new(
price,
&|value: u64| value > 20_120u64,
TopDocs::with_limit(2),
);
let top_docs = searcher.search(&query, &filter_some_collector).unwrap();
let top_docs = searcher.search(&query, &filter_some_collector)?;
assert_eq!(top_docs.len(), 1);
assert_eq!(top_docs[0].1, DocAddress::new(0, 1));
@@ -67,9 +67,10 @@ pub fn test_filter_collector() {
}
let filter_dates_collector = FilterCollector::new(date, &date_filter, TopDocs::with_limit(5));
let filtered_date_docs = searcher.search(&query, &filter_dates_collector).unwrap();
let filtered_date_docs = searcher.search(&query, &filter_dates_collector)?;
assert_eq!(filtered_date_docs.len(), 2);
Ok(())
}
/// Stores all of the doc ids.
@@ -274,8 +275,8 @@ fn make_test_searcher() -> crate::Result<crate::LeasedItem<Searcher>> {
let schema = Schema::builder().build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(Document::default());
index_writer.add_document(Document::default());
index_writer.add_document(Document::default())?;
index_writer.add_document(Document::default())?;
index_writer.commit()?;
Ok(index.reader()?.searcher())
}

4
src/collector/top_collector.rs
View File

@@ -70,9 +70,7 @@ where
/// # Panics
/// The method panics if limit is 0
pub fn with_limit(limit: usize) -> TopCollector<T> {
if limit < 1 {
panic!("Limit must be strictly greater than 0.");
}
assert!(limit >= 1, "Limit must be strictly greater than 0.");
Self {
limit,
offset: 0,

196
src/collector/top_score_collector.rs
View File

@@ -94,27 +94,30 @@ where
/// use tantivy::schema::{Schema, TEXT};
/// use tantivy::{doc, DocAddress, Index};
///
/// # fn main() -> tantivy::Result<()> {
/// let mut schema_builder = Schema::builder();
/// let title = schema_builder.add_text_field("title", TEXT);
/// let schema = schema_builder.build();
/// let index = Index::create_in_ram(schema);
///
/// let mut index_writer = index.writer_with_num_threads(1, 10_000_000).unwrap();
/// index_writer.add_document(doc!(title => "The Name of the Wind"));
/// index_writer.add_document(doc!(title => "The Diary of Muadib"));
/// index_writer.add_document(doc!(title => "A Dairy Cow"));
/// index_writer.add_document(doc!(title => "The Diary of a Young Girl"));
/// assert!(index_writer.commit().is_ok());
/// let mut index_writer = index.writer_with_num_threads(1, 10_000_000)?;
/// index_writer.add_document(doc!(title => "The Name of the Wind"))?;
/// index_writer.add_document(doc!(title => "The Diary of Muadib"))?;
/// index_writer.add_document(doc!(title => "A Dairy Cow"))?;
/// index_writer.add_document(doc!(title => "The Diary of a Young Girl"))?;
/// index_writer.commit()?;
///
/// let reader = index.reader().unwrap();
/// let reader = index.reader()?;
/// let searcher = reader.searcher();
///
/// let query_parser = QueryParser::for_index(&index, vec![title]);
/// let query = query_parser.parse_query("diary").unwrap();
/// let top_docs = searcher.search(&query, &TopDocs::with_limit(2)).unwrap();
/// let query = query_parser.parse_query("diary")?;
/// let top_docs = searcher.search(&query, &TopDocs::with_limit(2))?;
///
/// assert_eq!(top_docs[0].1, DocAddress::new(0, 1));
/// assert_eq!(top_docs[1].1, DocAddress::new(0, 3));
/// # Ok(())
/// # }
/// ```
pub struct TopDocs(TopCollector<Score>);
@@ -180,29 +183,32 @@ impl TopDocs {
/// use tantivy::schema::{Schema, TEXT};
/// use tantivy::{doc, DocAddress, Index};
///
/// # fn main() -> tantivy::Result<()> {
/// let mut schema_builder = Schema::builder();
/// let title = schema_builder.add_text_field("title", TEXT);
/// let schema = schema_builder.build();
/// let index = Index::create_in_ram(schema);
///
/// let mut index_writer = index.writer_with_num_threads(1, 10_000_000).unwrap();
/// index_writer.add_document(doc!(title => "The Name of the Wind"));
/// index_writer.add_document(doc!(title => "The Diary of Muadib"));
/// index_writer.add_document(doc!(title => "A Dairy Cow"));
/// index_writer.add_document(doc!(title => "The Diary of a Young Girl"));
/// index_writer.add_document(doc!(title => "The Diary of Lena Mukhina"));
/// assert!(index_writer.commit().is_ok());
/// let mut index_writer = index.writer_with_num_threads(1, 10_000_000)?;
/// index_writer.add_document(doc!(title => "The Name of the Wind"))?;
/// index_writer.add_document(doc!(title => "The Diary of Muadib"))?;
/// index_writer.add_document(doc!(title => "A Dairy Cow"))?;
/// index_writer.add_document(doc!(title => "The Diary of a Young Girl"))?;
/// index_writer.add_document(doc!(title => "The Diary of Lena Mukhina"))?;
/// index_writer.commit()?;
///
/// let reader = index.reader().unwrap();
/// let reader = index.reader()?;
/// let searcher = reader.searcher();
///
/// let query_parser = QueryParser::for_index(&index, vec![title]);
/// let query = query_parser.parse_query("diary").unwrap();
/// let top_docs = searcher.search(&query, &TopDocs::with_limit(2).and_offset(1)).unwrap();
/// let query = query_parser.parse_query("diary")?;
/// let top_docs = searcher.search(&query, &TopDocs::with_limit(2).and_offset(1))?;
///
/// assert_eq!(top_docs.len(), 2);
/// assert_eq!(top_docs[0].1, DocAddress::new(0, 4));
/// assert_eq!(top_docs[1].1, DocAddress::new(0, 3));
/// Ok(())
/// # }
/// ```
pub fn and_offset(self, offset: usize) -> TopDocs {
TopDocs(self.0.and_offset(offset))
@@ -234,11 +240,11 @@ impl TopDocs {
/// #
/// # let index = Index::create_in_ram(schema);
/// # let mut index_writer = index.writer_with_num_threads(1, 10_000_000)?;
/// # index_writer.add_document(doc!(title => "The Name of the Wind", rating => 92u64));
/// # index_writer.add_document(doc!(title => "The Diary of Muadib", rating => 97u64));
/// # index_writer.add_document(doc!(title => "A Dairy Cow", rating => 63u64));
/// # index_writer.add_document(doc!(title => "The Diary of a Young Girl", rating => 80u64));
/// # assert!(index_writer.commit().is_ok());
/// # index_writer.add_document(doc!(title => "The Name of the Wind", rating => 92u64))?;
/// # index_writer.add_document(doc!(title => "The Diary of Muadib", rating => 97u64))?;
/// # index_writer.add_document(doc!(title => "A Dairy Cow", rating => 63u64))?;
/// # index_writer.add_document(doc!(title => "The Diary of a Young Girl", rating => 80u64))?;
/// # index_writer.commit()?;
/// # let reader = index.reader()?;
/// # let query = QueryParser::for_index(&index, vec![title]).parse_query("diary")?;
/// # let top_docs = docs_sorted_by_rating(&reader.searcher(), &query, rating)?;
@@ -316,9 +322,9 @@ impl TopDocs {
/// #
/// # let index = Index::create_in_ram(schema);
/// # let mut index_writer = index.writer_with_num_threads(1, 10_000_000)?;
/// # index_writer.add_document(doc!(title => "MadCow Inc.", rating => 92_000_000i64));
/// # index_writer.add_document(doc!(title => "Zozo Cow KKK", rating => 119_000_000i64));
/// # index_writer.add_document(doc!(title => "Declining Cow", rating => -63_000_000i64));
/// # index_writer.add_document(doc!(title => "MadCow Inc.", rating => 92_000_000i64))?;
/// # index_writer.add_document(doc!(title => "Zozo Cow KKK", rating => 119_000_000i64))?;
/// # index_writer.add_document(doc!(title => "Declining Cow", rating => -63_000_000i64))?;
/// # assert!(index_writer.commit().is_ok());
/// # let reader = index.reader()?;
/// # let top_docs = docs_sorted_by_revenue(&reader.searcher(), &AllQuery, rating)?;
@@ -417,9 +423,9 @@ impl TopDocs {
/// let mut index_writer = index.writer_with_num_threads(1, 10_000_000)?;
/// let product_name = index.schema().get_field("product_name").unwrap();
/// let popularity: Field = index.schema().get_field("popularity").unwrap();
/// index_writer.add_document(doc!(product_name => "The Diary of Muadib", popularity => 1u64));
/// index_writer.add_document(doc!(product_name => "A Dairy Cow", popularity => 10u64));
/// index_writer.add_document(doc!(product_name => "The Diary of a Young Girl", popularity => 15u64));
/// index_writer.add_document(doc!(product_name => "The Diary of Muadib", popularity => 1u64))?;
/// index_writer.add_document(doc!(product_name => "A Dairy Cow", popularity => 10u64))?;
/// index_writer.add_document(doc!(product_name => "The Diary of a Young Girl", popularity => 15u64))?;
/// index_writer.commit()?;
/// Ok(index)
/// }
@@ -527,9 +533,9 @@ impl TopDocs {
/// #
/// let popularity: Field = index.schema().get_field("popularity").unwrap();
/// let boosted: Field = index.schema().get_field("boosted").unwrap();
/// # index_writer.add_document(doc!(boosted=>1u64, product_name => "The Diary of Muadib", popularity => 1u64));
/// # index_writer.add_document(doc!(boosted=>0u64, product_name => "A Dairy Cow", popularity => 10u64));
/// # index_writer.add_document(doc!(boosted=>0u64, product_name => "The Diary of a Young Girl", popularity => 15u64));
/// # index_writer.add_document(doc!(boosted=>1u64, product_name => "The Diary of Muadib", popularity => 1u64))?;
/// # index_writer.add_document(doc!(boosted=>0u64, product_name => "A Dairy Cow", popularity => 10u64))?;
/// # index_writer.add_document(doc!(boosted=>0u64, product_name => "The Diary of a Young Girl", popularity => 15u64))?;
/// # index_writer.commit()?;
/// // ...
/// # let user_query = "diary";
@@ -629,10 +635,10 @@ impl Collector for TopDocs {
let heap_len = self.0.limit + self.0.offset;
let mut heap: BinaryHeap<ComparableDoc<Score, DocId>> = BinaryHeap::with_capacity(heap_len);
if let Some(delete_bitset) = reader.delete_bitset() {
if let Some(alive_bitset) = reader.alive_bitset() {
let mut threshold = Score::MIN;
weight.for_each_pruning(threshold, reader, &mut |doc, score| {
if delete_bitset.is_deleted(doc) {
if alive_bitset.is_deleted(doc) {
return threshold;
}
let heap_item = ComparableDoc {
@@ -713,20 +719,18 @@ mod tests {
use crate::Score;
use crate::{DocAddress, DocId, SegmentReader};
fn make_index() -> Index {
fn make_index() -> crate::Result<Index> {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
// writing the segment
let mut index_writer = index.writer_with_num_threads(1, 10_000_000).unwrap();
index_writer.add_document(doc!(text_field=>"Hello happy tax payer."));
index_writer.add_document(doc!(text_field=>"Droopy says hello happy tax payer"));
index_writer.add_document(doc!(text_field=>"I like Droopy"));
assert!(index_writer.commit().is_ok());
}
index
// writing the segment
let mut index_writer = index.writer_with_num_threads(1, 10_000_000)?;
index_writer.add_document(doc!(text_field=>"Hello happy tax payer."))?;
index_writer.add_document(doc!(text_field=>"Droopy says hello happy tax payer"))?;
index_writer.add_document(doc!(text_field=>"I like Droopy"))?;
index_writer.commit()?;
Ok(index)
}
fn assert_results_equals(results: &[(Score, DocAddress)], expected: &[(Score, DocAddress)]) {
@@ -737,17 +741,15 @@ mod tests {
}
#[test]
fn test_top_collector_not_at_capacity_without_offset() {
let index = make_index();
fn test_top_collector_not_at_capacity_without_offset() -> crate::Result<()> {
let index = make_index()?;
let field = index.schema().get_field("text").unwrap();
let query_parser = QueryParser::for_index(&index, vec![field]);
let text_query = query_parser.parse_query("droopy tax").unwrap();
let text_query = query_parser.parse_query("droopy tax")?;
let score_docs: Vec<(Score, DocAddress)> = index
.reader()
.unwrap()
.reader()?
.searcher()
.search(&text_query, &TopDocs::with_limit(4))
.unwrap();
.search(&text_query, &TopDocs::with_limit(4))?;
assert_results_equals(
&score_docs,
&[
@@ -756,11 +758,12 @@ mod tests {
(0.48527452, DocAddress::new(0, 0)),
],
);
Ok(())
}
#[test]
fn test_top_collector_not_at_capacity_with_offset() {
let index = make_index();
let index = make_index().unwrap();
let field = index.schema().get_field("text").unwrap();
let query_parser = QueryParser::for_index(&index, vec![field]);
let text_query = query_parser.parse_query("droopy tax").unwrap();
@@ -775,7 +778,7 @@ mod tests {
#[test]
fn test_top_collector_at_capacity() {
let index = make_index();
let index = make_index().unwrap();
let field = index.schema().get_field("text").unwrap();
let query_parser = QueryParser::for_index(&index, vec![field]);
let text_query = query_parser.parse_query("droopy tax").unwrap();
@@ -796,7 +799,7 @@ mod tests {
#[test]
fn test_top_collector_at_capacity_with_offset() {
let index = make_index();
let index = make_index().unwrap();
let field = index.schema().get_field("text").unwrap();
let query_parser = QueryParser::for_index(&index, vec![field]);
let text_query = query_parser.parse_query("droopy tax").unwrap();
@@ -817,7 +820,7 @@ mod tests {
#[test]
fn test_top_collector_stable_sorting() {
let index = make_index();
let index = make_index().unwrap();
// using AllQuery to get a constant score
let searcher = index.reader().unwrap().searcher();
@@ -848,29 +851,35 @@ mod tests {
const SIZE: &str = "size";
#[test]
fn test_top_field_collector_not_at_capacity() {
fn test_top_field_collector_not_at_capacity() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let title = schema_builder.add_text_field(TITLE, TEXT);
let size = schema_builder.add_u64_field(SIZE, FAST);
let schema = schema_builder.build();
let (index, query) = index("beer", title, schema, |index_writer| {
index_writer.add_document(doc!(
title => "bottle of beer",
size => 12u64,
));
index_writer.add_document(doc!(
title => "growler of beer",
size => 64u64,
));
index_writer.add_document(doc!(
title => "pint of beer",
size => 16u64,
));
index_writer
.add_document(doc!(
title => "bottle of beer",
size => 12u64,
))
.unwrap();
index_writer
.add_document(doc!(
title => "growler of beer",
size => 64u64,
))
.unwrap();
index_writer
.add_document(doc!(
title => "pint of beer",
size => 16u64,
))
.unwrap();
});
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
let top_collector = TopDocs::with_limit(4).order_by_u64_field(size);
let top_docs: Vec<(u64, DocAddress)> = searcher.search(&query, &top_collector).unwrap();
let top_docs: Vec<(u64, DocAddress)> = searcher.search(&query, &top_collector)?;
assert_eq!(
&top_docs[..],
&[
@@ -879,6 +888,7 @@ mod tests {
(12, DocAddress::new(0, 0))
]
);
Ok(())
}
#[test]
@@ -894,12 +904,12 @@ mod tests {
index_writer.add_document(doc!(
name => "Paul Robeson",
birthday => pr_birthday
));
))?;
let mr_birthday = crate::DateTime::from_str("1947-11-08T00:00:00+00:00")?;
index_writer.add_document(doc!(
name => "Minnie Riperton",
birthday => mr_birthday
));
))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let top_collector = TopDocs::with_limit(3).order_by_fast_field(birthday);
@@ -926,11 +936,11 @@ mod tests {
index_writer.add_document(doc!(
city => "georgetown",
altitude => -1i64,
));
))?;
index_writer.add_document(doc!(
city => "tokyo",
altitude => 40i64,
));
))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let top_collector = TopDocs::with_limit(3).order_by_fast_field(altitude);
@@ -956,11 +966,11 @@ mod tests {
index_writer.add_document(doc!(
city => "georgetown",
altitude => -1.0f64,
));
))?;
index_writer.add_document(doc!(
city => "tokyo",
altitude => 40f64,
));
))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let top_collector = TopDocs::with_limit(3).order_by_fast_field(altitude);
@@ -983,10 +993,12 @@ mod tests {
let size = schema_builder.add_u64_field(SIZE, FAST);
let schema = schema_builder.build();
let (index, _) = index("beer", title, schema, |index_writer| {
index_writer.add_document(doc!(
title => "bottle of beer",
size => 12u64,
));
index_writer
.add_document(doc!(
title => "bottle of beer",
size => 12u64,
))
.unwrap();
});
let searcher = index.reader().unwrap().searcher();
let top_collector = TopDocs::with_limit(4).order_by_u64_field(Field::from_field_id(2));
@@ -1003,7 +1015,7 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(size=>1u64));
index_writer.add_document(doc!(size=>1u64))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let segment = searcher.segment_reader(0);
@@ -1020,7 +1032,7 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(size=>1u64));
index_writer.add_document(doc!(size=>1u64))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let segment = searcher.segment_reader(0);
@@ -1033,30 +1045,26 @@ mod tests {
}
#[test]
fn test_tweak_score_top_collector_with_offset() {
let index = make_index();
fn test_tweak_score_top_collector_with_offset() -> crate::Result<()> {
let index = make_index()?;
let field = index.schema().get_field("text").unwrap();
let query_parser = QueryParser::for_index(&index, vec![field]);
let text_query = query_parser.parse_query("droopy tax").unwrap();
let text_query = query_parser.parse_query("droopy tax")?;
let collector = TopDocs::with_limit(2).and_offset(1).tweak_score(
move |_segment_reader: &SegmentReader| move |doc: DocId, _original_score: Score| doc,
);
let score_docs: Vec<(u32, DocAddress)> = index
.reader()
.unwrap()
.searcher()
.search(&text_query, &collector)
.unwrap();
let score_docs: Vec<(u32, DocAddress)> =
index.reader()?.searcher().search(&text_query, &collector)?;
assert_eq!(
score_docs,
vec![(1, DocAddress::new(0, 1)), (0, DocAddress::new(0, 0)),]
);
Ok(())
}
#[test]
fn test_custom_score_top_collector_with_offset() {
let index = make_index();
let index = make_index().unwrap();
let field = index.schema().get_field("text").unwrap();
let query_parser = QueryParser::for_index(&index, vec![field]);
let text_query = query_parser.parse_query("droopy tax").unwrap();

94
src/core/index.rs
View File

@@ -123,8 +123,8 @@ impl IndexBuilder {
/// If a previous index was in this directory, it returns an `IndexAlreadyExists` error.
#[cfg(feature = "mmap")]
pub fn create_in_dir<P: AsRef<Path>>(self, directory_path: P) -> crate::Result<Index> {
let mmap_directory = MmapDirectory::open(directory_path)?;
if Index::exists(&mmap_directory)? {
let mmap_directory: Box<dyn Directory> = Box::new(MmapDirectory::open(directory_path)?);
if Index::exists(&*mmap_directory)? {
return Err(TantivyError::IndexAlreadyExists);
}
self.create(mmap_directory)
@@ -139,7 +139,7 @@ impl IndexBuilder {
/// For other unit tests, prefer the `RAMDirectory`, see: `create_in_ram`.
#[cfg(feature = "mmap")]
pub fn create_from_tempdir(self) -> crate::Result<Index> {
let mmap_directory = MmapDirectory::create_from_tempdir()?;
let mmap_directory: Box<dyn Directory> = Box::new(MmapDirectory::create_from_tempdir()?);
self.create(mmap_directory)
}
fn get_expect_schema(&self) -> crate::Result<Schema> {
@@ -149,8 +149,9 @@ impl IndexBuilder {
.ok_or(TantivyError::IndexBuilderMissingArgument("schema"))
}
/// Opens or creates a new index in the provided directory
pub fn open_or_create<Dir: Directory>(self, dir: Dir) -> crate::Result<Index> {
if !Index::exists(&dir)? {
pub fn open_or_create<T: Into<Box<dyn Directory>>>(self, dir: T) -> crate::Result<Index> {
let dir = dir.into();
if !Index::exists(&*dir)? {
return self.create(dir);
}
let index = Index::open(dir)?;
@@ -165,7 +166,8 @@ impl IndexBuilder {
/// Creates a new index given an implementation of the trait `Directory`.
///
/// If a directory previously existed, it will be erased.
fn create<Dir: Directory>(self, dir: Dir) -> crate::Result<Index> {
fn create<T: Into<Box<dyn Directory>>>(self, dir: T) -> crate::Result<Index> {
let dir = dir.into();
let directory = ManagedDirectory::wrap(dir)?;
save_new_metas(
self.get_expect_schema()?,
@@ -198,7 +200,7 @@ impl Index {
/// Examines the directory to see if it contains an index.
///
/// Effectively, it only checks for the presence of the `meta.json` file.
pub fn exists<Dir: Directory>(dir: &Dir) -> Result<bool, OpenReadError> {
pub fn exists(dir: &dyn Directory) -> Result<bool, OpenReadError> {
dir.exists(&META_FILEPATH)
}
@@ -250,7 +252,11 @@ impl Index {
}
/// Opens or creates a new index in the provided directory
pub fn open_or_create<Dir: Directory>(dir: Dir, schema: Schema) -> crate::Result<Index> {
pub fn open_or_create<T: Into<Box<dyn Directory>>>(
dir: T,
schema: Schema,
) -> crate::Result<Index> {
let dir = dir.into();
IndexBuilder::new().schema(schema).open_or_create(dir)
}
@@ -270,11 +276,12 @@ impl Index {
/// Creates a new index given an implementation of the trait `Directory`.
///
/// If a directory previously existed, it will be erased.
pub fn create<Dir: Directory>(
dir: Dir,
pub fn create<T: Into<Box<dyn Directory>>>(
dir: T,
schema: Schema,
settings: IndexSettings,
) -> crate::Result<Index> {
let dir: Box<dyn Directory> = dir.into();
let mut builder = IndexBuilder::new().schema(schema);
builder = builder.settings(settings);
builder.create(dir)
@@ -365,7 +372,8 @@ impl Index {
}
/// Open the index using the provided directory
pub fn open<D: Directory>(directory: D) -> crate::Result<Index> {
pub fn open<T: Into<Box<dyn Directory>>>(directory: T) -> crate::Result<Index> {
let directory = directory.into();
let directory = ManagedDirectory::wrap(directory)?;
let inventory = SegmentMetaInventory::default();
let metas = load_metas(&directory, &inventory)?;
@@ -395,9 +403,7 @@ impl Index {
///
/// # Errors
/// If the lockfile already exists, returns `Error::DirectoryLockBusy` or an `Error::IoError`.
///
/// # Panics
/// If the heap size per thread is too small, panics.
/// If the heap size per thread is too small or too big, returns `TantivyError::InvalidArgument`
pub fn writer_with_num_threads(
&self,
num_threads: usize,
@@ -439,14 +445,13 @@ impl Index {
/// Creates a multithreaded writer
///
/// Tantivy will automatically define the number of threads to use, but
/// no more than [`MAX_NUM_THREAD`] threads.
/// no more than 8 threads.
/// `overall_heap_size_in_bytes` is the total target memory usage that will be split
/// between a given number of threads.
///
/// # Errors
/// If the lockfile already exists, returns `Error::FileAlreadyExists`.
/// # Panics
/// If the heap size per thread is too small, panics.
/// If the heap size per thread is too small or too big, returns `TantivyError::InvalidArgument`
pub fn writer(&self, overall_heap_size_in_bytes: usize) -> crate::Result<IndexWriter> {
let mut num_threads = std::cmp::min(num_cpus::get(), MAX_NUM_THREAD);
let heap_size_in_bytes_per_thread = overall_heap_size_in_bytes / num_threads;
@@ -577,15 +582,15 @@ mod tests {
#[test]
fn test_index_exists() {
let directory = RamDirectory::create();
assert!(!Index::exists(&directory).unwrap());
let directory: Box<dyn Directory> = Box::new(RamDirectory::create());
assert!(!Index::exists(directory.as_ref()).unwrap());
assert!(Index::create(
directory.clone(),
throw_away_schema(),
IndexSettings::default()
)
.is_ok());
assert!(Index::exists(&directory).unwrap());
assert!(Index::exists(directory.as_ref()).unwrap());
}
#[test]
@@ -598,27 +603,27 @@ mod tests {
#[test]
fn open_or_create_should_open() {
let directory = RamDirectory::create();
let directory: Box<dyn Directory> = Box::new(RamDirectory::create());
assert!(Index::create(
directory.clone(),
throw_away_schema(),
IndexSettings::default()
)
.is_ok());
assert!(Index::exists(&directory).unwrap());
assert!(Index::exists(directory.as_ref()).unwrap());
assert!(Index::open_or_create(directory, throw_away_schema()).is_ok());
}
#[test]
fn create_should_wipeoff_existing() {
let directory = RamDirectory::create();
let directory: Box<dyn Directory> = Box::new(RamDirectory::create());
assert!(Index::create(
directory.clone(),
throw_away_schema(),
IndexSettings::default()
)
.is_ok());
assert!(Index::exists(&directory).unwrap());
assert!(Index::exists(directory.as_ref()).unwrap());
assert!(Index::create(
directory,
Schema::builder().build(),
@@ -652,7 +657,7 @@ mod tests {
}
#[test]
fn test_index_on_commit_reload_policy() {
fn test_index_on_commit_reload_policy() -> crate::Result<()> {
let schema = throw_away_schema();
let field = schema.get_field("num_likes").unwrap();
let index = Index::create_in_ram(schema);
@@ -662,7 +667,7 @@ mod tests {
.try_into()
.unwrap();
assert_eq!(reader.searcher().num_docs(), 0);
test_index_on_commit_reload_policy_aux(field, &index, &reader);
test_index_on_commit_reload_policy_aux(field, &index, &reader)
}
#[cfg(feature = "mmap")]
@@ -674,7 +679,7 @@ mod tests {
use tempfile::TempDir;
#[test]
fn test_index_on_commit_reload_policy_mmap() {
fn test_index_on_commit_reload_policy_mmap() -> crate::Result<()> {
let schema = throw_away_schema();
let field = schema.get_field("num_likes").unwrap();
let tempdir = TempDir::new().unwrap();
@@ -686,7 +691,7 @@ mod tests {
.try_into()
.unwrap();
assert_eq!(reader.searcher().num_docs(), 0);
test_index_on_commit_reload_policy_aux(field, &index, &reader);
test_index_on_commit_reload_policy_aux(field, &index, &reader)
}
#[test]
@@ -701,7 +706,7 @@ mod tests {
.reload_policy(ReloadPolicy::Manual)
.try_into()?;
assert_eq!(reader.searcher().num_docs(), 0);
writer.add_document(doc!(field=>1u64));
writer.add_document(doc!(field=>1u64))?;
let (sender, receiver) = crossbeam::channel::unbounded();
let _handle = index.directory_mut().watch(WatchCallback::new(move || {
let _ = sender.send(());
@@ -715,7 +720,7 @@ mod tests {
}
#[test]
fn test_index_on_commit_reload_policy_different_directories() {
fn test_index_on_commit_reload_policy_different_directories() -> crate::Result<()> {
let schema = throw_away_schema();
let field = schema.get_field("num_likes").unwrap();
let tempdir = TempDir::new().unwrap();
@@ -728,10 +733,14 @@ mod tests {
.try_into()
.unwrap();
assert_eq!(reader.searcher().num_docs(), 0);
test_index_on_commit_reload_policy_aux(field, &write_index, &reader);
test_index_on_commit_reload_policy_aux(field, &write_index, &reader)
}
}
fn test_index_on_commit_reload_policy_aux(field: Field, index: &Index, reader: &IndexReader) {
fn test_index_on_commit_reload_policy_aux(
field: Field,
index: &Index,
reader: &IndexReader,
) -> crate::Result<()> {
let mut reader_index = reader.index();
let (sender, receiver) = crossbeam::channel::unbounded();
let _watch_handle = reader_index
@@ -739,9 +748,9 @@ mod tests {
.watch(WatchCallback::new(move || {
let _ = sender.send(());
}));
let mut writer = index.writer_for_tests().unwrap();
let mut writer = index.writer_for_tests()?;
assert_eq!(reader.searcher().num_docs(), 0);
writer.add_document(doc!(field=>1u64));
writer.add_document(doc!(field=>1u64))?;
writer.commit().unwrap();
// We need a loop here because it is possible for notify to send more than
// one modify event. It was observed on CI on MacOS.
@@ -751,7 +760,7 @@ mod tests {
break;
}
}
writer.add_document(doc!(field=>2u64));
writer.add_document(doc!(field=>2u64))?;
writer.commit().unwrap();
// ... Same as above
loop {
@@ -760,37 +769,37 @@ mod tests {
break;
}
}
Ok(())
}
// This test will not pass on windows, because windows
// prevent deleting files that are MMapped.
#[cfg(not(target_os = "windows"))]
#[test]
fn garbage_collect_works_as_intended() {
fn garbage_collect_works_as_intended() -> crate::Result<()> {
let directory = RamDirectory::create();
let schema = throw_away_schema();
let field = schema.get_field("num_likes").unwrap();
let index = Index::create(directory.clone(), schema, IndexSettings::default()).unwrap();
let index = Index::create(directory.clone(), schema, IndexSettings::default())?;
let mut writer = index.writer_with_num_threads(8, 24_000_000).unwrap();
for i in 0u64..8_000u64 {
writer.add_document(doc!(field => i));
writer.add_document(doc!(field => i))?;
}
let (sender, receiver) = crossbeam::channel::unbounded();
let _handle = directory.watch(WatchCallback::new(move || {
let _ = sender.send(());
}));
writer.commit().unwrap();
writer.commit()?;
let mem_right_after_commit = directory.total_mem_usage();
assert!(receiver.recv().is_ok());
let reader = index
.reader_builder()
.reload_policy(ReloadPolicy::Manual)
.try_into()
.unwrap();
.try_into()?;
assert_eq!(reader.searcher().num_docs(), 8_000);
writer.wait_merging_threads().unwrap();
writer.wait_merging_threads()?;
let mem_right_after_merge_finished = directory.total_mem_usage();
reader.reload().unwrap();
@@ -802,5 +811,6 @@ mod tests {
mem_right_after_merge_finished,
mem_right_after_commit
);
Ok(())
}
}

2
src/core/searcher.rs
View File

@@ -88,7 +88,7 @@ impl Searcher {
&self.segment_readers
}
/// Returns the segment_reader associated with the given segment_ordinal
/// Returns the segment_reader associated with the given segment_ord
pub fn segment_reader(&self, segment_ord: u32) -> &SegmentReader {
&self.segment_readers[segment_ord as usize]
}

91
src/core/segment_reader.rs
View File

@@ -5,7 +5,8 @@ use crate::core::SegmentId;
use crate::directory::CompositeFile;
use crate::directory::FileSlice;
use crate::error::DataCorruption;
use crate::fastfield::DeleteBitSet;
use crate::fastfield::intersect_alive_bitsets;
use crate::fastfield::AliveBitSet;
use crate::fastfield::FacetReader;
use crate::fastfield::FastFieldReaders;
use crate::fieldnorm::{FieldNormReader, FieldNormReaders};
@@ -47,7 +48,7 @@ pub struct SegmentReader {
fieldnorm_readers: FieldNormReaders,
store_file: FileSlice,
delete_bitset_opt: Option<DeleteBitSet>,
alive_bitset_opt: Option<AliveBitSet>,
schema: Schema,
}
@@ -72,14 +73,12 @@ impl SegmentReader {
/// Return the number of documents that have been
/// deleted in the segment.
pub fn num_deleted_docs(&self) -> DocId {
self.delete_bitset()
.map(|delete_set| delete_set.num_deleted() as DocId)
.unwrap_or(0u32)
self.max_doc - self.num_docs
}
/// Returns true iff some of the documents of the segment have been deleted.
pub fn has_deletes(&self) -> bool {
self.delete_bitset().is_some()
self.num_deleted_docs() > 0
}
/// Accessor to a segment's fast field reader given a field.
@@ -142,6 +141,14 @@ impl SegmentReader {
/// Open a new segment for reading.
pub fn open(segment: &Segment) -> crate::Result<SegmentReader> {
Self::open_with_custom_alive_set(segment, None)
}
/// Open a new segment for reading.
pub fn open_with_custom_alive_set(
segment: &Segment,
custom_bitset: Option<AliveBitSet>,
) -> crate::Result<SegmentReader> {
let termdict_file = segment.open_read(SegmentComponent::Terms)?;
let termdict_composite = CompositeFile::open(&termdict_file)?;
@@ -166,29 +173,36 @@ impl SegmentReader {
let fast_fields_composite = CompositeFile::open(&fast_fields_data)?;
let fast_field_readers =
Arc::new(FastFieldReaders::new(schema.clone(), fast_fields_composite));
let fieldnorm_data = segment.open_read(SegmentComponent::FieldNorms)?;
let fieldnorm_readers = FieldNormReaders::open(fieldnorm_data)?;
let delete_bitset_opt = if segment.meta().has_deletes() {
let delete_data = segment.open_read(SegmentComponent::Delete)?;
let delete_bitset = DeleteBitSet::open(delete_data)?;
Some(delete_bitset)
let original_bitset = if segment.meta().has_deletes() {
let delete_file_slice = segment.open_read(SegmentComponent::Delete)?;
let delete_data = delete_file_slice.read_bytes()?;
Some(AliveBitSet::open(delete_data))
} else {
None
};
let alive_bitset_opt = intersect_alive_bitset(original_bitset, custom_bitset);
let max_doc = segment.meta().max_doc();
let num_docs = alive_bitset_opt
.as_ref()
.map(|alive_bitset| alive_bitset.num_alive_docs() as u32)
.unwrap_or(max_doc);
Ok(SegmentReader {
inv_idx_reader_cache: Default::default(),
max_doc: segment.meta().max_doc(),
num_docs: segment.meta().num_docs(),
num_docs,
max_doc,
termdict_composite,
postings_composite,
fast_fields_readers: fast_field_readers,
fieldnorm_readers,
segment_id: segment.id(),
store_file,
delete_bitset_opt,
alive_bitset_opt,
positions_composite,
schema,
})
@@ -274,21 +288,25 @@ impl SegmentReader {
/// Returns the bitset representing
/// the documents that have been deleted.
pub fn delete_bitset(&self) -> Option<&DeleteBitSet> {
self.delete_bitset_opt.as_ref()
pub fn alive_bitset(&self) -> Option<&AliveBitSet> {
self.alive_bitset_opt.as_ref()
}
/// Returns true iff the `doc` is marked
/// as deleted.
pub fn is_deleted(&self, doc: DocId) -> bool {
self.delete_bitset()
self.alive_bitset()
.map(|delete_set| delete_set.is_deleted(doc))
.unwrap_or(false)
}
/// Returns an iterator that will iterate over the alive document ids
pub fn doc_ids_alive(&self) -> impl Iterator<Item = DocId> + '_ {
(0u32..self.max_doc).filter(move |doc| !self.is_deleted(*doc))
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 {
Box::new(0u32..self.max_doc)
}
}
/// Summarize total space usage of this segment.
@@ -301,14 +319,29 @@ impl SegmentReader {
self.fast_fields_readers.space_usage(),
self.fieldnorm_readers.space_usage(),
self.get_store_reader()?.space_usage(),
self.delete_bitset_opt
self.alive_bitset_opt
.as_ref()
.map(DeleteBitSet::space_usage)
.map(AliveBitSet::space_usage)
.unwrap_or(0),
))
}
}
fn intersect_alive_bitset(
left_opt: Option<AliveBitSet>,
right_opt: Option<AliveBitSet>,
) -> Option<AliveBitSet> {
match (left_opt, right_opt) {
(Some(left), Some(right)) => {
assert_eq!(left.bitset().max_value(), right.bitset().max_value());
Some(intersect_alive_bitsets(left, right))
}
(Some(left), None) => Some(left),
(None, Some(right)) => Some(right),
(None, None) => None,
}
}
impl fmt::Debug for SegmentReader {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
write!(f, "SegmentReader({:?})", self.segment_id)
@@ -331,10 +364,10 @@ mod test {
{
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(name => "tantivy"));
index_writer.add_document(doc!(name => "horse"));
index_writer.add_document(doc!(name => "jockey"));
index_writer.add_document(doc!(name => "cap"));
index_writer.add_document(doc!(name => "tantivy"))?;
index_writer.add_document(doc!(name => "horse"))?;
index_writer.add_document(doc!(name => "jockey"))?;
index_writer.add_document(doc!(name => "cap"))?;
// we should now have one segment with two docs
index_writer.delete_term(Term::from_field_text(name, "horse"));
index_writer.delete_term(Term::from_field_text(name, "cap"));
@@ -357,10 +390,10 @@ mod test {
{
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(name => "tantivy"));
index_writer.add_document(doc!(name => "horse"));
index_writer.add_document(doc!(name => "jockey"));
index_writer.add_document(doc!(name => "cap"));
index_writer.add_document(doc!(name => "tantivy"))?;
index_writer.add_document(doc!(name => "horse"))?;
index_writer.add_document(doc!(name => "jockey"))?;
index_writer.add_document(doc!(name => "cap"))?;
// we should now have one segment with two docs
index_writer.commit()?;
}

12
src/directory/directory.rs
View File

@@ -230,3 +230,15 @@ where
Box::new(self.clone())
}
}
impl Clone for Box<dyn Directory> {
fn clone(&self) -> Self {
self.box_clone()
}
}
impl<T: Directory + 'static> From<T> for Box<dyn Directory> {
fn from(t: T) -> Self {
Box::new(t)
}
}

12
src/directory/managed_directory.rs
View File

@@ -64,7 +64,7 @@ fn save_managed_paths(
impl ManagedDirectory {
/// Wraps a directory as managed directory.
pub fn wrap<Dir: Directory>(directory: Dir) -> crate::Result<ManagedDirectory> {
pub fn wrap(directory: Box<dyn Directory>) -> crate::Result<ManagedDirectory> {
match directory.atomic_read(&MANAGED_FILEPATH) {
Ok(data) => {
let managed_files_json = String::from_utf8_lossy(&data);
@@ -76,14 +76,14 @@ impl ManagedDirectory {
)
})?;
Ok(ManagedDirectory {
directory: Box::new(directory),
directory,
meta_informations: Arc::new(RwLock::new(MetaInformation {
managed_paths: managed_files,
})),
})
}
Err(OpenReadError::FileDoesNotExist(_)) => Ok(ManagedDirectory {
directory: Box::new(directory),
directory,
meta_informations: Arc::default(),
}),
io_err @ Err(OpenReadError::IoError { .. }) => Err(io_err.err().unwrap().into()),
@@ -340,7 +340,7 @@ mod tests_mmap_specific {
let test_path2: &'static Path = Path::new("some_path_for_test_2");
{
let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
let mut managed_directory = ManagedDirectory::wrap(mmap_directory).unwrap();
let mut managed_directory = ManagedDirectory::wrap(Box::new(mmap_directory)).unwrap();
let write_file = managed_directory.open_write(test_path1).unwrap();
write_file.terminate().unwrap();
managed_directory
@@ -355,7 +355,7 @@ mod tests_mmap_specific {
}
{
let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
let mut managed_directory = ManagedDirectory::wrap(mmap_directory).unwrap();
let mut managed_directory = ManagedDirectory::wrap(Box::new(mmap_directory)).unwrap();
assert!(managed_directory.exists(test_path1).unwrap());
assert!(!managed_directory.exists(test_path2).unwrap());
let living_files: HashSet<PathBuf> = HashSet::new();
@@ -374,7 +374,7 @@ mod tests_mmap_specific {
let living_files = HashSet::new();
let mmap_directory = MmapDirectory::open(&tempdir_path).unwrap();
let mut managed_directory = ManagedDirectory::wrap(mmap_directory).unwrap();
let mut managed_directory = ManagedDirectory::wrap(Box::new(mmap_directory)).unwrap();
let mut write = managed_directory.open_write(test_path1).unwrap();
write.write_all(&[0u8, 1u8]).unwrap();
write.terminate().unwrap();

44
src/directory/mmap_directory.rs
View File

@@ -74,20 +74,12 @@ pub struct CacheInfo {
pub mmapped: Vec<PathBuf>,
}
#[derive(Default)]
struct MmapCache {
counters: CacheCounters,
cache: HashMap<PathBuf, WeakArcBytes>,
}
impl Default for MmapCache {
fn default() -> MmapCache {
MmapCache {
counters: CacheCounters::default(),
cache: HashMap::new(),
}
}
}
impl MmapCache {
fn get_info(&self) -> CacheInfo {
let paths: Vec<PathBuf> = self.cache.keys().cloned().collect();
@@ -242,7 +234,7 @@ impl MmapDirectory {
}
let fd = open_opts.open(&self.inner.root_path)?;
fd.sync_all()?;
fd.sync_data()?;
Ok(())
}
@@ -296,8 +288,7 @@ impl Write for SafeFileWriter {
}
fn flush(&mut self) -> io::Result<()> {
self.0.flush()?;
self.0.sync_all()
Ok(())
}
}
@@ -309,7 +300,9 @@ impl Seek for SafeFileWriter {
impl TerminatingWrite for SafeFileWriter {
fn terminate_ref(&mut self, _: AntiCallToken) -> io::Result<()> {
self.flush()
self.0.flush()?;
self.0.sync_data()?;
Ok(())
}
}
@@ -339,6 +332,7 @@ pub(crate) fn atomic_write(path: &Path, content: &[u8]) -> io::Result<()> {
let mut tempfile = tempfile::Builder::new().tempfile_in(&parent_path)?;
tempfile.write_all(content)?;
tempfile.flush()?;
tempfile.as_file_mut().sync_data()?;
tempfile.into_temp_path().persist(path)?;
Ok(())
}
@@ -582,8 +576,8 @@ mod tests {
}
#[test]
fn test_mmap_released() {
let mmap_directory = MmapDirectory::create_from_tempdir().unwrap();
fn test_mmap_released() -> crate::Result<()> {
let mmap_directory = MmapDirectory::create_from_tempdir()?;
let mut schema_builder: SchemaBuilder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
@@ -592,31 +586,30 @@ mod tests {
let index =
Index::create(mmap_directory.clone(), schema, IndexSettings::default()).unwrap();
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
let mut log_merge_policy = LogMergePolicy::default();
log_merge_policy.set_min_num_segments(3);
index_writer.set_merge_policy(Box::new(log_merge_policy));
for _num_commits in 0..10 {
for _ in 0..10 {
index_writer.add_document(doc!(text_field=>"abc"));
index_writer.add_document(doc!(text_field=>"abc"))?;
}
index_writer.commit().unwrap();
index_writer.commit()?;
}
let reader = index
.reader_builder()
.reload_policy(ReloadPolicy::Manual)
.try_into()
.unwrap();
.try_into()?;
for _ in 0..4 {
index_writer.add_document(doc!(text_field=>"abc"));
index_writer.commit().unwrap();
reader.reload().unwrap();
index_writer.add_document(doc!(text_field=>"abc"))?;
index_writer.commit()?;
reader.reload()?;
}
index_writer.wait_merging_threads().unwrap();
index_writer.wait_merging_threads()?;
reader.reload().unwrap();
reader.reload()?;
let num_segments = reader.searcher().segment_readers().len();
assert!(num_segments <= 4);
let num_components_except_deletes_and_tempstore =
@@ -627,5 +620,6 @@ mod tests {
);
}
assert!(mmap_directory.get_cache_info().mmapped.is_empty());
Ok(())
}
}

15
src/directory/ram_directory.rs
View File

@@ -18,13 +18,6 @@ use super::FileHandle;
/// Writer associated with the `RamDirectory`
///
/// The Writer just writes a buffer.
///
/// # Panics
///
/// On drop, if the writer was left in a *dirty* state.
/// That is, if flush was not called after the last call
/// to write.
///
struct VecWriter {
path: PathBuf,
shared_directory: RamDirectory,
@@ -46,7 +39,7 @@ impl VecWriter {
impl Drop for VecWriter {
fn drop(&mut self) {
if !self.is_flushed {
panic!(
warn!(
"You forgot to flush {:?} before its writter got Drop. Do not rely on drop. This also occurs when the indexer crashed, so you may want to check the logs for the root cause.",
self.path
)
@@ -221,14 +214,8 @@ impl Directory for RamDirectory {
}
fn atomic_write(&self, path: &Path, data: &[u8]) -> io::Result<()> {
fail_point!("RamDirectory::atomic_write", |msg| Err(io::Error::new(
io::ErrorKind::Other,
msg.unwrap_or_else(|| "Undefined".to_string())
)));
let path_buf = PathBuf::from(path);
self.fs.write().unwrap().write(path_buf, data);
if path == *META_FILEPATH {
let _ = self.fs.write().unwrap().watch_router.broadcast();
}

9
src/directory/tests.rs
View File

@@ -118,15 +118,6 @@ mod ram_directory_tests {
}
}
#[test]
#[should_panic]
fn ram_directory_panics_if_flush_forgotten() {
let test_path: &'static Path = Path::new("some_path_for_test");
let ram_directory = RamDirectory::create();
let mut write_file = ram_directory.open_write(test_path).unwrap();
assert!(write_file.write_all(&[4]).is_ok());
}
fn test_simple(directory: &dyn Directory) -> crate::Result<()> {
let test_path: &'static Path = Path::new("some_path_for_test");
let mut write_file = directory.open_write(test_path)?;

14
src/docset.rs
View File

@@ -1,4 +1,4 @@
use crate::fastfield::DeleteBitSet;
use crate::fastfield::AliveBitSet;
use crate::DocId;
use std::borrow::Borrow;
use std::borrow::BorrowMut;
@@ -85,11 +85,11 @@ pub trait DocSet: Send {
/// Returns the number documents matching.
/// Calling this method consumes the `DocSet`.
fn count(&mut self, delete_bitset: &DeleteBitSet) -> u32 {
fn count(&mut self, alive_bitset: &AliveBitSet) -> u32 {
let mut count = 0u32;
let mut doc = self.doc();
while doc != TERMINATED {
if !delete_bitset.is_deleted(doc) {
if alive_bitset.is_alive(doc) {
count += 1u32;
}
doc = self.advance();
@@ -130,8 +130,8 @@ impl<'a> DocSet for &'a mut dyn DocSet {
(**self).size_hint()
}
fn count(&mut self, delete_bitset: &DeleteBitSet) -> u32 {
(**self).count(delete_bitset)
fn count(&mut self, alive_bitset: &AliveBitSet) -> u32 {
(**self).count(alive_bitset)
}
fn count_including_deleted(&mut self) -> u32 {
@@ -160,9 +160,9 @@ impl<TDocSet: DocSet + ?Sized> DocSet for Box<TDocSet> {
unboxed.size_hint()
}
fn count(&mut self, delete_bitset: &DeleteBitSet) -> u32 {
fn count(&mut self, alive_bitset: &AliveBitSet) -> u32 {
let unboxed: &mut TDocSet = self.borrow_mut();
unboxed.count(delete_bitset)
unboxed.count(alive_bitset)
}
fn count_including_deleted(&mut self) -> u32 {

224
src/fastfield/alive_bitset.rs Normal file
View File

@@ -0,0 +1,224 @@
use crate::space_usage::ByteCount;
use crate::DocId;
use common::intersect_bitsets;
use common::BitSet;
use common::ReadOnlyBitSet;
use ownedbytes::OwnedBytes;
use std::io;
use std::io::Write;
/// Write a alive `BitSet`
///
/// where `alive_bitset` is the set of alive `DocId`.
/// Warning: this function does not call terminate. The caller is in charge of
/// closing the writer properly.
pub fn write_alive_bitset<T: Write>(alive_bitset: &BitSet, writer: &mut T) -> io::Result<()> {
alive_bitset.serialize(writer)?;
Ok(())
}
/// Set of alive `DocId`s.
#[derive(Clone)]
pub struct AliveBitSet {
num_alive_docs: usize,
bitset: ReadOnlyBitSet,
}
/// Intersects two AliveBitSets in a new one.
/// The two bitsets need to have the same max_value.
pub fn intersect_alive_bitsets(left: AliveBitSet, right: AliveBitSet) -> AliveBitSet {
assert_eq!(left.bitset().max_value(), right.bitset().max_value());
let bitset = intersect_bitsets(left.bitset(), right.bitset());
let num_alive_docs = bitset.len();
AliveBitSet {
num_alive_docs,
bitset,
}
}
impl AliveBitSet {
#[cfg(test)]
pub(crate) fn for_test_from_deleted_docs(deleted_docs: &[DocId], max_doc: u32) -> AliveBitSet {
assert!(deleted_docs.iter().all(|&doc| doc < max_doc));
let mut bitset = BitSet::with_max_value_and_full(max_doc);
for &doc in deleted_docs {
bitset.remove(doc);
}
let mut alive_bitset_buffer = Vec::new();
write_alive_bitset(&bitset, &mut alive_bitset_buffer).unwrap();
let alive_bitset_bytes = OwnedBytes::new(alive_bitset_buffer);
Self::open(alive_bitset_bytes)
}
pub(crate) fn from_bitset(bitset: &BitSet) -> AliveBitSet {
let readonly_bitset = ReadOnlyBitSet::from(bitset);
AliveBitSet::from(readonly_bitset)
}
/// Opens a delete bitset given its file.
pub fn open(bytes: OwnedBytes) -> AliveBitSet {
let bitset = ReadOnlyBitSet::open(bytes);
AliveBitSet::from(bitset)
}
/// Returns true iff the document is still "alive". In other words, if it has not been deleted.
#[inline]
pub fn is_alive(&self, doc: DocId) -> bool {
self.bitset.contains(doc)
}
/// Returns true iff the document has been marked as deleted.
#[inline]
pub fn is_deleted(&self, doc: DocId) -> bool {
!self.is_alive(doc)
}
/// Iterate over the alive doc_ids.
#[inline]
pub fn iter_alive(&self) -> impl Iterator<Item = DocId> + '_ {
self.bitset.iter()
}
/// Get underlying bitset
#[inline]
pub fn bitset(&self) -> &ReadOnlyBitSet {
&self.bitset
}
/// The number of deleted docs
pub fn num_alive_docs(&self) -> usize {
self.num_alive_docs
}
/// Summarize total space usage of this bitset.
pub fn space_usage(&self) -> ByteCount {
self.bitset().num_bytes()
}
}
impl From<ReadOnlyBitSet> for AliveBitSet {
fn from(bitset: ReadOnlyBitSet) -> AliveBitSet {
let num_alive_docs = bitset.len();
AliveBitSet {
num_alive_docs,
bitset,
}
}
}
#[cfg(test)]
mod tests {
use super::AliveBitSet;
#[test]
fn test_alive_bitset_empty() {
let alive_bitset = AliveBitSet::for_test_from_deleted_docs(&[], 10);
for doc in 0..10 {
assert_eq!(alive_bitset.is_deleted(doc), !alive_bitset.is_alive(doc));
assert!(!alive_bitset.is_deleted(doc));
}
assert_eq!(alive_bitset.num_alive_docs(), 10);
}
#[test]
fn test_alive_bitset() {
let alive_bitset = AliveBitSet::for_test_from_deleted_docs(&[1, 9], 10);
assert!(alive_bitset.is_alive(0));
assert!(alive_bitset.is_deleted(1));
assert!(alive_bitset.is_alive(2));
assert!(alive_bitset.is_alive(3));
assert!(alive_bitset.is_alive(4));
assert!(alive_bitset.is_alive(5));
assert!(alive_bitset.is_alive(6));
assert!(alive_bitset.is_alive(6));
assert!(alive_bitset.is_alive(7));
assert!(alive_bitset.is_alive(8));
assert!(alive_bitset.is_deleted(9));
for doc in 0..10 {
assert_eq!(alive_bitset.is_deleted(doc), !alive_bitset.is_alive(doc));
}
assert_eq!(alive_bitset.num_alive_docs(), 8);
}
#[test]
fn test_alive_bitset_iter_minimal() {
let alive_bitset = AliveBitSet::for_test_from_deleted_docs(&[7], 8);
let data: Vec<_> = alive_bitset.iter_alive().collect();
assert_eq!(data, vec![0, 1, 2, 3, 4, 5, 6]);
}
#[test]
fn test_alive_bitset_iter_small() {
let alive_bitset = AliveBitSet::for_test_from_deleted_docs(&[0, 2, 3, 6], 7);
let data: Vec<_> = alive_bitset.iter_alive().collect();
assert_eq!(data, vec![1, 4, 5]);
}
#[test]
fn test_alive_bitset_iter() {
let alive_bitset = AliveBitSet::for_test_from_deleted_docs(&[0, 1, 1000], 1001);
let data: Vec<_> = alive_bitset.iter_alive().collect();
assert_eq!(data, (2..=999).collect::<Vec<_>>());
}
}
#[cfg(all(test, feature = "unstable"))]
mod bench {
use super::AliveBitSet;
use rand::prelude::IteratorRandom;
use rand::thread_rng;
use test::Bencher;
fn get_alive() -> Vec<u32> {
let mut data = (0..1_000_000_u32).collect::<Vec<u32>>();
for _ in 0..(1_000_000) * 1 / 8 {
remove_rand(&mut data);
}
data
}
fn remove_rand(raw: &mut Vec<u32>) {
let i = (0..raw.len()).choose(&mut thread_rng()).unwrap();
raw.remove(i);
}
#[bench]
fn bench_deletebitset_iter_deser_on_fly(bench: &mut Bencher) {
let alive_bitset = AliveBitSet::for_test_from_deleted_docs(&[0, 1, 1000, 10000], 1_000_000);
bench.iter(|| alive_bitset.iter_alive().collect::<Vec<_>>());
}
#[bench]
fn bench_deletebitset_access(bench: &mut Bencher) {
let alive_bitset = AliveBitSet::for_test_from_deleted_docs(&[0, 1, 1000, 10000], 1_000_000);
bench.iter(|| {
(0..1_000_000_u32)
.filter(|doc| alive_bitset.is_alive(*doc))
.collect::<Vec<_>>()
});
}
#[bench]
fn bench_deletebitset_iter_deser_on_fly_1_8_alive(bench: &mut Bencher) {
let alive_bitset = AliveBitSet::for_test_from_deleted_docs(&get_alive(), 1_000_000);
bench.iter(|| alive_bitset.iter_alive().collect::<Vec<_>>());
}
#[bench]
fn bench_deletebitset_access_1_8_alive(bench: &mut Bencher) {
let alive_bitset = AliveBitSet::for_test_from_deleted_docs(&get_alive(), 1_000_000);
bench.iter(|| {
(0..1_000_000_u32)
.filter(|doc| alive_bitset.is_alive(*doc))
.collect::<Vec<_>>()
});
}
}

12
src/fastfield/bytes/mod.rs
View File

@@ -18,11 +18,11 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(bytes_field=>vec![0u8, 1, 2, 3]));
index_writer.add_document(doc!(bytes_field=>vec![]));
index_writer.add_document(doc!(bytes_field=>vec![255u8]));
index_writer.add_document(doc!(bytes_field=>vec![1u8, 3, 5, 7, 9]));
index_writer.add_document(doc!(bytes_field=>vec![0u8; 1000]));
index_writer.add_document(doc!(bytes_field=>vec![0u8, 1, 2, 3]))?;
index_writer.add_document(doc!(bytes_field=>vec![]))?;
index_writer.add_document(doc!(bytes_field=>vec![255u8]))?;
index_writer.add_document(doc!(bytes_field=>vec![1u8, 3, 5, 7, 9]))?;
index_writer.add_document(doc!(bytes_field=>vec![0u8; 1000]))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let segment_reader = searcher.segment_reader(0);
@@ -47,7 +47,7 @@ mod tests {
index_writer.add_document(doc!(
field => b"tantivy".as_ref(),
field => b"lucene".as_ref()
));
))?;
index_writer.commit()?;
Ok(index.reader()?.searcher())
}

144
src/fastfield/delete.rs
View File

@@ -1,144 +0,0 @@
use crate::directory::FileSlice;
use crate::directory::OwnedBytes;
use crate::directory::WritePtr;
use crate::space_usage::ByteCount;
use crate::DocId;
use common::BitSet;
use common::HasLen;
use std::io;
use std::io::Write;
/// Write a delete `BitSet`
///
/// where `delete_bitset` is the set of deleted `DocId`.
/// Warning: this function does not call terminate. The caller is in charge of
/// closing the writer properly.
pub fn write_delete_bitset(
delete_bitset: &BitSet,
max_doc: u32,
writer: &mut WritePtr,
) -> io::Result<()> {
let mut byte = 0u8;
let mut shift = 0u8;
for doc in 0..max_doc {
if delete_bitset.contains(doc) {
byte |= 1 << shift;
}
if shift == 7 {
writer.write_all(&[byte])?;
shift = 0;
byte = 0;
} else {
shift += 1;
}
}
if max_doc % 8 > 0 {
writer.write_all(&[byte])?;
}
Ok(())
}
/// Set of deleted `DocId`s.
#[derive(Clone)]
pub struct DeleteBitSet {
data: OwnedBytes,
num_deleted: usize,
}
impl DeleteBitSet {
#[cfg(test)]
pub(crate) fn for_test(docs: &[DocId], max_doc: u32) -> DeleteBitSet {
use crate::directory::{Directory, RamDirectory, TerminatingWrite};
use std::path::Path;
assert!(docs.iter().all(|&doc| doc < max_doc));
let mut bitset = BitSet::with_max_value(max_doc);
for &doc in docs {
bitset.insert(doc);
}
let directory = RamDirectory::create();
let path = Path::new("dummydeletebitset");
let mut wrt = directory.open_write(path).unwrap();
write_delete_bitset(&bitset, max_doc, &mut wrt).unwrap();
wrt.terminate().unwrap();
let file = directory.open_read(path).unwrap();
Self::open(file).unwrap()
}
/// Opens a delete bitset given its file.
pub fn open(file: FileSlice) -> crate::Result<DeleteBitSet> {
let bytes = file.read_bytes()?;
let num_deleted: usize = bytes
.as_slice()
.iter()
.map(|b| b.count_ones() as usize)
.sum();
Ok(DeleteBitSet {
data: bytes,
num_deleted,
})
}
/// Returns true iff the document is still "alive". In other words, if it has not been deleted.
pub fn is_alive(&self, doc: DocId) -> bool {
!self.is_deleted(doc)
}
/// Returns true iff the document has been marked as deleted.
#[inline]
pub fn is_deleted(&self, doc: DocId) -> bool {
let byte_offset = doc / 8u32;
let b: u8 = self.data.as_slice()[byte_offset as usize];
let shift = (doc & 7u32) as u8;
b & (1u8 << shift) != 0
}
/// The number of deleted docs
pub fn num_deleted(&self) -> usize {
self.num_deleted
}
/// Summarize total space usage of this bitset.
pub fn space_usage(&self) -> ByteCount {
self.data.len()
}
}
impl HasLen for DeleteBitSet {
fn len(&self) -> usize {
self.num_deleted
}
}
#[cfg(test)]
mod tests {
use super::DeleteBitSet;
use common::HasLen;
#[test]
fn test_delete_bitset_empty() {
let delete_bitset = DeleteBitSet::for_test(&[], 10);
for doc in 0..10 {
assert_eq!(delete_bitset.is_deleted(doc), !delete_bitset.is_alive(doc));
}
assert_eq!(delete_bitset.len(), 0);
}
#[test]
fn test_delete_bitset() {
let delete_bitset = DeleteBitSet::for_test(&[1, 9], 10);
assert!(delete_bitset.is_alive(0));
assert!(delete_bitset.is_deleted(1));
assert!(delete_bitset.is_alive(2));
assert!(delete_bitset.is_alive(3));
assert!(delete_bitset.is_alive(4));
assert!(delete_bitset.is_alive(5));
assert!(delete_bitset.is_alive(6));
assert!(delete_bitset.is_alive(6));
assert!(delete_bitset.is_alive(7));
assert!(delete_bitset.is_alive(8));
assert!(delete_bitset.is_deleted(9));
for doc in 0..10 {
assert_eq!(delete_bitset.is_deleted(doc), !delete_bitset.is_alive(doc));
}
assert_eq!(delete_bitset.len(), 2);
}
}

16
src/fastfield/facet_reader.rs
View File

@@ -95,7 +95,7 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(facet_field=>Facet::from_text("/a/b").unwrap()));
index_writer.add_document(doc!(facet_field=>Facet::from_text("/a/b").unwrap()))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let facet_reader = searcher
@@ -118,7 +118,7 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(facet_field=>Facet::from_text("/a/b").unwrap()));
index_writer.add_document(doc!(facet_field=>Facet::from_text("/a/b").unwrap()))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let facet_reader = searcher
@@ -141,7 +141,7 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(facet_field=>Facet::from_text("/a/b").unwrap()));
index_writer.add_document(doc!(facet_field=>Facet::from_text("/a/b").unwrap()))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let facet_reader = searcher
@@ -164,7 +164,7 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(facet_field=>Facet::from_text("/a/b").unwrap()));
index_writer.add_document(doc!(facet_field=>Facet::from_text("/a/b").unwrap()))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let facet_reader = searcher
@@ -187,8 +187,8 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(facet_field=>Facet::from_text("/a/b").unwrap()));
index_writer.add_document(Document::default());
index_writer.add_document(doc!(facet_field=>Facet::from_text("/a/b").unwrap()))?;
index_writer.add_document(Document::default())?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let facet_reader = searcher
@@ -210,8 +210,8 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(Document::default());
index_writer.add_document(Document::default());
index_writer.add_document(Document::default())?;
index_writer.add_document(Document::default())?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let facet_reader = searcher

40
src/fastfield/mod.rs
View File

@@ -23,9 +23,10 @@ values stored.
Read access performance is comparable to that of an array lookup.
*/
pub use self::alive_bitset::intersect_alive_bitsets;
pub use self::alive_bitset::write_alive_bitset;
pub use self::alive_bitset::AliveBitSet;
pub use self::bytes::{BytesFastFieldReader, BytesFastFieldWriter};
pub use self::delete::write_delete_bitset;
pub use self::delete::DeleteBitSet;
pub use self::error::{FastFieldNotAvailableError, Result};
pub use self::facet_reader::FacetReader;
pub use self::multivalued::{MultiValuedFastFieldReader, MultiValuedFastFieldWriter};
@@ -46,8 +47,8 @@ use crate::{
schema::Type,
};
mod alive_bitset;
mod bytes;
mod delete;
mod error;
mod facet_reader;
mod multivalued;
@@ -496,18 +497,18 @@ mod tests {
}
#[test]
fn test_merge_missing_date_fast_field() {
fn test_merge_missing_date_fast_field() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let date_field = schema_builder.add_date_field("date", FAST);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.set_merge_policy(Box::new(NoMergePolicy));
index_writer.add_document(doc!(date_field =>crate::chrono::prelude::Utc::now()));
index_writer.commit().unwrap();
index_writer.add_document(doc!());
index_writer.commit().unwrap();
let reader = index.reader().unwrap();
index_writer.add_document(doc!(date_field =>crate::chrono::prelude::Utc::now()))?;
index_writer.commit()?;
index_writer.add_document(doc!())?;
index_writer.commit()?;
let reader = index.reader()?;
let segment_ids: Vec<SegmentId> = reader
.searcher()
.segment_readers()
@@ -516,10 +517,10 @@ mod tests {
.collect();
assert_eq!(segment_ids.len(), 2);
let merge_future = index_writer.merge(&segment_ids[..]);
let merge_res = futures::executor::block_on(merge_future);
assert!(merge_res.is_ok());
assert!(reader.reload().is_ok());
futures::executor::block_on(merge_future)?;
reader.reload()?;
assert_eq!(reader.searcher().segment_readers().len(), 1);
Ok(())
}
#[test]
@@ -528,7 +529,7 @@ mod tests {
}
#[test]
fn test_datefastfield() {
fn test_datefastfield() -> crate::Result<()> {
use crate::fastfield::FastValue;
let mut schema_builder = Schema::builder();
let date_field = schema_builder.add_date_field("date", FAST);
@@ -538,22 +539,22 @@ mod tests {
);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
index_writer.set_merge_policy(Box::new(NoMergePolicy));
index_writer.add_document(doc!(
date_field => crate::DateTime::from_u64(1i64.to_u64()),
multi_date_field => crate::DateTime::from_u64(2i64.to_u64()),
multi_date_field => crate::DateTime::from_u64(3i64.to_u64())
));
))?;
index_writer.add_document(doc!(
date_field => crate::DateTime::from_u64(4i64.to_u64())
));
))?;
index_writer.add_document(doc!(
multi_date_field => crate::DateTime::from_u64(5i64.to_u64()),
multi_date_field => crate::DateTime::from_u64(6i64.to_u64())
));
index_writer.commit().unwrap();
let reader = index.reader().unwrap();
))?;
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
assert_eq!(searcher.segment_readers().len(), 1);
let segment_reader = searcher.segment_reader(0);
@@ -580,6 +581,7 @@ mod tests {
assert_eq!(dates[0].timestamp(), 5i64);
assert_eq!(dates[1].timestamp(), 6i64);
}
Ok(())
}
}

126
src/fastfield/multivalued/mod.rs
View File

@@ -26,7 +26,7 @@ mod tests {
use test_env_log::test;
#[test]
fn test_multivalued_u64() {
fn test_multivalued_u64() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let field = schema_builder.add_u64_field(
"multifield",
@@ -34,17 +34,17 @@ mod tests {
);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(field=>1u64, field=>3u64));
index_writer.add_document(doc!());
index_writer.add_document(doc!(field=>4u64));
index_writer.add_document(doc!(field=>5u64, field=>20u64,field=>1u64));
assert!(index_writer.commit().is_ok());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(field=>1u64, field=>3u64))?;
index_writer.add_document(doc!())?;
index_writer.add_document(doc!(field=>4u64))?;
index_writer.add_document(doc!(field=>5u64, field=>20u64,field=>1u64))?;
index_writer.commit()?;
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
let segment_reader = searcher.segment_reader(0);
let mut vals = Vec::new();
let multi_value_reader = segment_reader.fast_fields().u64s(field).unwrap();
let multi_value_reader = segment_reader.fast_fields().u64s(field)?;
{
multi_value_reader.get_vals(2, &mut vals);
assert_eq!(&vals, &[4u64]);
@@ -57,10 +57,11 @@ mod tests {
multi_value_reader.get_vals(1, &mut vals);
assert!(vals.is_empty());
}
Ok(())
}
#[test]
fn test_multivalued_date() {
fn test_multivalued_date() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let date_field = schema_builder.add_date_field(
"multi_date_field",
@@ -73,40 +74,37 @@ mod tests {
schema_builder.add_i64_field("time_stamp_i", IntOptions::default().set_stored());
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
let first_time_stamp = chrono::Utc::now();
index_writer.add_document(
doc!(date_field=>first_time_stamp, date_field=>first_time_stamp, time_i=>1i64),
);
index_writer.add_document(doc!(time_i=>0i64));
)?;
index_writer.add_document(doc!(time_i=>0i64))?;
// add one second
index_writer
.add_document(doc!(date_field=>first_time_stamp + Duration::seconds(1), time_i=>2i64));
index_writer.add_document(
doc!(date_field=>first_time_stamp + Duration::seconds(1), time_i=>2i64),
)?;
// add another second
let two_secs_ahead = first_time_stamp + Duration::seconds(2);
index_writer.add_document(doc!(date_field=>two_secs_ahead, date_field=>two_secs_ahead,date_field=>two_secs_ahead, time_i=>3i64));
index_writer.add_document(doc!(date_field=>two_secs_ahead, date_field=>two_secs_ahead,date_field=>two_secs_ahead, time_i=>3i64))?;
// add three seconds
index_writer
.add_document(doc!(date_field=>first_time_stamp + Duration::seconds(3), time_i=>4i64));
assert!(index_writer.commit().is_ok());
index_writer.add_document(
doc!(date_field=>first_time_stamp + Duration::seconds(3), time_i=>4i64),
)?;
index_writer.commit()?;
let reader = index.reader().unwrap();
let reader = index.reader()?;
let searcher = reader.searcher();
let reader = searcher.segment_reader(0);
assert_eq!(reader.num_docs(), 5);
{
let parser = QueryParser::for_index(&index, vec![date_field]);
let query = parser
.parse_query(&format!("\"{}\"", first_time_stamp.to_rfc3339()))
.expect("could not parse query");
let results = searcher
.search(&query, &TopDocs::with_limit(5))
.expect("could not query index");
let query = parser.parse_query(&format!("\"{}\"", first_time_stamp.to_rfc3339()))?;
let results = searcher.search(&query, &TopDocs::with_limit(5))?;
assert_eq!(results.len(), 1);
for (_score, doc_address) in results {
let retrieved_doc = searcher.doc(doc_address).expect("cannot fetch doc");
let retrieved_doc = searcher.doc(doc_address)?;
assert_eq!(
retrieved_doc
.get_first(date_field)
@@ -128,12 +126,8 @@ mod tests {
{
let parser = QueryParser::for_index(&index, vec![date_field]);
let query = parser
.parse_query(&format!("\"{}\"", two_secs_ahead.to_rfc3339()))
.expect("could not parse query");
let results = searcher
.search(&query, &TopDocs::with_limit(5))
.expect("could not query index");
let query = parser.parse_query(&format!("\"{}\"", two_secs_ahead.to_rfc3339()))?;
let results = searcher.search(&query, &TopDocs::with_limit(5))?;
assert_eq!(results.len(), 1);
@@ -165,10 +159,8 @@ mod tests {
(first_time_stamp + Duration::seconds(1)).to_rfc3339(),
(first_time_stamp + Duration::seconds(3)).to_rfc3339()
);
let query = parser.parse_query(&range_q).expect("could not parse query");
let results = searcher
.search(&query, &TopDocs::with_limit(5))
.expect("could not query index");
let query = parser.parse_query(&range_q)?;
let results = searcher.search(&query, &TopDocs::with_limit(5))?;
assert_eq!(results.len(), 2);
for (i, doc_pair) in results.iter().enumerate() {
@@ -196,16 +188,16 @@ mod tests {
retrieved_doc
.get_first(time_i)
.expect("cannot find value")
.i64_value()
.expect("value not of i64 type"),
time_i_val
.i64_value(),
Some(time_i_val)
);
}
}
Ok(())
}
#[test]
fn test_multivalued_i64() {
fn test_multivalued_i64() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let field = schema_builder.add_i64_field(
"multifield",
@@ -213,14 +205,14 @@ mod tests {
);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(field=> 1i64, field => 3i64));
index_writer.add_document(doc!());
index_writer.add_document(doc!(field=> -4i64));
index_writer.add_document(doc!(field=> -5i64, field => -20i64, field=>1i64));
assert!(index_writer.commit().is_ok());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(field=> 1i64, field => 3i64))?;
index_writer.add_document(doc!())?;
index_writer.add_document(doc!(field=> -4i64))?;
index_writer.add_document(doc!(field=> -5i64, field => -20i64, field=>1i64))?;
index_writer.commit()?;
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
let segment_reader = searcher.segment_reader(0);
let mut vals = Vec::new();
let multi_value_reader = segment_reader.fast_fields().i64s(field).unwrap();
@@ -232,9 +224,10 @@ mod tests {
assert!(vals.is_empty());
multi_value_reader.get_vals(3, &mut vals);
assert_eq!(&vals, &[-5i64, -20i64, 1i64]);
Ok(())
}
fn test_multivalued_no_panic(ops: &[IndexingOp]) {
fn test_multivalued_no_panic(ops: &[IndexingOp]) -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let field = schema_builder.add_u64_field(
"multifield",
@@ -244,7 +237,7 @@ mod tests {
);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
index_writer.set_merge_policy(Box::new(NoMergePolicy));
for &op in ops {
@@ -252,19 +245,19 @@ mod tests {
IndexingOp::AddDoc { id } => {
match id % 3 {
0 => {
index_writer.add_document(doc!());
index_writer.add_document(doc!())?;
}
1 => {
let mut doc = Document::new();
for _ in 0..5001 {
doc.add_u64(field, id as u64);
}
index_writer.add_document(doc);
index_writer.add_document(doc)?;
}
_ => {
let mut doc = Document::new();
doc.add_u64(field, id as u64);
index_writer.add_document(doc);
index_writer.add_document(doc)?;
}
};
}
@@ -275,18 +268,16 @@ mod tests {
index_writer.commit().unwrap();
}
IndexingOp::Merge => {
let segment_ids = index
.searchable_segment_ids()
.expect("Searchable segments failed.");
let segment_ids = index.searchable_segment_ids()?;
if segment_ids.len() >= 2 {
block_on(index_writer.merge(&segment_ids)).unwrap();
assert!(index_writer.segment_updater().wait_merging_thread().is_ok());
block_on(index_writer.merge(&segment_ids))?;
index_writer.segment_updater().wait_merging_thread()?;
}
}
}
}
assert!(index_writer.commit().is_ok());
index_writer.commit()?;
// Merging the segments
{
@@ -298,6 +289,7 @@ mod tests {
assert!(index_writer.wait_merging_threads().is_ok());
}
}
Ok(())
}
#[derive(Debug, Clone, Copy)]
@@ -320,7 +312,7 @@ mod tests {
proptest! {
#[test]
fn test_multivalued_proptest(ops in proptest::collection::vec(operation_strategy(), 1..10)) {
test_multivalued_no_panic(&ops[..]);
assert!(test_multivalued_no_panic(&ops[..]).is_ok());
}
}
@@ -335,20 +327,22 @@ mod tests {
Merge,
];
test_multivalued_no_panic(&ops[..]);
assert!(test_multivalued_no_panic(&ops[..]).is_ok());
}
#[test]
#[ignore]
fn test_many_facets() {
fn test_many_facets() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let field = schema_builder.add_facet_field("facetfield", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
for i in 0..100_000 {
index_writer.add_document(doc!(field=> Facet::from(format!("/lang/{}", i).as_str())));
index_writer
.add_document(doc!(field=> Facet::from(format!("/lang/{}", i).as_str())))?;
}
assert!(index_writer.commit().is_ok());
index_writer.commit()?;
Ok(())
}
}

34
src/fastfield/multivalued/reader.rs
View File

@@ -94,24 +94,22 @@ mod tests {
use crate::schema::{Cardinality, Facet, IntOptions, Schema, INDEXED};
#[test]
fn test_multifastfield_reader() {
fn test_multifastfield_reader() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let facet_field = schema_builder.add_facet_field("facets", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index
.writer_for_tests()
.expect("Failed to create index writer.");
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(
facet_field => Facet::from("/category/cat2"),
facet_field => Facet::from("/category/cat1"),
));
index_writer.add_document(doc!(facet_field => Facet::from("/category/cat2")));
index_writer.add_document(doc!(facet_field => Facet::from("/category/cat3")));
index_writer.commit().expect("Commit failed");
let searcher = index.reader().unwrap().searcher();
))?;
index_writer.add_document(doc!(facet_field => Facet::from("/category/cat2")))?;
index_writer.add_document(doc!(facet_field => Facet::from("/category/cat3")))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let segment_reader = searcher.segment_reader(0);
let mut facet_reader = segment_reader.facet_reader(facet_field).unwrap();
let mut facet_reader = segment_reader.facet_reader(facet_field)?;
let mut facet = Facet::root();
{
@@ -145,10 +143,11 @@ mod tests {
facet_reader.facet_ords(2, &mut vals);
assert_eq!(&vals[..], &[4]);
}
Ok(())
}
#[test]
fn test_multifastfield_reader_min_max() {
fn test_multifastfield_reader_min_max() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let field_options = IntOptions::default()
.set_indexed()
@@ -163,15 +162,16 @@ mod tests {
item_field => 2i64,
item_field => 3i64,
item_field => -2i64,
));
index_writer.add_document(doc!(item_field => 6i64, item_field => 3i64));
index_writer.add_document(doc!(item_field => 4i64));
index_writer.commit().expect("Commit failed");
let searcher = index.reader().unwrap().searcher();
))?;
index_writer.add_document(doc!(item_field => 6i64, item_field => 3i64))?;
index_writer.add_document(doc!(item_field => 4i64))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let segment_reader = searcher.segment_reader(0);
let field_reader = segment_reader.fast_fields().i64s(item_field).unwrap();
let field_reader = segment_reader.fast_fields().i64s(item_field)?;
assert_eq!(field_reader.min_value(), -2);
assert_eq!(field_reader.max_value(), 6);
Ok(())
}
}

6
src/functional_test.rs
View File

@@ -49,7 +49,7 @@ fn test_functional_store() -> crate::Result<()> {
}
for _ in 0..num_docs {
doc_set.push(doc_id);
index_writer.add_document(doc!(id_field=>doc_id));
index_writer.add_document(doc!(id_field=>doc_id))?;
doc_id += 1;
}
index_writer.commit()?;
@@ -124,7 +124,7 @@ fn test_functional_indexing_sorted() -> crate::Result<()> {
doc.add_u64(multiples_field, random_val * i);
}
doc.add_text(text_field, get_text());
index_writer.add_document(doc);
index_writer.add_document(doc)?;
}
}
Ok(())
@@ -201,7 +201,7 @@ fn test_functional_indexing_unsorted() -> crate::Result<()> {
doc.add_u64(multiples_field, random_val * i);
}
doc.add_text(text_field, get_text());
index_writer.add_document(doc);
index_writer.add_document(doc)?;
}
}
Ok(())

324
src/indexer/demuxer.rs Normal file
View File

@@ -0,0 +1,324 @@
use common::BitSet;
use itertools::Itertools;
use crate::fastfield::AliveBitSet;
use crate::{merge_filtered_segments, Directory, Index, IndexSettings, Segment, SegmentOrdinal};
/// DemuxMapping can be used to reorganize data from multiple segments.
///
/// DemuxMapping is useful in a multitenant settings, in which each document might actually belong to a different tenant.
/// It allows to reorganize documents as follows:
///
/// e.g. if you have two tenant ids TENANT_A and TENANT_B and two segments with
/// the documents (simplified)
/// Seg 1 [TENANT_A, TENANT_B]
/// Seg 2 [TENANT_A, TENANT_B]
///
/// You may want to group your documents to
/// Seg 1 [TENANT_A, TENANT_A]
/// Seg 2 [TENANT_B, TENANT_B]
///
/// Demuxing is the tool for that.
/// Semantically you can define a mapping from [old segment ordinal, old doc_id] -> [new segment ordinal].
#[derive(Debug, Default)]
pub struct DemuxMapping {
/// [index old segment ordinal] -> [index doc_id] = new segment ordinal
mapping: Vec<DocIdToSegmentOrdinal>,
}
/// DocIdToSegmentOrdinal maps from doc_id within a segment to the new segment ordinal for demuxing.
///
/// For every source segment there is a `DocIdToSegmentOrdinal` to distribute its doc_ids.
#[derive(Debug, Default)]
pub struct DocIdToSegmentOrdinal {
doc_id_index_to_segment_ord: Vec<SegmentOrdinal>,
}
impl DocIdToSegmentOrdinal {
/// Creates a new DocIdToSegmentOrdinal with size of num_doc_ids.
/// Initially all doc_ids point to segment ordinal 0 and need to be set
/// the via `set` method.
pub fn with_max_doc(max_doc: usize) -> Self {
DocIdToSegmentOrdinal {
doc_id_index_to_segment_ord: vec![0; max_doc],
}
}
/// Returns the number of documents in this mapping.
/// It should be equal to the `max_doc` of the segment it targets.
pub fn max_doc(&self) -> u32 {
self.doc_id_index_to_segment_ord.len() as u32
}
/// Associates a doc_id with an output `SegmentOrdinal`.
pub fn set(&mut self, doc_id: u32, segment_ord: SegmentOrdinal) {
self.doc_id_index_to_segment_ord[doc_id as usize] = segment_ord;
}
/// Iterates over the new SegmentOrdinal in the order of the doc_id.
pub fn iter(&self) -> impl Iterator<Item = SegmentOrdinal> + '_ {
self.doc_id_index_to_segment_ord.iter().cloned()
}
}
impl DemuxMapping {
/// Adds a DocIdToSegmentOrdinal. The order of the pus calls
/// defines the old segment ordinal. e.g. first push = ordinal 0.
pub fn add(&mut self, segment_mapping: DocIdToSegmentOrdinal) {
self.mapping.push(segment_mapping);
}
/// Returns the old number of segments.
pub fn get_old_num_segments(&self) -> usize {
self.mapping.len()
}
}
fn docs_for_segment_ord(
doc_id_to_segment_ord: &DocIdToSegmentOrdinal,
target_segment_ord: SegmentOrdinal,
) -> AliveBitSet {
let mut bitset = BitSet::with_max_value(doc_id_to_segment_ord.max_doc());
for doc_id in doc_id_to_segment_ord
.iter()
.enumerate()
.filter(|(_doc_id, new_segment_ord)| *new_segment_ord == target_segment_ord)
.map(|(doc_id, _)| doc_id)
{
// add document if segment ordinal = target segment ordinal
bitset.insert(doc_id as u32);
}
AliveBitSet::from_bitset(&bitset)
}
fn get_alive_bitsets(
demux_mapping: &DemuxMapping,
target_segment_ord: SegmentOrdinal,
) -> Vec<AliveBitSet> {
demux_mapping
.mapping
.iter()
.map(|doc_id_to_segment_ord| {
docs_for_segment_ord(doc_id_to_segment_ord, target_segment_ord)
})
.collect_vec()
}
/// Demux the segments according to `demux_mapping`. See `DemuxMapping`.
/// The number of output_directories need to match max new segment ordinal from `demux_mapping`.
///
/// The ordinal of `segments` need to match the ordinals provided in `demux_mapping`.
pub fn demux(
segments: &[Segment],
demux_mapping: &DemuxMapping,
target_settings: IndexSettings,
output_directories: Vec<Box<dyn Directory>>,
) -> crate::Result<Vec<Index>> {
let mut indices = vec![];
for (target_segment_ord, output_directory) in output_directories.into_iter().enumerate() {
let delete_bitsets = get_alive_bitsets(demux_mapping, target_segment_ord as u32)
.into_iter()
.map(Some)
.collect_vec();
let index = merge_filtered_segments(
segments,
target_settings.clone(),
delete_bitsets,
output_directory,
)?;
indices.push(index);
}
Ok(indices)
}
#[cfg(test)]
mod tests {
use crate::{
collector::TopDocs,
directory::RamDirectory,
query::QueryParser,
schema::{Schema, TEXT},
DocAddress, Term,
};
use super::*;
#[test]
fn test_demux_map_to_deletebitset() {
let max_value = 2;
let mut demux_mapping = DemuxMapping::default();
//segment ordinal 0 mapping
let mut doc_id_to_segment = DocIdToSegmentOrdinal::with_max_doc(max_value);
doc_id_to_segment.set(0, 1);
doc_id_to_segment.set(1, 0);
demux_mapping.add(doc_id_to_segment);
//segment ordinal 1 mapping
let mut doc_id_to_segment = DocIdToSegmentOrdinal::with_max_doc(max_value);
doc_id_to_segment.set(0, 1);
doc_id_to_segment.set(1, 1);
demux_mapping.add(doc_id_to_segment);
{
let bit_sets_for_demuxing_to_segment_ord_0 = get_alive_bitsets(&demux_mapping, 0);
assert_eq!(
bit_sets_for_demuxing_to_segment_ord_0[0].is_deleted(0),
true
);
assert_eq!(
bit_sets_for_demuxing_to_segment_ord_0[0].is_deleted(1),
false
);
assert_eq!(
bit_sets_for_demuxing_to_segment_ord_0[1].is_deleted(0),
true
);
assert_eq!(
bit_sets_for_demuxing_to_segment_ord_0[1].is_deleted(1),
true
);
}
{
let bit_sets_for_demuxing_to_segment_ord_1 = get_alive_bitsets(&demux_mapping, 1);
assert_eq!(
bit_sets_for_demuxing_to_segment_ord_1[0].is_deleted(0),
false
);
assert_eq!(
bit_sets_for_demuxing_to_segment_ord_1[0].is_deleted(1),
true
);
assert_eq!(
bit_sets_for_demuxing_to_segment_ord_1[1].is_deleted(0),
false
);
assert_eq!(
bit_sets_for_demuxing_to_segment_ord_1[1].is_deleted(1),
false
);
}
}
#[test]
fn test_demux_segments() -> crate::Result<()> {
let first_index = {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"texto1"))?;
index_writer.add_document(doc!(text_field=>"texto2"))?;
index_writer.commit()?;
index
};
let second_index = {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"texto3"))?;
index_writer.add_document(doc!(text_field=>"texto4"))?;
index_writer.delete_term(Term::from_field_text(text_field, "4"));
index_writer.commit()?;
index
};
let mut segments: Vec<Segment> = Vec::new();
segments.extend(first_index.searchable_segments()?);
segments.extend(second_index.searchable_segments()?);
let target_settings = first_index.settings().clone();
let mut demux_mapping = DemuxMapping::default();
{
let max_value = 2;
//segment ordinal 0 mapping
let mut doc_id_to_segment = DocIdToSegmentOrdinal::with_max_doc(max_value);
doc_id_to_segment.set(0, 1);
doc_id_to_segment.set(1, 0);
demux_mapping.add(doc_id_to_segment);
//segment ordinal 1 mapping
let mut doc_id_to_segment = DocIdToSegmentOrdinal::with_max_doc(max_value);
doc_id_to_segment.set(0, 1);
doc_id_to_segment.set(1, 1);
demux_mapping.add(doc_id_to_segment);
}
assert_eq!(demux_mapping.get_old_num_segments(), 2);
let demuxed_indices = demux(
&segments,
&demux_mapping,
target_settings,
vec![
Box::new(RamDirectory::default()),
Box::new(RamDirectory::default()),
],
)?;
{
let index = &demuxed_indices[0];
let segments = index.searchable_segments()?;
assert_eq!(segments.len(), 1);
let segment_metas = segments[0].meta();
assert_eq!(segment_metas.num_deleted_docs(), 0);
assert_eq!(segment_metas.num_docs(), 1);
let searcher = index.reader().unwrap().searcher();
{
let text_field = index.schema().get_field("text").unwrap();
let do_search = |term: &str| {
let query = QueryParser::for_index(&index, vec![text_field])
.parse_query(term)
.unwrap();
let top_docs: Vec<(f32, DocAddress)> =
searcher.search(&query, &TopDocs::with_limit(3)).unwrap();
top_docs.iter().map(|el| el.1.doc_id).collect::<Vec<_>>()
};
assert_eq!(do_search("texto1"), vec![] as Vec<u32>);
assert_eq!(do_search("texto2"), vec![0]);
}
}
{
let index = &demuxed_indices[1];
let segments = index.searchable_segments()?;
assert_eq!(segments.len(), 1);
let segment_metas = segments[0].meta();
assert_eq!(segment_metas.num_deleted_docs(), 0);
assert_eq!(segment_metas.num_docs(), 3);
let searcher = index.reader().unwrap().searcher();
{
let text_field = index.schema().get_field("text").unwrap();
let do_search = |term: &str| {
let query = QueryParser::for_index(&index, vec![text_field])
.parse_query(term)
.unwrap();
let top_docs: Vec<(f32, DocAddress)> =
searcher.search(&query, &TopDocs::with_limit(3)).unwrap();
top_docs.iter().map(|el| el.1.doc_id).collect::<Vec<_>>()
};
assert_eq!(do_search("texto1"), vec![0]);
assert_eq!(do_search("texto2"), vec![] as Vec<u32>);
assert_eq!(do_search("texto3"), vec![1]);
assert_eq!(do_search("texto4"), vec![2]);
}
}
Ok(())
}
}

65
src/indexer/doc_id_mapping.rs
View File

@@ -2,23 +2,23 @@
//! to get mappings from old doc_id to new doc_id and vice versa, after sorting
//!
use super::{merger::SegmentReaderWithOrdinal, SegmentWriter};
use super::SegmentWriter;
use crate::{
schema::{Field, Schema},
DocId, IndexSortByField, Order, TantivyError,
DocId, IndexSortByField, Order, SegmentOrdinal, TantivyError,
};
use std::{cmp::Reverse, ops::Index};
/// Struct to provide mapping from new doc_id to old doc_id and segment.
#[derive(Clone)]
pub(crate) struct SegmentDocidMapping<'a> {
new_doc_id_to_old_and_segment: Vec<(DocId, SegmentReaderWithOrdinal<'a>)>,
pub(crate) struct SegmentDocIdMapping {
new_doc_id_to_old_and_segment: Vec<(DocId, SegmentOrdinal)>,
is_trivial: bool,
}
impl<'a> SegmentDocidMapping<'a> {
impl SegmentDocIdMapping {
pub(crate) fn new(
new_doc_id_to_old_and_segment: Vec<(DocId, SegmentReaderWithOrdinal<'a>)>,
new_doc_id_to_old_and_segment: Vec<(DocId, SegmentOrdinal)>,
is_trivial: bool,
) -> Self {
Self {
@@ -26,7 +26,7 @@ impl<'a> SegmentDocidMapping<'a> {
is_trivial,
}
}
pub(crate) fn iter(&self) -> impl Iterator<Item = &(DocId, SegmentReaderWithOrdinal)> {
pub(crate) fn iter(&self) -> impl Iterator<Item = &(DocId, SegmentOrdinal)> {
self.new_doc_id_to_old_and_segment.iter()
}
pub(crate) fn len(&self) -> usize {
@@ -40,15 +40,15 @@ impl<'a> SegmentDocidMapping<'a> {
self.is_trivial
}
}
impl<'a> Index<usize> for SegmentDocidMapping<'a> {
type Output = (DocId, SegmentReaderWithOrdinal<'a>);
impl Index<usize> for SegmentDocIdMapping {
type Output = (DocId, SegmentOrdinal);
fn index(&self, idx: usize) -> &Self::Output {
&self.new_doc_id_to_old_and_segment[idx]
}
}
impl<'a> IntoIterator for SegmentDocidMapping<'a> {
type Item = (DocId, SegmentReaderWithOrdinal<'a>);
impl IntoIterator for SegmentDocIdMapping {
type Item = (DocId, SegmentOrdinal);
type IntoIter = std::vec::IntoIter<Self::Item>;
fn into_iter(self) -> Self::IntoIter {
@@ -146,7 +146,7 @@ mod tests_indexsorting {
fn create_test_index(
index_settings: Option<IndexSettings>,
text_field_options: TextOptions,
) -> Index {
) -> crate::Result<Index> {
let mut schema_builder = Schema::builder();
let my_text_field = schema_builder.add_text_field("text_field", text_field_options);
@@ -166,19 +166,20 @@ mod tests_indexsorting {
if let Some(settings) = index_settings {
index_builder = index_builder.settings(settings);
}
let index = index_builder.create_in_ram().unwrap();
let index = index_builder.create_in_ram()?;
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(my_number=>40_u64));
index_writer
.add_document(doc!(my_number=>20_u64, multi_numbers => 5_u64, multi_numbers => 6_u64));
index_writer.add_document(doc!(my_number=>100_u64));
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(my_number=>40_u64))?;
index_writer.add_document(
doc!(my_number=>20_u64, multi_numbers => 5_u64, multi_numbers => 6_u64),
)?;
index_writer.add_document(doc!(my_number=>100_u64))?;
index_writer.add_document(
doc!(my_number=>10_u64, my_string_field=> "blublub", my_text_field => "some text"),
);
index_writer.add_document(doc!(my_number=>30_u64, multi_numbers => 3_u64 ));
index_writer.commit().unwrap();
index
)?;
index_writer.add_document(doc!(my_number=>30_u64, multi_numbers => 3_u64 ))?;
index_writer.commit()?;
Ok(index)
}
fn get_text_options() -> TextOptions {
TextOptions::default().set_indexing_options(
@@ -203,7 +204,7 @@ mod tests_indexsorting {
for option in options {
//let options = get_text_options();
// no index_sort
let index = create_test_index(None, option.clone());
let index = create_test_index(None, option.clone())?;
let my_text_field = index.schema().get_field("text_field").unwrap();
let searcher = index.reader()?.searcher();
@@ -225,7 +226,7 @@ mod tests_indexsorting {
..Default::default()
}),
option.clone(),
);
)?;
let my_text_field = index.schema().get_field("text_field").unwrap();
let reader = index.reader()?;
let searcher = reader.searcher();
@@ -257,7 +258,7 @@ mod tests_indexsorting {
..Default::default()
}),
option.clone(),
);
)?;
let my_string_field = index.schema().get_field("text_field").unwrap();
let searcher = index.reader()?.searcher();
@@ -287,7 +288,7 @@ mod tests_indexsorting {
#[test]
fn test_sort_index_get_documents() -> crate::Result<()> {
// default baseline
let index = create_test_index(None, get_text_options());
let index = create_test_index(None, get_text_options())?;
let my_string_field = index.schema().get_field("string_field").unwrap();
let searcher = index.reader()?.searcher();
{
@@ -316,7 +317,7 @@ mod tests_indexsorting {
..Default::default()
}),
get_text_options(),
);
)?;
let my_string_field = index.schema().get_field("string_field").unwrap();
let searcher = index.reader()?.searcher();
{
@@ -341,7 +342,7 @@ mod tests_indexsorting {
..Default::default()
}),
get_text_options(),
);
)?;
let my_string_field = index.schema().get_field("string_field").unwrap();
let searcher = index.reader()?.searcher();
{
@@ -356,7 +357,7 @@ mod tests_indexsorting {
#[test]
fn test_sort_index_test_string_field() -> crate::Result<()> {
let index = create_test_index(None, get_text_options());
let index = create_test_index(None, get_text_options())?;
let my_string_field = index.schema().get_field("string_field").unwrap();
let searcher = index.reader()?.searcher();
@@ -376,7 +377,7 @@ mod tests_indexsorting {
..Default::default()
}),
get_text_options(),
);
)?;
let my_string_field = index.schema().get_field("string_field").unwrap();
let reader = index.reader()?;
let searcher = reader.searcher();
@@ -407,7 +408,7 @@ mod tests_indexsorting {
..Default::default()
}),
get_text_options(),
);
)?;
let my_string_field = index.schema().get_field("string_field").unwrap();
let searcher = index.reader()?.searcher();
@@ -443,7 +444,7 @@ mod tests_indexsorting {
..Default::default()
}),
get_text_options(),
);
)?;
assert_eq!(
index.settings().sort_by_field.as_ref().unwrap().field,
"my_number".to_string()

486
src/indexer/index_writer.rs
View File

@@ -11,9 +11,10 @@ use crate::directory::TerminatingWrite;
use crate::directory::{DirectoryLock, GarbageCollectionResult};
use crate::docset::{DocSet, TERMINATED};
use crate::error::TantivyError;
use crate::fastfield::write_delete_bitset;
use crate::fastfield::write_alive_bitset;
use crate::indexer::delete_queue::{DeleteCursor, DeleteQueue};
use crate::indexer::doc_opstamp_mapping::DocToOpstampMapping;
use crate::indexer::index_writer_status::IndexWriterStatus;
use crate::indexer::operation::DeleteOperation;
use crate::indexer::stamper::Stamper;
use crate::indexer::MergePolicy;
@@ -28,13 +29,13 @@ use crossbeam::channel;
use futures::executor::block_on;
use futures::future::Future;
use smallvec::smallvec;
use smallvec::SmallVec;
use std::mem;
use std::ops::Range;
use std::sync::Arc;
use std::thread;
use std::thread::JoinHandle;
use super::{AddBatch, AddBatchReceiver, AddBatchSender};
// Size of the margin for the heap. A segment is closed when the remaining memory
// in the heap goes below MARGIN_IN_BYTES.
pub const MARGIN_IN_BYTES: usize = 1_000_000;
@@ -50,15 +51,12 @@ pub const MAX_NUM_THREAD: usize = 8;
// reaches `PIPELINE_MAX_SIZE_IN_DOCS`
const PIPELINE_MAX_SIZE_IN_DOCS: usize = 10_000;
// Group of operations.
// Most of the time, users will send operation one-by-one, but it can be useful to
// send them as a small block to ensure that
// - all docs in the operation will happen on the same segment and continuous doc_ids.
// - all operations in the group are committed at the same time, making the group
// atomic.
type OperationGroup = SmallVec<[AddOperation; 4]>;
type OperationSender = channel::Sender<OperationGroup>;
type OperationReceiver = channel::Receiver<OperationGroup>;
fn error_in_index_worker_thread(context: &str) -> TantivyError {
TantivyError::ErrorInThread(format!(
"{}. A worker thread encounterred an error (io::Error most likely) or panicked.",
context
))
}
/// `IndexWriter` is the user entry-point to add document to an index.
///
@@ -77,8 +75,8 @@ pub struct IndexWriter {
workers_join_handle: Vec<JoinHandle<crate::Result<()>>>,
operation_receiver: OperationReceiver,
operation_sender: OperationSender,
index_writer_status: IndexWriterStatus,
operation_sender: AddBatchSender,
segment_updater: SegmentUpdater,
@@ -93,7 +91,7 @@ pub struct IndexWriter {
}
fn compute_deleted_bitset(
delete_bitset: &mut BitSet,
alive_bitset: &mut BitSet,
segment_reader: &SegmentReader,
delete_cursor: &mut DeleteCursor,
doc_opstamps: &DocToOpstampMapping,
@@ -114,7 +112,7 @@ fn compute_deleted_bitset(
let mut doc_matching_deleted_term = docset.doc();
while doc_matching_deleted_term != TERMINATED {
if doc_opstamps.is_deleted(doc_matching_deleted_term, delete_op.opstamp) {
delete_bitset.insert(doc_matching_deleted_term);
alive_bitset.remove(doc_matching_deleted_term);
might_have_changed = true;
}
doc_matching_deleted_term = docset.advance();
@@ -141,7 +139,7 @@ pub(crate) fn advance_deletes(
return Ok(());
}
if segment_entry.delete_bitset().is_none() && segment_entry.delete_cursor().get().is_none() {
if segment_entry.alive_bitset().is_none() && segment_entry.delete_cursor().get().is_none() {
// There has been no `DeleteOperation` between the segment status and `target_opstamp`.
return Ok(());
}
@@ -149,38 +147,32 @@ pub(crate) fn advance_deletes(
let segment_reader = SegmentReader::open(&segment)?;
let max_doc = segment_reader.max_doc();
let mut delete_bitset: BitSet = match segment_entry.delete_bitset() {
Some(previous_delete_bitset) => (*previous_delete_bitset).clone(),
None => BitSet::with_max_value(max_doc),
let mut alive_bitset: BitSet = match segment_entry.alive_bitset() {
Some(previous_alive_bitset) => (*previous_alive_bitset).clone(),
None => BitSet::with_max_value_and_full(max_doc),
};
let num_deleted_docs_before = segment.meta().num_deleted_docs();
compute_deleted_bitset(
&mut delete_bitset,
&mut alive_bitset,
&segment_reader,
segment_entry.delete_cursor(),
&DocToOpstampMapping::None,
target_opstamp,
)?;
// TODO optimize
// It should be possible to do something smarter by manipulation bitsets directly
// to compute this union.
if let Some(seg_delete_bitset) = segment_reader.delete_bitset() {
for doc in 0u32..max_doc {
if seg_delete_bitset.is_deleted(doc) {
delete_bitset.insert(doc);
}
}
if let Some(seg_alive_bitset) = segment_reader.alive_bitset() {
alive_bitset.intersect_update(seg_alive_bitset.bitset());
}
let num_deleted_docs: u32 = delete_bitset.len() as u32;
let num_alive_docs: u32 = alive_bitset.len() as u32;
let num_deleted_docs = max_doc - num_alive_docs;
if num_deleted_docs > num_deleted_docs_before {
// There are new deletes. We need to write a new delete file.
segment = segment.with_delete_meta(num_deleted_docs as u32, target_opstamp);
let mut delete_file = segment.open_write(SegmentComponent::Delete)?;
write_delete_bitset(&delete_bitset, max_doc, &mut delete_file)?;
write_alive_bitset(&alive_bitset, &mut delete_file)?;
delete_file.terminate()?;
}
@@ -191,10 +183,10 @@ pub(crate) fn advance_deletes(
fn index_documents(
memory_budget: usize,
segment: Segment,
grouped_document_iterator: &mut dyn Iterator<Item = OperationGroup>,
grouped_document_iterator: &mut dyn Iterator<Item = AddBatch>,
segment_updater: &mut SegmentUpdater,
mut delete_cursor: DeleteCursor,
) -> crate::Result<bool> {
) -> crate::Result<()> {
let schema = segment.schema();
let mut segment_writer = SegmentWriter::for_segment(memory_budget, segment.clone(), &schema)?;
@@ -213,7 +205,7 @@ fn index_documents(
}
if !segment_updater.is_alive() {
return Ok(false);
return Ok(());
}
let max_doc = segment_writer.max_doc();
@@ -226,21 +218,20 @@ fn index_documents(
let segment_with_max_doc = segment.with_max_doc(max_doc);
let delete_bitset_opt =
apply_deletes(&segment_with_max_doc, &mut delete_cursor, &doc_opstamps)?;
let alive_bitset_opt = apply_deletes(&segment_with_max_doc, &mut delete_cursor, &doc_opstamps)?;
let meta = segment_with_max_doc.meta().clone();
meta.untrack_temp_docstore();
// update segment_updater inventory to remove tempstore
let segment_entry = SegmentEntry::new(meta, delete_cursor, delete_bitset_opt);
let segment_entry = SegmentEntry::new(meta, delete_cursor, alive_bitset_opt);
block_on(segment_updater.schedule_add_segment(segment_entry))?;
Ok(true)
Ok(())
}
/// `doc_opstamps` is required to be non-empty.
fn apply_deletes(
segment: &Segment,
mut delete_cursor: &mut DeleteCursor,
delete_cursor: &mut DeleteCursor,
doc_opstamps: &[Opstamp],
) -> crate::Result<Option<BitSet>> {
if delete_cursor.get().is_none() {
@@ -259,11 +250,11 @@ fn apply_deletes(
let doc_to_opstamps = DocToOpstampMapping::WithMap(doc_opstamps);
let max_doc = segment.meta().max_doc();
let mut deleted_bitset = BitSet::with_max_value(max_doc);
let mut deleted_bitset = BitSet::with_max_value_and_full(max_doc);
let may_have_deletes = compute_deleted_bitset(
&mut deleted_bitset,
&segment_reader,
&mut delete_cursor,
delete_cursor,
&doc_to_opstamps,
max_doc_opstamp,
)?;
@@ -287,8 +278,7 @@ impl IndexWriter {
/// should work at the same time.
/// # Errors
/// If the lockfile already exists, returns `Error::FileAlreadyExists`.
/// # Panics
/// If the heap size per thread is too small, panics.
/// If the heap size per thread is too small or too big, returns `TantivyError::InvalidArgument`
pub(crate) fn new(
index: &Index,
num_threads: usize,
@@ -306,7 +296,7 @@ impl IndexWriter {
let err_msg = format!("The heap size per thread cannot exceed {}", HEAP_SIZE_MAX);
return Err(TantivyError::InvalidArgument(err_msg));
}
let (document_sender, document_receiver): (OperationSender, OperationReceiver) =
let (document_sender, document_receiver): (AddBatchSender, AddBatchReceiver) =
channel::bounded(PIPELINE_MAX_SIZE_IN_DOCS);
let delete_queue = DeleteQueue::new();
@@ -324,7 +314,7 @@ impl IndexWriter {
heap_size_in_bytes_per_thread,
index: index.clone(),
operation_receiver: document_receiver,
index_writer_status: IndexWriterStatus::from(document_receiver),
operation_sender: document_sender,
segment_updater,
@@ -348,6 +338,11 @@ impl IndexWriter {
self.operation_sender = sender;
}
/// Accessor to the index.
pub fn index(&self) -> &Index {
&self.index
}
/// If there are some merging threads, blocks until they all finish their work and
/// then drop the `IndexWriter`.
pub fn wait_merging_threads(mut self) -> crate::Result<()> {
@@ -359,16 +354,14 @@ impl IndexWriter {
for join_handle in former_workers_handles {
join_handle
.join()
.expect("Indexing Worker thread panicked")
.map_err(|_| {
TantivyError::ErrorInThread("Error in indexing worker thread.".into())
})?;
.map_err(|_| error_in_index_worker_thread("Worker thread panicked."))?
.map_err(|_| error_in_index_worker_thread("Worker thread failed."))?;
}
let result = self
.segment_updater
.wait_merging_thread()
.map_err(|_| TantivyError::ErrorInThread("Failed to join merging thread.".into()));
.map_err(|_| error_in_index_worker_thread("Failed to join merging thread."));
if let Err(ref e) = result {
error!("Some merging thread failed {:?}", e);
@@ -396,10 +389,18 @@ impl IndexWriter {
self.index.new_segment()
}
fn operation_receiver(&self) -> crate::Result<AddBatchReceiver> {
self.index_writer_status
.operation_receiver()
.ok_or_else(|| crate::TantivyError::ErrorInThread("The index writer was killed. It can happen if an indexing worker encounterred an Io error for instance.".to_string()))
}
/// Spawns a new worker thread for indexing.
/// The thread consumes documents from the pipeline.
fn add_indexing_worker(&mut self) -> crate::Result<()> {
let document_receiver_clone = self.operation_receiver.clone();
let document_receiver_clone = self.operation_receiver()?;
let index_writer_bomb = self.index_writer_status.create_bomb();
let mut segment_updater = self.segment_updater.clone();
let mut delete_cursor = self.delete_queue.cursor();
@@ -410,32 +411,31 @@ impl IndexWriter {
.name(format!("thrd-tantivy-index{}", self.worker_id))
.spawn(move || {
loop {
let mut document_iterator =
document_receiver_clone.clone().into_iter().peekable();
let mut document_iterator = document_receiver_clone
.clone()
.into_iter()
.filter(|batch| !batch.is_empty())
.peekable();
// the peeking here is to avoid
// creating a new segment's files
// The peeking here is to avoid creating a new segment's files
// if no document are available.
//
// this is a valid guarantee as the
// peeked document now belongs to
// This is a valid guarantee as the peeked document now belongs to
// our local iterator.
if let Some(operations) = document_iterator.peek() {
if let Some(first) = operations.first() {
delete_cursor.skip_to(first.opstamp);
} else {
return Ok(());
}
if let Some(batch) = document_iterator.peek() {
assert!(!batch.is_empty());
delete_cursor.skip_to(batch[0].opstamp);
} else {
// No more documents.
// Happens when there is a commit, or if the `IndexWriter`
// It happens when there is a commit, or if the `IndexWriter`
// was dropped.
index_writer_bomb.defuse();
return Ok(());
}
let segment = index.new_segment();
index_documents(
mem_budget,
segment,
index.new_segment(),
&mut document_iterator,
&mut segment_updater,
delete_cursor.clone(),
@@ -465,10 +465,8 @@ impl IndexWriter {
}
/// Detects and removes the files that are not used by the index anymore.
pub fn garbage_collect_files(
&self,
) -> impl Future<Output = crate::Result<GarbageCollectionResult>> {
self.segment_updater.schedule_garbage_collect()
pub async fn garbage_collect_files(&self) -> crate::Result<GarbageCollectionResult> {
self.segment_updater.schedule_garbage_collect().await
}
/// Deletes all documents from the index
@@ -491,7 +489,7 @@ impl IndexWriter {
/// let index = Index::create_in_ram(schema.clone());
///
/// let mut index_writer = index.writer_with_num_threads(1, 50_000_000)?;
/// index_writer.add_document(doc!(title => "The modern Promotheus"));
/// index_writer.add_document(doc!(title => "The modern Promotheus"))?;
/// index_writer.commit()?;
///
/// let clear_res = index_writer.delete_all_documents().unwrap();
@@ -535,12 +533,11 @@ impl IndexWriter {
/// when no documents are remaining.
///
/// Returns the former segment_ready channel.
#[allow(unused_must_use)]
fn recreate_document_channel(&mut self) -> OperationReceiver {
let (document_sender, document_receiver): (OperationSender, OperationReceiver) =
fn recreate_document_channel(&mut self) {
let (document_sender, document_receiver): (AddBatchSender, AddBatchReceiver) =
channel::bounded(PIPELINE_MAX_SIZE_IN_DOCS);
mem::replace(&mut self.operation_sender, document_sender);
mem::replace(&mut self.operation_receiver, document_receiver)
self.operation_sender = document_sender;
self.index_writer_status = IndexWriterStatus::from(document_receiver);
}
/// Rollback to the last commit
@@ -556,7 +553,7 @@ impl IndexWriter {
// marks the segment updater as killed. From now on, all
// segment updates will be ignored.
self.segment_updater.kill();
let document_receiver = self.operation_receiver.clone();
let document_receiver = self.operation_receiver();
// take the directory lock to create a new index_writer.
let directory_lock = self
@@ -696,14 +693,10 @@ impl IndexWriter {
/// The opstamp is an increasing `u64` that can
/// be used by the client to align commits with its own
/// document queue.
pub fn add_document(&self, document: Document) -> Opstamp {
pub fn add_document(&self, document: Document) -> crate::Result<Opstamp> {
let opstamp = self.stamper.stamp();
let add_operation = AddOperation { opstamp, document };
let send_result = self.operation_sender.send(smallvec![add_operation]);
if let Err(e) = send_result {
panic!("Failed to index document. Sending to indexing channel failed. This probably means all of the indexing threads have panicked. {:?}", e);
}
opstamp
self.send_add_documents_batch(smallvec![AddOperation { opstamp, document }])?;
Ok(opstamp)
}
/// Gets a range of stamps from the stamper and "pops" the last stamp
@@ -716,11 +709,7 @@ impl IndexWriter {
fn get_batch_opstamps(&self, count: Opstamp) -> (Opstamp, Range<Opstamp>) {
let Range { start, end } = self.stamper.stamps(count + 1u64);
let last_opstamp = end - 1;
let stamps = Range {
start,
end: last_opstamp,
};
(last_opstamp, stamps)
(last_opstamp, start..last_opstamp)
}
/// Runs a group of document operations ensuring that the operations are
@@ -739,16 +728,20 @@ impl IndexWriter {
/// Like adds and deletes (see `IndexWriter.add_document` and
/// `IndexWriter.delete_term`), the changes made by calling `run` will be
/// visible to readers only after calling `commit()`.
pub fn run(&self, user_operations: Vec<UserOperation>) -> Opstamp {
let count = user_operations.len() as u64;
pub fn run<I>(&self, user_operations: I) -> crate::Result<Opstamp>
where
I: IntoIterator<Item = UserOperation>,
I::IntoIter: ExactSizeIterator,
{
let user_operations_it = user_operations.into_iter();
let count = user_operations_it.len() as u64;
if count == 0 {
return self.stamper.stamp();
return Ok(self.stamper.stamp());
}
let (batch_opstamp, stamps) = self.get_batch_opstamps(count);
let mut adds = OperationGroup::default();
for (user_op, opstamp) in user_operations.into_iter().zip(stamps) {
let mut adds = AddBatch::default();
for (user_op, opstamp) in user_operations_it.zip(stamps) {
match user_op {
UserOperation::Delete(term) => {
let delete_operation = DeleteOperation { opstamp, term };
@@ -760,12 +753,16 @@ impl IndexWriter {
}
}
}
let send_result = self.operation_sender.send(adds);
if let Err(e) = send_result {
panic!("Failed to index document. Sending to indexing channel failed. This probably means all of the indexing threads have panicked. {:?}", e);
};
self.send_add_documents_batch(adds)?;
Ok(batch_opstamp)
}
batch_opstamp
fn send_add_documents_batch(&self, add_ops: AddBatch) -> crate::Result<()> {
if self.index_writer_status.is_alive() && self.operation_sender.send(add_ops).is_ok() {
Ok(())
} else {
Err(error_in_index_worker_thread("An index writer was killed."))
}
}
}
@@ -831,7 +828,7 @@ mod tests {
UserOperation::Add(doc!(text_field=>"a")),
UserOperation::Add(doc!(text_field=>"b")),
];
let batch_opstamp1 = index_writer.run(operations);
let batch_opstamp1 = index_writer.run(operations).unwrap();
assert_eq!(batch_opstamp1, 2u64);
}
@@ -842,14 +839,18 @@ mod tests {
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(text_field => "hello1"));
index_writer.add_document(doc!(text_field => "hello2"));
index_writer
.add_document(doc!(text_field => "hello1"))
.unwrap();
index_writer
.add_document(doc!(text_field => "hello2"))
.unwrap();
assert!(index_writer.commit().is_ok());
let reader = index.reader().unwrap();
let searcher = reader.searcher();
assert_eq!(searcher.segment_readers().len(), 1);
assert_eq!(searcher.segment_reader(0u32).num_deleted_docs(), 0);
assert_eq!(searcher.segment_reader(0u32).num_docs(), 2);
index_writer.delete_term(Term::from_field_text(text_field, "hello1"));
assert!(index_writer.commit().is_ok());
@@ -857,7 +858,7 @@ mod tests {
assert!(reader.reload().is_ok());
let searcher = reader.searcher();
assert_eq!(searcher.segment_readers().len(), 1);
assert_eq!(searcher.segment_reader(0u32).num_deleted_docs(), 1);
assert_eq!(searcher.segment_reader(0u32).num_docs(), 1);
let previous_delete_opstamp = index.load_metas().unwrap().segments[0].delete_opstamp();
@@ -869,7 +870,7 @@ mod tests {
assert!(reader.reload().is_ok());
let searcher = reader.searcher();
assert_eq!(searcher.segment_readers().len(), 1);
assert_eq!(searcher.segment_reader(0u32).num_deleted_docs(), 1);
assert_eq!(searcher.segment_reader(0u32).num_docs(), 1);
let after_delete_opstamp = index.load_metas().unwrap().segments[0].delete_opstamp();
assert_eq!(after_delete_opstamp, previous_delete_opstamp);
@@ -900,7 +901,7 @@ mod tests {
UserOperation::Delete(b_term),
];
index_writer.run(operations);
index_writer.run(operations).unwrap();
index_writer.commit().expect("failed to commit");
reader.reload().expect("failed to load searchers");
@@ -930,10 +931,10 @@ mod tests {
let index = Index::create_in_ram(schema_builder.build());
let index_writer = index.writer(3_000_000).unwrap();
let operations1 = vec![];
let batch_opstamp1 = index_writer.run(operations1);
let batch_opstamp1 = index_writer.run(operations1).unwrap();
assert_eq!(batch_opstamp1, 0u64);
let operations2 = vec![];
let batch_opstamp2 = index_writer.run(operations2);
let batch_opstamp2 = index_writer.run(operations2).unwrap();
assert_eq!(batch_opstamp2, 1u64);
}
@@ -993,15 +994,14 @@ mod tests {
}
#[test]
fn test_commit_and_rollback() {
fn test_commit_and_rollback() -> crate::Result<()> {
let mut schema_builder = schema::Schema::builder();
let text_field = schema_builder.add_text_field("text", schema::TEXT);
let index = Index::create_in_ram(schema_builder.build());
let reader = index
.reader_builder()
.reload_policy(ReloadPolicy::Manual)
.try_into()
.unwrap();
.try_into()?;
let num_docs_containing = |s: &str| {
let searcher = reader.searcher();
let term = Term::from_field_text(text_field, s);
@@ -1010,136 +1010,127 @@ mod tests {
{
// writing the segment
let mut index_writer = index.writer(3_000_000).unwrap();
index_writer.add_document(doc!(text_field=>"a"));
index_writer.rollback().unwrap();
let mut index_writer = index.writer(3_000_000)?;
index_writer.add_document(doc!(text_field=>"a"))?;
index_writer.rollback()?;
assert_eq!(index_writer.commit_opstamp(), 0u64);
assert_eq!(num_docs_containing("a"), 0);
{
index_writer.add_document(doc!(text_field=>"b"));
index_writer.add_document(doc!(text_field=>"c"));
}
assert!(index_writer.commit().is_ok());
reader.reload().unwrap();
index_writer.add_document(doc!(text_field=>"b"))?;
index_writer.add_document(doc!(text_field=>"c"))?;
index_writer.commit()?;
reader.reload()?;
assert_eq!(num_docs_containing("a"), 0);
assert_eq!(num_docs_containing("b"), 1);
assert_eq!(num_docs_containing("c"), 1);
}
reader.reload().unwrap();
reader.reload()?;
reader.searcher();
Ok(())
}
#[test]
fn test_with_merges() {
fn test_with_merges() -> crate::Result<()> {
let mut schema_builder = schema::Schema::builder();
let text_field = schema_builder.add_text_field("text", schema::TEXT);
let index = Index::create_in_ram(schema_builder.build());
let reader = index
.reader_builder()
.reload_policy(ReloadPolicy::Manual)
.try_into()
.unwrap();
.try_into()?;
let num_docs_containing = |s: &str| {
let term_a = Term::from_field_text(text_field, s);
reader.searcher().doc_freq(&term_a).unwrap()
};
{
// writing the segment
let mut index_writer = index.writer(12_000_000).unwrap();
// create 8 segments with 100 tiny docs
for _doc in 0..100 {
index_writer.add_document(doc!(text_field=>"a"));
}
index_writer.commit().expect("commit failed");
for _doc in 0..100 {
index_writer.add_document(doc!(text_field=>"a"));
}
// this should create 8 segments and trigger a merge.
index_writer.commit().expect("commit failed");
index_writer
.wait_merging_threads()
.expect("waiting merging thread failed");
reader.reload().unwrap();
assert_eq!(num_docs_containing("a"), 200);
assert!(index.searchable_segments().unwrap().len() < 8);
// writing the segment
let mut index_writer = index.writer(12_000_000).unwrap();
// create 8 segments with 100 tiny docs
for _doc in 0..100 {
index_writer.add_document(doc!(text_field=>"a"))?;
}
index_writer.commit()?;
for _doc in 0..100 {
index_writer.add_document(doc!(text_field=>"a"))?;
}
// this should create 8 segments and trigger a merge.
index_writer.commit()?;
index_writer.wait_merging_threads()?;
reader.reload()?;
assert_eq!(num_docs_containing("a"), 200);
assert!(index.searchable_segments()?.len() < 8);
Ok(())
}
#[test]
fn test_prepare_with_commit_message() {
fn test_prepare_with_commit_message() -> crate::Result<()> {
let mut schema_builder = schema::Schema::builder();
let text_field = schema_builder.add_text_field("text", schema::TEXT);
let index = Index::create_in_ram(schema_builder.build());
// writing the segment
let mut index_writer = index.writer(12_000_000)?;
// create 8 segments with 100 tiny docs
for _doc in 0..100 {
index_writer.add_document(doc!(text_field => "a"))?;
}
{
let mut prepared_commit = index_writer.prepare_commit()?;
prepared_commit.set_payload("first commit");
prepared_commit.commit()?;
}
{
let metas = index.load_metas()?;
assert_eq!(metas.payload.unwrap(), "first commit");
}
for _doc in 0..100 {
index_writer.add_document(doc!(text_field => "a"))?;
}
index_writer.commit()?;
{
let metas = index.load_metas()?;
assert!(metas.payload.is_none());
}
Ok(())
}
#[test]
fn test_prepare_but_rollback() -> crate::Result<()> {
let mut schema_builder = schema::Schema::builder();
let text_field = schema_builder.add_text_field("text", schema::TEXT);
let index = Index::create_in_ram(schema_builder.build());
{
// writing the segment
let mut index_writer = index.writer(12_000_000).unwrap();
let mut index_writer = index.writer_with_num_threads(4, 12_000_000)?;
// create 8 segments with 100 tiny docs
for _doc in 0..100 {
index_writer.add_document(doc!(text_field => "a"));
index_writer.add_document(doc!(text_field => "a"))?;
}
{
let mut prepared_commit = index_writer.prepare_commit().expect("commit failed");
let mut prepared_commit = index_writer.prepare_commit()?;
prepared_commit.set_payload("first commit");
prepared_commit.commit().expect("commit failed");
prepared_commit.abort()?;
}
{
let metas = index.load_metas().unwrap();
assert_eq!(metas.payload.unwrap(), "first commit");
}
for _doc in 0..100 {
index_writer.add_document(doc!(text_field => "a"));
}
index_writer.commit().unwrap();
{
let metas = index.load_metas().unwrap();
assert!(metas.payload.is_none());
}
}
}
#[test]
fn test_prepare_but_rollback() {
let mut schema_builder = schema::Schema::builder();
let text_field = schema_builder.add_text_field("text", schema::TEXT);
let index = Index::create_in_ram(schema_builder.build());
{
// writing the segment
let mut index_writer = index.writer_with_num_threads(4, 12_000_000).unwrap();
// create 8 segments with 100 tiny docs
for _doc in 0..100 {
index_writer.add_document(doc!(text_field => "a"));
}
{
let mut prepared_commit = index_writer.prepare_commit().expect("commit failed");
prepared_commit.set_payload("first commit");
prepared_commit.abort().expect("commit failed");
}
{
let metas = index.load_metas().unwrap();
let metas = index.load_metas()?;
assert!(metas.payload.is_none());
}
for _doc in 0..100 {
index_writer.add_document(doc!(text_field => "b"));
index_writer.add_document(doc!(text_field => "b"))?;
}
index_writer.commit().unwrap();
index_writer.commit()?;
}
let num_docs_containing = |s: &str| {
let term_a = Term::from_field_text(text_field, s);
index
.reader_builder()
.reload_policy(ReloadPolicy::Manual)
.try_into()
.unwrap()
.try_into()?
.searcher()
.doc_freq(&term_a)
.unwrap()
};
assert_eq!(num_docs_containing("a"), 0);
assert_eq!(num_docs_containing("b"), 100);
assert_eq!(num_docs_containing("a")?, 0);
assert_eq!(num_docs_containing("b")?, 100);
Ok(())
}
#[test]
@@ -1160,7 +1151,7 @@ mod tests {
};
let mut index_writer = index.writer_with_num_threads(4, 12_000_000).unwrap();
let add_tstamp = index_writer.add_document(doc!(text_field => "a"));
let add_tstamp = index_writer.add_document(doc!(text_field => "a")).unwrap();
let commit_tstamp = index_writer.commit().unwrap();
assert!(commit_tstamp > add_tstamp);
index_writer.delete_all_documents().unwrap();
@@ -1177,7 +1168,7 @@ mod tests {
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_with_num_threads(4, 12_000_000).unwrap();
let add_tstamp = index_writer.add_document(doc!(text_field => "a"));
let add_tstamp = index_writer.add_document(doc!(text_field => "a")).unwrap();
// commit documents - they are now available
let first_commit = index_writer.commit();
@@ -1196,7 +1187,7 @@ mod tests {
// add new documents again
for _ in 0..100 {
index_writer.add_document(doc!(text_field => "b"));
index_writer.add_document(doc!(text_field => "b")).unwrap();
}
// rollback to last commit, when index was empty
@@ -1230,7 +1221,7 @@ mod tests {
assert!(index_writer.commit().is_ok());
// add one simple doc
index_writer.add_document(doc!(text_field => "a"));
index_writer.add_document(doc!(text_field => "a")).unwrap();
assert!(index_writer.commit().is_ok());
let term_a = Term::from_field_text(text_field, "a");
@@ -1254,7 +1245,7 @@ mod tests {
let mut index_writer = index.writer_with_num_threads(4, 12_000_000).unwrap();
// add one simple doc
index_writer.add_document(doc!(text_field => "a"));
assert!(index_writer.add_document(doc!(text_field => "a")).is_ok());
let comm = index_writer.commit();
assert!(comm.is_ok());
let commit_tstamp = comm.unwrap();
@@ -1330,13 +1321,13 @@ mod tests {
// create and delete docs in same commit
for id in 0u64..5u64 {
index_writer.add_document(doc!(id_field => id));
index_writer.add_document(doc!(id_field => id))?;
}
for id in 2u64..4u64 {
index_writer.delete_term(Term::from_field_u64(id_field, id));
}
for id in 5u64..10u64 {
index_writer.add_document(doc!(id_field => id));
index_writer.add_document(doc!(id_field => id))?;
}
index_writer.commit()?;
index_reader.reload()?;
@@ -1364,15 +1355,24 @@ mod tests {
Merge,
}
fn operation_strategy() -> impl Strategy<Value = IndexingOp> {
fn balanced_operation_strategy() -> impl Strategy<Value = IndexingOp> {
prop_oneof![
(0u64..10u64).prop_map(|id| IndexingOp::DeleteDoc { id }),
(0u64..10u64).prop_map(|id| IndexingOp::AddDoc { id }),
(0u64..2u64).prop_map(|_| IndexingOp::Commit),
(0u64..20u64).prop_map(|id| IndexingOp::DeleteDoc { id }),
(0u64..20u64).prop_map(|id| IndexingOp::AddDoc { id }),
(0u64..1u64).prop_map(|_| IndexingOp::Commit),
(0u64..1u64).prop_map(|_| IndexingOp::Merge),
]
}
fn adding_operation_strategy() -> impl Strategy<Value = IndexingOp> {
prop_oneof![
10 => (0u64..100u64).prop_map(|id| IndexingOp::DeleteDoc { id }),
50 => (0u64..100u64).prop_map(|id| IndexingOp::AddDoc { id }),
2 => (0u64..1u64).prop_map(|_| IndexingOp::Commit),
1 => (0u64..1u64).prop_map(|_| IndexingOp::Merge),
]
}
fn expected_ids(ops: &[IndexingOp]) -> (HashMap<u64, u64>, HashSet<u64>) {
let mut existing_ids = HashMap::new();
let mut deleted_ids = HashSet::new();
@@ -1439,12 +1439,14 @@ mod tests {
let mut index_writer = index.writer_for_tests()?;
index_writer.set_merge_policy(Box::new(NoMergePolicy));
let old_reader = index.reader()?;
for &op in ops {
match op {
IndexingOp::AddDoc { id } => {
let facet = Facet::from(&("/cola/".to_string() + &id.to_string()));
index_writer
.add_document(doc!(id_field=>id, multi_numbers=> id, multi_numbers => id, text_field => id.to_string(), facet_field => facet, large_text_field=> LOREM));
.add_document(doc!(id_field=>id, multi_numbers=> id, multi_numbers => id, text_field => id.to_string(), facet_field => facet, large_text_field=> LOREM))?;
}
IndexingOp::DeleteDoc { id } => {
index_writer.delete_term(Term::from_field_u64(id_field, id));
@@ -1477,6 +1479,21 @@ mod tests {
assert!(index_writer.wait_merging_threads().is_ok());
}
}
old_reader.reload()?;
let old_searcher = old_reader.searcher();
let ids_old_searcher: HashSet<u64> = old_searcher
.segment_readers()
.iter()
.flat_map(|segment_reader| {
let ff_reader = segment_reader.fast_fields().u64(id_field).unwrap();
segment_reader
.doc_ids_alive()
.map(move |doc| ff_reader.get(doc))
})
.collect();
let ids: HashSet<u64> = searcher
.segment_readers()
.iter()
@@ -1489,6 +1506,19 @@ mod tests {
.collect();
let (expected_ids_and_num_occurences, deleted_ids) = expected_ids(ops);
let num_docs_expected = expected_ids_and_num_occurences
.iter()
.map(|(_, id_occurences)| *id_occurences as usize)
.sum::<usize>();
assert_eq!(searcher.num_docs() as usize, num_docs_expected);
assert_eq!(old_searcher.num_docs() as usize, num_docs_expected);
assert_eq!(
ids_old_searcher,
expected_ids_and_num_occurences
.keys()
.cloned()
.collect::<HashSet<_>>()
);
assert_eq!(
ids,
expected_ids_and_num_occurences
@@ -1513,7 +1543,7 @@ mod tests {
for segment_reader in searcher.segment_readers().iter() {
let store_reader = segment_reader.get_store_reader().unwrap();
// test store iterator
for doc in store_reader.iter(segment_reader.delete_bitset()) {
for doc in store_reader.iter(segment_reader.alive_bitset()) {
let id = doc
.unwrap()
.get_first(id_field)
@@ -1583,22 +1613,42 @@ mod tests {
}
proptest! {
#![proptest_config(ProptestConfig::with_cases(20))]
#[test]
fn test_delete_with_sort_proptest(ops in proptest::collection::vec(operation_strategy(), 1..10)) {
fn test_delete_with_sort_proptest_adding(ops in proptest::collection::vec(adding_operation_strategy(), 1..100)) {
assert!(test_operation_strategy(&ops[..], true, false).is_ok());
}
#[test]
fn test_delete_without_sort_proptest(ops in proptest::collection::vec(operation_strategy(), 1..10)) {
fn test_delete_without_sort_proptest_adding(ops in proptest::collection::vec(adding_operation_strategy(), 1..100)) {
assert!(test_operation_strategy(&ops[..], false, false).is_ok());
}
#[test]
fn test_delete_with_sort_proptest_with_merge(ops in proptest::collection::vec(operation_strategy(), 1..10)) {
fn test_delete_with_sort_proptest_with_merge_adding(ops in proptest::collection::vec(adding_operation_strategy(), 1..100)) {
assert!(test_operation_strategy(&ops[..], true, true).is_ok());
}
#[test]
fn test_delete_without_sort_proptest_with_merge(ops in proptest::collection::vec(operation_strategy(), 1..10)) {
fn test_delete_without_sort_proptest_with_merge_adding(ops in proptest::collection::vec(adding_operation_strategy(), 1..100)) {
assert!(test_operation_strategy(&ops[..], false, true).is_ok());
}
#[test]
fn test_delete_with_sort_proptest(ops in proptest::collection::vec(balanced_operation_strategy(), 1..10)) {
assert!(test_operation_strategy(&ops[..], true, false).is_ok());
}
#[test]
fn test_delete_without_sort_proptest(ops in proptest::collection::vec(balanced_operation_strategy(), 1..10)) {
assert!(test_operation_strategy(&ops[..], false, false).is_ok());
}
#[test]
fn test_delete_with_sort_proptest_with_merge(ops in proptest::collection::vec(balanced_operation_strategy(), 1..10)) {
assert!(test_operation_strategy(&ops[..], true, true).is_ok());
}
#[test]
fn test_delete_without_sort_proptest_with_merge(ops in proptest::collection::vec(balanced_operation_strategy(), 1..100)) {
assert!(test_operation_strategy(&ops[..], false, true).is_ok());
}
}
#[test]
@@ -1623,11 +1673,11 @@ mod tests {
let mut index_writer = index.writer_for_tests()?;
// We add a doc...
index_writer.add_document(doc!(sort_by_field => 2u64, id_field => 0u64));
index_writer.add_document(doc!(sort_by_field => 2u64, id_field => 0u64))?;
// And remove it.
index_writer.delete_term(Term::from_field_u64(id_field, 0u64));
// We add another doc.
index_writer.add_document(doc!(sort_by_field=>1u64, id_field => 0u64));
index_writer.add_document(doc!(sort_by_field=>1u64, id_field => 0u64))?;
// The expected result is a segment with
// maxdoc = 2
@@ -1639,19 +1689,19 @@ mod tests {
let segment_reader = searcher.segment_reader(0);
assert_eq!(segment_reader.max_doc(), 2);
assert_eq!(segment_reader.num_deleted_docs(), 1);
assert_eq!(segment_reader.num_docs(), 1);
Ok(())
}
#[test]
fn test_index_doc_missing_field() {
fn test_index_doc_missing_field() -> crate::Result<()> {
let mut schema_builder = schema::Schema::builder();
let idfield = schema_builder.add_text_field("id", STRING);
schema_builder.add_text_field("optfield", STRING);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(idfield=>"myid"));
let commit = index_writer.commit();
assert!(commit.is_ok());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(idfield=>"myid"))?;
index_writer.commit()?;
Ok(())
}
}

118
src/indexer/index_writer_status.rs Normal file
View File

@@ -0,0 +1,118 @@
use std::sync::atomic::{AtomicBool, Ordering};
use std::sync::{Arc, RwLock};
use super::AddBatchReceiver;
#[derive(Clone)]
pub(crate) struct IndexWriterStatus {
inner: Arc<Inner>,
}
impl IndexWriterStatus {
/// Returns true iff the index writer is alive.
pub fn is_alive(&self) -> bool {
self.inner.as_ref().is_alive()
}
/// Returns a copy of the operation receiver.
/// If the index writer was killed, returns None.
pub fn operation_receiver(&self) -> Option<AddBatchReceiver> {
let rlock = self
.inner
.receive_channel
.read()
.expect("This lock should never be poisoned");
rlock.as_ref().map(|receiver| receiver.clone())
}
/// Create an index writer bomb.
/// If dropped, the index writer status will be killed.
pub(crate) fn create_bomb(&self) -> IndexWriterBomb {
IndexWriterBomb {
inner: Some(self.inner.clone()),
}
}
}
struct Inner {
is_alive: AtomicBool,
receive_channel: RwLock<Option<AddBatchReceiver>>,
}
impl Inner {
fn is_alive(&self) -> bool {
self.is_alive.load(Ordering::Relaxed)
}
fn kill(&self) {
self.is_alive.store(false, Ordering::Relaxed);
self.receive_channel
.write()
.expect("This lock should never be poisoned")
.take();
}
}
impl From<AddBatchReceiver> for IndexWriterStatus {
fn from(receiver: AddBatchReceiver) -> Self {
IndexWriterStatus {
inner: Arc::new(Inner {
is_alive: AtomicBool::new(true),
receive_channel: RwLock::new(Some(receiver)),
}),
}
}
}
/// If dropped, the index writer will be killed.
/// To prevent this, clients can call `.defuse()`.
pub(crate) struct IndexWriterBomb {
inner: Option<Arc<Inner>>,
}
impl IndexWriterBomb {
/// Defuses the bomb.
///
/// This is the only way to drop the bomb without killing
/// the index writer.
pub fn defuse(mut self) {
self.inner = None;
}
}
impl Drop for IndexWriterBomb {
fn drop(&mut self) {
if let Some(inner) = self.inner.take() {
inner.kill();
}
}
}
#[cfg(test)]
mod tests {
use super::IndexWriterStatus;
use crossbeam::channel;
use std::mem;
#[test]
fn test_bomb_goes_boom() {
let (_tx, rx) = channel::bounded(10);
let index_writer_status: IndexWriterStatus = IndexWriterStatus::from(rx);
assert!(index_writer_status.operation_receiver().is_some());
let bomb = index_writer_status.create_bomb();
assert!(index_writer_status.operation_receiver().is_some());
mem::drop(bomb);
// boom!
assert!(index_writer_status.operation_receiver().is_none());
}
#[test]
fn test_bomb_defused() {
let (_tx, rx) = channel::bounded(10);
let index_writer_status: IndexWriterStatus = IndexWriterStatus::from(rx);
assert!(index_writer_status.operation_receiver().is_some());
let bomb = index_writer_status.create_bomb();
bomb.defuse();
assert!(index_writer_status.operation_receiver().is_some());
}
}

37
src/indexer/log_merge_policy.rs
View File

@@ -114,7 +114,7 @@ mod tests {
use crate::Index;
#[test]
fn create_index_test_max_merge_issue_1035() {
fn create_index_test_max_merge_issue_1035() -> crate::Result<()> {
let mut schema_builder = schema::Schema::builder();
let int_field = schema_builder.add_u64_field("intval", INDEXED);
let schema = schema_builder.build();
@@ -127,34 +127,34 @@ mod tests {
log_merge_policy.set_max_docs_before_merge(1);
log_merge_policy.set_min_layer_size(0);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
index_writer.set_merge_policy(Box::new(log_merge_policy));
// after every commit the merge checker is started, it will merge only segments with 1
// element in it because of the max_merge_size.
index_writer.add_document(doc!(int_field=>1_u64));
assert!(index_writer.commit().is_ok());
index_writer.add_document(doc!(int_field=>1_u64))?;
index_writer.commit()?;
index_writer.add_document(doc!(int_field=>2_u64));
assert!(index_writer.commit().is_ok());
index_writer.add_document(doc!(int_field=>2_u64))?;
index_writer.commit()?;
index_writer.add_document(doc!(int_field=>3_u64));
assert!(index_writer.commit().is_ok());
index_writer.add_document(doc!(int_field=>3_u64))?;
index_writer.commit()?;
index_writer.add_document(doc!(int_field=>4_u64));
assert!(index_writer.commit().is_ok());
index_writer.add_document(doc!(int_field=>4_u64))?;
index_writer.commit()?;
index_writer.add_document(doc!(int_field=>5_u64));
assert!(index_writer.commit().is_ok());
index_writer.add_document(doc!(int_field=>5_u64))?;
index_writer.commit()?;
index_writer.add_document(doc!(int_field=>6_u64));
assert!(index_writer.commit().is_ok());
index_writer.add_document(doc!(int_field=>6_u64))?;
index_writer.commit()?;
index_writer.add_document(doc!(int_field=>7_u64));
assert!(index_writer.commit().is_ok());
index_writer.add_document(doc!(int_field=>7_u64))?;
index_writer.commit()?;
index_writer.add_document(doc!(int_field=>8_u64));
assert!(index_writer.commit().is_ok());
index_writer.add_document(doc!(int_field=>8_u64))?;
index_writer.commit()?;
}
let _segment_ids = index
@@ -169,6 +169,7 @@ mod tests {
panic!("segment can't have more than two segments");
} // don't know how to wait for the merge, then it could be a simple eq
}
Ok(())
}
fn test_merge_policy() -> LogMergePolicy {

467
src/indexer/merger.rs
View File

@@ -1,6 +1,6 @@
use crate::error::DataCorruption;
use crate::fastfield::AliveBitSet;
use crate::fastfield::CompositeFastFieldSerializer;
use crate::fastfield::DeleteBitSet;
use crate::fastfield::DynamicFastFieldReader;
use crate::fastfield::FastFieldDataAccess;
use crate::fastfield::FastFieldReader;
@@ -10,7 +10,7 @@ use crate::fastfield::MultiValuedFastFieldReader;
use crate::fieldnorm::FieldNormsSerializer;
use crate::fieldnorm::FieldNormsWriter;
use crate::fieldnorm::{FieldNormReader, FieldNormReaders};
use crate::indexer::doc_id_mapping::SegmentDocidMapping;
use crate::indexer::doc_id_mapping::SegmentDocIdMapping;
use crate::indexer::SegmentSerializer;
use crate::postings::Postings;
use crate::postings::{InvertedIndexSerializer, SegmentPostings};
@@ -29,7 +29,6 @@ use crate::{
SegmentOrdinal,
};
use crate::{DocId, InvertedIndexReader, SegmentComponent};
use common::HasLen;
use itertools::Itertools;
use measure_time::debug_time;
use std::cmp;
@@ -69,58 +68,33 @@ fn compute_total_num_tokens(readers: &[SegmentReader], field: Field) -> crate::R
.sum::<u64>())
}
/// `ReaderWithOrdinal` is used to be able to easier associate
/// data with a `SegmentReader`. The ordinal is supposed to be
/// used as an index access.
///
/// The ordinal identifies the position within `Merger` readers.
#[derive(Clone, Copy)]
pub(crate) struct SegmentReaderWithOrdinal<'a> {
pub reader: &'a SegmentReader,
pub ordinal: SegmentOrdinal,
}
impl<'a> From<(usize, &'a SegmentReader)> for SegmentReaderWithOrdinal<'a> {
fn from(data: (usize, &'a SegmentReader)) -> Self {
SegmentReaderWithOrdinal {
reader: data.1,
ordinal: data.0 as u32,
}
}
}
pub struct IndexMerger {
index_settings: IndexSettings,
schema: Schema,
readers: Vec<SegmentReader>,
pub(crate) readers: Vec<SegmentReader>,
max_doc: u32,
}
fn compute_min_max_val(
u64_reader: &impl FastFieldReader<u64>,
max_doc: DocId,
delete_bitset_opt: Option<&DeleteBitSet>,
segment_reader: &SegmentReader,
) -> Option<(u64, u64)> {
if max_doc == 0 {
None
} else {
match delete_bitset_opt {
Some(delete_bitset) => {
// some deleted documents,
// we need to recompute the max / min
minmax(
(0..max_doc)
.filter(|doc_id| delete_bitset.is_alive(*doc_id))
.map(|doc_id| u64_reader.get(doc_id)),
)
}
None => {
// no deleted documents,
// we can use the previous min_val, max_val.
Some((u64_reader.min_value(), u64_reader.max_value()))
}
}
if segment_reader.max_doc() == 0 {
return None;
}
if segment_reader.alive_bitset().is_none() {
// no deleted documents,
// we can use the previous min_val, max_val.
return Some((u64_reader.min_value(), u64_reader.max_value()));
}
// some deleted documents,
// we need to recompute the max / min
minmax(
segment_reader
.doc_ids_alive()
.map(|doc_id| u64_reader.get(doc_id)),
)
}
struct TermOrdinalMapping {
@@ -183,16 +157,38 @@ impl IndexMerger {
schema: Schema,
index_settings: IndexSettings,
segments: &[Segment],
) -> crate::Result<IndexMerger> {
let delete_bitsets = segments.iter().map(|_| None).collect_vec();
Self::open_with_custom_alive_set(schema, index_settings, segments, delete_bitsets)
}
// Create merge with a custom delete set.
// For every Segment, a delete bitset can be provided, which
// will be merged with the existing bit set. Make sure the index
// corresponds to the segment index.
//
// If `None` is provided for custom alive set, the regular alive set will be used.
// If a delete_bitsets is provided, the union between the provided and regular
// alive set will be used.
//
// This can be used to merge but also apply an additional filter.
// One use case is demux, which is basically taking a list of
// segments and partitions them e.g. by a value in a field.
pub fn open_with_custom_alive_set(
schema: Schema,
index_settings: IndexSettings,
segments: &[Segment],
alive_bitset_opt: Vec<Option<AliveBitSet>>,
) -> crate::Result<IndexMerger> {
let mut readers = vec![];
let mut max_doc: u32 = 0u32;
for segment in segments {
for (segment, new_alive_bitset_opt) in segments.iter().zip(alive_bitset_opt.into_iter()) {
if segment.meta().num_docs() > 0 {
let reader = SegmentReader::open(segment)?;
max_doc += reader.num_docs();
let reader =
SegmentReader::open_with_custom_alive_set(segment, new_alive_bitset_opt)?;
readers.push(reader);
}
}
let max_doc = readers.iter().map(|reader| reader.num_docs()).sum();
if let Some(sort_by_field) = index_settings.sort_by_field.as_ref() {
readers = Self::sort_readers_by_min_sort_field(readers, sort_by_field)?;
}
@@ -240,7 +236,7 @@ impl IndexMerger {
fn write_fieldnorms(
&self,
mut fieldnorms_serializer: FieldNormsSerializer,
doc_id_mapping: &SegmentDocidMapping,
doc_id_mapping: &SegmentDocIdMapping,
) -> crate::Result<()> {
let fields = FieldNormsWriter::fields_with_fieldnorm(&self.schema);
let mut fieldnorms_data = Vec::with_capacity(self.max_doc as usize);
@@ -252,8 +248,8 @@ impl IndexMerger {
.iter()
.map(|reader| reader.get_fieldnorms_reader(field))
.collect::<Result<_, _>>()?;
for (doc_id, reader_with_ordinal) in doc_id_mapping.iter() {
let fieldnorms_reader = &fieldnorms_readers[reader_with_ordinal.ordinal as usize];
for (doc_id, reader_ordinal) in doc_id_mapping.iter() {
let fieldnorms_reader = &fieldnorms_readers[*reader_ordinal as usize];
let fieldnorm_id = fieldnorms_reader.fieldnorm_id(*doc_id);
fieldnorms_data.push(fieldnorm_id);
}
@@ -268,7 +264,7 @@ impl IndexMerger {
&self,
fast_field_serializer: &mut CompositeFastFieldSerializer,
mut term_ord_mappings: HashMap<Field, TermOrdinalMapping>,
doc_id_mapping: &SegmentDocidMapping,
doc_id_mapping: &SegmentDocIdMapping,
) -> crate::Result<()> {
debug_time!("write_fast_fields");
@@ -319,14 +315,14 @@ impl IndexMerger {
&self,
field: Field,
fast_field_serializer: &mut CompositeFastFieldSerializer,
doc_id_mapping: &SegmentDocidMapping,
doc_id_mapping: &SegmentDocIdMapping,
) -> crate::Result<()> {
let (min_value, max_value) = self.readers.iter().map(|reader|{
let u64_reader: DynamicFastFieldReader<u64> = reader
.fast_fields()
.typed_fast_field_reader(field)
.expect("Failed to find a reader for single fast field. This is a tantivy bug and it should never happen.");
compute_min_max_val(&u64_reader, reader.max_doc(), reader.delete_bitset())
compute_min_max_val(&u64_reader, reader)
})
.flatten()
.reduce(|a, b| {
@@ -351,26 +347,26 @@ impl IndexMerger {
num_vals: doc_id_mapping.len() as u64,
};
#[derive(Clone)]
struct SortedDocidFieldAccessProvider<'a> {
doc_id_mapping: &'a SegmentDocidMapping<'a>,
struct SortedDocIdFieldAccessProvider<'a> {
doc_id_mapping: &'a SegmentDocIdMapping,
fast_field_readers: &'a Vec<DynamicFastFieldReader<u64>>,
}
impl<'a> FastFieldDataAccess for SortedDocidFieldAccessProvider<'a> {
impl<'a> FastFieldDataAccess for SortedDocIdFieldAccessProvider<'a> {
fn get_val(&self, doc: u64) -> u64 {
let (doc_id, reader_with_ordinal) = self.doc_id_mapping[doc as usize];
self.fast_field_readers[reader_with_ordinal.ordinal as usize].get(doc_id)
let (doc_id, reader_ordinal) = self.doc_id_mapping[doc as usize];
self.fast_field_readers[reader_ordinal as usize].get(doc_id)
}
}
let fastfield_accessor = SortedDocidFieldAccessProvider {
let fastfield_accessor = SortedDocIdFieldAccessProvider {
doc_id_mapping,
fast_field_readers: &fast_field_readers,
};
let iter1 = doc_id_mapping.iter().map(|(doc_id, reader_with_ordinal)| {
let fast_field_reader = &fast_field_readers[reader_with_ordinal.ordinal as usize];
let iter1 = doc_id_mapping.iter().map(|(doc_id, reader_ordinal)| {
let fast_field_reader = &fast_field_readers[*reader_ordinal as usize];
fast_field_reader.get(*doc_id)
});
let iter2 = doc_id_mapping.iter().map(|(doc_id, reader_with_ordinal)| {
let fast_field_reader = &fast_field_readers[reader_with_ordinal.ordinal as usize];
let iter2 = doc_id_mapping.iter().map(|(doc_id, reader_ordinal)| {
let fast_field_reader = &fast_field_readers[*reader_ordinal as usize];
fast_field_reader.get(*doc_id)
});
fast_field_serializer.create_auto_detect_u64_fast_field(
@@ -390,9 +386,10 @@ impl IndexMerger {
&self,
sort_by_field: &IndexSortByField,
) -> crate::Result<bool> {
let reader_and_field_accessors = self.get_reader_with_sort_field_accessor(sort_by_field)?;
let reader_ordinal_and_field_accessors =
self.get_reader_with_sort_field_accessor(sort_by_field)?;
let everything_is_in_order = reader_and_field_accessors
let everything_is_in_order = reader_ordinal_and_field_accessors
.into_iter()
.map(|reader| reader.1)
.tuple_windows()
@@ -415,27 +412,24 @@ impl IndexMerger {
Ok(value_accessor)
}
/// Collecting value_accessors into a vec to bind the lifetime.
pub(crate) fn get_reader_with_sort_field_accessor<'a, 'b>(
&'a self,
sort_by_field: &'b IndexSortByField,
) -> crate::Result<
Vec<(
SegmentReaderWithOrdinal<'a>,
impl FastFieldReader<u64> + Clone,
)>,
> {
let reader_and_field_accessors = self
pub(crate) fn get_reader_with_sort_field_accessor(
&self,
sort_by_field: &IndexSortByField,
) -> crate::Result<Vec<(SegmentOrdinal, impl FastFieldReader<u64> + Clone)>> {
let reader_ordinal_and_field_accessors = self
.readers
.iter()
.enumerate()
.map(Into::into)
.map(|reader_with_ordinal: SegmentReaderWithOrdinal| {
let value_accessor =
Self::get_sort_field_accessor(reader_with_ordinal.reader, sort_by_field)?;
Ok((reader_with_ordinal, value_accessor))
.map(|(reader_ordinal, _)| reader_ordinal as SegmentOrdinal)
.map(|reader_ordinal: SegmentOrdinal| {
let value_accessor = Self::get_sort_field_accessor(
&self.readers[reader_ordinal as usize],
sort_by_field,
)?;
Ok((reader_ordinal, value_accessor))
})
.collect::<crate::Result<Vec<_>>>()?;
Ok(reader_and_field_accessors)
Ok(reader_ordinal_and_field_accessors)
}
/// Generates the doc_id mapping where position in the vec=new
@@ -445,55 +439,59 @@ impl IndexMerger {
pub(crate) fn generate_doc_id_mapping(
&self,
sort_by_field: &IndexSortByField,
) -> crate::Result<SegmentDocidMapping> {
let reader_and_field_accessors = self.get_reader_with_sort_field_accessor(sort_by_field)?;
) -> crate::Result<SegmentDocIdMapping> {
let reader_ordinal_and_field_accessors =
self.get_reader_with_sort_field_accessor(sort_by_field)?;
// Loading the field accessor on demand causes a 15x regression
// create iterators over segment/sort_accessor/doc_id tuple
let doc_id_reader_pair =
reader_and_field_accessors
reader_ordinal_and_field_accessors
.iter()
.map(|reader_and_field_accessor| {
reader_and_field_accessor
.0
.reader
.doc_ids_alive()
.map(move |doc_id| {
(
doc_id,
reader_and_field_accessor.0,
&reader_and_field_accessor.1,
)
})
let reader = &self.readers[reader_and_field_accessor.0 as usize];
reader.doc_ids_alive().map(move |doc_id| {
(
doc_id,
reader_and_field_accessor.0,
&reader_and_field_accessor.1,
)
})
});
let total_num_new_docs = self
.readers
.iter()
.map(|reader| reader.num_docs() as usize)
.sum();
let mut sorted_doc_ids = Vec::with_capacity(total_num_new_docs);
// create iterator tuple of (old doc_id, reader) in order of the new doc_ids
let sorted_doc_ids: Vec<(DocId, SegmentReaderWithOrdinal)> = doc_id_reader_pair
.into_iter()
.kmerge_by(|a, b| {
let val1 = a.2.get(a.0);
let val2 = b.2.get(b.0);
if sort_by_field.order == Order::Asc {
val1 < val2
} else {
val1 > val2
}
})
.map(|(doc_id, reader_with_id, _)| (doc_id, reader_with_id))
.collect::<Vec<_>>();
Ok(SegmentDocidMapping::new(sorted_doc_ids, false))
sorted_doc_ids.extend(
doc_id_reader_pair
.into_iter()
.kmerge_by(|a, b| {
let val1 = a.2.get(a.0);
let val2 = b.2.get(b.0);
if sort_by_field.order == Order::Asc {
val1 < val2
} else {
val1 > val2
}
})
.map(|(doc_id, reader_with_id, _)| (doc_id, reader_with_id)),
);
Ok(SegmentDocIdMapping::new(sorted_doc_ids, false))
}
// Creating the index file to point into the data, generic over `BytesFastFieldReader` and
// `MultiValuedFastFieldReader`
//
// Important: reader_and_field_accessor needs
// to have the same order as self.readers since ReaderWithOrdinal
// is used to index the reader_and_field_accessors vec.
fn write_1_n_fast_field_idx_generic<T: MultiValueLength>(
field: Field,
fast_field_serializer: &mut CompositeFastFieldSerializer,
doc_id_mapping: &SegmentDocidMapping,
doc_id_mapping: &SegmentDocIdMapping,
reader_and_field_accessors: &[(&SegmentReader, T)],
) -> crate::Result<Vec<u64>> {
let mut total_num_vals = 0u64;
@@ -503,13 +501,11 @@ impl IndexMerger {
// what should be the bit length use for bitpacking.
let mut num_docs = 0;
for (reader, u64s_reader) in reader_and_field_accessors.iter() {
if let Some(delete_bitset) = reader.delete_bitset() {
num_docs += reader.max_doc() as u64 - delete_bitset.len() as u64;
for doc in 0u32..reader.max_doc() {
if delete_bitset.is_alive(doc) {
let num_vals = u64s_reader.get_len(doc) as u64;
total_num_vals += num_vals;
}
if let Some(alive_bitset) = reader.alive_bitset() {
num_docs += alive_bitset.num_alive_docs() as u64;
for doc in reader.doc_ids_alive() {
let num_vals = u64s_reader.get_len(doc) as u64;
total_num_vals += num_vals;
}
} else {
num_docs += reader.max_doc() as u64;
@@ -531,10 +527,10 @@ impl IndexMerger {
// acccess on the fly or 2. change the codec api to make random access optional, but
// they both have also major drawbacks.
let mut offsets = vec![];
let mut offsets = Vec::with_capacity(doc_id_mapping.len());
let mut offset = 0;
for (doc_id, reader) in doc_id_mapping.iter() {
let reader = &reader_and_field_accessors[reader.ordinal as usize].1;
let reader = &reader_and_field_accessors[*reader as usize].1;
offsets.push(offset);
offset += reader.get_len(*doc_id) as u64;
}
@@ -554,9 +550,9 @@ impl IndexMerger {
&self,
field: Field,
fast_field_serializer: &mut CompositeFastFieldSerializer,
doc_id_mapping: &SegmentDocidMapping,
doc_id_mapping: &SegmentDocIdMapping,
) -> crate::Result<Vec<u64>> {
let reader_and_field_accessors = self.readers.iter().map(|reader|{
let reader_ordinal_and_field_accessors = self.readers.iter().map(|reader|{
let u64s_reader: MultiValuedFastFieldReader<u64> = reader.fast_fields()
.typed_fast_field_multi_reader(field)
.expect("Failed to find index for multivalued field. This is a bug in tantivy, please report.");
@@ -567,7 +563,7 @@ impl IndexMerger {
field,
fast_field_serializer,
doc_id_mapping,
&reader_and_field_accessors,
&reader_ordinal_and_field_accessors,
)
}
@@ -576,7 +572,7 @@ impl IndexMerger {
field: Field,
term_ordinal_mappings: &TermOrdinalMapping,
fast_field_serializer: &mut CompositeFastFieldSerializer,
doc_id_mapping: &SegmentDocidMapping,
doc_id_mapping: &SegmentDocIdMapping,
) -> crate::Result<()> {
debug_time!("write_hierarchical_facet_field");
@@ -606,11 +602,11 @@ impl IndexMerger {
fast_field_serializer.new_u64_fast_field_with_idx(field, 0u64, max_term_ord, 1)?;
let mut vals = Vec::with_capacity(100);
for (old_doc_id, reader_with_ordinal) in doc_id_mapping.iter() {
for (old_doc_id, reader_ordinal) in doc_id_mapping.iter() {
let term_ordinal_mapping: &[TermOrdinal] =
term_ordinal_mappings.get_segment(reader_with_ordinal.ordinal as usize);
term_ordinal_mappings.get_segment(*reader_ordinal as usize);
let ff_reader = &fast_field_reader[reader_with_ordinal.ordinal as usize];
let ff_reader = &fast_field_reader[*reader_ordinal as usize];
ff_reader.get_vals(*old_doc_id, &mut vals);
for &prev_term_ord in &vals {
let new_term_ord = term_ordinal_mapping[prev_term_ord as usize];
@@ -625,29 +621,33 @@ impl IndexMerger {
/// Creates a mapping if the segments are stacked. this is helpful to merge codelines between index
/// sorting and the others
pub(crate) fn get_doc_id_from_concatenated_data(&self) -> crate::Result<SegmentDocidMapping> {
let mapping: Vec<_> = self
pub(crate) fn get_doc_id_from_concatenated_data(&self) -> crate::Result<SegmentDocIdMapping> {
let total_num_new_docs = self
.readers
.iter()
.enumerate()
.map(|(ordinal, reader)| {
let reader_with_ordinal = SegmentReaderWithOrdinal {
ordinal: ordinal as u32,
reader,
};
reader
.doc_ids_alive()
.map(move |doc_id| (doc_id, reader_with_ordinal))
})
.flatten()
.collect();
Ok(SegmentDocidMapping::new(mapping, true))
.map(|reader| reader.num_docs() as usize)
.sum();
let mut mapping = Vec::with_capacity(total_num_new_docs);
mapping.extend(
self.readers
.iter()
.enumerate()
.map(|(reader_ordinal, reader)| {
reader
.doc_ids_alive()
.map(move |doc_id| (doc_id, reader_ordinal as SegmentOrdinal))
})
.flatten(),
);
Ok(SegmentDocIdMapping::new(mapping, true))
}
fn write_multi_fast_field(
&self,
field: Field,
fast_field_serializer: &mut CompositeFastFieldSerializer,
doc_id_mapping: &SegmentDocidMapping,
doc_id_mapping: &SegmentDocIdMapping,
) -> crate::Result<()> {
// Multifastfield consists in 2 fastfields.
// The first serves as an index into the second one and is stricly increasing.
@@ -703,16 +703,16 @@ impl IndexMerger {
min_value,
};
struct SortedDocidMultiValueAccessProvider<'a> {
doc_id_mapping: &'a SegmentDocidMapping<'a>,
struct SortedDocIdMultiValueAccessProvider<'a> {
doc_id_mapping: &'a SegmentDocIdMapping,
fast_field_readers: &'a Vec<MultiValuedFastFieldReader<u64>>,
offsets: Vec<u64>,
}
impl<'a> FastFieldDataAccess for SortedDocidMultiValueAccessProvider<'a> {
impl<'a> FastFieldDataAccess for SortedDocIdMultiValueAccessProvider<'a> {
fn get_val(&self, pos: u64) -> u64 {
// use the offsets index to find the doc_id which will contain the position.
// the offsets are stricly increasing so we can do a simple search on it.
let new_docid = self
let new_doc_id = self
.offsets
.iter()
.position(|&offset| offset > pos)
@@ -720,29 +720,27 @@ impl IndexMerger {
- 1;
// now we need to find the position of `pos` in the multivalued bucket
let num_pos_covered_until_now = self.offsets[new_docid];
let num_pos_covered_until_now = self.offsets[new_doc_id];
let pos_in_values = pos - num_pos_covered_until_now;
let (old_doc_id, reader_with_ordinal) = self.doc_id_mapping[new_docid as usize];
let num_vals = self.fast_field_readers[reader_with_ordinal.ordinal as usize]
.get_len(old_doc_id);
let (old_doc_id, reader_ordinal) = self.doc_id_mapping[new_doc_id as usize];
let num_vals = self.fast_field_readers[reader_ordinal as usize].get_len(old_doc_id);
assert!(num_vals >= pos_in_values);
let mut vals = vec![];
self.fast_field_readers[reader_with_ordinal.ordinal as usize]
.get_vals(old_doc_id, &mut vals);
self.fast_field_readers[reader_ordinal as usize].get_vals(old_doc_id, &mut vals);
vals[pos_in_values as usize]
}
}
let fastfield_accessor = SortedDocidMultiValueAccessProvider {
let fastfield_accessor = SortedDocIdMultiValueAccessProvider {
doc_id_mapping,
fast_field_readers: &ff_readers,
offsets,
};
let iter1 = doc_id_mapping
.iter()
.map(|(doc_id, reader_with_ordinal)| {
let ff_reader = &ff_readers[reader_with_ordinal.ordinal as usize];
.map(|(doc_id, reader_ordinal)| {
let ff_reader = &ff_readers[*reader_ordinal as usize];
let mut vals = vec![];
ff_reader.get_vals(*doc_id, &mut vals);
vals.into_iter()
@@ -750,8 +748,8 @@ impl IndexMerger {
.flatten();
let iter2 = doc_id_mapping
.iter()
.map(|(doc_id, reader_with_ordinal)| {
let ff_reader = &ff_readers[reader_with_ordinal.ordinal as usize];
.map(|(doc_id, reader_ordinal)| {
let ff_reader = &ff_readers[*reader_ordinal as usize];
let mut vals = vec![];
ff_reader.get_vals(*doc_id, &mut vals);
vals.into_iter()
@@ -773,7 +771,7 @@ impl IndexMerger {
&self,
field: Field,
fast_field_serializer: &mut CompositeFastFieldSerializer,
doc_id_mapping: &SegmentDocidMapping,
doc_id_mapping: &SegmentDocIdMapping,
) -> crate::Result<()> {
let reader_and_field_accessors = self
.readers
@@ -793,8 +791,8 @@ impl IndexMerger {
)?;
let mut serialize_vals = fast_field_serializer.new_bytes_fast_field_with_idx(field, 1);
for (doc_id, reader_with_ordinal) in doc_id_mapping.iter() {
let bytes_reader = &reader_and_field_accessors[reader_with_ordinal.ordinal as usize].1;
for (doc_id, reader_ordinal) in doc_id_mapping.iter() {
let bytes_reader = &reader_and_field_accessors[*reader_ordinal as usize].1;
let val = bytes_reader.get_bytes(*doc_id);
serialize_vals.write_all(val)?;
}
@@ -809,7 +807,7 @@ impl IndexMerger {
field_type: &FieldType,
serializer: &mut InvertedIndexSerializer,
fieldnorm_reader: Option<FieldNormReader>,
doc_id_mapping: &SegmentDocidMapping,
doc_id_mapping: &SegmentDocIdMapping,
) -> crate::Result<Option<TermOrdinalMapping>> {
debug_time!("write_postings_for_field");
let mut positions_buffer: Vec<u32> = Vec::with_capacity(1_000);
@@ -848,8 +846,8 @@ impl IndexMerger {
segment_local_map
})
.collect();
for (new_doc_id, (old_doc_id, segment_and_ordinal)) in doc_id_mapping.iter().enumerate() {
let segment_map = &mut merged_doc_id_map[segment_and_ordinal.ordinal as usize];
for (new_doc_id, (old_doc_id, segment_ord)) in doc_id_mapping.iter().enumerate() {
let segment_map = &mut merged_doc_id_map[*segment_ord as usize];
segment_map[*old_doc_id as usize] = Some(new_doc_id as DocId);
}
@@ -891,14 +889,14 @@ impl IndexMerger {
let mut total_doc_freq = 0;
// Let's compute the list of non-empty posting lists
for (segment_ord, term_info) in merged_terms.current_segment_ordinals_and_term_infos() {
for (segment_ord, term_info) in merged_terms.current_segment_ords_and_term_infos() {
let segment_reader = &self.readers[segment_ord];
let inverted_index: &InvertedIndexReader = &*field_readers[segment_ord];
let segment_postings = inverted_index
.read_postings_from_terminfo(&term_info, segment_postings_option)?;
let delete_bitset_opt = segment_reader.delete_bitset();
let doc_freq = if let Some(delete_bitset) = delete_bitset_opt {
segment_postings.doc_freq_given_deletes(delete_bitset)
let alive_bitset_opt = segment_reader.alive_bitset();
let doc_freq = if let Some(alive_bitset) = alive_bitset_opt {
segment_postings.doc_freq_given_deletes(alive_bitset)
} else {
segment_postings.doc_freq()
};
@@ -941,9 +939,9 @@ impl IndexMerger {
// there is at least one document.
let term_freq = segment_postings.term_freq();
segment_postings.positions(&mut positions_buffer);
// if doc_id_mapping exists, the docids are reordered, they are
// if doc_id_mapping exists, the doc_ids are reordered, they are
// not just stacked. The field serializer expects monotonically increasing
// docids, so we collect and sort them first, before writing.
// doc_ids, so we collect and sort them first, before writing.
//
// I think this is not strictly necessary, it would be possible to
// avoid the loading into a vec via some form of kmerge, but then the merge
@@ -983,7 +981,7 @@ impl IndexMerger {
&self,
serializer: &mut InvertedIndexSerializer,
fieldnorm_readers: FieldNormReaders,
doc_id_mapping: &SegmentDocidMapping,
doc_id_mapping: &SegmentDocIdMapping,
) -> crate::Result<HashMap<Field, TermOrdinalMapping>> {
let mut term_ordinal_mappings = HashMap::new();
for (field, field_entry) in self.schema.fields() {
@@ -1006,7 +1004,7 @@ impl IndexMerger {
fn write_storable_fields(
&self,
store_writer: &mut StoreWriter,
doc_id_mapping: &SegmentDocidMapping,
doc_id_mapping: &SegmentDocIdMapping,
) -> crate::Result<()> {
debug_time!("write_storable_fields");
@@ -1018,11 +1016,11 @@ impl IndexMerger {
let mut document_iterators: Vec<_> = store_readers
.iter()
.enumerate()
.map(|(i, store)| store.iter_raw(self.readers[i].delete_bitset()))
.map(|(i, store)| store.iter_raw(self.readers[i].alive_bitset()))
.collect();
if !doc_id_mapping.is_trivial() {
for (old_doc_id, reader_with_ordinal) in doc_id_mapping.iter() {
let doc_bytes_it = &mut document_iterators[reader_with_ordinal.ordinal as usize];
for (old_doc_id, reader_ordinal) in doc_id_mapping.iter() {
let doc_bytes_it = &mut document_iterators[*reader_ordinal as usize];
if let Some(doc_bytes_res) = doc_bytes_it.next() {
let doc_bytes = doc_bytes_res?;
store_writer.store_bytes(&doc_bytes)?;
@@ -1037,7 +1035,7 @@ impl IndexMerger {
} else {
for reader in &self.readers {
let store_reader = reader.get_store_reader()?;
if reader.num_deleted_docs() > 0
if reader.has_deletes()
// If there is not enough data in the store, we avoid stacking in order to
// avoid creating many small blocks in the doc store. Once we have 5 full blocks,
// we start stacking. In the worst case 2/7 of the blocks would be very small.
@@ -1048,13 +1046,13 @@ impl IndexMerger {
// the doc stores would be on average half full, given total randomness (which
// is not the case here, but not sure how it behaves exactly).
//
// https://github.com/tantivy-search/tantivy/issues/1053
// https://github.com/quickwit-inc/tantivy/issues/1053
//
// take 7 in order to not walk over all checkpoints.
|| store_reader.block_checkpoints().take(7).count() < 6
|| store_reader.compressor() != store_writer.compressor()
{
for doc_bytes_res in store_reader.iter_raw(reader.delete_bitset()) {
for doc_bytes_res in store_reader.iter_raw(reader.alive_bitset()) {
let doc_bytes = doc_bytes_res?;
store_writer.store_bytes(&doc_bytes)?;
}
@@ -1164,18 +1162,17 @@ mod tests {
score_field => 3u64,
date_field => curr_time,
bytes_score_field => 3u32.to_be_bytes().as_ref()
));
))?;
index_writer.add_document(doc!(
text_field => "a b c",
score_field => 5u64,
bytes_score_field => 5u32.to_be_bytes().as_ref()
));
))?;
index_writer.add_document(doc!(
text_field => "a b c d",
score_field => 7u64,
bytes_score_field => 7u32.to_be_bytes().as_ref()
));
))?;
index_writer.commit()?;
// writing the segment
index_writer.add_document(doc!(
@@ -1183,12 +1180,12 @@ mod tests {
date_field => curr_time,
score_field => 11u64,
bytes_score_field => 11u32.to_be_bytes().as_ref()
));
))?;
index_writer.add_document(doc!(
text_field => "a b c g",
score_field => 13u64,
bytes_score_field => 13u32.to_be_bytes().as_ref()
));
))?;
index_writer.commit()?;
}
{
@@ -1322,18 +1319,18 @@ mod tests {
text_field => "a b d",
score_field => 1u64,
bytes_score_field => vec![0u8, 0, 0, 1],
));
))?;
index_writer.add_document(doc!(
text_field => "b c",
score_field => 2u64,
bytes_score_field => vec![0u8, 0, 0, 2],
));
))?;
index_writer.delete_term(Term::from_field_text(text_field, "c"));
index_writer.add_document(doc!(
text_field => "c d",
score_field => 3u64,
bytes_score_field => vec![0u8, 0, 0, 3],
));
))?;
index_writer.commit()?;
reader.reload()?;
let searcher = reader.searcher();
@@ -1363,24 +1360,24 @@ mod tests {
text_field => "a d e",
score_field => 4_000u64,
bytes_score_field => vec![0u8, 0, 0, 4],
));
))?;
index_writer.add_document(doc!(
text_field => "e f",
score_field => 5_000u64,
bytes_score_field => vec![0u8, 0, 0, 5],
));
))?;
index_writer.delete_term(Term::from_field_text(text_field, "a"));
index_writer.delete_term(Term::from_field_text(text_field, "f"));
index_writer.add_document(doc!(
text_field => "f g",
score_field => 6_000u64,
bytes_score_field => vec![0u8, 0, 23, 112],
));
))?;
index_writer.add_document(doc!(
text_field => "g h",
score_field => 7_000u64,
bytes_score_field => vec![0u8, 0, 27, 88],
));
))?;
index_writer.commit()?;
reader.reload()?;
let searcher = reader.searcher();
@@ -1598,7 +1595,7 @@ mod tests {
#[test]
fn test_merge_facets_sort_asc() {
// In the merge case this will go through the docid mapping code
// In the merge case this will go through the doc_id mapping code
test_merge_facets(
Some(IndexSettings {
sort_by_field: Some(IndexSortByField {
@@ -1609,7 +1606,7 @@ mod tests {
}),
true,
);
// In the merge case this will not go through the docid mapping code, because the data is
// In the merge case this will not go through the doc_id mapping code, because the data is
// sorted and disjunct
test_merge_facets(
Some(IndexSettings {
@@ -1625,7 +1622,7 @@ mod tests {
#[test]
fn test_merge_facets_sort_desc() {
// In the merge case this will go through the docid mapping code
// In the merge case this will go through the doc_id mapping code
test_merge_facets(
Some(IndexSettings {
sort_by_field: Some(IndexSortByField {
@@ -1636,7 +1633,7 @@ mod tests {
}),
true,
);
// In the merge case this will not go through the docid mapping code, because the data is
// In the merge case this will not go through the doc_id mapping code, because the data is
// sorted and disjunct
test_merge_facets(
Some(IndexSettings {
@@ -1676,7 +1673,7 @@ mod tests {
}
doc.add_u64(int_field, *int_val);
*int_val += 1;
index_writer.add_document(doc);
index_writer.add_document(doc).unwrap();
};
index_doc(
@@ -1789,70 +1786,69 @@ mod tests {
}
#[test]
fn test_bug_merge() {
fn test_bug_merge() -> crate::Result<()> {
let mut schema_builder = schema::Schema::builder();
let int_field = schema_builder.add_u64_field("intvals", INDEXED);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(int_field => 1u64));
index_writer.add_document(doc!(int_field => 1u64))?;
index_writer.commit().expect("commit failed");
index_writer.add_document(doc!(int_field => 1u64));
index_writer.add_document(doc!(int_field => 1u64))?;
index_writer.commit().expect("commit failed");
let reader = index.reader().unwrap();
let reader = index.reader()?;
let searcher = reader.searcher();
assert_eq!(searcher.num_docs(), 2);
index_writer.delete_term(Term::from_field_u64(int_field, 1));
let segment_ids = index
.searchable_segment_ids()
.expect("Searchable segments failed.");
block_on(index_writer.merge(&segment_ids)).expect("Merging failed");
reader.reload().unwrap();
block_on(index_writer.merge(&segment_ids))?;
reader.reload()?;
// commit has not been called yet. The document should still be
// there.
assert_eq!(reader.searcher().num_docs(), 2);
Ok(())
}
#[test]
fn test_merge_multivalued_int_fields_all_deleted() {
fn test_merge_multivalued_int_fields_all_deleted() -> crate::Result<()> {
let mut schema_builder = schema::Schema::builder();
let int_options = IntOptions::default()
.set_fast(Cardinality::MultiValues)
.set_indexed();
let int_field = schema_builder.add_u64_field("intvals", int_options);
let index = Index::create_in_ram(schema_builder.build());
let reader = index.reader().unwrap();
let reader = index.reader()?;
{
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
let mut doc = Document::default();
doc.add_u64(int_field, 1);
index_writer.add_document(doc.clone());
assert!(index_writer.commit().is_ok());
index_writer.add_document(doc);
assert!(index_writer.commit().is_ok());
index_writer.add_document(doc.clone())?;
index_writer.commit()?;
index_writer.add_document(doc)?;
index_writer.commit()?;
index_writer.delete_term(Term::from_field_u64(int_field, 1));
let segment_ids = index
.searchable_segment_ids()
.expect("Searchable segments failed.");
assert!(block_on(index_writer.merge(&segment_ids)).is_ok());
let segment_ids = index.searchable_segment_ids()?;
block_on(index_writer.merge(&segment_ids))?;
// assert delete has not been committed
assert!(reader.reload().is_ok());
reader.reload()?;
let searcher = reader.searcher();
assert_eq!(searcher.num_docs(), 2);
index_writer.commit().unwrap();
index_writer.commit()?;
index_writer.wait_merging_threads().unwrap();
index_writer.wait_merging_threads()?;
}
reader.reload().unwrap();
reader.reload()?;
let searcher = reader.searcher();
assert_eq!(searcher.num_docs(), 0);
Ok(())
}
#[test]
fn test_merge_multivalued_int_fields_simple() {
fn test_merge_multivalued_int_fields_simple() -> crate::Result<()> {
let mut schema_builder = schema::Schema::builder();
let int_options = IntOptions::default()
.set_fast(Cardinality::MultiValues)
@@ -1861,13 +1857,13 @@ mod tests {
let index = Index::create_in_ram(schema_builder.build());
{
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
let index_doc = |index_writer: &mut IndexWriter, int_vals: &[u64]| {
let mut doc = Document::default();
for &val in int_vals {
doc.add_u64(int_field, val);
}
index_writer.add_document(doc);
index_writer.add_document(doc).unwrap();
};
index_doc(&mut index_writer, &[1, 2]);
index_doc(&mut index_writer, &[1, 2, 3]);
@@ -1883,7 +1879,7 @@ mod tests {
index_doc(&mut index_writer, &[1_000]);
assert!(index_writer.commit().is_ok());
}
let reader = index.reader().unwrap();
let reader = index.reader()?;
let searcher = reader.searcher();
let mut vals: Vec<u64> = Vec::new();
@@ -1932,14 +1928,12 @@ mod tests {
// Merging the segments
{
let segment_ids = index
.searchable_segment_ids()
.expect("Searchable segments failed.");
let mut index_writer = index.writer_for_tests().unwrap();
assert!(block_on(index_writer.merge(&segment_ids)).is_ok());
assert!(index_writer.wait_merging_threads().is_ok());
let segment_ids = index.searchable_segment_ids()?;
let mut index_writer = index.writer_for_tests()?;
block_on(index_writer.merge(&segment_ids))?;
index_writer.wait_merging_threads()?;
}
assert!(reader.reload().is_ok());
reader.reload()?;
{
let searcher = reader.searcher();
@@ -1976,6 +1970,7 @@ mod tests {
ff_reader.get_vals(9, &mut vals);
assert_eq!(&vals, &[20]);
}
Ok(())
}
#[test]
@@ -2001,7 +1996,7 @@ mod tests {
doc.add_f64(field, 42.0);
doc.add_f64(multi_field, 0.24);
doc.add_f64(multi_field, 0.27);
writer.add_document(doc);
writer.add_document(doc)?;
if i % 5 == 0 {
writer.commit()?;
}
@@ -2025,7 +2020,7 @@ mod tests {
let happy_term = Term::from_field_text(text, "happy");
let term_query = TermQuery::new(happy_term, IndexRecordOption::WithFreqs);
for _ in 0..62 {
writer.add_document(doc!(text=>"hello happy tax payer"));
writer.add_document(doc!(text=>"hello happy tax payer"))?;
}
writer.commit()?;
let reader = index.reader()?;
@@ -2037,7 +2032,7 @@ mod tests {
assert_nearly_equals!(term_scorer.block_max_score(), 0.0079681855);
assert_nearly_equals!(term_scorer.score(), 0.0079681855);
for _ in 0..81 {
writer.add_document(doc!(text=>"hello happy tax payer"));
writer.add_document(doc!(text=>"hello happy tax payer"))?;
}
writer.commit()?;
reader.reload()?;

93
src/indexer/merger_sorted_index_test.rs
View File

@@ -1,6 +1,6 @@
#[cfg(test)]
mod tests {
use crate::fastfield::{DeleteBitSet, FastFieldReader};
use crate::fastfield::{AliveBitSet, FastFieldReader};
use crate::schema::IndexRecordOption;
use crate::{
collector::TopDocs,
@@ -39,14 +39,17 @@ mod tests {
{
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(int_field=>3_u64, facet_field=> Facet::from("/crime")));
index_writer.add_document(doc!(int_field=>6_u64, facet_field=> Facet::from("/crime")));
assert!(index_writer.commit().is_ok());
index_writer.add_document(doc!(int_field=>5_u64, facet_field=> Facet::from("/fanta")));
assert!(index_writer.commit().is_ok());
index_writer
.add_document(doc!(int_field=>3_u64, facet_field=> Facet::from("/crime")))
.unwrap();
index_writer
.add_document(doc!(int_field=>6_u64, facet_field=> Facet::from("/crime")))
.unwrap();
index_writer.commit().unwrap();
index_writer
.add_document(doc!(int_field=>5_u64, facet_field=> Facet::from("/fanta")))
.unwrap();
index_writer.commit().unwrap();
}
// Merging the segments
@@ -66,7 +69,7 @@ mod tests {
fn create_test_index(
index_settings: Option<IndexSettings>,
force_disjunct_segment_sort_values: bool,
) -> Index {
) -> crate::Result<Index> {
let mut schema_builder = schema::Schema::builder();
let int_options = IntOptions::default()
.set_fast(Cardinality::SingleValue)
@@ -95,34 +98,34 @@ mod tests {
if let Some(settings) = index_settings {
index_builder = index_builder.settings(settings);
}
let index = index_builder.create_in_ram().unwrap();
let index = index_builder.create_in_ram()?;
{
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
// segment 1 - range 1-3
index_writer.add_document(doc!(int_field=>1_u64));
index_writer.add_document(doc!(int_field=>1_u64))?;
index_writer.add_document(
doc!(int_field=>3_u64, multi_numbers => 3_u64, multi_numbers => 4_u64, bytes_field => vec![1, 2, 3], text_field => "some text", facet_field=> Facet::from("/book/crime")),
);
)?;
index_writer.add_document(
doc!(int_field=>1_u64, text_field=> "deleteme", text_field => "ok text more text"),
);
)?;
index_writer.add_document(
doc!(int_field=>2_u64, multi_numbers => 2_u64, multi_numbers => 3_u64, text_field => "ok text more text"),
);
)?;
assert!(index_writer.commit().is_ok());
index_writer.commit()?;
// segment 2 - range 1-20 , with force_disjunct_segment_sort_values 10-20
index_writer.add_document(doc!(int_field=>20_u64, multi_numbers => 20_u64));
index_writer.add_document(doc!(int_field=>20_u64, multi_numbers => 20_u64))?;
let in_val = if force_disjunct_segment_sort_values {
10_u64
} else {
1
};
index_writer.add_document(doc!(int_field=>in_val, text_field=> "deleteme" , text_field => "ok text more text", facet_field=> Facet::from("/book/crime")));
assert!(index_writer.commit().is_ok());
index_writer.add_document(doc!(int_field=>in_val, text_field=> "deleteme" , text_field => "ok text more text", facet_field=> Facet::from("/book/crime")))?;
index_writer.commit()?;
// segment 3 - range 5-1000, with force_disjunct_segment_sort_values 50-1000
let int_vals = if force_disjunct_segment_sort_values {
[100_u64, 50]
@@ -131,26 +134,24 @@ mod tests {
};
index_writer.add_document( // position of this doc after delete in desc sorting = [2], in disjunct case [1]
doc!(int_field=>int_vals[0], multi_numbers => 10_u64, multi_numbers => 11_u64, text_field=> "blubber", facet_field=> Facet::from("/book/fantasy")),
);
index_writer.add_document(doc!(int_field=>int_vals[1], text_field=> "deleteme"));
)?;
index_writer.add_document(doc!(int_field=>int_vals[1], text_field=> "deleteme"))?;
index_writer.add_document(
doc!(int_field=>1_000u64, multi_numbers => 1001_u64, multi_numbers => 1002_u64, bytes_field => vec![5, 5],text_field => "the biggest num")
);
)?;
index_writer.delete_term(Term::from_field_text(text_field, "deleteme"));
assert!(index_writer.commit().is_ok());
index_writer.commit()?;
}
// Merging the segments
{
let segment_ids = index
.searchable_segment_ids()
.expect("Searchable segments failed.");
let mut index_writer = index.writer_for_tests().unwrap();
assert!(block_on(index_writer.merge(&segment_ids)).is_ok());
assert!(index_writer.wait_merging_threads().is_ok());
let segment_ids = index.searchable_segment_ids()?;
let mut index_writer = index.writer_for_tests()?;
block_on(index_writer.merge(&segment_ids))?;
index_writer.wait_merging_threads()?;
}
index
Ok(index)
}
#[test]
@@ -183,7 +184,8 @@ mod tests {
..Default::default()
}),
force_disjunct_segment_sort_values,
);
)
.unwrap();
let int_field = index.schema().get_field("intval").unwrap();
let reader = index.reader().unwrap();
@@ -257,10 +259,10 @@ mod tests {
.unwrap();
assert_eq!(postings.doc_freq(), 2);
let fallback_bitset = DeleteBitSet::for_test(&[0], 100);
let fallback_bitset = AliveBitSet::for_test_from_deleted_docs(&[0], 100);
assert_eq!(
postings.doc_freq_given_deletes(
segment_reader.delete_bitset().unwrap_or(&fallback_bitset)
segment_reader.alive_bitset().unwrap_or(&fallback_bitset)
),
2
);
@@ -300,7 +302,8 @@ mod tests {
..Default::default()
}),
false,
);
)
.unwrap();
let reader = index.reader().unwrap();
let searcher = reader.searcher();
@@ -336,10 +339,10 @@ mod tests {
.unwrap()
.unwrap();
assert_eq!(postings.doc_freq(), 2);
let fallback_bitset = DeleteBitSet::for_test(&[0], 100);
let fallback_bitset = AliveBitSet::for_test_from_deleted_docs(&[0], 100);
assert_eq!(
postings.doc_freq_given_deletes(
segment_reader.delete_bitset().unwrap_or(&fallback_bitset)
segment_reader.alive_bitset().unwrap_or(&fallback_bitset)
),
2
);
@@ -367,7 +370,8 @@ mod tests {
..Default::default()
}),
false,
);
)
.unwrap();
let int_field = index.schema().get_field("intval").unwrap();
let multi_numbers = index.schema().get_field("multi_numbers").unwrap();
@@ -446,10 +450,10 @@ mod tests {
.unwrap();
assert_eq!(postings.doc_freq(), 2);
let fallback_bitset = DeleteBitSet::for_test(&[0], 100);
let fallback_bitset = AliveBitSet::for_test_from_deleted_docs(&[0], 100);
assert_eq!(
postings.doc_freq_given_deletes(
segment_reader.delete_bitset().unwrap_or(&fallback_bitset)
segment_reader.alive_bitset().unwrap_or(&fallback_bitset)
),
2
);
@@ -546,14 +550,15 @@ mod bench_sorted_index_merge {
let doc_id_mapping = merger.generate_doc_id_mapping(&sort_by_field).unwrap();
b.iter(|| {
let sorted_doc_ids = doc_id_mapping.iter().map(|(doc_id, reader)|{
let u64_reader: DynamicFastFieldReader<u64> = reader.reader
let sorted_doc_ids = doc_id_mapping.iter().map(|(doc_id, ordinal)|{
let reader = &merger.readers[*ordinal as usize];
let u64_reader: DynamicFastFieldReader<u64> = reader
.fast_fields()
.typed_fast_field_reader(field)
.expect("Failed to find a reader for single fast field. This is a tantivy bug and it should never happen.");
(doc_id, reader, u64_reader)
});
// add values in order of the new docids
// add values in order of the new doc_ids
let mut val = 0;
for (doc_id, _reader, field_reader) in sorted_doc_ids {
val = field_reader.get(*doc_id);
@@ -566,7 +571,7 @@ mod bench_sorted_index_merge {
Ok(())
}
#[bench]
fn create_sorted_index_create_docid_mapping(b: &mut Bencher) -> crate::Result<()> {
fn create_sorted_index_create_doc_id_mapping(b: &mut Bencher) -> crate::Result<()> {
let sort_by_field = IndexSortByField {
field: "intval".to_string(),
order: Order::Desc,

37
src/indexer/mod.rs
View File

@@ -1,15 +1,17 @@
pub mod delete_queue;
pub mod demuxer;
pub mod doc_id_mapping;
mod doc_opstamp_mapping;
pub mod index_writer;
mod index_writer_status;
mod log_merge_policy;
mod merge_operation;
pub mod merge_policy;
pub mod merger;
mod merger_sorted_index_test;
pub mod operation;
mod prepared_commit;
pub mod prepared_commit;
mod segment_entry;
mod segment_manager;
mod segment_register;
@@ -18,6 +20,11 @@ pub mod segment_updater;
mod segment_writer;
mod stamper;
use crossbeam::channel;
use smallvec::SmallVec;
use crate::indexer::operation::AddOperation;
pub use self::index_writer::IndexWriter;
pub use self::log_merge_policy::LogMergePolicy;
pub use self::merge_operation::MergeOperation;
@@ -26,12 +33,23 @@ pub use self::prepared_commit::PreparedCommit;
pub use self::segment_entry::SegmentEntry;
pub use self::segment_manager::SegmentManager;
pub use self::segment_serializer::SegmentSerializer;
pub use self::segment_updater::merge_segments;
pub use self::segment_updater::merge_filtered_segments;
pub use self::segment_updater::merge_indices;
pub use self::segment_writer::SegmentWriter;
/// Alias for the default merge policy, which is the `LogMergePolicy`.
pub type DefaultMergePolicy = LogMergePolicy;
// Batch of documents.
// Most of the time, users will send operation one-by-one, but it can be useful to
// send them as a small block to ensure that
// - all docs in the operation will happen on the same segment and continuous doc_ids.
// - all operations in the group are committed at the same time, making the group
// atomic.
type AddBatch = SmallVec<[AddOperation; 4]>;
type AddBatchSender = channel::Sender<AddBatch>;
type AddBatchReceiver = channel::Receiver<AddBatch>;
#[cfg(feature = "mmap")]
#[cfg(test)]
mod tests_mmap {
@@ -39,19 +57,20 @@ mod tests_mmap {
use crate::{Index, Term};
#[test]
fn test_advance_delete_bug() {
fn test_advance_delete_bug() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", schema::TEXT);
let index = Index::create_from_tempdir(schema_builder.build()).unwrap();
let mut index_writer = index.writer_for_tests().unwrap();
let index = Index::create_from_tempdir(schema_builder.build())?;
let mut index_writer = index.writer_for_tests()?;
// there must be one deleted document in the segment
index_writer.add_document(doc!(text_field=>"b"));
index_writer.add_document(doc!(text_field=>"b"))?;
index_writer.delete_term(Term::from_field_text(text_field, "b"));
// we need enough data to trigger the bug (at least 32 documents)
for _ in 0..32 {
index_writer.add_document(doc!(text_field=>"c"));
index_writer.add_document(doc!(text_field=>"c"))?;
}
index_writer.commit().unwrap();
index_writer.commit().unwrap();
index_writer.commit()?;
index_writer.commit()?;
Ok(())
}
}

23
src/indexer/prepared_commit.rs
View File

@@ -18,25 +18,38 @@ impl<'a> PreparedCommit<'a> {
}
}
/// Returns the opstamp associated to the prepared commit.
pub fn opstamp(&self) -> Opstamp {
self.opstamp
}
/// Adds an arbitrary payload to the commit.
pub fn set_payload(&mut self, payload: &str) {
self.payload = Some(payload.to_string())
}
/// Rollbacks any change.
pub fn abort(self) -> crate::Result<Opstamp> {
self.index_writer.rollback()
}
/// Proceeds to commit.
/// See `.commit_async()`.
pub fn commit(self) -> crate::Result<Opstamp> {
block_on(self.commit_async())
}
/// Proceeds to commit.
///
/// Unfortunately, contrary to what `PrepareCommit` may suggests,
/// this operation is not at all really light.
/// At this point deletes have not been flushed yet.
pub async fn commit_async(self) -> crate::Result<Opstamp> {
info!("committing {}", self.opstamp);
let _ = block_on(
self.index_writer
.segment_updater()
.schedule_commit(self.opstamp, self.payload),
);
self.index_writer
.segment_updater()
.schedule_commit(self.opstamp, self.payload)
.await?;
Ok(self.opstamp)
}
}

18
src/indexer/segment_entry.rs
View File

@@ -9,18 +9,16 @@ use std::fmt;
///
/// In addition to segment `meta`,
/// it contains a few transient states
/// - `state` expresses whether the segment is already in the
/// middle of a merge
/// - `delete_bitset` is a bitset describing
/// documents that were deleted during the commit
/// - `alive_bitset` is a bitset describing
/// documents that were alive during the commit
/// itself.
/// - `delete_cursor` is the position in the delete queue.
/// Deletes happening before the cursor are reflected either
/// in the .del file or in the `delete_bitset`.
/// in the .del file or in the `alive_bitset`.
#[derive(Clone)]
pub struct SegmentEntry {
meta: SegmentMeta,
delete_bitset: Option<BitSet>,
alive_bitset: Option<BitSet>,
delete_cursor: DeleteCursor,
}
@@ -29,11 +27,11 @@ impl SegmentEntry {
pub fn new(
segment_meta: SegmentMeta,
delete_cursor: DeleteCursor,
delete_bitset: Option<BitSet>,
alive_bitset: Option<BitSet>,
) -> SegmentEntry {
SegmentEntry {
meta: segment_meta,
delete_bitset,
alive_bitset,
delete_cursor,
}
}
@@ -41,8 +39,8 @@ impl SegmentEntry {
/// Return a reference to the segment entry deleted bitset.
///
/// `DocId` in this bitset are flagged as deleted.
pub fn delete_bitset(&self) -> Option<&BitSet> {
self.delete_bitset.as_ref()
pub fn alive_bitset(&self) -> Option<&BitSet> {
self.alive_bitset.as_ref()
}
/// Set the `SegmentMeta` for this segment.

7
src/indexer/segment_register.rs
View File

@@ -66,13 +66,10 @@ impl SegmentRegister {
}
pub fn segment_metas(&self) -> Vec<SegmentMeta> {
let mut segment_ids: Vec<SegmentMeta> = self
.segment_states
self.segment_states
.values()
.map(|segment_entry| segment_entry.meta().clone())
.collect();
segment_ids.sort_by_key(SegmentMeta::id);
segment_ids
.collect()
}
pub fn contains_all(&self, segment_ids: &[SegmentId]) -> bool {

432
src/indexer/segment_updater.rs
View File

@@ -7,6 +7,7 @@ use crate::core::SegmentId;
use crate::core::SegmentMeta;
use crate::core::META_FILEPATH;
use crate::directory::{Directory, DirectoryClone, GarbageCollectionResult};
use crate::fastfield::AliveBitSet;
use crate::indexer::delete_queue::DeleteCursor;
use crate::indexer::index_writer::advance_deletes;
use crate::indexer::merge_operation::MergeOperationInventory;
@@ -19,12 +20,15 @@ use crate::indexer::{DefaultMergePolicy, MergePolicy};
use crate::indexer::{MergeCandidate, MergeOperation};
use crate::schema::Schema;
use crate::Opstamp;
use crate::TantivyError;
use fail::fail_point;
use futures::channel::oneshot;
use futures::executor::{ThreadPool, ThreadPoolBuilder};
use futures::future::Future;
use futures::future::TryFutureExt;
use std::borrow::BorrowMut;
use std::collections::HashSet;
use std::io;
use std::io::Write;
use std::ops::Deref;
use std::path::PathBuf;
@@ -74,6 +78,10 @@ fn save_metas(metas: &IndexMeta, directory: &dyn Directory) -> crate::Result<()>
let mut buffer = serde_json::to_vec_pretty(metas)?;
// Just adding a new line at the end of the buffer.
writeln!(&mut buffer)?;
fail_point!("save_metas", |msg| Err(TantivyError::from(io::Error::new(
io::ErrorKind::Other,
msg.unwrap_or_else(|| "Undefined".to_string())
))));
directory.atomic_write(&META_FILEPATH, &buffer[..])?;
debug!("Saved metas {:?}", serde_json::to_string_pretty(&metas));
Ok(())
@@ -159,9 +167,9 @@ fn merge(
/// meant to work if you have an IndexWriter running for the origin indices, or
/// the destination Index.
#[doc(hidden)]
pub fn merge_segments<Dir: Directory>(
pub fn merge_indices<T: Into<Box<dyn Directory>>>(
indices: &[Index],
output_directory: Dir,
output_directory: T,
) -> crate::Result<Index> {
if indices.is_empty() {
// If there are no indices to merge, there is no need to do anything.
@@ -170,19 +178,8 @@ pub fn merge_segments<Dir: Directory>(
));
}
let target_schema = indices[0].schema();
let target_settings = indices[0].settings().clone();
// let's check that all of the indices have the same schema
if indices
.iter()
.skip(1)
.any(|index| index.schema() != target_schema)
{
return Err(crate::TantivyError::InvalidArgument(
"Attempt to merge different schema indices".to_string(),
));
}
// let's check that all of the indices have the same index settings
if indices
.iter()
@@ -199,13 +196,61 @@ pub fn merge_segments<Dir: Directory>(
segments.extend(index.searchable_segments()?);
}
let mut merged_index = Index::create(output_directory, target_schema.clone(), target_settings)?;
let non_filter = segments.iter().map(|_| None).collect::<Vec<_>>();
merge_filtered_segments(&segments, target_settings, non_filter, output_directory)
}
/// Advanced: Merges a list of segments from different indices in a new index.
/// Additional you can provide a delete bitset for each segment to ignore doc_ids.
///
/// Returns `TantivyError` if the the indices list is empty or their
/// schemas don't match.
///
/// `output_directory`: is assumed to be empty.
///
/// # Warning
/// This function does NOT check or take the `IndexWriter` is running. It is not
/// meant to work if you have an IndexWriter running for the origin indices, or
/// the destination Index.
#[doc(hidden)]
pub fn merge_filtered_segments<T: Into<Box<dyn Directory>>>(
segments: &[Segment],
target_settings: IndexSettings,
filter_doc_ids: Vec<Option<AliveBitSet>>,
output_directory: T,
) -> crate::Result<Index> {
if segments.is_empty() {
// If there are no indices to merge, there is no need to do anything.
return Err(crate::TantivyError::InvalidArgument(
"No segments given to marge".to_string(),
));
}
let target_schema = segments[0].schema();
// let's check that all of the indices have the same schema
if segments
.iter()
.skip(1)
.any(|index| index.schema() != target_schema)
{
return Err(crate::TantivyError::InvalidArgument(
"Attempt to merge different schema indices".to_string(),
));
}
let mut merged_index = Index::create(
output_directory,
target_schema.clone(),
target_settings.clone(),
)?;
let merged_segment = merged_index.new_segment();
let merged_segment_id = merged_segment.id();
let merger: IndexMerger = IndexMerger::open(
let merger: IndexMerger = IndexMerger::open_with_custom_alive_set(
merged_index.schema(),
merged_index.settings().clone(),
&segments[..],
segments,
filter_doc_ids,
)?;
let segment_serializer = SegmentSerializer::for_segment(merged_segment, true)?;
let num_docs = merger.write(segment_serializer)?;
@@ -225,7 +270,7 @@ pub fn merge_segments<Dir: Directory>(
);
let index_meta = IndexMeta {
index_settings: indices[0].load_metas()?.index_settings, // index_settings of all segments should be the same
index_settings: target_settings, // index_settings of all segments should be the same
segments: vec![segment_meta],
schema: target_schema,
opstamp: 0u64,
@@ -306,37 +351,39 @@ impl SegmentUpdater {
*self.merge_policy.write().unwrap() = arc_merge_policy;
}
fn schedule_future<T: 'static + Send, F: Future<Output = crate::Result<T>> + 'static + Send>(
async fn schedule_task<
T: 'static + Send,
F: Future<Output = crate::Result<T>> + 'static + Send,
>(
&self,
f: F,
) -> impl Future<Output = crate::Result<T>> {
let (sender, receiver) = oneshot::channel();
if self.is_alive() {
self.pool.spawn_ok(async move {
let _ = sender.send(f.await);
});
} else {
let _ = sender.send(Err(crate::TantivyError::SystemError(
task: F,
) -> crate::Result<T> {
if !self.is_alive() {
return Err(crate::TantivyError::SystemError(
"Segment updater killed".to_string(),
)));
));
}
receiver.unwrap_or_else(|_| {
let (sender, receiver) = oneshot::channel();
self.pool.spawn_ok(async move {
let task_result = task.await;
let _ = sender.send(task_result);
});
let task_result = receiver.await;
task_result.unwrap_or_else(|_| {
let err_msg =
"A segment_updater future did not success. This should never happen.".to_string();
Err(crate::TantivyError::SystemError(err_msg))
})
}
pub fn schedule_add_segment(
&self,
segment_entry: SegmentEntry,
) -> impl Future<Output = crate::Result<()>> {
pub async fn schedule_add_segment(&self, segment_entry: SegmentEntry) -> crate::Result<()> {
let segment_updater = self.clone();
self.schedule_future(async move {
self.schedule_task(async move {
segment_updater.segment_manager.add_segment(segment_entry);
segment_updater.consider_merge_options().await;
Ok(())
})
.await
}
/// Orders `SegmentManager` to remove all segments
@@ -403,11 +450,9 @@ impl SegmentUpdater {
Ok(())
}
pub fn schedule_garbage_collect(
&self,
) -> impl Future<Output = crate::Result<GarbageCollectionResult>> {
pub async fn schedule_garbage_collect(&self) -> crate::Result<GarbageCollectionResult> {
let garbage_collect_future = garbage_collect_files(self.clone());
self.schedule_future(garbage_collect_future)
self.schedule_task(garbage_collect_future).await
}
/// List the files that are useful to the index.
@@ -425,13 +470,13 @@ impl SegmentUpdater {
files
}
pub fn schedule_commit(
pub(crate) async fn schedule_commit(
&self,
opstamp: Opstamp,
payload: Option<String>,
) -> impl Future<Output = crate::Result<()>> {
) -> crate::Result<()> {
let segment_updater: SegmentUpdater = self.clone();
self.schedule_future(async move {
self.schedule_task(async move {
let segment_entries = segment_updater.purge_deletes(opstamp)?;
segment_updater.segment_manager.commit(segment_entries);
segment_updater.save_metas(opstamp, payload)?;
@@ -439,6 +484,7 @@ impl SegmentUpdater {
segment_updater.consider_merge_options().await;
Ok(())
})
.await
}
fn store_meta(&self, index_meta: &IndexMeta) {
@@ -513,9 +559,7 @@ impl SegmentUpdater {
e
);
// ... cancel merge
if cfg!(test) {
panic!("Merge failed.");
}
assert!(!cfg!(test), "Merge failed.");
}
}
});
@@ -568,14 +612,14 @@ impl SegmentUpdater {
}
}
fn end_merge(
async fn end_merge(
&self,
merge_operation: MergeOperation,
mut after_merge_segment_entry: SegmentEntry,
) -> impl Future<Output = crate::Result<SegmentMeta>> {
) -> crate::Result<SegmentMeta> {
let segment_updater = self.clone();
let after_merge_segment_meta = after_merge_segment_entry.meta().clone();
let end_merge_future = self.schedule_future(async move {
self.schedule_task(async move {
info!("End merge {:?}", after_merge_segment_entry.meta());
{
let mut delete_cursor = after_merge_segment_entry.delete_cursor().clone();
@@ -594,9 +638,8 @@ impl SegmentUpdater {
merge_operation.segment_ids(),
advance_deletes_err
);
if cfg!(test) {
panic!("Merge failed.");
}
assert!(!cfg!(test), "Merge failed.");
// ... cancel merge
// `merge_operations` are tracked. As it is dropped, the
// the segment_ids will be available again for merge.
@@ -619,8 +662,9 @@ impl SegmentUpdater {
let _ = garbage_collect_files(segment_updater).await;
Ok(())
});
end_merge_future.map_ok(|_| after_merge_segment_meta)
})
.await?;
Ok(after_merge_segment_meta)
}
/// Wait for current merging threads.
@@ -646,11 +690,19 @@ impl SegmentUpdater {
#[cfg(test)]
mod tests {
use super::merge_segments;
use super::merge_indices;
use crate::collector::TopDocs;
use crate::directory::RamDirectory;
use crate::fastfield::AliveBitSet;
use crate::indexer::merge_policy::tests::MergeWheneverPossible;
use crate::indexer::merger::IndexMerger;
use crate::indexer::segment_updater::merge_filtered_segments;
use crate::query::QueryParser;
use crate::schema::*;
use crate::Directory;
use crate::DocAddress;
use crate::Index;
use crate::Segment;
#[test]
fn test_delete_during_merge() -> crate::Result<()> {
@@ -663,19 +715,19 @@ mod tests {
index_writer.set_merge_policy(Box::new(MergeWheneverPossible));
for _ in 0..100 {
index_writer.add_document(doc!(text_field=>"a"));
index_writer.add_document(doc!(text_field=>"b"));
index_writer.add_document(doc!(text_field=>"a"))?;
index_writer.add_document(doc!(text_field=>"b"))?;
}
index_writer.commit()?;
for _ in 0..100 {
index_writer.add_document(doc!(text_field=>"c"));
index_writer.add_document(doc!(text_field=>"d"));
index_writer.add_document(doc!(text_field=>"c"))?;
index_writer.add_document(doc!(text_field=>"d"))?;
}
index_writer.commit()?;
index_writer.add_document(doc!(text_field=>"e"));
index_writer.add_document(doc!(text_field=>"f"));
index_writer.add_document(doc!(text_field=>"e"))?;
index_writer.add_document(doc!(text_field=>"f"))?;
index_writer.commit()?;
let term = Term::from_field_text(text_field, "a");
@@ -693,6 +745,50 @@ mod tests {
Ok(())
}
#[test]
fn delete_all_docs_min() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
// writing the segment
let mut index_writer = index.writer_for_tests()?;
for _ in 0..10 {
index_writer.add_document(doc!(text_field=>"a"))?;
index_writer.add_document(doc!(text_field=>"b"))?;
}
index_writer.commit()?;
let seg_ids = index.searchable_segment_ids()?;
// docs exist, should have at least 1 segment
assert!(!seg_ids.is_empty());
let term = Term::from_field_text(text_field, "a");
index_writer.delete_term(term);
index_writer.commit()?;
let term = Term::from_field_text(text_field, "b");
index_writer.delete_term(term);
index_writer.commit()?;
index_writer.wait_merging_threads()?;
let reader = index.reader()?;
assert_eq!(reader.searcher().num_docs(), 0);
let seg_ids = index.searchable_segment_ids()?;
assert!(seg_ids.is_empty());
reader.reload()?;
assert_eq!(reader.searcher().num_docs(), 0);
// empty segments should be erased
assert!(index.searchable_segment_metas()?.is_empty());
assert!(reader.searcher().segment_readers().is_empty());
Ok(())
}
#[test]
fn delete_all_docs() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
@@ -703,19 +799,19 @@ mod tests {
let mut index_writer = index.writer_for_tests()?;
for _ in 0..100 {
index_writer.add_document(doc!(text_field=>"a"));
index_writer.add_document(doc!(text_field=>"b"));
index_writer.add_document(doc!(text_field=>"a"))?;
index_writer.add_document(doc!(text_field=>"b"))?;
}
index_writer.commit()?;
for _ in 0..100 {
index_writer.add_document(doc!(text_field=>"c"));
index_writer.add_document(doc!(text_field=>"d"));
index_writer.add_document(doc!(text_field=>"c"))?;
index_writer.add_document(doc!(text_field=>"d"))?;
}
index_writer.commit()?;
index_writer.add_document(doc!(text_field=>"e"));
index_writer.add_document(doc!(text_field=>"f"));
index_writer.add_document(doc!(text_field=>"e"))?;
index_writer.add_document(doc!(text_field=>"f"))?;
index_writer.commit()?;
let seg_ids = index.searchable_segment_ids()?;
@@ -755,8 +851,8 @@ mod tests {
// writing the segment
let mut index_writer = index.writer_for_tests()?;
for _ in 0..100 {
index_writer.add_document(doc!(text_field=>"a"));
index_writer.add_document(doc!(text_field=>"b"));
index_writer.add_document(doc!(text_field=>"a"))?;
index_writer.add_document(doc!(text_field=>"b"))?;
}
index_writer.commit()?;
@@ -782,22 +878,22 @@ mod tests {
// writing two segments
let mut index_writer = index.writer_for_tests()?;
for _ in 0..100 {
index_writer.add_document(doc!(text_field=>"fizz"));
index_writer.add_document(doc!(text_field=>"buzz"));
index_writer.add_document(doc!(text_field=>"fizz"))?;
index_writer.add_document(doc!(text_field=>"buzz"))?;
}
index_writer.commit()?;
for _ in 0..1000 {
index_writer.add_document(doc!(text_field=>"foo"));
index_writer.add_document(doc!(text_field=>"bar"));
index_writer.add_document(doc!(text_field=>"foo"))?;
index_writer.add_document(doc!(text_field=>"bar"))?;
}
index_writer.commit()?;
indices.push(index);
}
assert_eq!(indices.len(), 3);
let output_directory = RamDirectory::default();
let index = merge_segments(&indices, output_directory)?;
let output_directory: Box<dyn Directory> = Box::new(RamDirectory::default());
let index = merge_indices(&indices, output_directory)?;
assert_eq!(index.schema(), schema);
let segments = index.searchable_segments()?;
@@ -811,7 +907,7 @@ mod tests {
#[test]
fn test_merge_empty_indices_array() {
let merge_result = merge_segments(&[], RamDirectory::default());
let merge_result = merge_indices(&[], RamDirectory::default());
assert!(merge_result.is_err());
}
@@ -822,7 +918,7 @@ mod tests {
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"some text"));
index_writer.add_document(doc!(text_field=>"some text"))?;
index_writer.commit()?;
index
};
@@ -832,15 +928,197 @@ mod tests {
let body_field = schema_builder.add_text_field("body", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(body_field=>"some body"));
index_writer.add_document(doc!(body_field=>"some body"))?;
index_writer.commit()?;
index
};
// mismatched schema index list
let result = merge_segments(&[first_index, second_index], RamDirectory::default());
let result = merge_indices(&[first_index, second_index], RamDirectory::default());
assert!(result.is_err());
Ok(())
}
#[test]
fn test_merge_filtered_segments() -> crate::Result<()> {
let first_index = {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"some text 1"))?;
index_writer.add_document(doc!(text_field=>"some text 2"))?;
index_writer.commit()?;
index
};
let second_index = {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"some text 3"))?;
index_writer.add_document(doc!(text_field=>"some text 4"))?;
index_writer.delete_term(Term::from_field_text(text_field, "4"));
index_writer.commit()?;
index
};
let mut segments: Vec<Segment> = Vec::new();
segments.extend(first_index.searchable_segments()?);
segments.extend(second_index.searchable_segments()?);
let target_settings = first_index.settings().clone();
let filter_segment_1 = AliveBitSet::for_test_from_deleted_docs(&[1], 2);
let filter_segment_2 = AliveBitSet::for_test_from_deleted_docs(&[0], 2);
let filter_segments = vec![Some(filter_segment_1), Some(filter_segment_2)];
let merged_index = merge_filtered_segments(
&segments,
target_settings,
filter_segments,
RamDirectory::default(),
)?;
let segments = merged_index.searchable_segments()?;
assert_eq!(segments.len(), 1);
let segment_metas = segments[0].meta();
assert_eq!(segment_metas.num_deleted_docs(), 0);
assert_eq!(segment_metas.num_docs(), 1);
Ok(())
}
#[test]
fn test_merge_single_filtered_segments() -> crate::Result<()> {
let first_index = {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"test text"))?;
index_writer.add_document(doc!(text_field=>"some text 2"))?;
index_writer.add_document(doc!(text_field=>"some text 3"))?;
index_writer.add_document(doc!(text_field=>"some text 4"))?;
index_writer.delete_term(Term::from_field_text(text_field, "4"));
index_writer.commit()?;
index
};
let mut segments: Vec<Segment> = Vec::new();
segments.extend(first_index.searchable_segments()?);
let target_settings = first_index.settings().clone();
let filter_segment = AliveBitSet::for_test_from_deleted_docs(&[0], 4);
let filter_segments = vec![Some(filter_segment)];
let index = merge_filtered_segments(
&segments,
target_settings,
filter_segments,
RamDirectory::default(),
)?;
let segments = index.searchable_segments()?;
assert_eq!(segments.len(), 1);
let segment_metas = segments[0].meta();
assert_eq!(segment_metas.num_deleted_docs(), 0);
assert_eq!(segment_metas.num_docs(), 2);
let searcher = index.reader()?.searcher();
{
let text_field = index.schema().get_field("text").unwrap();
let do_search = |term: &str| {
let query = QueryParser::for_index(&index, vec![text_field])
.parse_query(term)
.unwrap();
let top_docs: Vec<(f32, DocAddress)> =
searcher.search(&query, &TopDocs::with_limit(3)).unwrap();
top_docs.iter().map(|el| el.1.doc_id).collect::<Vec<_>>()
};
assert_eq!(do_search("test"), vec![] as Vec<u32>);
assert_eq!(do_search("text"), vec![0, 1]);
}
Ok(())
}
#[test]
fn test_apply_doc_id_filter_in_merger() -> crate::Result<()> {
let first_index = {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"some text 1"))?;
index_writer.add_document(doc!(text_field=>"some text 2"))?;
index_writer.add_document(doc!(text_field=>"some text 3"))?;
index_writer.add_document(doc!(text_field=>"some text 4"))?;
index_writer.delete_term(Term::from_field_text(text_field, "4"));
index_writer.commit()?;
index
};
let mut segments: Vec<Segment> = Vec::new();
segments.extend(first_index.searchable_segments()?);
let target_settings = first_index.settings().clone();
{
let filter_segment = AliveBitSet::for_test_from_deleted_docs(&[1], 4);
let filter_segments = vec![Some(filter_segment)];
let target_schema = segments[0].schema();
let merged_index = Index::create(
RamDirectory::default(),
target_schema.clone(),
target_settings.clone(),
)?;
let merger: IndexMerger = IndexMerger::open_with_custom_alive_set(
merged_index.schema(),
merged_index.settings().clone(),
&segments[..],
filter_segments,
)?;
let doc_ids_alive: Vec<_> = merger.readers[0].doc_ids_alive().collect();
assert_eq!(doc_ids_alive, vec![0, 2]);
}
{
let filter_segments = vec![None];
let target_schema = segments[0].schema();
let merged_index = Index::create(
RamDirectory::default(),
target_schema.clone(),
target_settings.clone(),
)?;
let merger: IndexMerger = IndexMerger::open_with_custom_alive_set(
merged_index.schema(),
merged_index.settings().clone(),
&segments[..],
filter_segments,
)?;
let doc_ids_alive: Vec<_> = merger.readers[0].doc_ids_alive().collect();
assert_eq!(doc_ids_alive, vec![0, 1, 2]);
}
Ok(())
}
}

121
src/lib.rs
View File

@@ -10,6 +10,7 @@
)]
#![doc(test(attr(allow(unused_variables), deny(warnings))))]
#![warn(missing_docs)]
#![allow(clippy::len_without_is_empty)]
//! # `tantivy`
//!
@@ -62,7 +63,7 @@
//! body => "He was an old man who fished alone in a skiff in \
//! the Gulf Stream and he had gone eighty-four days \
//! now without taking a fish."
//! ));
//! ))?;
//!
//! // We need to call .commit() explicitly to force the
//! // index_writer to finish processing the documents in the queue,
@@ -103,7 +104,7 @@
//! A good place for you to get started is to check out
//! the example code (
//! [literate programming](https://tantivy-search.github.io/examples/basic_search.html) /
//! [source code](https://github.com/tantivy-search/tantivy/blob/main/examples/basic_search.rs))
//! [source code](https://github.com/quickwit-inc/tantivy/blob/main/examples/basic_search.rs))
#[cfg_attr(test, macro_use)]
extern crate serde_json;
@@ -169,9 +170,11 @@ pub use crate::core::{
};
pub use crate::core::{InvertedIndexReader, SegmentReader};
pub use crate::directory::Directory;
pub use crate::indexer::merge_segments;
pub use crate::indexer::demuxer::*;
pub use crate::indexer::merge_filtered_segments;
pub use crate::indexer::merge_indices;
pub use crate::indexer::operation::UserOperation;
pub use crate::indexer::IndexWriter;
pub use crate::indexer::{IndexWriter, PreparedCommit};
pub use crate::postings::Postings;
pub use crate::reader::LeasedItem;
pub use crate::schema::{Document, Term};
@@ -376,24 +379,22 @@ pub mod tests {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_from_tempdir(schema).unwrap();
let index = Index::create_from_tempdir(schema)?;
// writing the segment
let mut index_writer = index.writer_for_tests()?;
{
// writing the segment
let mut index_writer = index.writer_for_tests()?;
{
let doc = doc!(text_field=>"af b");
index_writer.add_document(doc);
}
{
let doc = doc!(text_field=>"a b c");
index_writer.add_document(doc);
}
{
let doc = doc!(text_field=>"a b c d");
index_writer.add_document(doc);
}
assert!(index_writer.commit().is_ok());
let doc = doc!(text_field=>"af b");
index_writer.add_document(doc)?;
}
{
let doc = doc!(text_field=>"a b c");
index_writer.add_document(doc)?;
}
{
let doc = doc!(text_field=>"a b c d");
index_writer.add_document(doc)?;
}
index_writer.commit()?;
Ok(())
}
@@ -403,12 +404,12 @@ pub mod tests {
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"a b c"));
index_writer.add_document(doc!(text_field=>"a b c"))?;
index_writer.commit()?;
index_writer.add_document(doc!(text_field=>"a"));
index_writer.add_document(doc!(text_field=>"a a"));
index_writer.add_document(doc!(text_field=>"a"))?;
index_writer.add_document(doc!(text_field=>"a a"))?;
index_writer.commit()?;
index_writer.add_document(doc!(text_field=>"c"));
index_writer.add_document(doc!(text_field=>"c"))?;
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
@@ -430,7 +431,7 @@ pub mod tests {
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"a b c"));
index_writer.add_document(doc!(text_field=>"a b c"))?;
index_writer.commit()?;
let index_reader = index.reader()?;
let searcher = index_reader.searcher();
@@ -452,9 +453,9 @@ pub mod tests {
let text_field = schema_builder.add_text_field("text", TEXT);
let index = Index::create_in_ram(schema_builder.build());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"a b c"));
index_writer.add_document(doc!());
index_writer.add_document(doc!(text_field=>"a b"));
index_writer.add_document(doc!(text_field=>"a b c"))?;
index_writer.add_document(doc!())?;
index_writer.add_document(doc!(text_field=>"a b"))?;
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
@@ -496,20 +497,20 @@ pub mod tests {
// writing the segment
let mut index_writer = index.writer_for_tests()?;
// 0
index_writer.add_document(doc!(text_field=>"a b"));
index_writer.add_document(doc!(text_field=>"a b"))?;
// 1
index_writer.add_document(doc!(text_field=>" a c"));
index_writer.add_document(doc!(text_field=>" a c"))?;
// 2
index_writer.add_document(doc!(text_field=>" b c"));
index_writer.add_document(doc!(text_field=>" b c"))?;
// 3
index_writer.add_document(doc!(text_field=>" b d"));
index_writer.add_document(doc!(text_field=>" b d"))?;
index_writer.delete_term(Term::from_field_text(text_field, "c"));
index_writer.delete_term(Term::from_field_text(text_field, "a"));
// 4
index_writer.add_document(doc!(text_field=>" b c"));
index_writer.add_document(doc!(text_field=>" b c"))?;
// 5
index_writer.add_document(doc!(text_field=>" a"));
index_writer.add_document(doc!(text_field=>" a"))?;
index_writer.commit()?;
}
{
@@ -543,7 +544,7 @@ pub mod tests {
// writing the segment
let mut index_writer = index.writer_for_tests()?;
// 0
index_writer.add_document(doc!(text_field=>"a b"));
index_writer.add_document(doc!(text_field=>"a b"))?;
// 1
index_writer.delete_term(Term::from_field_text(text_field, "c"));
index_writer.rollback()?;
@@ -579,7 +580,7 @@ pub mod tests {
{
// writing the segment
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"a b"));
index_writer.add_document(doc!(text_field=>"a b"))?;
index_writer.delete_term(Term::from_field_text(text_field, "c"));
index_writer.rollback()?;
index_writer.delete_term(Term::from_field_text(text_field, "a"));
@@ -629,7 +630,7 @@ pub mod tests {
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(field=>1u64));
index_writer.add_document(doc!(field=>1u64))?;
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
@@ -653,7 +654,7 @@ pub mod tests {
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
let negative_val = -1i64;
index_writer.add_document(doc!(value_field => negative_val));
index_writer.add_document(doc!(value_field => negative_val))?;
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
@@ -677,7 +678,7 @@ pub mod tests {
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
let val = std::f64::consts::PI;
index_writer.add_document(doc!(value_field => val));
index_writer.add_document(doc!(value_field => val))?;
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
@@ -700,7 +701,7 @@ pub mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"a"));
index_writer.add_document(doc!(text_field=>"a"))?;
assert!(index_writer.commit().is_ok());
let reader = index.reader()?;
let searcher = reader.searcher();
@@ -723,14 +724,14 @@ pub mod tests {
// writing the segment
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"63"));
index_writer.add_document(doc!(text_field=>"70"));
index_writer.add_document(doc!(text_field=>"34"));
index_writer.add_document(doc!(text_field=>"1"));
index_writer.add_document(doc!(text_field=>"38"));
index_writer.add_document(doc!(text_field=>"33"));
index_writer.add_document(doc!(text_field=>"40"));
index_writer.add_document(doc!(text_field=>"17"));
index_writer.add_document(doc!(text_field=>"63"))?;
index_writer.add_document(doc!(text_field=>"70"))?;
index_writer.add_document(doc!(text_field=>"34"))?;
index_writer.add_document(doc!(text_field=>"1"))?;
index_writer.add_document(doc!(text_field=>"38"))?;
index_writer.add_document(doc!(text_field=>"33"))?;
index_writer.add_document(doc!(text_field=>"40"))?;
index_writer.add_document(doc!(text_field=>"17"))?;
index_writer.delete_term(Term::from_field_text(text_field, "38"));
index_writer.delete_term(Term::from_field_text(text_field, "34"));
index_writer.commit()?;
@@ -748,7 +749,7 @@ pub mod tests {
{
// writing the segment
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"af af af bc bc"));
index_writer.add_document(doc!(text_field=>"af af af bc bc"))?;
index_writer.commit()?;
}
{
@@ -780,9 +781,9 @@ pub mod tests {
let reader = index.reader()?;
// writing the segment
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"af af af b"));
index_writer.add_document(doc!(text_field=>"a b c"));
index_writer.add_document(doc!(text_field=>"a b c d"));
index_writer.add_document(doc!(text_field=>"af af af b"))?;
index_writer.add_document(doc!(text_field=>"a b c"))?;
index_writer.add_document(doc!(text_field=>"a b c d"))?;
index_writer.commit()?;
reader.reload()?;
@@ -844,9 +845,9 @@ pub mod tests {
assert_eq!(reader.searcher().num_docs(), 0u64);
// writing the segment
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"af b"));
index_writer.add_document(doc!(text_field=>"a b c"));
index_writer.add_document(doc!(text_field=>"a b c d"));
index_writer.add_document(doc!(text_field=>"af b"))?;
index_writer.add_document(doc!(text_field=>"a b c"))?;
index_writer.add_document(doc!(text_field=>"a b c d"))?;
index_writer.commit()?;
reader.reload()?;
assert_eq!(reader.searcher().num_docs(), 3u64);
@@ -886,7 +887,7 @@ pub mod tests {
{
let document =
doc!(fast_field_unsigned => 4u64, fast_field_signed=>4i64, fast_field_float=>4f64);
index_writer.add_document(document);
index_writer.add_document(document)?;
index_writer.commit()?;
}
let reader = index.reader()?;
@@ -953,7 +954,7 @@ pub mod tests {
index_writer.set_merge_policy(Box::new(NoMergePolicy));
for doc_id in 0u64..DOC_COUNT {
index_writer.add_document(doc!(id => doc_id));
index_writer.add_document(doc!(id => doc_id))?;
}
index_writer.commit()?;
@@ -970,7 +971,7 @@ pub mod tests {
index_writer.delete_term(Term::from_field_u64(id, doc_id));
index_writer.commit()?;
index_reader.reload()?;
index_writer.add_document(doc!(id => doc_id));
index_writer.add_document(doc!(id => doc_id))?;
index_writer.commit()?;
index_reader.reload()?;
let searcher = index_reader.searcher();
@@ -1005,8 +1006,8 @@ pub mod tests {
let index = Index::create_in_dir(&index_path, schema)?;
let mut writer = index.writer(50_000_000)?;
for _ in 0..5000 {
writer.add_document(doc!(body => "foo"));
writer.add_document(doc!(body => "boo"));
writer.add_document(doc!(body => "foo"))?;
writer.add_document(doc!(body => "boo"))?;
}
writer.commit()?;
assert!(index.validate_checksum()?.is_empty());

29
src/postings/block_search.rs
View File

@@ -1,14 +1,5 @@
use crate::postings::compression::COMPRESSION_BLOCK_SIZE;
unsafe fn binary_search_step(ptr: *const u32, target: u32, half_size: isize) -> *const u32 {
let mid = ptr.offset(half_size);
if *mid < target {
mid.offset(1)
} else {
ptr
}
}
/// Search the first index containing an element greater or equal to
/// the target.
///
@@ -30,18 +21,16 @@ unsafe fn binary_search_step(ptr: *const u32, target: u32, half_size: isize) ->
/// end of the last block for instance.
/// - The target is assumed smaller or equal to the last element of the block.
pub fn branchless_binary_search(arr: &[u32; COMPRESSION_BLOCK_SIZE], target: u32) -> usize {
let start_ptr: *const u32 = &arr[0] as *const u32;
unsafe {
let mut ptr = start_ptr;
ptr = binary_search_step(ptr, target, 63);
ptr = binary_search_step(ptr, target, 31);
ptr = binary_search_step(ptr, target, 15);
ptr = binary_search_step(ptr, target, 7);
ptr = binary_search_step(ptr, target, 3);
ptr = binary_search_step(ptr, target, 1);
let extra = if *ptr < target { 1 } else { 0 };
(ptr.offset_from(start_ptr) as usize) + extra
let mut start = 0;
let mut len = arr.len();
for _ in 0..7 {
len /= 2;
let pivot = unsafe { *arr.get_unchecked(start + len - 1) };
if pivot < target {
start += len;
}
}
start
}
#[cfg(test)]

41
src/postings/block_segment_postings.rs
View File

@@ -393,8 +393,8 @@ mod tests {
}
#[test]
fn test_block_segment_postings() {
let mut block_segments = build_block_postings(&(0..100_000).collect::<Vec<u32>>());
fn test_block_segment_postings() -> crate::Result<()> {
let mut block_segments = build_block_postings(&(0..100_000).collect::<Vec<u32>>())?;
let mut offset: u32 = 0u32;
// checking that the `doc_freq` is correct
assert_eq!(block_segments.doc_freq(), 100_000);
@@ -409,16 +409,17 @@ mod tests {
offset += block.len() as u32;
block_segments.advance();
}
Ok(())
}
#[test]
fn test_skip_right_at_new_block() {
fn test_skip_right_at_new_block() -> crate::Result<()> {
let mut doc_ids = (0..128).collect::<Vec<u32>>();
// 128 is missing
doc_ids.push(129);
doc_ids.push(130);
{
let block_segments = build_block_postings(&doc_ids);
let block_segments = build_block_postings(&doc_ids)?;
let mut docset = SegmentPostings::from_block_postings(block_segments, None);
assert_eq!(docset.seek(128), 129);
assert_eq!(docset.doc(), 129);
@@ -427,7 +428,7 @@ mod tests {
assert_eq!(docset.advance(), TERMINATED);
}
{
let block_segments = build_block_postings(&doc_ids);
let block_segments = build_block_postings(&doc_ids).unwrap();
let mut docset = SegmentPostings::from_block_postings(block_segments, None);
assert_eq!(docset.seek(129), 129);
assert_eq!(docset.doc(), 129);
@@ -436,46 +437,47 @@ mod tests {
assert_eq!(docset.advance(), TERMINATED);
}
{
let block_segments = build_block_postings(&doc_ids);
let block_segments = build_block_postings(&doc_ids)?;
let mut docset = SegmentPostings::from_block_postings(block_segments, None);
assert_eq!(docset.doc(), 0);
assert_eq!(docset.seek(131), TERMINATED);
assert_eq!(docset.doc(), TERMINATED);
}
Ok(())
}
fn build_block_postings(docs: &[DocId]) -> BlockSegmentPostings {
fn build_block_postings(docs: &[DocId]) -> crate::Result<BlockSegmentPostings> {
let mut schema_builder = Schema::builder();
let int_field = schema_builder.add_u64_field("id", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
let mut last_doc = 0u32;
for &doc in docs {
for _ in last_doc..doc {
index_writer.add_document(doc!(int_field=>1u64));
index_writer.add_document(doc!(int_field=>1u64))?;
}
index_writer.add_document(doc!(int_field=>0u64));
index_writer.add_document(doc!(int_field=>0u64))?;
last_doc = doc + 1;
}
index_writer.commit().unwrap();
let searcher = index.reader().unwrap().searcher();
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let segment_reader = searcher.segment_reader(0);
let inverted_index = segment_reader.inverted_index(int_field).unwrap();
let term = Term::from_field_u64(int_field, 0u64);
let term_info = inverted_index.get_term_info(&term).unwrap().unwrap();
inverted_index
.read_block_postings_from_terminfo(&term_info, IndexRecordOption::Basic)
.unwrap()
let term_info = inverted_index.get_term_info(&term)?.unwrap();
let block_postings = inverted_index
.read_block_postings_from_terminfo(&term_info, IndexRecordOption::Basic)?;
Ok(block_postings)
}
#[test]
fn test_block_segment_postings_seek() {
fn test_block_segment_postings_seek() -> crate::Result<()> {
let mut docs = vec![0];
for i in 0..1300 {
docs.push((i * i / 100) + i);
}
let mut block_postings = build_block_postings(&docs[..]);
let mut block_postings = build_block_postings(&docs[..])?;
for i in &[0, 424, 10000] {
block_postings.seek(*i);
let docs = block_postings.docs();
@@ -484,6 +486,7 @@ mod tests {
}
block_postings.seek(100_000);
assert_eq!(block_postings.doc(COMPRESSION_BLOCK_SIZE - 1), TERMINATED);
Ok(())
}
#[test]
@@ -497,7 +500,7 @@ mod tests {
// the other containing odd numbers.
for i in 0..6 {
let doc = doc!(int_field=> (i % 2) as u64);
index_writer.add_document(doc);
index_writer.add_document(doc)?;
}
index_writer.commit()?;
let searcher = index.reader()?.searcher();

100
src/postings/mod.rs
View File

@@ -87,12 +87,12 @@ pub mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(title => r#"abc abc abc"#));
index_writer.add_document(doc!(title => r#"abc be be be be abc"#));
index_writer.add_document(doc!(title => r#"abc abc abc"#))?;
index_writer.add_document(doc!(title => r#"abc be be be be abc"#))?;
for _ in 0..1_000 {
index_writer.add_document(doc!(title => r#"abc abc abc"#));
index_writer.add_document(doc!(title => r#"abc abc abc"#))?;
}
index_writer.add_document(doc!(title => r#"abc be be be be abc"#));
index_writer.add_document(doc!(title => r#"abc be be be be abc"#))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
@@ -154,10 +154,7 @@ pub mod tests {
}
#[test]
pub fn test_drop_token_that_are_too_long() -> crate::Result<()> {
let ok_token_text: String = "A".repeat(MAX_TOKEN_LEN);
let mut exceeding_token_text: String = "A".repeat(MAX_TOKEN_LEN + 1);
exceeding_token_text.push_str(" hello");
pub fn test_index_max_length_token() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let text_options = TextOptions::default().set_indexing_options(
TextFieldIndexing::default()
@@ -170,33 +167,54 @@ pub mod tests {
index
.tokenizers()
.register("simple_no_truncation", SimpleTokenizer);
let reader = index.reader().unwrap();
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.set_merge_policy(Box::new(NoMergePolicy));
{
index_writer.add_document(doc!(text_field=>exceeding_token_text));
index_writer.commit().unwrap();
reader.reload().unwrap();
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(0u32);
let inverted_index = segment_reader.inverted_index(text_field)?;
assert_eq!(inverted_index.terms().num_terms(), 1);
let mut bytes = vec![];
assert!(inverted_index.terms().ord_to_term(0, &mut bytes)?);
assert_eq!(&bytes, b"hello");
}
{
index_writer.add_document(doc!(text_field=>ok_token_text.clone()));
index_writer.commit().unwrap();
reader.reload().unwrap();
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(1u32);
let inverted_index = segment_reader.inverted_index(text_field)?;
assert_eq!(inverted_index.terms().num_terms(), 1);
let mut bytes = vec![];
assert!(inverted_index.terms().ord_to_term(0, &mut bytes)?);
assert_eq!(&bytes[..], ok_token_text.as_bytes());
}
let reader = index.reader()?;
let mut index_writer = index.writer_for_tests()?;
let ok_token_text: String = "A".repeat(MAX_TOKEN_LEN);
index_writer.add_document(doc!(text_field=>ok_token_text.clone()))?;
index_writer.commit()?;
reader.reload()?;
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(0u32);
let inverted_index = segment_reader.inverted_index(text_field)?;
assert_eq!(inverted_index.terms().num_terms(), 1);
let mut bytes = vec![];
assert!(inverted_index.terms().ord_to_term(0, &mut bytes)?);
assert_eq!(&bytes[..], ok_token_text.as_bytes());
Ok(())
}
#[test]
pub fn test_drop_token_that_are_too_long() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let text_options = TextOptions::default().set_indexing_options(
TextFieldIndexing::default()
.set_index_option(IndexRecordOption::WithFreqsAndPositions)
.set_tokenizer("simple_no_truncation"),
);
let text_field = schema_builder.add_text_field("text", text_options);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
index
.tokenizers()
.register("simple_no_truncation", SimpleTokenizer);
let reader = index.reader()?;
let mut index_writer = index.writer_for_tests()?;
let mut exceeding_token_text: String = "A".repeat(MAX_TOKEN_LEN + 1);
exceeding_token_text.push_str(" hello");
index_writer.add_document(doc!(text_field=>exceeding_token_text))?;
index_writer.commit()?;
reader.reload()?;
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(0u32);
let inverted_index = segment_reader.inverted_index(text_field)?;
assert_eq!(inverted_index.terms().num_terms(), 1);
let mut bytes = vec![];
assert!(inverted_index.terms().ord_to_term(0, &mut bytes)?);
assert_eq!(&bytes, b"hello");
Ok(())
}
@@ -315,13 +333,13 @@ pub mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(text_field => "g b b d c g c"));
index_writer.add_document(doc!(text_field => "g a b b a d c g c"));
assert!(index_writer.commit().is_ok());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field => "g b b d c g c"))?;
index_writer.add_document(doc!(text_field => "g a b b a d c g c"))?;
index_writer.commit()?;
}
let term_a = Term::from_field_text(text_field, "a");
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
let segment_reader = searcher.segment_reader(0);
let mut postings = segment_reader
.inverted_index(text_field)?
@@ -350,7 +368,7 @@ pub mod tests {
let mut index_writer = index.writer_for_tests()?;
for i in 0u64..num_docs as u64 {
let doc = doc!(value_field => 2u64, value_field => i % 2u64);
index_writer.add_document(doc);
index_writer.add_document(doc)?;
}
assert!(index_writer.commit().is_ok());
}

5
src/postings/postings_writer.rs
View File

@@ -33,7 +33,7 @@ fn posting_from_field_entry(field_entry: &FieldEntry) -> Box<dyn PostingsWriter>
SpecializedPostingsWriter::<TfAndPositionRecorder>::new_boxed()
}
})
.unwrap_or_else(|| SpecializedPostingsWriter::<NothingRecorder>::new_boxed()),
.unwrap_or_else(SpecializedPostingsWriter::<NothingRecorder>::new_boxed),
FieldType::U64(_)
| FieldType::I64(_)
| FieldType::F64(_)
@@ -133,7 +133,8 @@ impl MultiFieldPostingsWriter {
doc_id_map: Option<&DocIdMapping>,
) -> crate::Result<HashMap<Field, FnvHashMap<UnorderedTermId, TermOrdinal>>> {
let mut term_offsets: Vec<(&[u8], Addr, UnorderedTermId)> =
self.term_index.iter().collect();
Vec::with_capacity(self.term_index.len());
term_offsets.extend(self.term_index.iter());
term_offsets.sort_unstable_by_key(|&(k, _, _)| k);
let mut unordered_term_mappings: HashMap<Field, FnvHashMap<UnorderedTermId, TermOrdinal>> =

15
src/postings/segment_postings.rs
View File

@@ -1,5 +1,5 @@
use crate::docset::DocSet;
use crate::fastfield::DeleteBitSet;
use crate::fastfield::AliveBitSet;
use crate::positions::PositionReader;
use crate::postings::branchless_binary_search;
use crate::postings::compression::COMPRESSION_BLOCK_SIZE;
@@ -34,7 +34,7 @@ impl SegmentPostings {
///
/// This method will clone and scan through the posting lists.
/// (this is a rather expensive operation).
pub fn doc_freq_given_deletes(&self, delete_bitset: &DeleteBitSet) -> u32 {
pub fn doc_freq_given_deletes(&self, alive_bitset: &AliveBitSet) -> u32 {
let mut docset = self.clone();
let mut doc_freq = 0;
loop {
@@ -42,7 +42,7 @@ impl SegmentPostings {
if doc == TERMINATED {
return doc_freq;
}
if delete_bitset.is_alive(doc) {
if alive_bitset.is_alive(doc) {
doc_freq += 1u32;
}
docset.advance();
@@ -268,7 +268,7 @@ mod tests {
use common::HasLen;
use crate::docset::{DocSet, TERMINATED};
use crate::fastfield::DeleteBitSet;
use crate::fastfield::AliveBitSet;
use crate::postings::postings::Postings;
#[test]
@@ -296,9 +296,10 @@ mod tests {
fn test_doc_freq() {
let docs = SegmentPostings::create_from_docs(&[0, 2, 10]);
assert_eq!(docs.doc_freq(), 3);
let delete_bitset = DeleteBitSet::for_test(&[2], 12);
assert_eq!(docs.doc_freq_given_deletes(&delete_bitset), 2);
let all_deleted = DeleteBitSet::for_test(&[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], 12);
let alive_bitset = AliveBitSet::for_test_from_deleted_docs(&[2], 12);
assert_eq!(docs.doc_freq_given_deletes(&alive_bitset), 2);
let all_deleted =
AliveBitSet::for_test_from_deleted_docs(&[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], 12);
assert_eq!(docs.doc_freq_given_deletes(&all_deleted), 0);
}
}

4
src/postings/serializer.rs
View File

@@ -13,6 +13,7 @@ use crate::termdict::{TermDictionaryBuilder, TermOrdinal};
use crate::{DocId, Score};
use common::CountingWriter;
use common::{BinarySerializable, VInt};
use fail::fail_point;
use std::cmp::Ordering;
use std::io::{self, Write};
@@ -212,6 +213,9 @@ impl<'a> FieldSerializer<'a> {
/// If the current block is incomplete, it need to be encoded
/// using `VInt` encoding.
pub fn close_term(&mut self) -> io::Result<()> {
fail_point!("FieldSerializer::close_term", |msg: Option<String>| {
Err(io::Error::new(io::ErrorKind::Other, format!("{:?}", msg)))
});
if self.term_open {
self.postings_serializer
.close_term(self.current_term_info.doc_freq)?;

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

@@ -148,6 +148,10 @@ impl TermHashMap {
unordered_term_id
}
pub fn len(&self) -> usize {
self.len
}
pub fn iter(&self) -> Iter<'_> {
Iter {
inner: self.occupied.iter(),

42
src/query/all_query.rs
View File

@@ -78,29 +78,29 @@ mod tests {
use crate::schema::{Schema, TEXT};
use crate::Index;
fn create_test_index() -> Index {
fn create_test_index() -> crate::Result<Index> {
let mut schema_builder = Schema::builder();
let field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(field=>"aaa"));
index_writer.add_document(doc!(field=>"bbb"));
index_writer.commit().unwrap();
index_writer.add_document(doc!(field=>"ccc"));
index_writer.commit().unwrap();
index
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(field=>"aaa"))?;
index_writer.add_document(doc!(field=>"bbb"))?;
index_writer.commit()?;
index_writer.add_document(doc!(field=>"ccc"))?;
index_writer.commit()?;
Ok(index)
}
#[test]
fn test_all_query() {
let index = create_test_index();
let reader = index.reader().unwrap();
fn test_all_query() -> crate::Result<()> {
let index = create_test_index()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let weight = AllQuery.weight(&searcher, false).unwrap();
let weight = AllQuery.weight(&searcher, false)?;
{
let reader = searcher.segment_reader(0);
let mut scorer = weight.scorer(reader, 1.0).unwrap();
let mut scorer = weight.scorer(reader, 1.0)?;
assert_eq!(scorer.doc(), 0u32);
assert_eq!(scorer.advance(), 1u32);
assert_eq!(scorer.doc(), 1u32);
@@ -108,28 +108,30 @@ mod tests {
}
{
let reader = searcher.segment_reader(1);
let mut scorer = weight.scorer(reader, 1.0).unwrap();
let mut scorer = weight.scorer(reader, 1.0)?;
assert_eq!(scorer.doc(), 0u32);
assert_eq!(scorer.advance(), TERMINATED);
}
Ok(())
}
#[test]
fn test_all_query_with_boost() {
let index = create_test_index();
let reader = index.reader().unwrap();
fn test_all_query_with_boost() -> crate::Result<()> {
let index = create_test_index()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let weight = AllQuery.weight(&searcher, false).unwrap();
let weight = AllQuery.weight(&searcher, false)?;
let reader = searcher.segment_reader(0);
{
let mut scorer = weight.scorer(reader, 2.0).unwrap();
let mut scorer = weight.scorer(reader, 2.0)?;
assert_eq!(scorer.doc(), 0u32);
assert_eq!(scorer.score(), 2.0);
}
{
let mut scorer = weight.scorer(reader, 1.5).unwrap();
let mut scorer = weight.scorer(reader, 1.5)?;
assert_eq!(scorer.doc(), 0u32);
assert_eq!(scorer.score(), 1.5);
}
Ok(())
}
}

36
src/query/automaton_weight.rs
View File

@@ -92,16 +92,16 @@ mod tests {
use crate::Index;
use tantivy_fst::Automaton;
fn create_index() -> Index {
fn create_index() -> crate::Result<Index> {
let mut schema = Schema::builder();
let title = schema.add_text_field("title", STRING);
let index = Index::create_in_ram(schema.build());
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(title=>"abc"));
index_writer.add_document(doc!(title=>"bcd"));
index_writer.add_document(doc!(title=>"abcd"));
assert!(index_writer.commit().is_ok());
index
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(title=>"abc"))?;
index_writer.add_document(doc!(title=>"bcd"))?;
index_writer.add_document(doc!(title=>"abcd"))?;
index_writer.commit()?;
Ok(index)
}
#[derive(Clone, Copy)]
@@ -140,34 +140,32 @@ mod tests {
}
#[test]
fn test_automaton_weight() {
let index = create_index();
fn test_automaton_weight() -> crate::Result<()> {
let index = create_index()?;
let field = index.schema().get_field("title").unwrap();
let automaton_weight = AutomatonWeight::new(field, PrefixedByA);
let reader = index.reader().unwrap();
let reader = index.reader()?;
let searcher = reader.searcher();
let mut scorer = automaton_weight
.scorer(searcher.segment_reader(0u32), 1.0)
.unwrap();
let mut scorer = automaton_weight.scorer(searcher.segment_reader(0u32), 1.0)?;
assert_eq!(scorer.doc(), 0u32);
assert_eq!(scorer.score(), 1.0);
assert_eq!(scorer.advance(), 2u32);
assert_eq!(scorer.doc(), 2u32);
assert_eq!(scorer.score(), 1.0);
assert_eq!(scorer.advance(), TERMINATED);
Ok(())
}
#[test]
fn test_automaton_weight_boost() {
let index = create_index();
fn test_automaton_weight_boost() -> crate::Result<()> {
let index = create_index()?;
let field = index.schema().get_field("title").unwrap();
let automaton_weight = AutomatonWeight::new(field, PrefixedByA);
let reader = index.reader().unwrap();
let reader = index.reader()?;
let searcher = reader.searcher();
let mut scorer = automaton_weight
.scorer(searcher.segment_reader(0u32), 1.32)
.unwrap();
let mut scorer = automaton_weight.scorer(searcher.segment_reader(0u32), 1.32)?;
assert_eq!(scorer.doc(), 0u32);
assert_eq!(scorer.score(), 1.32);
Ok(())
}
}

94
src/query/boolean_query/block_wand.rs
View File

@@ -42,27 +42,39 @@ fn find_pivot_doc(
Some((before_pivot_len, pivot_len, pivot_doc))
}
// Before and after calling this method, scorers need to be sorted by their `.doc()`.
/// Advance the scorer with best score among the scorers[..pivot_len] to
/// the next doc candidate defined by the min of `last_doc_in_block + 1` for
/// scorer in scorers[..pivot_len] and `scorer.doc()` for scorer in scorers[pivot_len..].
/// Note: before and after calling this method, scorers need to be sorted by their `.doc()`.
fn block_max_was_too_low_advance_one_scorer(
scorers: &mut Vec<TermScorerWithMaxScore>,
pivot_len: usize,
) {
debug_assert!(is_sorted(scorers.iter().map(|scorer| scorer.doc())));
let mut scorer_to_seek = pivot_len - 1;
let mut doc_to_seek_after = scorers[scorer_to_seek].doc();
let mut global_max_score = scorers[scorer_to_seek].max_score;
let mut doc_to_seek_after = scorers[scorer_to_seek].last_doc_in_block();
for scorer_ord in (0..pivot_len - 1).rev() {
let scorer = &scorers[scorer_ord];
if scorer.last_doc_in_block() <= doc_to_seek_after {
doc_to_seek_after = scorer.last_doc_in_block();
}
if scorers[scorer_ord].max_score > global_max_score {
global_max_score = scorers[scorer_ord].max_score;
scorer_to_seek = scorer_ord;
}
}
// Add +1 to go to the next block unless we are already at the end.
if doc_to_seek_after != TERMINATED {
doc_to_seek_after += 1;
}
for scorer in &scorers[pivot_len..] {
if scorer.doc() <= doc_to_seek_after {
doc_to_seek_after = scorer.doc();
}
}
scorers[scorer_to_seek].seek(doc_to_seek_after + 1);
scorers[scorer_to_seek].seek(doc_to_seek_after);
restore_ordering(scorers, scorer_to_seek);
debug_assert!(is_sorted(scorers.iter().map(|scorer| scorer.doc())));
}
@@ -130,6 +142,9 @@ fn advance_all_scorers_on_pivot(term_scorers: &mut Vec<TermScorerWithMaxScore>,
term_scorers.sort_by_key(|scorer| scorer.doc());
}
/// Implements the WAND (Weak AND) algorithm for dynamic pruning
/// described in the paper "Faster Top-k Document Retrieval Using Block-Max Indexes".
/// Link: http://engineering.nyu.edu/~suel/papers/bmw.pdf
pub fn block_wand(
mut scorers: Vec<TermScorer>,
mut threshold: Score,
@@ -187,6 +202,7 @@ pub fn block_wand(
.iter_mut()
.map(|scorer| scorer.score())
.sum();
if score > threshold {
threshold = callback(pivot_doc, score);
}
@@ -195,6 +211,56 @@ pub fn block_wand(
}
}
/// Specialized version of [`block_wand`] for a single scorer.
/// In this case, the algorithm is simple and readable and faster (~ x3)
/// than the generic algorithm.
/// The algorithm behaves as follows:
/// - While we don't hit the end of the docset:
/// - While the block max score is under the `threshold`, go to the
/// next block.
/// - On a block, advance until the end and execute `callback``
/// when the doc score is greater or equal to the `threshold`.
pub fn block_wand_single_scorer(
mut scorer: TermScorer,
mut threshold: Score,
callback: &mut dyn FnMut(u32, Score) -> Score,
) {
let mut doc = scorer.doc();
loop {
// We position the scorer on a block that can reach
// the threshold.
while scorer.block_max_score() < threshold {
let last_doc_in_block = scorer.last_doc_in_block();
if last_doc_in_block == TERMINATED {
return;
}
doc = last_doc_in_block + 1;
scorer.shallow_seek(doc);
}
// Seek will effectively load that block.
doc = scorer.seek(doc);
if doc == TERMINATED {
break;
}
loop {
let score = scorer.score();
if score > threshold {
threshold = callback(doc, score);
}
debug_assert!(doc <= scorer.last_doc_in_block());
if doc == scorer.last_doc_in_block() {
break;
}
doc = scorer.advance();
if doc == TERMINATED {
return;
}
}
doc += 1;
scorer.shallow_seek(doc);
}
}
struct TermScorerWithMaxScore<'a> {
scorer: &'a mut TermScorer,
max_score: Score,
@@ -272,13 +338,14 @@ mod tests {
}
fn compute_checkpoints_for_each_pruning(
term_scorers: Vec<TermScorer>,
mut term_scorers: Vec<TermScorer>,
n: usize,
) -> Vec<(DocId, Score)> {
let mut heap: BinaryHeap<Float> = BinaryHeap::with_capacity(n);
let mut checkpoints: Vec<(DocId, Score)> = Vec::new();
let mut limit: Score = 0.0;
super::block_wand(term_scorers, Score::MIN, &mut |doc, score| {
let callback = &mut |doc, score| {
heap.push(Float(score));
if heap.len() > n {
heap.pop().unwrap();
@@ -290,7 +357,14 @@ mod tests {
checkpoints.push((doc, score));
}
limit
});
};
if term_scorers.len() == 1 {
let scorer = term_scorers.pop().unwrap();
super::block_wand_single_scorer(scorer, Score::MIN, callback);
} else {
super::block_wand(term_scorers, Score::MIN, callback);
}
checkpoints
}
@@ -424,6 +498,14 @@ mod tests {
}
}
proptest! {
#![proptest_config(ProptestConfig::with_cases(500))]
#[test]
fn test_block_wand_single_term_scorer((posting_lists, fieldnorms) in gen_term_scorers(1)) {
test_block_wand_aux(&posting_lists[..], &fieldnorms[..]);
}
}
#[test]
fn test_fn_reproduce_proptest() {
let postings_lists = &[

22
src/query/boolean_query/boolean_query.rs
View File

@@ -41,22 +41,22 @@ use std::collections::BTreeMap;
/// let mut index_writer = index.writer(3_000_000)?;
/// index_writer.add_document(doc!(
/// title => "The Name of the Wind",
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "The Diary of Muadib",
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "A Dairy Cow",
/// body => "hidden",
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "A Dairy Cow",
/// body => "found",
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "The Diary of a Young Girl",
/// ));
/// index_writer.commit().unwrap();
/// ))?;
/// index_writer.commit()?;
/// }
///
/// let reader = index.reader()?;
@@ -217,11 +217,11 @@ mod tests {
let text = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut writer = index.writer_for_tests().unwrap();
writer.add_document(doc!(text=>"b c"));
writer.add_document(doc!(text=>"a c"));
writer.add_document(doc!(text=>"a b"));
writer.add_document(doc!(text=>"a d"));
let mut writer = index.writer_for_tests()?;
writer.add_document(doc!(text=>"b c"))?;
writer.add_document(doc!(text=>"a c"))?;
writer.add_document(doc!(text=>"a b"))?;
writer.add_document(doc!(text=>"a d"))?;
writer.commit()?;
Ok(index)
}

127
src/query/boolean_query/mod.rs
View File

@@ -3,6 +3,7 @@ mod boolean_query;
mod boolean_weight;
pub(crate) use self::block_wand::block_wand;
pub(crate) use self::block_wand::block_wand_single_scorer;
pub use self::boolean_query::BooleanQuery;
#[cfg(test)]
@@ -25,72 +26,75 @@ mod tests {
use crate::Index;
use crate::{DocAddress, DocId, Score};
fn aux_test_helper() -> (Index, Field) {
fn aux_test_helper() -> crate::Result<(Index, Field)> {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
// writing the segment
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(text_field => "a b c"));
index_writer.add_document(doc!(text_field => "a c"));
index_writer.add_document(doc!(text_field => "b c"));
index_writer.add_document(doc!(text_field => "a b c d"));
index_writer.add_document(doc!(text_field => "d"));
assert!(index_writer.commit().is_ok());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field => "a b c"))?;
index_writer.add_document(doc!(text_field => "a c"))?;
index_writer.add_document(doc!(text_field => "b c"))?;
index_writer.add_document(doc!(text_field => "a b c d"))?;
index_writer.add_document(doc!(text_field => "d"))?;
index_writer.commit()?;
}
(index, text_field)
Ok((index, text_field))
}
#[test]
pub fn test_boolean_non_all_term_disjunction() {
let (index, text_field) = aux_test_helper();
pub fn test_boolean_non_all_term_disjunction() -> crate::Result<()> {
let (index, text_field) = aux_test_helper()?;
let query_parser = QueryParser::for_index(&index, vec![text_field]);
let query = query_parser.parse_query("(+a +b) d").unwrap();
let searcher = index.reader().unwrap().searcher();
assert_eq!(query.count(&searcher).unwrap(), 3);
let query = query_parser.parse_query("(+a +b) d")?;
let searcher = index.reader()?.searcher();
assert_eq!(query.count(&searcher)?, 3);
Ok(())
}
#[test]
pub fn test_boolean_single_must_clause() {
let (index, text_field) = aux_test_helper();
pub fn test_boolean_single_must_clause() -> crate::Result<()> {
let (index, text_field) = aux_test_helper()?;
let query_parser = QueryParser::for_index(&index, vec![text_field]);
let query = query_parser.parse_query("+a").unwrap();
let searcher = index.reader().unwrap().searcher();
let weight = query.weight(&searcher, true).unwrap();
let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0).unwrap();
let query = query_parser.parse_query("+a")?;
let searcher = index.reader()?.searcher();
let weight = query.weight(&searcher, true)?;
let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0)?;
assert!(scorer.is::<TermScorer>());
Ok(())
}
#[test]
pub fn test_boolean_termonly_intersection() {
let (index, text_field) = aux_test_helper();
pub fn test_boolean_termonly_intersection() -> crate::Result<()> {
let (index, text_field) = aux_test_helper()?;
let query_parser = QueryParser::for_index(&index, vec![text_field]);
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
{
let query = query_parser.parse_query("+a +b +c").unwrap();
let weight = query.weight(&searcher, true).unwrap();
let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0).unwrap();
let query = query_parser.parse_query("+a +b +c")?;
let weight = query.weight(&searcher, true)?;
let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0)?;
assert!(scorer.is::<Intersection<TermScorer>>());
}
{
let query = query_parser.parse_query("+a +(b c)").unwrap();
let weight = query.weight(&searcher, true).unwrap();
let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0).unwrap();
let query = query_parser.parse_query("+a +(b c)")?;
let weight = query.weight(&searcher, true)?;
let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0)?;
assert!(scorer.is::<Intersection<Box<dyn Scorer>>>());
}
Ok(())
}
#[test]
pub fn test_boolean_reqopt() {
let (index, text_field) = aux_test_helper();
pub fn test_boolean_reqopt() -> crate::Result<()> {
let (index, text_field) = aux_test_helper()?;
let query_parser = QueryParser::for_index(&index, vec![text_field]);
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
{
let query = query_parser.parse_query("+a b").unwrap();
let weight = query.weight(&searcher, true).unwrap();
let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0).unwrap();
let query = query_parser.parse_query("+a b")?;
let weight = query.weight(&searcher, true)?;
let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0)?;
assert!(scorer.is::<RequiredOptionalScorer<
Box<dyn Scorer>,
Box<dyn Scorer>,
@@ -98,16 +102,17 @@ mod tests {
>>());
}
{
let query = query_parser.parse_query("+a b").unwrap();
let weight = query.weight(&searcher, false).unwrap();
let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0).unwrap();
let query = query_parser.parse_query("+a b")?;
let weight = query.weight(&searcher, false)?;
let scorer = weight.scorer(searcher.segment_reader(0u32), 1.0)?;
assert!(scorer.is::<TermScorer>());
}
Ok(())
}
#[test]
pub fn test_boolean_query() {
let (index, text_field) = aux_test_helper();
pub fn test_boolean_query() -> crate::Result<()> {
let (index, text_field) = aux_test_helper()?;
let make_term_query = |text: &str| {
let term_query = TermQuery::new(
@@ -118,7 +123,7 @@ mod tests {
query
};
let reader = index.reader().unwrap();
let reader = index.reader()?;
let matching_docs = |boolean_query: &dyn Query| {
reader
@@ -165,11 +170,12 @@ mod tests {
let boolean_query = BooleanQuery::new(vec![(Occur::MustNot, make_term_query("d"))]);
assert_eq!(matching_docs(&boolean_query), Vec::<u32>::new());
}
Ok(())
}
#[test]
pub fn test_boolean_query_two_excluded() {
let (index, text_field) = aux_test_helper();
pub fn test_boolean_query_two_excluded() -> crate::Result<()> {
let (index, text_field) = aux_test_helper()?;
let make_term_query = |text: &str| {
let term_query = TermQuery::new(
@@ -180,7 +186,7 @@ mod tests {
query
};
let reader = index.reader().unwrap();
let reader = index.reader()?;
let matching_topdocs = |query: &dyn Query| {
reader
@@ -213,20 +219,21 @@ mod tests {
assert_eq!(top_doc, DocAddress::new(0, 4));
assert_eq!(top_score, score_doc_4);
}
Ok(())
}
#[test]
pub fn test_boolean_query_with_weight() {
pub fn test_boolean_query_with_weight() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(text_field => "a b c"));
index_writer.add_document(doc!(text_field => "a c"));
index_writer.add_document(doc!(text_field => "b c"));
assert!(index_writer.commit().is_ok());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field => "a b c"))?;
index_writer.add_document(doc!(text_field => "a c"))?;
index_writer.add_document(doc!(text_field => "b c"))?;
index_writer.commit()?;
}
let term_a: Box<dyn Query> = Box::new(TermQuery::new(
Term::from_field_text(text_field, "a"),
@@ -242,24 +249,21 @@ mod tests {
BooleanQuery::new(vec![(Occur::Should, term_a), (Occur::Should, term_b)]);
let boolean_weight = boolean_query.weight(&searcher, true).unwrap();
{
let mut boolean_scorer = boolean_weight
.scorer(searcher.segment_reader(0u32), 1.0)
.unwrap();
let mut boolean_scorer = boolean_weight.scorer(searcher.segment_reader(0u32), 1.0)?;
assert_eq!(boolean_scorer.doc(), 0u32);
assert_nearly_equals!(boolean_scorer.score(), 0.84163445);
}
{
let mut boolean_scorer = boolean_weight
.scorer(searcher.segment_reader(0u32), 2.0)
.unwrap();
let mut boolean_scorer = boolean_weight.scorer(searcher.segment_reader(0u32), 2.0)?;
assert_eq!(boolean_scorer.doc(), 0u32);
assert_nearly_equals!(boolean_scorer.score(), 1.6832689);
}
Ok(())
}
#[test]
pub fn test_intersection_score() {
let (index, text_field) = aux_test_helper();
pub fn test_intersection_score() -> crate::Result<()> {
let (index, text_field) = aux_test_helper()?;
let make_term_query = |text: &str| {
let term_query = TermQuery::new(
@@ -269,7 +273,7 @@ mod tests {
let query: Box<dyn Query> = Box::new(term_query);
query
};
let reader = index.reader().unwrap();
let reader = index.reader()?;
let score_docs = |boolean_query: &dyn Query| {
let fruit = reader
.searcher()
@@ -287,6 +291,7 @@ mod tests {
assert_nearly_equals!(scores[0], 0.977973);
assert_nearly_equals!(scores[1], 0.84699446);
}
Ok(())
}
#[test]
@@ -296,8 +301,8 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_with_num_threads(1, 5_000_000)?;
index_writer.add_document(doc!(text=>"a"));
index_writer.add_document(doc!(text=>"b"));
index_writer.add_document(doc!(text=>"a"))?;
index_writer.add_document(doc!(text=>"b"))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let term_a: Box<dyn Query> = Box::new(TermQuery::new(

19
src/query/boost_query.rs
View File

@@ -1,4 +1,4 @@
use crate::fastfield::DeleteBitSet;
use crate::fastfield::AliveBitSet;
use crate::query::explanation::does_not_match;
use crate::query::{Explanation, Query, Scorer, Weight};
use crate::{DocId, DocSet, Score, Searcher, SegmentReader, Term};
@@ -118,8 +118,8 @@ impl<S: Scorer> DocSet for BoostScorer<S> {
self.underlying.size_hint()
}
fn count(&mut self, delete_bitset: &DeleteBitSet) -> u32 {
self.underlying.count(delete_bitset)
fn count(&mut self, alive_bitset: &AliveBitSet) -> u32 {
self.underlying.count(alive_bitset)
}
fn count_including_deleted(&mut self) -> u32 {
@@ -141,19 +141,20 @@ mod tests {
use crate::{DocAddress, Document, Index};
#[test]
fn test_boost_query_explain() {
fn test_boost_query_explain() -> crate::Result<()> {
let schema = Schema::builder().build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(Document::new());
assert!(index_writer.commit().is_ok());
let reader = index.reader().unwrap();
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(Document::new())?;
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let query = BoostQuery::new(Box::new(AllQuery), 0.2);
let explanation = query.explain(&searcher, DocAddress::new(0, 0u32)).unwrap();
assert_eq!(
explanation.to_pretty_json(),
"{\n \"value\": 0.2,\n \"description\": \"Boost x0.2 of ...\",\n \"details\": [\n {\n \"value\": 1.0,\n \"description\": \"AllQuery\",\n \"context\": []\n }\n ],\n \"context\": []\n}"
)
);
Ok(())
}
}

65
src/query/fuzzy_query.rs
View File

@@ -67,17 +67,17 @@ static LEV_BUILDER: Lazy<HashMap<(u8, bool), LevenshteinAutomatonBuilder>> = Laz
/// let mut index_writer = index.writer(3_000_000)?;
/// index_writer.add_document(doc!(
/// title => "The Name of the Wind",
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "The Diary of Muadib",
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "A Dairy Cow",
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "The Diary of a Young Girl",
/// ));
/// index_writer.commit().unwrap();
/// ))?;
/// index_writer.commit()?;
/// }
/// let reader = index.reader()?;
/// let searcher = reader.searcher();
@@ -129,7 +129,7 @@ impl FuzzyTermQuery {
fn specialized_weight(&self) -> crate::Result<AutomatonWeight<DfaWrapper>> {
// LEV_BUILDER is a HashMap, whose `get` method returns an Option
match LEV_BUILDER.get(&(self.distance, false)) {
match LEV_BUILDER.get(&(self.distance, self.transposition_cost_one)) {
// Unwrap the option and build the Ok(AutomatonWeight)
Some(automaton_builder) => {
let automaton = if self.prefix {
@@ -164,6 +164,7 @@ impl Query for FuzzyTermQuery {
mod test {
use super::FuzzyTermQuery;
use crate::assert_nearly_equals;
use crate::collector::Count;
use crate::collector::TopDocs;
use crate::schema::Schema;
use crate::schema::TEXT;
@@ -171,32 +172,29 @@ mod test {
use crate::Term;
#[test]
pub fn test_fuzzy_term() {
pub fn test_fuzzy_term() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let country_field = schema_builder.add_text_field("country", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(
country_field => "japan",
));
))?;
index_writer.add_document(doc!(
country_field => "korea",
));
index_writer.commit().unwrap();
))?;
index_writer.commit()?;
}
let reader = index.reader().unwrap();
let reader = index.reader()?;
let searcher = reader.searcher();
// passes because Levenshtein distance is 1 (substitute 'o' with 'a')
{
let term = Term::from_field_text(country_field, "japon");
let fuzzy_query = FuzzyTermQuery::new(term, 1, true);
let top_docs = searcher
.search(&fuzzy_query, &TopDocs::with_limit(2))
.unwrap();
let top_docs = searcher.search(&fuzzy_query, &TopDocs::with_limit(2))?;
assert_eq!(top_docs.len(), 1, "Expected only 1 document");
let (score, _) = top_docs[0];
assert_nearly_equals!(1.0, score);
@@ -207,23 +205,44 @@ mod test {
let term = Term::from_field_text(country_field, "jap");
let fuzzy_query = FuzzyTermQuery::new(term, 1, true);
let top_docs = searcher
.search(&fuzzy_query, &TopDocs::with_limit(2))
.unwrap();
let top_docs = searcher.search(&fuzzy_query, &TopDocs::with_limit(2))?;
assert_eq!(top_docs.len(), 0, "Expected no document");
}
// passes because prefix Levenshtein distance is 0
{
let term = Term::from_field_text(country_field, "jap");
let fuzzy_query = FuzzyTermQuery::new_prefix(term, 1, true);
let top_docs = searcher
.search(&fuzzy_query, &TopDocs::with_limit(2))
.unwrap();
let top_docs = searcher.search(&fuzzy_query, &TopDocs::with_limit(2))?;
assert_eq!(top_docs.len(), 1, "Expected only 1 document");
let (score, _) = top_docs[0];
assert_nearly_equals!(1.0, score);
}
Ok(())
}
#[test]
pub fn test_fuzzy_term_transposition_cost_one() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let country_field = schema_builder.add_text_field("country", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(country_field => "japan"))?;
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let term_jaapn = Term::from_field_text(country_field, "jaapn");
{
let fuzzy_query_transposition = FuzzyTermQuery::new(term_jaapn.clone(), 1, true);
let count = searcher.search(&fuzzy_query_transposition, &Count)?;
assert_eq!(count, 1);
}
{
let fuzzy_query_transposition = FuzzyTermQuery::new(term_jaapn, 1, false);
let count = searcher.search(&fuzzy_query_transposition, &Count)?;
assert_eq!(count, 0);
}
Ok(())
}
}

70
src/query/intersection_two.rs
View File

@@ -1,70 +0,0 @@
use docset::DocSet;
use query::Scorer;
use DocId;
use Score;
use SkipResult;
/// Creates a `DocSet` that iterate through the intersection of two `DocSet`s.
pub struct IntersectionTwoTerms<TDocSet> {
left: TDocSet,
right: TDocSet
}
impl<TDocSet: DocSet> IntersectionTwoTerms<TDocSet> {
pub fn new(left: TDocSet, right: TDocSet) -> IntersectionTwoTerms<TDocSet> {
IntersectionTwoTerms {
left,
right
}
}
}
impl<TDocSet: DocSet> DocSet for IntersectionTwoTerms<TDocSet> {
fn advance(&mut self) -> bool {
let (left, right) = (&mut self.left, &mut self.right);
if !left.advance() {
return false;
}
let mut candidate = left.doc();
loop {
match right.skip_next(candidate) {
SkipResult::Reached => {
return true;
}
SkipResult::End => {
return false;
}
SkipResult::OverStep => {
candidate = right.doc();
}
}
match left.skip_next(candidate) {
SkipResult::Reached => {
return true;
}
SkipResult::End => {
return false;
}
SkipResult::OverStep => {
candidate = left.doc();
}
}
}
}
fn doc(&self) -> DocId {
self.left.doc()
}
fn size_hint(&self) -> u32 {
self.left.size_hint().min(self.right.size_hint())
}
}
impl<TScorer: Scorer> Scorer for IntersectionTwoTerms<TScorer> {
fn score(&mut self) -> Score {
self.left.score() + self.right.score()
}
}

37
src/query/more_like_this/query.rs
View File

@@ -61,19 +61,11 @@ impl Query for MoreLikeThisQuery {
}
/// The builder for more-like-this query
#[derive(Debug, Clone)]
#[derive(Debug, Clone, Default)]
pub struct MoreLikeThisQueryBuilder {
mlt: MoreLikeThis,
}
impl Default for MoreLikeThisQueryBuilder {
fn default() -> Self {
Self {
mlt: MoreLikeThis::default(),
}
}
}
impl MoreLikeThisQueryBuilder {
/// Sets the minimum document frequency.
///
@@ -184,20 +176,20 @@ mod tests {
use crate::DocAddress;
use crate::Index;
fn create_test_index() -> Index {
fn create_test_index() -> crate::Result<Index> {
let mut schema_builder = Schema::builder();
let title = schema_builder.add_text_field("title", TEXT);
let body = schema_builder.add_text_field("body", TEXT | STORED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(title => "aaa", body => "the old man and the sea"));
index_writer.add_document(doc!(title => "bbb", body => "an old man sailing on the sea"));
index_writer.add_document(doc!(title => "ccc", body=> "send this message to alice"));
index_writer.add_document(doc!(title => "ddd", body=> "a lady was riding and old bike"));
index_writer.add_document(doc!(title => "eee", body=> "Yes, my lady."));
index_writer.commit().unwrap();
index
index_writer.add_document(doc!(title => "aaa", body => "the old man and the sea"))?;
index_writer.add_document(doc!(title => "bbb", body => "an old man sailing on the sea"))?;
index_writer.add_document(doc!(title => "ccc", body=> "send this message to alice"))?;
index_writer.add_document(doc!(title => "ddd", body=> "a lady was riding and old bike"))?;
index_writer.add_document(doc!(title => "eee", body=> "Yes, my lady."))?;
index_writer.commit()?;
Ok(index)
}
#[test]
@@ -243,9 +235,9 @@ mod tests {
}
#[test]
fn test_more_like_this_query() {
let index = create_test_index();
let reader = index.reader().unwrap();
fn test_more_like_this_query() -> crate::Result<()> {
let index = create_test_index()?;
let reader = index.reader()?;
let searcher = reader.searcher();
// search base 1st doc with words [sea, and] skipping [old]
@@ -258,7 +250,7 @@ mod tests {
.with_boost_factor(1.0)
.with_stop_words(vec!["old".to_string()])
.with_document(DocAddress::new(0, 0));
let top_docs = searcher.search(&query, &TopDocs::with_limit(5)).unwrap();
let top_docs = searcher.search(&query, &TopDocs::with_limit(5))?;
let mut doc_ids: Vec<_> = top_docs.iter().map(|item| item.1.doc_id).collect();
doc_ids.sort_unstable();
@@ -274,11 +266,12 @@ mod tests {
.with_max_word_length(5)
.with_boost_factor(1.0)
.with_document(DocAddress::new(0, 4));
let top_docs = searcher.search(&query, &TopDocs::with_limit(5)).unwrap();
let top_docs = searcher.search(&query, &TopDocs::with_limit(5))?;
let mut doc_ids: Vec<_> = top_docs.iter().map(|item| item.1.doc_id).collect();
doc_ids.sort_unstable();
assert_eq!(doc_ids.len(), 2);
assert_eq!(doc_ids, vec![3, 4]);
Ok(())
}
}

80
src/query/phrase_query/mod.rs
View File

@@ -18,34 +18,34 @@ pub mod tests {
use crate::DocId;
use crate::{DocAddress, TERMINATED};
pub fn create_index(texts: &[&'static str]) -> Index {
pub fn create_index(texts: &[&'static str]) -> crate::Result<Index> {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
for &text in texts {
let doc = doc!(text_field=>text);
index_writer.add_document(doc);
index_writer.add_document(doc)?;
}
assert!(index_writer.commit().is_ok());
index_writer.commit()?;
}
index
Ok(index)
}
#[test]
pub fn test_phrase_query() {
pub fn test_phrase_query() -> crate::Result<()> {
let index = create_index(&[
"b b b d c g c",
"a b b d c g c",
"a b a b c",
"c a b a d ga a",
"a b c",
]);
])?;
let schema = index.schema();
let text_field = schema.get_field("text").unwrap();
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
let test_query = |texts: Vec<&str>| {
let terms: Vec<Term> = texts
.iter()
@@ -54,7 +54,7 @@ pub mod tests {
let phrase_query = PhraseQuery::new(terms);
let test_fruits = searcher
.search(&phrase_query, &TEST_COLLECTOR_WITH_SCORE)
.expect("search should succeed");
.unwrap();
test_fruits
.docs()
.iter()
@@ -66,11 +66,12 @@ pub mod tests {
assert_eq!(test_query(vec!["b", "b"]), vec![0, 1]);
assert!(test_query(vec!["g", "ewrwer"]).is_empty());
assert!(test_query(vec!["g", "a"]).is_empty());
Ok(())
}
#[test]
pub fn test_phrase_query_simple() -> crate::Result<()> {
let index = create_index(&["a b b d c g c", "a b a b c"]);
let index = create_index(&["a b b d c g c", "a b a b c"])?;
let text_field = index.schema().get_field("text").unwrap();
let searcher = index.reader()?.searcher();
let terms: Vec<Term> = vec!["a", "b", "c"]
@@ -86,17 +87,17 @@ pub mod tests {
}
#[test]
pub fn test_phrase_query_no_score() {
pub fn test_phrase_query_no_score() -> crate::Result<()> {
let index = create_index(&[
"b b b d c g c",
"a b b d c g c",
"a b a b c",
"c a b a d ga a",
"a b c",
]);
])?;
let schema = index.schema();
let text_field = schema.get_field("text").unwrap();
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
let test_query = |texts: Vec<&str>| {
let terms: Vec<Term> = texts
.iter()
@@ -105,7 +106,7 @@ pub mod tests {
let phrase_query = PhraseQuery::new(terms);
let test_fruits = searcher
.search(&phrase_query, &TEST_COLLECTOR_WITHOUT_SCORE)
.expect("search should succeed");
.unwrap();
test_fruits
.docs()
.iter()
@@ -117,10 +118,11 @@ pub mod tests {
assert_eq!(test_query(vec!["b", "b"]), vec![0, 1]);
assert!(test_query(vec!["g", "ewrwer"]).is_empty());
assert!(test_query(vec!["g", "a"]).is_empty());
Ok(())
}
#[test]
pub fn test_phrase_query_no_positions() {
pub fn test_phrase_query_no_positions() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
use crate::schema::IndexRecordOption;
use crate::schema::TextFieldIndexing;
@@ -135,33 +137,34 @@ pub mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(text_field=>"a b c"));
assert!(index_writer.commit().is_ok());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"a b c"))?;
index_writer.commit()?;
}
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
let phrase_query = PhraseQuery::new(vec![
Term::from_field_text(text_field, "a"),
Term::from_field_text(text_field, "b"),
]);
let search_result = searcher
let search_error = searcher
.search(&phrase_query, &TEST_COLLECTOR_WITH_SCORE)
.map(|_| ());
.err();
assert!(matches!(
search_result,
Err(crate::TantivyError::SchemaError(msg))
search_error,
Some(crate::TantivyError::SchemaError(msg))
if msg == "Applied phrase query on field \"text\", which does not have positions \
indexed"
));
Ok(())
}
#[test]
pub fn test_phrase_score() {
let index = create_index(&["a b c", "a b c a b"]);
pub fn test_phrase_score() -> crate::Result<()> {
let index = create_index(&["a b c", "a b c a b"])?;
let schema = index.schema();
let text_field = schema.get_field("text").unwrap();
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
let test_query = |texts: Vec<&str>| {
let terms: Vec<Term> = texts
.iter()
@@ -177,23 +180,24 @@ pub mod tests {
let scores = test_query(vec!["a", "b"]);
assert_nearly_equals!(scores[0], 0.40618482);
assert_nearly_equals!(scores[1], 0.46844664);
Ok(())
}
#[test] // motivated by #234
pub fn test_phrase_query_docfreq_order() {
pub fn test_phrase_query_docfreq_order() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(text_field=>"b"));
index_writer.add_document(doc!(text_field=>"a b"));
index_writer.add_document(doc!(text_field=>"b a"));
assert!(index_writer.commit().is_ok());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"b"))?;
index_writer.add_document(doc!(text_field=>"a b"))?;
index_writer.add_document(doc!(text_field=>"b a"))?;
index_writer.commit()?;
}
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
let test_query = |texts: Vec<&str>| {
let terms: Vec<Term> = texts
.iter()
@@ -208,18 +212,19 @@ pub mod tests {
};
assert_eq!(test_query(vec!["a", "b"]), vec![DocAddress::new(0, 1)]);
assert_eq!(test_query(vec!["b", "a"]), vec![DocAddress::new(0, 2)]);
Ok(())
}
#[test] // motivated by #234
pub fn test_phrase_query_non_trivial_offsets() {
pub fn test_phrase_query_non_trivial_offsets() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(text_field=>"a b c d e f g h"));
assert!(index_writer.commit().is_ok());
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"a b c d e f g h"))?;
index_writer.commit()?;
}
let searcher = index.reader().unwrap().searcher();
let test_query = |texts: Vec<(usize, &str)>| {
@@ -245,5 +250,6 @@ pub mod tests {
assert_eq!(test_query(vec![(4, "e"), (0, "a"), (2, "c")]), vec![0]);
assert!(test_query(vec![(0, "a"), (2, "d")]).is_empty());
assert_eq!(test_query(vec![(1, "a"), (3, "c")]), vec![0]);
Ok(())
}
}

10
src/query/phrase_query/phrase_weight.rs
View File

@@ -116,19 +116,18 @@ mod tests {
use crate::{DocSet, Term};
#[test]
pub fn test_phrase_count() {
let index = create_index(&["a c", "a a b d a b c", " a b"]);
pub fn test_phrase_count() -> crate::Result<()> {
let index = create_index(&["a c", "a a b d a b c", " a b"])?;
let schema = index.schema();
let text_field = schema.get_field("text").unwrap();
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
let phrase_query = PhraseQuery::new(vec![
Term::from_field_text(text_field, "a"),
Term::from_field_text(text_field, "b"),
]);
let phrase_weight = phrase_query.phrase_weight(&searcher, true).unwrap();
let mut phrase_scorer = phrase_weight
.phrase_scorer(searcher.segment_reader(0u32), 1.0)
.unwrap()
.phrase_scorer(searcher.segment_reader(0u32), 1.0)?
.unwrap();
assert_eq!(phrase_scorer.doc(), 1);
assert_eq!(phrase_scorer.phrase_count(), 2);
@@ -136,5 +135,6 @@ mod tests {
assert_eq!(phrase_scorer.doc(), 2);
assert_eq!(phrase_scorer.phrase_count(), 1);
assert_eq!(phrase_scorer.advance(), TERMINATED);
Ok(())
}
}

35
src/query/range_query.rs
View File

@@ -51,7 +51,7 @@ fn map_bound<TFrom, TTo, Transform: Fn(&TFrom) -> TTo>(
/// for year in 1950u64..2017u64 {
/// let num_docs_within_year = 10 + (year - 1950) * (year - 1950);
/// for _ in 0..num_docs_within_year {
/// index_writer.add_document(doc!(year_field => year));
/// index_writer.add_document(doc!(year_field => year))?;
/// }
/// }
/// index_writer.commit()?;
@@ -338,34 +338,35 @@ mod tests {
use std::ops::Bound;
#[test]
fn test_range_query_simple() {
fn test_range_query_simple() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let year_field = schema_builder.add_u64_field("year", INDEXED);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
for year in 1950u64..2017u64 {
let num_docs_within_year = 10 + (year - 1950) * (year - 1950);
for _ in 0..num_docs_within_year {
index_writer.add_document(doc!(year_field => year));
index_writer.add_document(doc!(year_field => year))?;
}
}
index_writer.commit().unwrap();
index_writer.commit()?;
}
let reader = index.reader().unwrap();
let reader = index.reader()?;
let searcher = reader.searcher();
let docs_in_the_sixties = RangeQuery::new_u64(year_field, 1960u64..1970u64);
// ... or `1960..=1969` if inclusive range is enabled.
let count = searcher.search(&docs_in_the_sixties, &Count).unwrap();
let count = searcher.search(&docs_in_the_sixties, &Count)?;
assert_eq!(count, 2285);
Ok(())
}
#[test]
fn test_range_query() {
fn test_range_query() -> crate::Result<()> {
let int_field: Field;
let schema = {
let mut schema_builder = Schema::builder();
@@ -375,7 +376,7 @@ mod tests {
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_with_num_threads(2, 6_000_000).unwrap();
let mut index_writer = index.writer_with_num_threads(2, 6_000_000)?;
for i in 1..100 {
let mut doc = Document::new();
@@ -384,10 +385,10 @@ mod tests {
doc.add_i64(int_field, j as i64);
}
}
index_writer.add_document(doc);
index_writer.add_document(doc)?;
}
index_writer.commit().unwrap();
index_writer.commit()?;
}
let reader = index.reader().unwrap();
let searcher = reader.searcher();
@@ -419,10 +420,11 @@ mod tests {
)),
91
);
Ok(())
}
#[test]
fn test_range_float() {
fn test_range_float() -> crate::Result<()> {
let float_field: Field;
let schema = {
let mut schema_builder = Schema::builder();
@@ -441,12 +443,12 @@ mod tests {
doc.add_f64(float_field, j as f64);
}
}
index_writer.add_document(doc);
index_writer.add_document(doc)?;
}
index_writer.commit().unwrap();
index_writer.commit()?;
}
let reader = index.reader().unwrap();
let reader = index.reader()?;
let searcher = reader.searcher();
let count_multiples =
|range_query: RangeQuery| searcher.search(&range_query, &Count).unwrap();
@@ -479,6 +481,7 @@ mod tests {
)),
91
);
Ok(())
}
#[test]
@@ -494,7 +497,7 @@ mod tests {
index_writer.add_document(doc!(
title => "hemoglobin blood",
year => 1990_i64
));
))?;
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();

42
src/query/regex_query.rs
View File

@@ -28,17 +28,17 @@ use tantivy_fst::Regex;
/// let mut index_writer = index.writer(3_000_000)?;
/// index_writer.add_document(doc!(
/// title => "The Name of the Wind",
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "The Diary of Muadib",
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "A Dairy Cow",
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "The Diary of a Young Girl",
/// ));
/// index_writer.commit().unwrap();
/// ))?;
/// index_writer.commit()?;
/// }
///
/// let reader = index.reader()?;
@@ -100,7 +100,7 @@ mod test {
use std::sync::Arc;
use tantivy_fst::Regex;
fn build_test_index() -> (IndexReader, Field) {
fn build_test_index() -> crate::Result<(IndexReader, Field)> {
let mut schema_builder = Schema::builder();
let country_field = schema_builder.add_text_field("country", TEXT);
let schema = schema_builder.build();
@@ -109,15 +109,15 @@ mod test {
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(
country_field => "japan",
));
))?;
index_writer.add_document(doc!(
country_field => "korea",
));
index_writer.commit().unwrap();
))?;
index_writer.commit()?;
}
let reader = index.reader().unwrap();
let reader = index.reader()?;
(reader, country_field)
Ok((reader, country_field))
}
fn verify_regex_query(
@@ -141,31 +141,32 @@ mod test {
}
#[test]
pub fn test_regex_query() {
let (reader, field) = build_test_index();
let matching_one = RegexQuery::from_pattern("jap[ao]n", field).unwrap();
let matching_zero = RegexQuery::from_pattern("jap[A-Z]n", field).unwrap();
pub fn test_regex_query() -> crate::Result<()> {
let (reader, field) = build_test_index()?;
let matching_one = RegexQuery::from_pattern("jap[ao]n", field)?;
let matching_zero = RegexQuery::from_pattern("jap[A-Z]n", field)?;
verify_regex_query(matching_one, matching_zero, reader);
Ok(())
}
#[test]
pub fn test_construct_from_regex() {
let (reader, field) = build_test_index();
pub fn test_construct_from_regex() -> crate::Result<()> {
let (reader, field) = build_test_index()?;
let matching_one = RegexQuery::from_regex(Regex::new("jap[ao]n").unwrap(), field);
let matching_zero = RegexQuery::from_regex(Regex::new("jap[A-Z]n").unwrap(), field);
verify_regex_query(matching_one, matching_zero, reader);
Ok(())
}
#[test]
pub fn test_construct_from_reused_regex() {
pub fn test_construct_from_reused_regex() -> crate::Result<()> {
let r1 = Arc::new(Regex::new("jap[ao]n").unwrap());
let r2 = Arc::new(Regex::new("jap[A-Z]n").unwrap());
let (reader, field) = build_test_index();
let (reader, field) = build_test_index()?;
let matching_one = RegexQuery::from_regex(r1.clone(), field);
let matching_zero = RegexQuery::from_regex(r2.clone(), field);
@@ -176,5 +177,6 @@ mod test {
let matching_zero = RegexQuery::from_regex(r2, field);
verify_regex_query(matching_one, matching_zero, reader);
Ok(())
}
}

75
src/query/term_query/mod.rs
View File

@@ -18,28 +18,29 @@ mod tests {
use crate::{Index, Term, TERMINATED};
#[test]
pub fn test_term_query_no_freq() {
pub fn test_term_query_no_freq() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", STRING);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
// writing the segment
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
let doc = doc!(text_field => "a");
index_writer.add_document(doc);
assert!(index_writer.commit().is_ok());
index_writer.add_document(doc)?;
index_writer.commit()?;
}
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
let term_query = TermQuery::new(
Term::from_field_text(text_field, "a"),
IndexRecordOption::Basic,
);
let term_weight = term_query.weight(&searcher, true).unwrap();
let term_weight = term_query.weight(&searcher, true)?;
let segment_reader = searcher.segment_reader(0);
let mut term_scorer = term_weight.scorer(segment_reader, 1.0).unwrap();
let mut term_scorer = term_weight.scorer(segment_reader, 1.0)?;
assert_eq!(term_scorer.doc(), 0);
assert_nearly_equals!(term_scorer.score(), 0.28768212);
Ok(())
}
#[test]
@@ -53,7 +54,7 @@ mod tests {
let mut index_writer = index.writer_for_tests()?;
for _ in 0..COMPRESSION_BLOCK_SIZE {
let doc = doc!(text_field => "a");
index_writer.add_document(doc);
index_writer.add_document(doc)?;
}
index_writer.commit()?;
}
@@ -78,7 +79,7 @@ mod tests {
}
#[test]
pub fn test_term_weight() {
pub fn test_term_weight() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let left_field = schema_builder.add_text_field("left", TEXT);
let right_field = schema_builder.add_text_field("right", TEXT);
@@ -86,22 +87,20 @@ mod tests {
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
{
let mut index_writer = index.writer_for_tests().unwrap();
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(
left_field => "left1 left2 left2 left2f2 left2f2 left3 abcde abcde abcde abcde abcde abcde abcde abcde abcde abcewde abcde abcde",
right_field => "right1 right2",
large_field => "large0 large1 large2 large3 large4 large5 large6 large7 large8 large9 large10 large11 large12 large13 large14 large15 large16 large17 large18 large19 large20 large21 large22 large23 large24 large25 large26 large27 large28 large29 large30 large31 large32 large33 large34 large35 large36 large37 large38 large39 large40 large41 large42 large43 large44 large45 large46 large47 large48 large49 large50 large51 large52 large53 large54 large55 large56 large57 large58 large59 large60 large61 large62 large63 large64 large65 large66 large67 large68 large69 large70 large71 large72 large73 large74 large75 large76 large77 large78 large79 large80 large81 large82 large83 large84 large85 large86 large87 large88 large89 large90 large91 large92 large93 large94 large95 large96 large97 large98 large99 large100 large101 large102 large103 large104 large105 large106 large107 large108 large109 large110 large111 large112 large113 large114 large115 large116 large117 large118 large119 large120 large121 large122 large123 large124 large125 large126 large127 large128 large129 large130 large131 large132 large133 large134 large135 large136 large137 large138 large139 large140 large141 large142 large143 large144 large145 large146 large147 large148 large149 large150 large151 large152 large153 large154 large155 large156 large157 large158 large159 large160 large161 large162 large163 large164 large165 large166 large167 large168 large169 large170 large171 large172 large173 large174 large175 large176 large177 large178 large179 large180 large181 large182 large183 large184 large185 large186 large187 large188 large189 large190 large191 large192 large193 large194 large195 large196 large197 large198 large199 large200 large201 large202 large203 large204 large205 large206 large207 large208 large209 large210 large211 large212 large213 large214 large215 large216 large217 large218 large219 large220 large221 large222 large223 large224 large225 large226 large227 large228 large229 large230 large231 large232 large233 large234 large235 large236 large237 large238 large239 large240 large241 large242 large243 large244 large245 large246 large247 large248 large249 large250 large251 large252 large253 large254 large255 large256 large257 large258 large259 large260 large261 large262 large263 large264 large265 large266 large267 large268 large269 large270 large271 large272 large273 large274 large275 large276 large277 large278 large279 large280 large281 large282 large283 large284 large285 large286"
));
index_writer.add_document(doc!(left_field => "left4 left1"));
index_writer.commit().unwrap();
))?;
index_writer.add_document(doc!(left_field => "left4 left1"))?;
index_writer.commit()?;
}
let searcher = index.reader().unwrap().searcher();
let searcher = index.reader()?.searcher();
{
let term = Term::from_field_text(left_field, "left2");
let term_query = TermQuery::new(term, IndexRecordOption::WithFreqs);
let topdocs = searcher
.search(&term_query, &TopDocs::with_limit(2))
.unwrap();
let topdocs = searcher.search(&term_query, &TopDocs::with_limit(2))?;
assert_eq!(topdocs.len(), 1);
let (score, _) = topdocs[0];
assert_nearly_equals!(0.77802235, score);
@@ -109,9 +108,7 @@ mod tests {
{
let term = Term::from_field_text(left_field, "left1");
let term_query = TermQuery::new(term, IndexRecordOption::WithFreqs);
let top_docs = searcher
.search(&term_query, &TopDocs::with_limit(2))
.unwrap();
let top_docs = searcher.search(&term_query, &TopDocs::with_limit(2))?;
assert_eq!(top_docs.len(), 2);
let (score1, _) = top_docs[0];
assert_nearly_equals!(0.27101856, score1);
@@ -119,32 +116,34 @@ mod tests {
assert_nearly_equals!(0.13736556, score2);
}
{
let query_parser = QueryParser::for_index(&index, vec![]);
let query = query_parser.parse_query("left:left2 left:left1").unwrap();
let top_docs = searcher.search(&query, &TopDocs::with_limit(2)).unwrap();
let query_parser = QueryParser::for_index(&index, Vec::new());
let query = query_parser.parse_query("left:left2 left:left1")?;
let top_docs = searcher.search(&query, &TopDocs::with_limit(2))?;
assert_eq!(top_docs.len(), 2);
let (score1, _) = top_docs[0];
assert_nearly_equals!(0.9153879, score1);
let (score2, _) = top_docs[1];
assert_nearly_equals!(0.27101856, score2);
}
Ok(())
}
#[test]
fn test_term_query_count_when_there_are_deletes() {
fn test_term_query_count_when_there_are_deletes() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(text_field=>"a b"));
index_writer.add_document(doc!(text_field=>"a c"));
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"a b"))?;
index_writer.add_document(doc!(text_field=>"a c"))?;
index_writer.delete_term(Term::from_field_text(text_field, "b"));
index_writer.commit().unwrap();
index_writer.commit()?;
let term_a = Term::from_field_text(text_field, "a");
let term_query = TermQuery::new(term_a, IndexRecordOption::Basic);
let reader = index.reader().unwrap();
assert_eq!(term_query.count(&*reader.searcher()).unwrap(), 1);
let reader = index.reader()?;
assert_eq!(term_query.count(&*reader.searcher())?, 1);
Ok(())
}
#[test]
@@ -153,9 +152,9 @@ mod tests {
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(text_field=>"a"));
index_writer.add_document(doc!(text_field=>"a"));
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"a"))?;
index_writer.add_document(doc!(text_field=>"a"))?;
index_writer.commit()?;
let term_a = Term::from_field_text(text_field, "a");
let term_query = TermQuery::new(term_a, IndexRecordOption::Basic);
@@ -186,11 +185,11 @@ mod tests {
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests().unwrap();
index_writer.add_document(doc!(text_field=>"b"));
index_writer.add_document(doc!(text_field=>"a"));
index_writer.add_document(doc!(text_field=>"a"));
index_writer.add_document(doc!(text_field=>"b"));
let mut index_writer = index.writer_for_tests()?;
index_writer.add_document(doc!(text_field=>"b"))?;
index_writer.add_document(doc!(text_field=>"a"))?;
index_writer.add_document(doc!(text_field=>"a"))?;
index_writer.add_document(doc!(text_field=>"b"))?;
index_writer.commit()?;
let term_a = Term::from_field_text(text_field, "a");
let term_query = TermQuery::new(term_a, IndexRecordOption::Basic);

8
src/query/term_query/term_query.rs
View File

@@ -32,16 +32,16 @@ use std::fmt;
/// let mut index_writer = index.writer(3_000_000)?;
/// index_writer.add_document(doc!(
/// title => "The Name of the Wind",
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "The Diary of Muadib",
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "A Dairy Cow",
/// ));
/// ))?;
/// index_writer.add_document(doc!(
/// title => "The Diary of a Young Girl",
/// ));
/// ))?;
/// index_writer.commit()?;
/// }
/// let reader = index.reader()?;

2
src/query/term_query/term_scorer.rs
View File

@@ -305,7 +305,7 @@ mod tests {
let term_freq = rng.gen_range(1..10000);
let words: Vec<&str> = std::iter::repeat("bbbb").take(term_freq).collect();
let text = words.join(" ");
writer.add_document(doc!(text_field=>text));
writer.add_document(doc!(text_field=>text))?;
}
writer.commit()?;
let term_query = TermQuery::new(

6
src/query/term_query/term_weight.rs
View File

@@ -40,8 +40,8 @@ impl Weight for TermWeight {
}
fn count(&self, reader: &SegmentReader) -> crate::Result<u32> {
if let Some(delete_bitset) = reader.delete_bitset() {
Ok(self.scorer(reader, 1.0)?.count(delete_bitset))
if let Some(alive_bitset) = reader.alive_bitset() {
Ok(self.scorer(reader, 1.0)?.count(alive_bitset))
} else {
let field = self.term.field();
let inv_index = reader.inverted_index(field)?;
@@ -79,7 +79,7 @@ impl Weight for TermWeight {
callback: &mut dyn FnMut(DocId, Score) -> Score,
) -> crate::Result<()> {
let scorer = self.specialized_scorer(reader, 1.0)?;
crate::query::boolean_query::block_wand(vec![scorer], threshold, callback);
crate::query::boolean_query::block_wand_single_scorer(scorer, threshold, callback);
Ok(())
}
}

4
src/query/weight.rs
View File

@@ -59,8 +59,8 @@ pub trait Weight: Send + Sync + 'static {
/// Returns the number documents within the given `SegmentReader`.
fn count(&self, reader: &SegmentReader) -> crate::Result<u32> {
let mut scorer = self.scorer(reader, 1.0)?;
if let Some(delete_bitset) = reader.delete_bitset() {
Ok(scorer.count(delete_bitset))
if let Some(alive_bitset) = reader.alive_bitset() {
Ok(scorer.count(alive_bitset))
} else {
Ok(scorer.count_including_deleted())
}

4
src/reader/mod.rs
View File

@@ -94,7 +94,7 @@ impl IndexReaderBuilder {
}
Ok(IndexReader {
inner: inner_reader_arc,
watch_handle_opt,
_watch_handle_opt: watch_handle_opt,
})
}
@@ -167,7 +167,7 @@ impl InnerIndexReader {
#[derive(Clone)]
pub struct IndexReader {
inner: Arc<InnerIndexReader>,
watch_handle_opt: Option<WatchHandle>,
_watch_handle_opt: Option<WatchHandle>,
}
impl IndexReader {

12
src/schema/bytes_options.rs
View File

@@ -3,7 +3,7 @@ use std::ops::BitOr;
use super::flags::{FastFlag, IndexedFlag, SchemaFlagList, StoredFlag};
/// Define how an a bytes field should be handled by tantivy.
#[derive(Clone, Debug, PartialEq, Eq, Serialize, Deserialize)]
#[derive(Clone, Debug, PartialEq, Eq, Serialize, Deserialize, Default)]
pub struct BytesOptions {
indexed: bool,
fast: bool,
@@ -56,16 +56,6 @@ impl BytesOptions {
}
}
impl Default for BytesOptions {
fn default() -> BytesOptions {
BytesOptions {
indexed: false,
fast: false,
stored: false,
}
}
}
impl<T: Into<BytesOptions>> BitOr<T> for BytesOptions {
type Output = BytesOptions;

11
src/schema/facet_options.rs
View File

@@ -3,7 +3,7 @@ use serde::{Deserialize, Serialize};
use std::ops::BitOr;
/// Define how a facet field should be handled by tantivy.
#[derive(Clone, Debug, PartialEq, Eq, Serialize, Deserialize)]
#[derive(Clone, Debug, PartialEq, Eq, Serialize, Deserialize, Default)]
pub struct FacetOptions {
indexed: bool,
stored: bool,
@@ -39,15 +39,6 @@ impl FacetOptions {
}
}
impl Default for FacetOptions {
fn default() -> FacetOptions {
FacetOptions {
indexed: false,
stored: false,
}
}
}
impl From<()> for FacetOptions {
fn from(_: ()) -> FacetOptions {
FacetOptions::default()

162
src/schema/field_entry.rs
View File

@@ -4,10 +4,7 @@ use crate::schema::{is_valid_field_name, IntOptions};
use crate::schema::bytes_options::BytesOptions;
use crate::schema::FieldType;
use serde::de::{self, MapAccess, Visitor};
use serde::ser::SerializeStruct;
use serde::{Deserialize, Deserializer, Serialize, Serializer};
use std::fmt;
use serde::{Deserialize, Serialize};
/// A `FieldEntry` represents a field and its configuration.
/// `Schema` are a collection of `FieldEntry`
@@ -16,9 +13,10 @@ use std::fmt;
/// - a field name
/// - a field type, itself wrapping up options describing
/// how the field should be indexed.
#[derive(Clone, Debug, PartialEq)]
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct FieldEntry {
name: String,
#[serde(flatten)]
field_type: FieldType,
}
@@ -141,140 +139,6 @@ impl FieldEntry {
}
}
impl Serialize for FieldEntry {
fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>
where
S: Serializer,
{
let mut s = serializer.serialize_struct("field_entry", 3)?;
s.serialize_field("name", &self.name)?;
match self.field_type {
FieldType::Str(ref options) => {
s.serialize_field("type", "text")?;
s.serialize_field("options", options)?;
}
FieldType::U64(ref options) => {
s.serialize_field("type", "u64")?;
s.serialize_field("options", options)?;
}
FieldType::I64(ref options) => {
s.serialize_field("type", "i64")?;
s.serialize_field("options", options)?;
}
FieldType::F64(ref options) => {
s.serialize_field("type", "f64")?;
s.serialize_field("options", options)?;
}
FieldType::Date(ref options) => {
s.serialize_field("type", "date")?;
s.serialize_field("options", options)?;
}
FieldType::HierarchicalFacet(ref options) => {
s.serialize_field("type", "hierarchical_facet")?;
s.serialize_field("options", options)?;
}
FieldType::Bytes(ref options) => {
s.serialize_field("type", "bytes")?;
s.serialize_field("options", options)?;
}
}
s.end()
}
}
impl<'de> Deserialize<'de> for FieldEntry {
fn deserialize<D>(deserializer: D) -> Result<Self, D::Error>
where
D: Deserializer<'de>,
{
#[derive(Deserialize)]
#[serde(field_identifier, rename_all = "lowercase")]
enum Field {
Name,
Type,
Options,
}
const FIELDS: &[&str] = &["name", "type", "options"];
struct FieldEntryVisitor;
impl<'de> Visitor<'de> for FieldEntryVisitor {
type Value = FieldEntry;
fn expecting(&self, formatter: &mut fmt::Formatter<'_>) -> fmt::Result {
formatter.write_str("struct FieldEntry")
}
fn visit_map<V>(self, mut map: V) -> Result<FieldEntry, V::Error>
where
V: MapAccess<'de>,
{
let mut name = None;
let mut ty = None;
let mut field_type = None;
while let Some(key) = map.next_key()? {
match key {
Field::Name => {
if name.is_some() {
return Err(de::Error::duplicate_field("name"));
}
name = Some(map.next_value()?);
}
Field::Type => {
if ty.is_some() {
return Err(de::Error::duplicate_field("type"));
}
let type_string = map.next_value::<String>()?;
match type_string.as_str() {
"text" | "u64" | "i64" | "f64" | "date" | "bytes"
| "hierarchical_facet" => {
// These types require additional options to create a field_type
}
_ => panic!("unhandled type"),
}
ty = Some(type_string);
}
Field::Options => match ty {
None => {
let msg = "The `type` field must be \
specified before `options`";
return Err(de::Error::custom(msg));
}
Some(ref ty) => match ty.as_str() {
"text" => field_type = Some(FieldType::Str(map.next_value()?)),
"u64" => field_type = Some(FieldType::U64(map.next_value()?)),
"i64" => field_type = Some(FieldType::I64(map.next_value()?)),
"f64" => field_type = Some(FieldType::F64(map.next_value()?)),
"date" => field_type = Some(FieldType::Date(map.next_value()?)),
"bytes" => field_type = Some(FieldType::Bytes(map.next_value()?)),
"hierarchical_facet" => {
field_type =
Some(FieldType::HierarchicalFacet(map.next_value()?))
}
_ => {
let msg = format!("Unrecognised type {}", ty);
return Err(de::Error::custom(msg));
}
},
},
}
}
let name = name.ok_or_else(|| de::Error::missing_field("name"))?;
ty.ok_or_else(|| de::Error::missing_field("ty"))?;
let field_type = field_type.ok_or_else(|| de::Error::missing_field("options"))?;
Ok(FieldEntry { name, field_type })
}
}
deserializer.deserialize_struct("field_entry", FIELDS, FieldEntryVisitor)
}
}
#[cfg(test)]
mod tests {
use super::*;
@@ -315,4 +179,24 @@ mod tests {
_ => panic!("expected FieldType::Str"),
}
}
#[test]
fn test_json_deserialization() {
let json_str = r#"{
"name": "title",
"options": {
"indexing": {
"record": "position",
"tokenizer": "default"
},
"stored": false
},
"type": "text"
}"#;
let field_entry: FieldEntry = serde_json::from_str(json_str).unwrap();
match field_entry.field_type {
FieldType::Str(_) => {}
_ => panic!("expected FieldType::Str"),
}
}
}

6
src/schema/field_type.rs
View File

@@ -7,6 +7,7 @@ use crate::schema::Value;
use crate::schema::{IntOptions, TextOptions};
use crate::tokenizer::PreTokenizedString;
use chrono::{FixedOffset, Utc};
use serde::{Deserialize, Serialize};
use serde_json::Value as JsonValue;
/// Possible error that may occur while parsing a field value
@@ -48,9 +49,12 @@ pub enum Type {
/// A `FieldType` describes the type (text, u64) of a field as well as
/// how it should be handled by tantivy.
#[derive(Clone, Debug, PartialEq)]
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
#[serde(tag = "type", content = "options")]
#[serde(rename_all = "snake_case")]
pub enum FieldType {
/// String field type configuration
#[serde(rename = "text")]
Str(TextOptions),
/// Unsigned 64-bits integers field type configuration
U64(IntOptions),

12
src/schema/int_options.rs
View File

@@ -15,7 +15,7 @@ pub enum Cardinality {
}
/// Define how an u64, i64, of f64 field should be handled by tantivy.
#[derive(Clone, Debug, PartialEq, Eq, Serialize, Deserialize)]
#[derive(Clone, Debug, PartialEq, Eq, Serialize, Deserialize, Default)]
pub struct IntOptions {
indexed: bool,
#[serde(skip_serializing_if = "Option::is_none")]
@@ -79,16 +79,6 @@ impl IntOptions {
}
}
impl Default for IntOptions {
fn default() -> IntOptions {
IntOptions {
indexed: false,
stored: false,
fast: None,
}
}
}
impl From<()> for IntOptions {
fn from(_: ()) -> IntOptions {
IntOptions::default()

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