rust/tantivy
1
0
Fork 0
mirror of https://github.com/quickwit-oss/tantivy.git synced 2025-12-28 04:52:55 +00:00
Code Issues Packages Projects Releases Wiki Activity

Compare commits

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

291 Commits
remove_ran ... 0.18

Author SHA1 Message Date
Paul Masurel
f0a2b1cc44 Bumped tantivy and subcrate versions. 2022-05-25 22:50:33 +09:00
Paul Masurel
fcfdc44c61 Bumped tantivy-grammar version 2022-05-25 21:52:46 +09:00
Paul Masurel
3171f0b9ba Added ZSTD support in CHANGELOG 2022-05-25 21:51:46 +09:00
PSeitz
89e19f14b5 Merge pull request #1374 from kryesh/main 
Add Zstd compression support, Make block size configurable via IndexSettings
 2022-05-25 07:39:46 +02:00
PSeitz
1a6a1396cd Merge pull request #1376 from saroh/json-example 
Add examples to explain default field handling in the json example
 2022-05-24 07:09:37 +02:00
saroh
e766375700 remove useless example 2022-05-23 19:49:31 +02:00
PSeitz
496b4a4fdb Update examples/json_field.rs 2022-05-23 12:24:36 +02:00
PSeitz
93cc8498b3 Update examples/json_field.rs 2022-05-23 11:59:42 +02:00
PSeitz
0aa3d63a9f Update examples/json_field.rs 2022-05-23 11:39:45 +02:00
PSeitz
4e2a053b69 Update examples/json_field.rs 2022-05-23 11:27:05 +02:00
Paul Masurel
71c4393ec4 Clippy 2022-05-23 10:20:37 +09:00
saroh
b2e97e266a more examples to explain default field handling 2022-05-21 17:36:39 +02:00
Antoine G
9ee4772140 Fix deps for unicode regex compiling (#1373) 
* lint doc warning

* fix regex build
 2022-05-20 10:18:44 +09:00
Kryesh
c95013b11e Add zstd-compression feature to github workflow tests 2022-05-19 22:15:18 +10:00
Kryesh
fc045e6bf9 Cleanup imports, remove unneeded error mapping 2022-05-19 10:34:02 +10:00
Kryesh
6837a4d468 Fix bench 2022-05-18 20:35:29 +10:00
Kryesh
0759bf9448 Cleanup zstd structure and serialise to u32 in line with lz4 2022-05-18 20:31:22 +10:00
Kryesh
152e8238d7 Fix silly errors from running tests without feature flag 2022-05-18 19:49:10 +10:00
Kryesh
d4e5b48437 Apply feedback - standardise on u64 and fix correct compression bounds 2022-05-18 19:37:28 +10:00
Kryesh
03040ed81d Add Zstd compression support 2022-05-18 14:04:43 +10:00
Kryesh
aaa22ad225 Make block size configurable to allow for better compression ratios on large documents 2022-05-18 11:13:15 +10:00
Antoine G
3223bdf254 Refactorize PhraseScorer::compute_phrase_match (#1364) 
* Refactorize PhraseScorer::compute_phrase_match
* implem optim for slop
 2022-05-13 09:57:21 +09:00
dependabot[bot]
cbd06ab189 Update pprof requirement from 0.8.0 to 0.9.0 (#1365) 
Updates the requirements on [pprof](https://github.com/tikv/pprof-rs) to permit the latest version.
- [Release notes](https://github.com/tikv/pprof-rs/releases)
- [Changelog](https://github.com/tikv/pprof-rs/blob/master/CHANGELOG.md)
- [Commits](https://github.com/tikv/pprof-rs/commits)

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

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

Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2022-05-11 11:42:04 +09:00
Paul Masurel
749395bbb8 Added rustdoc for MultiFruit extract function (#1369) 2022-05-11 11:41:39 +09:00
Paul Masurel
617ba1f0c0 Bugfix in the document deserialization. (#1368) 
Deserializing a json field does not expect the
end of the document anymore.

This behavior is well documented in serde_json.
https://docs.serde.rs/serde_json/fn.from_reader.html

Closes #1366
 2022-05-11 11:38:10 +09:00
Paul Masurel
2f1cd7e7f0 Bugfix in the document deserialization. (#1367) 
Deserializing a json field does not expect the
end of the document anymore.

This behavior is well documented in serde_json.
https://docs.serde.rs/serde_json/fn.from_reader.html

Closes #1366
 2022-05-11 11:27:04 +09:00
PSeitz
58c0cb5fc4 Merge pull request #1357 from saroh/1302-json-term-writer-API 
Expose helpers to generate json field writer terms
 2022-05-10 11:02:05 +08:00
PSeitz
7f45a6ac96 allow setting tokenizer manager on index (#1362) 
handle json in tokenizer_for_field
 2022-05-09 18:15:45 +09:00
saroh
0ade871126 rename constructor to be more explicit 2022-05-06 13:29:07 +02:00
PSeitz
aab65490c9 Merge pull request #1358 from quickwit-oss/fix_docs 
add alias shard_size to split_size for quickwit
 2022-05-06 18:41:34 +08:00
Pascal Seitz
d77e8de36a flip alias variable name 2022-05-06 17:52:36 +08:00
Pascal Seitz
d11a8cce26 minor docs fix 2022-05-06 17:52:36 +08:00
Pascal Seitz
bc607a921b add alias shard_size split_size for quickwit 
improve some docs
 2022-05-06 17:52:36 +08:00
Paul Masurel
1273f33338 Fixed comment. 2022-05-06 18:35:25 +09:00
Paul Masurel
e30449743c Shortens blocks' last_key in the SSTable block index. (#1361) 
Right now we store last key in the blocks of the SSTable index.
This PR replaces the last key by a shorter string that is greater or
equal and still lesser than the next key.
This property is sufficiently to ensure the block index
works properly.

Related to quickwit#1366
 2022-05-06 16:29:06 +08:00
Paul Masurel
ed26552296 Minor changes in query parsing for quickwit#1334. (#1356) 
Quickwit's still heavily relies on generating field names
containing a '.' for nested object, yet allows for
user defined field names to contain a dot.

In order to reuse tantivy query parser, we will end up
using quickwit field names directly into tantivy.
Only '.' will be escaped.

This PR makes minor changes in how tantivy query parser parses
a field name and resolves it to a field.
Some of the new edge case behavior is hacky.

Closes #1355
 2022-05-06 13:20:10 +09:00
Saroh
65d129afbd better function names 2022-05-05 10:12:28 +02:00
Antoine G
386ffab76c Fix documentation regression (#1359) 
This breaks the doc on doc.rs as the type seems to shadow the struct https://docs.rs/tantivy/latest/tantivy/termdict/type.TermDictionary.html
introduced by #1293 which may not have been up to date with what was done in #1242
 2022-05-05 14:59:25 +09:00
Pasha Podolsky
57a8d0359c Make FruitHandle and MultiFruit public (#1360) 
* Make `FruitHandle` and `MultiFruit` public

* Add docs for `MultiFruit` and `FruitHandle`
 2022-05-05 14:58:33 +09:00
Saroh
14cb66ee00 move helper to indexer module 2022-05-04 18:01:57 +02:00
Saroh
9e38343352 expose helpers for json field writer manipulation 
closes #1302
 2022-05-04 18:01:45 +02:00
PSeitz
944302ae2f Merge pull request #1350 from quickwit-oss/update_edition 
update edition
 2022-05-04 11:02:52 +02:00
Paul Masurel
be70804d17 Removed AtomicUsize. 2022-05-04 16:45:24 +09:00
PSeitz
a1afc80600 Update src/core/executor.rs 
Co-authored-by: Paul Masurel <paul@quickwit.io>
 2022-05-04 08:39:44 +02:00
Paul Masurel
02e24fda52 Clippy fix 2022-05-04 12:24:07 +09:00
PSeitz
7e3c0c5392 Merge pull request #1353 from quickwit-oss/fix_docs 
minor docs fixes
 2022-05-02 07:48:25 +02:00
Pascal Seitz
fdb2524f9e minor docs fixes 2022-05-02 12:26:12 +08:00
Pascal Seitz
4db655ae82 update dependencies, update edition 2022-04-28 22:50:55 +08:00
Pascal Seitz
bb44cc84c4 update dependencies 2022-04-28 20:55:36 +08:00
PSeitz
8c1e1cf1ad Merge pull request #1349 from quickwit-oss/fix_error_message 
print whole query on syntax error
 2022-04-28 09:31:45 +02:00
Pascal Seitz
b5b16948b0 print whole query on syntax error 2022-04-27 12:48:30 +08:00
PSeitz
c305d3a2a2 Merge pull request #1346 from quickwit-oss/term_agg 
term agg
 2022-04-26 07:08:07 +02:00
PSeitz
038d234ff1 Merge pull request #1347 from quickwit-oss/query_parser_error 
fix query parser error field not found
 2022-04-26 07:01:48 +02:00
Pascal Seitz
c45eb9a9fa improve readability, add json test 2022-04-26 11:22:34 +08:00
Pascal Seitz
824d6f96fe return query on parse error 2022-04-22 16:11:36 +08:00
Pascal Seitz
7cf821bac0 fix query parser error field not found 2022-04-22 12:40:00 +08:00
PSeitz
ae83fc8298 bump uuid to 1.0 (#1345) 2022-04-22 10:02:24 +09:00
dependabot[bot]
a7bc361145 Update pprof requirement from 0.7 to 0.8 (#1343) 
Updates the requirements on [pprof](https://github.com/tikv/pprof-rs) to permit the latest version.
- [Release notes](https://github.com/tikv/pprof-rs/releases)
- [Changelog](https://github.com/tikv/pprof-rs/blob/master/CHANGELOG.md)
- [Commits](https://github.com/tikv/pprof-rs/commits)

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

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

Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2022-04-21 09:35:13 +09:00
Pascal Seitz
2805291400 minor fixes 2022-04-20 14:22:44 +08:00
Pascal Seitz
6614a2cba0 fix is_fast for bytes field 2022-04-20 12:02:38 +08:00
Pascal Seitz
6f4d203d1b return error on missing sub aggregation 2022-04-20 11:19:36 +08:00
Pascal Seitz
1be6c6111c support order property on term aggregations 
support order property on term aggregations
order can be by doc_count, key, or a metric sub_aggregation
 2022-04-20 00:34:38 +08:00
PSeitz
c7c3eab256 Merge pull request #1340 from PSeitz/term_agg 
fix collecting term_dict field names
 2022-04-18 08:21:27 +02:00
Pascal Seitz
ec69875d15 fix collecting term_dict field names 
fix collecting term_dict field names for sub_aggregations, minor refactoring
 2022-04-15 17:49:20 +08:00
PSeitz
d832cfcfd8 Merge pull request #1329 from quickwit-oss/term_agg 
add term aggregation
 2022-04-14 14:45:21 +08:00
Pascal Seitz
ab6b532cc4 add comments 2022-04-14 12:06:36 +08:00
Pascal Seitz
4b6047f7d7 return Option from as_ methods 2022-04-14 10:48:36 +08:00
Pascal Seitz
5ca04beb94 add min_doc_count test 2022-04-13 19:51:18 +08:00
Pascal Seitz
902d05ebec refactor getffreader function 2022-04-13 19:51:18 +08:00
Pascal Seitz
f1b298642a remove unnecessary benchmarks 2022-04-13 19:51:18 +08:00
Pascal Seitz
dd13dedaeb forward errors, remove unwrap 2022-04-13 19:51:18 +08:00
Pascal Seitz
46724b4a05 add segment_size, add get term dict fields, add tests 2022-04-13 19:51:18 +08:00
Pascal Seitz
24432bf523 add term aggregation 2022-04-13 19:51:18 +08:00
PSeitz
31d3bcfff2 Merge pull request #1334 from PSeitz/minor_fixes 
fix DateTime naming, fix docs, cleanup
 2022-04-13 13:13:57 +08:00
Pascal Seitz
706fbd6886 fix DateTime naming, fix docs, cleanup 2022-04-13 13:01:00 +08:00
PSeitz
8a8a048015 fix coverage (#1335) 2022-04-13 13:47:47 +09:00
dependabot[bot]
c72549cb9a Bump codecov/codecov-action from 2 to 3 (#1328) 
Bumps [codecov/codecov-action](https://github.com/codecov/codecov-action) from 2 to 3.
- [Release notes](https://github.com/codecov/codecov-action/releases)
- [Changelog](https://github.com/codecov/codecov-action/blob/master/CHANGELOG.md)
- [Commits](https://github.com/codecov/codecov-action/compare/v2...v3)

---
updated-dependencies:
- dependency-name: codecov/codecov-action
  dependency-type: direct:production
  update-type: version-update:semver-major
...

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

Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2022-04-11 21:26:52 +09:00
PSeitz
d6f803212c Merge pull request #1325 from quickwit-oss/term_agg 
fast field on string
 2022-04-04 15:34:31 +08:00
Pascal Seitz
dac73537d2 update changelog 2022-04-04 14:15:40 +08:00
Pascal Seitz
bb5254de12 always serialize, use enum as param 2022-04-04 13:50:23 +08:00
Maxim Kraynyuchenko
be5218c2f6 Company Logos were not visible in Dark Theme. (#1326) 2022-04-04 11:53:31 +09:00
Pascal Seitz
ec9478830a add text test 
move get multiple values to test code
remove sorting term ids per docidi for non facets
 2022-03-30 11:31:33 +08:00
Pascal Seitz
8807bfd13d fast field on string 
enables FAST on string fields, which creates a fastfield containing the term ordinals
 2022-03-29 12:40:10 +08:00
Maxim Kraynyuchenko
447811c111 Update README following sections: features, benchmark illustration & FAQ. (#1318) 
* Updated features, benchmark illustration & FAQ.
* Updated README: Feat,Graph,Non-Feat,Companies,FAQ
 2022-03-23 10:02:09 +09:00
PSeitz
f29acf5d8c fix clippy (#1321) 2022-03-22 12:48:23 +09:00
Uwe Klotz
125707dbe0 Replace chrono with time (#1307) 
For date values `chrono` has been replaced with `time` 
- The `time` crate is re-exported as `tantivy::time` instead of `tantivy::chrono`.
- The type alias `tantivy::DateTime` has been removed.
- `Value::Date` wraps `time::PrimitiveDateTime` without time zone information.
- Internally date/time values are stored as seconds since UNIX epoch in UTC.
- Converting a `time::OffsetDateTime` to `Value::Date` implicitly converts the value into UTC.
If this is not desired do the time zone conversion yourself and use `time::PrimitiveDateTime`
directly instead.

Closes #1304
 2022-03-21 10:50:19 +09:00
Paul Masurel
46d5de920d Removes all usage of block_on, and use a oneshot channel instead. (#1315) 
* Removes all usage of block_on, and use a oneshot channel instead.

Calling `block_on` panics in certain context.
For instance, it panics when it is called in a the context of another
call to block.

Using it in tantivy is unnecessary. We replace it by a thin wrapper
around a oneshot channel that supports both async/sync.

* Removing needless uses of async in the API.

Co-authored-by: PSeitz <PSeitz@users.noreply.github.com>
 2022-03-18 16:54:58 +09:00
PSeitz
d2a7bcf217 fix fmt (#1317) 2022-03-18 15:53:27 +09:00
PSeitz
141b9aa245 Merge pull request #1306 from PSeitz/histogram 
add Histogram aggregation
 2022-03-18 05:03:46 +01:00
PSeitz
c5a6282fa8 Update src/aggregation/bucket/histogram/histogram.rs 
Co-authored-by: Paul Masurel <paul@quickwit.io>
 2022-03-18 04:55:31 +01:00
PSeitz
c0f524e1a3 Update src/aggregation/bucket/histogram/histogram.rs 
Co-authored-by: Paul Masurel <paul@quickwit.io>
 2022-03-18 04:55:25 +01:00
Paul Masurel
958b2bee08 Clippy comments (#1316) 2022-03-17 18:57:55 +09:00
Pascal Seitz
f619658e2c rename 2022-03-17 16:37:57 +08:00
Pascal Seitz
aa391bf843 refactor parameters 2022-03-17 16:28:37 +08:00
Pascal Seitz
47dcbdbeae handle empty results, empty indices, add tests 2022-03-17 10:24:34 +08:00
Pascal Seitz
691245bf20 make code more concise 2022-03-16 14:21:58 +08:00
Pascal Seitz
90798d4b39 address comments, add single bucket test 2022-03-16 13:58:13 +08:00
Pascal Seitz
0b6d9f90cf improve docs 2022-03-16 12:39:26 +08:00
PSeitz
8a5a12d961 add setter to json object options (#1311) 2022-03-16 10:36:30 +09:00
Pascal Seitz
e73542e2e8 Elasticsearch behaviour on hard/extended_bounds 2022-03-15 16:46:45 +08:00
Pascal Seitz
0262e44bbd merge_fruits pass by value 2022-03-15 12:59:22 +08:00
Pascal Seitz
613aad7a8a vec optional, improve performance 2022-03-14 21:29:07 +08:00
Pascal Seitz
1aa88b0c51 improve performance 2022-03-14 20:28:08 +08:00
Pascal Seitz
564fa38085 move sub_aggregations to own vec, use itertools minmax 2022-03-14 16:20:26 +08:00
dependabot[bot]
59ec21479f Update pprof requirement from 0.6 to 0.7 (#1305) 
Updates the requirements on [pprof](https://github.com/tikv/pprof-rs) to permit the latest version.
- [Release notes](https://github.com/tikv/pprof-rs/releases)
- [Changelog](https://github.com/tikv/pprof-rs/blob/master/CHANGELOG.md)
- [Commits](https://github.com/tikv/pprof-rs/commits)

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

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

Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2022-03-14 13:57:22 +09:00
PSeitz
42283f9e91 fix error message UnknownTokenizer (#1308) 
closes #1303
 2022-03-14 13:54:47 +09:00
PSeitz
b105bf72e1 use defaults in meta.json (#1310) 
This change allows to have unset fields in meta.json and fall back to their defaults
Currently it is required to explicitly put e.g. fieldnorms: false
 2022-03-14 13:54:06 +09:00
Pascal Seitz
226f577803 Add Histogram aggregation 2022-03-11 21:52:07 +08:00
Paul Masurel
2e255c4bef Preparing for release 2022-03-09 09:59:08 +09:00
Paul Masurel
387592809f Updated CHANGELOG 2022-03-07 15:31:35 +09:00
Halvor Fladsrud Bø
cedced5bb0 Slop support for phrase queries (#1241) 
Closes #1068
 2022-03-07 15:29:18 +09:00
dependabot[bot]
d31f045872 Bump actions/checkout from 2 to 3 (#1300) 
Bumps [actions/checkout](https://github.com/actions/checkout) from 2 to 3.
- [Release notes](https://github.com/actions/checkout/releases)
- [Changelog](https://github.com/actions/checkout/blob/main/CHANGELOG.md)
- [Commits](https://github.com/actions/checkout/compare/v2...v3)

---
updated-dependencies:
- dependency-name: actions/checkout
  dependency-type: direct:production
  update-type: version-update:semver-major
...

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

Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2022-03-07 11:54:26 +09:00
PSeitz
6656a70d1b Merge pull request #1301 from saroh/1232-doc-fastfield 
update fastfield doc
 2022-03-04 08:18:21 +01:00
saroh
d36e0a9549 fix fastfield doc 2022-03-03 17:43:18 +01:00
Antoine G
8771b2673f Update src/fastfield/writer.rs 
Co-authored-by: PSeitz <PSeitz@users.noreply.github.com>
 2022-03-03 11:25:24 +01:00
Antoine G
a41d3d51a4 Update fastfield_codecs/src/lib.rs 2022-03-03 11:25:06 +01:00
Saroh
cae34ffe47 update fastfield doc 2022-03-02 16:04:15 +01:00
PSeitz
4b62f7907d Merge pull request #1297 from PSeitz/fix_clippy 
fix clippy issues
 2022-03-02 10:11:56 +01:00
Pascal Seitz
7fa6a0b665 cargo fmt 2022-03-02 09:24:14 +01:00
PSeitz
458ed29a31 Merge pull request #1299 from saroh/1232-doc-lint 
doc lint for errors and aggregations
 2022-03-02 09:22:07 +01:00
Antoine G
e37775fe21 iff->if or if and only if (#1298) 
* has_xxx is_xxx -> if, these function usualy define equivalence
xxx returns bool -> specify equivalence when appropriate

* fix doc
 2022-03-02 11:00:00 +09:00
Saroh
1cd2434a32 fix(aggregations) Readme 2022-03-01 20:37:48 +01:00
Saroh
de2cba6d1e error definitions 2022-03-01 20:13:59 +01:00
Paul Masurel
c0b1a58d27 Apply suggestions from code review 2022-03-01 18:41:58 +09:00
Paul Masurel
848b795b9f Apply suggestions from code review 2022-03-01 18:37:51 +09:00
Pascal Seitz
091b668624 fix clippy issues 2022-03-01 08:58:51 +01:00
Paul Masurel
5004290daa Return an error on certain type of corruption. (#1296) 2022-03-01 11:35:56 +09:00
StyMaar
5d2c2b804c Fix link to RamDirectory and MMapDirectory in Directory's documentation (#1295) 2022-03-01 09:46:53 +09:00
PSeitz
1a92b588e0 Merge pull request #1294 from PSeitz/aggregation 
fix intermediate result de/serialization
 2022-02-28 08:39:23 +01:00
Pascal Seitz
010e92c118 fix intermediate result de/serialization 
return None for empty average/stats metric
add test for de/serialization of intermediate result
add test for metric on empty result
 2022-02-25 16:39:57 +01:00
Paul Masurel
2ead010c83 Tantivy quickwit (#1293) 
* Added sstable and enabling it by default, and parallel boolean query.
* Added async API for FileSlice.
* Added async get_doc
* Reduce blocksize to 32_000
* Added debug logs

Quickwit specific feature a hidden behind the quickwit feature flag.
 2022-02-25 17:32:49 +09:00
PSeitz
c4f66eb185 improve validation in aggregation, extend invalid field test (#1292) 
* improve validation in aggregation, extend invalid field test

improve validation in aggregation
extend invalid field test
Fixes #1291

* collect fast field names on request structure

* fix visibility of AggregationSegmentCollector
 2022-02-25 15:21:19 +09:00
Paul Masurel
d7b46d2137 Added JSON Type (#1270) 
- Removed useless copy when ingesting JSON.
- Bugfix in phrase query with a missing field norms.
- Disabled range query on default fields

Closes #1251
 2022-02-24 16:25:22 +09:00
PSeitz
d042ce74c7 Merge pull request #1289 from PSeitz/numeric_options 
rename IntOptions to NumericOptions
 2022-02-23 14:04:40 +01:00
PSeitz
7ba9e662b8 Merge pull request #1290 from PSeitz/improve_docs 
improve aggregation docs
 2022-02-23 14:04:20 +01:00
Pascal Seitz
fdd5ef85e5 improve aggregation docs 2022-02-22 10:37:54 +01:00
Pascal Seitz
704498a1ac rename IntOptions to NumericOptions 
keep IntOptions with deprecation warning
Fixes #1286
 2022-02-21 22:20:07 +01:00
PSeitz
1232af7928 fix docs (#1288) 2022-02-21 23:15:58 +09:00
Paul Masurel
d37633e034 Minor changes in indexing. (#1285) 2022-02-21 17:16:52 +09:00
Paul Masurel
9815067171 Minor changes 2022-02-21 13:55:01 +09:00
PSeitz
972cb6c26d Aggregation (#1276) 
Added support for aggregation compatible with Elasticsearch's API.
 2022-02-21 09:59:11 +09:00
Paul Masurel
4dc80cfa25 Removes TokenStream chain. (#1283) 
This change is mostly motivated by the introduction of json object.

We need to be able to inject a position object to make the position
shift.
 2022-02-21 09:51:27 +09:00
PSeitz
cef145790c Fix opening bytes index with dynamic codec (#1279) 
* Fix opening bytes index with dynamic codec

Fix #1278

* extend proptest to cover bytes field codec bug
 2022-02-18 20:44:21 +09:00
Paul Masurel
e05e2a0c51 Added profiling to indexing bench (#1282) 2022-02-18 20:43:28 +09:00
Paul Masurel
e028515caf Simplified expull code. (#1281) 2022-02-18 18:57:10 +09:00
Paul Masurel
850b9eaea4 added a bench to measure the perf of indexing logs (#1275) 2022-02-18 16:48:29 +09:00
Shikhar Bhushan
505e6a440c Remove test assertion sensitive to background segment merging (#1274) 2022-02-17 10:59:46 +09:00
Koichi Akabe
fcd651f6a9 Add Vaporetto tokenizer to README (#1271) 
* Add Vaporetto tokenizer to README

* Update README.md
 2022-02-14 18:19:57 +09:00
Paul Masurel
e6653228a9 Renamed github workflows (#1269) 2022-02-04 15:10:24 +09:00
Paul Masurel
bdedefe07d Adding an IndexingContext object (#1268) 2022-02-04 15:08:01 +09:00
Paul Masurel
13a4473faa Removing obsolete clippy allow thingy. 2022-02-01 11:54:01 +09:00
Paul Masurel
2069e3e52b Fixing clippy comments 2022-02-01 10:24:05 +09:00
Paul Masurel
0d8263cba1 Using nightly to format 2022-01-31 16:10:11 +09:00
Paul Masurel
65b365b81c Fixing all-features build. 2022-01-31 14:41:14 +09:00
dependabot[bot]
4c1366da87 Update fastdivide requirement from 0.3 to 0.4 (#1265) 
Updates the requirements on fastdivide to permit the latest version.

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

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

Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2022-01-31 11:26:50 +09:00
Paul Masurel
eca6628b3c Minor refactoring (#1266) 2022-01-28 15:55:55 +09:00
Paul Masurel
9679c5f306 Rename quickwit-inc -> quickwit-oss 2022-01-27 15:37:09 +09:00
Shikhar Bhushan
5a2497b6fd Avoid exposing TrackedObject from Warmer API (#1264) 2022-01-25 10:04:08 +09:00
Shikhar Bhushan
99d4b1a177 Searcher Warming API (#1261) 
Adds an API to register Warmers in the IndexReader.

Co-authored-by: Paul Masurel <paul@quickwit.io>
 2022-01-20 23:40:25 +09:00
Paul Masurel
732f6847c0 Field type with codes (#1255) 
* Term are now typed.

This change is backward compatible:
While the Term has a byte representation that is modified, a Term itself
is a transient object that is not serialized as is in the index.

Its .field() and .value_bytes() on the other hand are unchanged.
This change offers better Debug information for terms.

While not necessary it also will help in the support for JSON types.

* Renamed Hierarchical Facet -> Facet
 2022-01-07 20:49:00 +09:00
Paul Masurel
1c6d9bdc6a Comparison of Value based on serialization. (#1250) 2022-01-07 20:31:26 +09:00
Paul Masurel
3ea6800ac5 Pleasing clippy (#1253) 2022-01-06 16:41:24 +09:00
Antoine G
395303b644 Collector + directory doc fixes (#1247) 
* doc(collector)

* doc(directory)

* doc(misc)

* wording
 2022-01-04 09:22:58 +09:00
Daniel Müller
2c200b46cb Use test-log instead of test-env-log (#1248) 
The test-env-log crate has been renamed to test-log to better reflect
its intent of not only catering to env_logger specific initialization
but also tracing (and potentially others in the future).
This change updates the crate to use test-log instead of the now
deprecated test-env-log.
 2022-01-04 09:20:30 +09:00
Liam Warfield
17e00df112 Change Snippet.fragments -> Snippet.fragment (#1243) 
* Change Snippet.fragments -> Snippet.fragment
* Apply suggestions from code review

Co-authored-by: Liam Warfield <lwarfield@arista.com>
 2022-01-03 22:23:51 +09:00
Antoine G
3129d86743 doc(termdict) expose structs (#1242) 
* doc(termdict) expose structs
also add merger doc + lint
refs #1232
 2022-01-03 22:20:31 +09:00
Shikhar Bhushan
e5e252cbc0 LogMergePolicy knob del_docs_percentage_before_merge (#1238) 
Add a knob to LogMergePolicy to always merge segments that exceed a threshold of deleted docs

Closes #115
 2021-12-20 13:14:56 +09:00
Paul Masurel
b2da82f151 Making MergeCandidate public in order to allow the usage of custom merge (#1237) 
policies.

Closes #1235
 2021-12-13 09:54:21 +09:00
Paul Masurel
c81b3030fa Issue/922b (#1233) 
* Add a NORMED options on field

Make fieldnorm indexation optional:

* for all types except text => added a NORMED options
* for text field
** if STRING, field has not fieldnorm retained
** if TEXT, field has fieldnorm computed

* Finalize making fieldnorm optional for all field types.

- Using Option for fieldnorm readers.
 2021-12-10 21:12:29 +09:00
Paul Masurel
9e66c75fc6 Using stable in CI as rustc nightly seems broken 2021-12-10 18:45:23 +09:00
Paul Masurel
ebdbb6bd2e Fixing compilation warnings & clippy comments. 2021-12-10 16:47:59 +09:00
Antoine G
c980b19dd9 canonicalize path when opening MmapDirectory (#1231) 
* canonicalize path when opening `MmapDirectory`
fixes #1229
 2021-12-09 10:19:52 +09:00
Paul Masurel
098eea843a Reducing the number of call to fsync on the directory. (#1228) 
This work by introducing a new API method in the Directory
trait. The user needs to explicitely call this method.
(In particular, once before a commmit)

Closes #1225
 2021-12-03 03:10:52 +00:00
Paul Masurel
466dc8233c Cargo fmt 2021-12-02 18:46:28 +09:00
Paul Masurel
03c2f6ece2 We are missing 4 bytes in the LZ4 compression buffer. (#1226) 
Closes #831
 2021-12-02 16:00:29 +09:00
Paul Masurel
1d4e9a29db Cargo fmt 2021-12-02 15:51:44 +09:00
Paul Masurel
f378d9a57b Pleasing clippy 2021-12-02 14:48:33 +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
Paul Masurel
46b86a7976 Bounced version and edited changelog 2021-09-10 23:05:09 +09:00
PSeitz
3bc177e69d fix #1151 (#1152) 
* fix #1151

Fixes a off by one error in the stats for the index fast field in the multi value fast field.
When retrieving the data range for a docid, `get(doc)..get(docid+1)` is requested. On creation
the num_vals statistic was set to doc instead of docid + 1. In the multivaluelinearinterpol fast
field the last value was therefore not serialized (and would return 0 instead in most cases).
So the last document get(lastdoc)..get(lastdoc + 1) would return the invalid range `value..0`.

This PR adds a proptest to cover this scenario. A combination of a large number values, since multilinear
interpolation is only active for more than 5_000 values, and a merge is required.
 2021-09-10 23:00:37 +09:00
PSeitz
319609e9c1 test cargo-llvm-cov (#1149) 2021-09-03 22:00:43 +09:00
Kanji Yomoda
9d87b89718 Fix incorrect comment for Index::create_in_dir (#1148) 
* Fix incorrect comment for Index::create_in_dir
 2021-09-03 10:37:16 +09:00
Tomoko Uchida
dd81e38e53 Add WhitespaceTokenizer (#1147) 
* Add WhitespaceTokenizer.
 2021-08-29 18:20:49 +09:00
Paul Masurel
9f32b22602 Preparing for release. 2021-08-26 09:07:08 +09:00
sigaloid
096ce7488e Resolve some clippys, format (#1144) 
* cargo +nightly clippy --fix -Z unstable-options
 2021-08-26 08:46:00 +09:00
PSeitz
a1782dd172 Update index_sorting.md 2021-08-25 07:55:50 +01:00
PSeitz
000d76b11a Update index_sorting.md 2021-08-24 19:28:06 +01:00
PSeitz
abd29f6646 Update index_sorting.md 2021-08-24 19:26:19 +01:00
PSeitz
b4ecf0ab2f Merge pull request #1146 from tantivy-search/sorting_doc 
add sorting to book
 2021-08-23 17:37:54 +01:00
Pascal Seitz
798f7dbf67 add sorting to book 2021-08-23 17:36:41 +01:00
PSeitz
06a2e47c8d Merge pull request #1145 from tantivy-search/blub2 
cargo fmt
 2021-08-21 18:52:50 +01:00
Pascal Seitz
e0b83eb291 cargo fmt 2021-08-21 18:52:10 +01:00
PSeitz
13401f46ea add wildcard mention 2021-08-21 18:10:33 +01:00
PSeitz
1a45b030dc Merge pull request #1141 from tantivy-search/tantivy_common 
dissolve common module
 2021-08-20 08:03:37 +01:00
Pascal Seitz
62052bcc2d add missing test function 
closes #1139
 2021-08-20 07:26:22 +01:00
Pascal Seitz
3265f7bec3 dissolve common module 2021-08-19 23:26:34 +01:00
Pascal Seitz
ee0881712a move bitset to common crate, move composite file to directory 2021-08-19 17:45:09 +01:00
PSeitz
483e0336b6 Merge pull request #1140 from tantivy-search/tantivy_common 
rename common to tantivy-common
 2021-08-19 13:02:54 +01:00
Pascal Seitz
3e8f267e33 rename common to tantivy-common 2021-08-19 10:27:20 +01:00
Paul Masurel
3b247fd968 Version bump 2021-08-19 10:12:30 +09:00
Paul Masurel
750f6e6479 Removed obsolete unit test (#1138) 2021-08-19 10:07:49 +09:00
Evance Soumaoro
5b475e6603 Checksum validation using active files (#1130) 
* now validate checksum uses segment files not managed files
 2021-08-19 10:03:20 +09:00
PSeitz
0ca7f73dc5 add docs badge, fix build badge 2021-08-13 19:40:33 +01:00
PSeitz
47ed18845e Merge pull request #1136 from tantivy-search/minor_fixes 
more docs detail
 2021-08-13 18:11:47 +01:00
Pascal Seitz
dc141cdb29 more docs detail 
remove code duplicate
 2021-08-13 17:40:13 +01:00
PSeitz
f6cf6e889b Merge pull request #1133 from tantivy-search/merge_overflow 
test doc_freq and term_freq in sorted index
 2021-08-05 07:53:46 +01:00
Pascal Seitz
f379a80233 test doc_freq and term_freq in sorted index 2021-08-03 11:38:05 +01:00
PSeitz
4a320fd1ff fix delta position in merge and index sorting (#1132) 
fixes #1125
 2021-08-03 18:06:36 +09:00
PSeitz
85d23e8e3b Merge pull request #1129 from tantivy-search/merge_overflow 
add long running test in ci
 2021-08-02 15:54:31 +01:00
Pascal Seitz
022ab9d298 don't run as pr 2021-08-02 15:44:00 +01:00
Pascal Seitz
605e8603dc add positions to long running test 2021-08-02 15:29:49 +01:00
Pascal Seitz
70f160b329 add long running test in ci 2021-08-02 11:35:39 +01:00
PSeitz
6d265e6bed fix gh action name 2021-08-02 10:38:01 +01:00
PSeitz
fdc512391b Merge pull request #1128 from tantivy-search/merge_overflow 
add sort to functional test, add env for iterations
 2021-08-02 10:29:16 +01:00
Pascal Seitz
108714c934 add sort to functional test, add env for iterations 2021-08-02 10:11:17 +01:00
Paul Masurel
44e8cf98a5 Cargo fmt 2021-07-30 15:30:01 +09:00
Paul Masurel
f0ee69d9e9 Remove the complicated block search logic for a simpler branchless (#1124) 
binary search

The code is simpler and faster.

Before
test postings::bench::bench_segment_intersection                                                                         ... bench:   2,093,697 ns/iter (+/- 115,509)
test postings::bench::bench_skip_next_p01                                                                                ... bench:      58,585 ns/iter (+/- 796)
test postings::bench::bench_skip_next_p1                                                                                 ... bench:     160,872 ns/iter (+/- 5,164)
test postings::bench::bench_skip_next_p10                                                                                ... bench:     615,229 ns/iter (+/- 25,108)
test postings::bench::bench_skip_next_p90                                                                                ... bench:   1,120,509 ns/iter (+/- 22,271)

After
test postings::bench::bench_segment_intersection                                                                         ... bench:   1,747,726 ns/iter (+/- 52,867)
test postings::bench::bench_skip_next_p01                                                                                ... bench:      55,205 ns/iter (+/- 714)
test postings::bench::bench_skip_next_p1                                                                                 ... bench:     131,433 ns/iter (+/- 2,814)
test postings::bench::bench_skip_next_p10                                                                                ... bench:     478,830 ns/iter (+/- 12,794)
test postings::bench::bench_skip_next_p90                                                                                ... bench:     931,082 ns/iter (+/- 31,468)
 2021-07-30 14:38:42 +09:00
Evance Soumaoro
b8a10c8406 switched to memmap2-rs (#1120) 2021-07-27 18:40:41 +09:00
PSeitz
ff4813529e add comments on compression (#1119) 2021-07-26 22:54:22 +09:00
PSeitz
470bc18e9b Merge pull request #1118 from tantivy-search/remove_rand 
move rand to optional dependencies
 2021-07-21 18:01:22 +01:00
293 changed files with 123869 additions and 8033 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

33
.github/workflows/coverage.yml vendored
View File

@@ -1,27 +1,24 @@
name: coverage
name: Coverage
on:
push:
branches: [ main ]
pull_request:
branches: [ main ]
jobs:
test:
name: coverage
runs-on: ubuntu-latest
container:
image: xd009642/tarpaulin:develop-nightly
options: --security-opt seccomp=unconfined
coverage:
runs-on: ubuntu-latest
steps:
- name: Checkout repository
uses: actions/checkout@v2
- name: Generate code coverage
run: |
cargo +nightly tarpaulin --verbose --all-features --workspace --timeout 120 --out Xml
- name: Upload to codecov.io
uses: codecov/codecov-action@v1
- uses: actions/checkout@v3
- name: Install Rust
run: rustup toolchain install nightly --component llvm-tools-preview
- uses: taiki-e/install-action@cargo-llvm-cov
- name: Generate code coverage
run: cargo +nightly llvm-cov --all-features --workspace --lcov --output-path lcov.info
- name: Upload coverage to Codecov
uses: codecov/codecov-action@v3
with:
# token: ${{secrets.CODECOV_TOKEN}} # not required for public repos
fail_ci_if_error: true
token: ${{ secrets.CODECOV_TOKEN }} # not required for public repos
files: lcov.info
fail_ci_if_error: true

24
.github/workflows/long_running.yml vendored Normal file
View File

@@ -0,0 +1,24 @@
name: Long running tests
on:
push:
branches: [ main ]
env:
CARGO_TERM_COLOR: always
NUM_FUNCTIONAL_TEST_ITERATIONS: 20000
jobs:
functional_test_unsorted:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- name: Run indexing_unsorted
run: cargo test indexing_unsorted -- --ignored
functional_test_sorted:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v3
- name: Run indexing_sorted
run: cargo test indexing_sorted -- --ignored

29
.github/workflows/test.yml vendored
View File

@@ -1,4 +1,4 @@
name: Rust
name: Unit tests
on:
push:
@@ -10,12 +10,12 @@ env:
CARGO_TERM_COLOR: always
jobs:
build:
test:
runs-on: ubuntu-latest
steps:
- uses: actions/checkout@v2
- uses: actions/checkout@v3
- name: Build
run: cargo build --verbose --workspace
- name: Install latest nightly to test also against unstable feature flag
@@ -24,7 +24,26 @@ jobs:
toolchain: nightly
override: true
components: rustfmt
- name: Install latest nightly to test also against unstable feature flag
uses: actions-rs/toolchain@v1
with:
toolchain: stable
override: true
components: rustfmt, clippy
- name: Run tests
run: cargo test --all-features --verbose --workspace
run: cargo +stable test --features mmap,brotli-compression,lz4-compression,snappy-compression,zstd-compression,failpoints --verbose --workspace
- name: Run tests quickwit feature
run: cargo +stable test --features mmap,quickwit,failpoints --verbose --workspace
- name: Check Formatting
run: cargo fmt --all -- --check
run: cargo +nightly fmt --all -- --check
- uses: actions-rs/clippy-check@v1
with:
toolchain: stable
token: ${{ secrets.GITHUB_TOKEN }}
args: --tests

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

45
CHANGELOG.md
View File

@@ -1,3 +1,46 @@
Tantivy 0.18
================================
- For date values `chrono` has been replaced with `time` (@uklotzde) #1304 :
- The `time` crate is re-exported as `tantivy::time` instead of `tantivy::chrono`.
- The type alias `tantivy::DateTime` has been removed.
- `Value::Date` wraps `time::PrimitiveDateTime` without time zone information.
- Internally date/time values are stored as seconds since UNIX epoch in UTC.
- Converting a `time::OffsetDateTime` to `Value::Date` implicitly converts the value into UTC.
If this is not desired do the time zone conversion yourself and use `time::PrimitiveDateTime`
directly instead.
- Add [histogram](https://github.com/quickwit-oss/tantivy/pull/1306) aggregation (@PSeitz)
- Add support for fastfield on text fields (@PSeitz)
- Add terms aggregation (@PSeitz)
- Add support for zstd compression (@kryesh)
Tantivy 0.17
================================
- LogMergePolicy now triggers merges if the ratio of deleted documents reaches a threshold (@shikhar @fulmicoton) [#115](https://github.com/quickwit-oss/tantivy/issues/115)
- Adds a searcher Warmer API (@shikhar @fulmicoton)
- Change to non-strict schema. Ignore fields in data which are not defined in schema. Previously this returned an error. #1211
- Facets are necessarily indexed. Existing index with indexed facets should work out of the box. Index without facets that are marked with index: false should be broken (but they were already broken in a sense). (@fulmicoton) #1195 .
- Bugfix that could in theory impact durability in theory on some filesystems [#1224](https://github.com/quickwit-oss/tantivy/issues/1224)
- Schema now offers not indexing fieldnorms (@lpouget) [#922](https://github.com/quickwit-oss/tantivy/issues/922)
- Reduce the number of fsync calls [#1225](https://github.com/quickwit-oss/tantivy/issues/1225)
- Fix opening bytes index with dynamic codec (@PSeitz) [#1278](https://github.com/quickwit-oss/tantivy/issues/1278)
- Added an aggregation collector for range, average and stats compatible with Elasticsearch. (@PSeitz)
- Added a JSON schema type @fulmicoton [#1251](https://github.com/quickwit-oss/tantivy/issues/1251)
- Added support for slop in phrase queries @halvorboe [#1068](https://github.com/quickwit-oss/tantivy/issues/1068)
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
=========================
- Bugfix in the filesum check. (@evanxg852000) #1127
- Bugfix in positions when the index is sorted by a field. (@appaquet) #1125
Tantivy 0.15.3
=========================
- Major bugfix. Deleting documents was broken when the index was sorted by a field. (@appaquet, @fulmicoton) #1101
@@ -104,7 +147,7 @@ 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-oss/tantivy/blob/main/examples/custom_tokenizer.rs
to check for some code sample.
Tantivy 0.11.3

112
Cargo.toml
View File

@@ -1,73 +1,81 @@
[package]
name = "tantivy"
version = "0.16.0-dev"
version = "0.18.0"
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-oss/tantivy"
repository = "https://github.com/quickwit-oss/tantivy"
readme = "README.md"
keywords = ["search", "information", "retrieval"]
edition = "2018"
edition = "2021"
[dependencies]
base64 = "0.13"
oneshot = "0.1.3"
base64 = "0.13.0"
byteorder = "1.4.3"
crc32fast = "1.2.1"
once_cell = "1.7.2"
regex ={ version = "1.5.4", default-features = false, features = ["std"] }
tantivy-fst = "0.3"
memmap = {version = "0.7", optional=true}
lz4_flex = { version = "0.8.0", default-features = false, features = ["checked-decode"], optional = true }
brotli = { version = "3.3", optional = true }
crc32fast = "1.3.2"
once_cell = "1.10.0"
regex = { version = "1.5.5", default-features = false, features = ["std", "unicode"] }
tantivy-fst = "0.3.0"
memmap2 = { version = "0.5.3", optional = true }
lz4_flex = { version = "0.9.2", default-features = false, features = ["checked-decode"], optional = true }
brotli = { version = "3.3.4", optional = true }
zstd = { version = "0.11", optional = true }
snap = { version = "1.0.5", optional = true }
tempfile = { version = "3.2", optional = true }
log = "0.4.14"
serde = { version = "1.0.126", features = ["derive"] }
serde_json = "1.0.64"
num_cpus = "1.13"
tempfile = { version = "3.3.0", optional = true }
log = "0.4.16"
serde = { version = "1.0.136", features = ["derive"] }
serde_json = "1.0.79"
num_cpus = "1.13.1"
fs2={ version = "0.4.3", optional = true }
levenshtein_automata = "0.2"
uuid = { version = "0.8.2", features = ["v4", "serde"] }
crossbeam = "0.8"
futures = { version = "0.3.15", features = ["thread-pool"] }
tantivy-query-grammar = { version="0.15.0", path="./query-grammar" }
tantivy-bitpacker = { version="0.1", path="./bitpacker" }
common = { version="0.1", path="./common" }
fastfield_codecs = { version="0.1", path="./fastfield_codecs", default-features = false }
ownedbytes = { version="0.1", path="./ownedbytes" }
stable_deref_trait = "1.2"
rust-stemmers = "1.2"
downcast-rs = "1.2"
levenshtein_automata = "0.2.1"
uuid = { version = "1.0.0", features = ["v4", "serde"] }
crossbeam-channel = "0.5.4"
tantivy-query-grammar = { version="0.18.0", path="./query-grammar" }
tantivy-bitpacker = { version="0.2", path="./bitpacker" }
common = { version = "0.3", path = "./common/", package = "tantivy-common" }
fastfield_codecs = { version="0.2", path="./fastfield_codecs", default-features = false }
ownedbytes = { version="0.3", path="./ownedbytes" }
stable_deref_trait = "1.2.0"
rust-stemmers = "1.2.0"
downcast-rs = "1.2.0"
bitpacking = { version = "0.8.4", default-features = false, features = ["bitpacker4x"] }
census = "0.4"
census = "0.4.0"
fnv = "1.0.7"
thiserror = "1.0.24"
thiserror = "1.0.30"
htmlescape = "0.3.1"
fail = "0.4"
murmurhash32 = "0.2"
chrono = "0.4.19"
smallvec = "1.6.1"
rayon = "1.5"
lru = "0.6.5"
fastdivide = "0.3"
itertools = "0.10.0"
measure_time = "0.7.0"
fail = "0.5.0"
murmurhash32 = "0.2.0"
time = { version = "0.3.9", features = ["serde-well-known"] }
smallvec = "1.8.0"
rayon = "1.5.2"
lru = "0.7.5"
fastdivide = "0.4.0"
itertools = "0.10.3"
measure_time = "0.8.2"
pretty_assertions = "1.2.1"
serde_cbor = { version = "0.11.2", optional = true }
async-trait = "0.1.53"
[target.'cfg(windows)'.dependencies]
winapi = "0.3.9"
[dev-dependencies]
rand = "0.8.3"
rand = "0.8.5"
maplit = "1.0.2"
matches = "0.1.8"
proptest = "1.0"
criterion = "0.3.4"
matches = "0.1.9"
proptest = "1.0.0"
criterion = "0.3.5"
test-log = "0.2.10"
env_logger = "0.9.0"
pprof = { version = "0.9.0", features = ["flamegraph", "criterion"] }
futures = "0.3.21"
[dev-dependencies.fail]
version = "0.4"
version = "0.5.0"
features = ["failpoints"]
[profile.release]
@@ -81,22 +89,21 @@ overflow-checks = true
[features]
default = ["mmap", "lz4-compression" ]
mmap = ["fs2", "tempfile", "memmap"]
mmap = ["fs2", "tempfile", "memmap2"]
brotli-compression = ["brotli"]
lz4-compression = ["lz4_flex"]
snappy-compression = ["snap"]
zstd-compression = ["zstd"]
failpoints = ["fail/failpoints"]
unstable = [] # useful for benches.
wasm-bindgen = ["uuid/wasm-bindgen"]
quickwit = ["serde_cbor"]
[workspace]
members = ["query-grammar", "bitpacker", "common", "fastfield_codecs", "ownedbytes"]
[badges]
travis-ci = { repository = "tantivy-search/tantivy" }
# Following the "fail" crate best practises, we isolate
# tests that define specific behavior in fail check points
# in a different binary.
@@ -112,3 +119,8 @@ required-features = ["fail/failpoints"]
[[bench]]
name = "analyzer"
harness = false
[[bench]]
name = "index-bench"
harness = false

3
Makefile
View File

@@ -1,3 +1,6 @@
test:
echo "Run test only... No examples."
cargo test --tests --lib
fmt:
cargo +nightly fmt --all

87
README.md
View File

@@ -1,26 +1,14 @@
[![Build Status](https://travis-ci.org/tantivy-search/tantivy.svg?branch=main)](https://travis-ci.org/tantivy-search/tantivy)
[![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)
[![Docs](https://docs.rs/tantivy/badge.svg)](https://docs.rs/crate/tantivy/)
[![Build Status](https://github.com/quickwit-oss/tantivy/actions/workflows/test.yml/badge.svg)](https://github.com/quickwit-oss/tantivy/actions/workflows/test.yml)
[![codecov](https://codecov.io/gh/quickwit-oss/tantivy/branch/main/graph/badge.svg)](https://codecov.io/gh/quickwit-oss/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)
[![Build status](https://ci.appveyor.com/api/projects/status/r7nb13kj23u8m9pj/branch/main?svg=true)](https://ci.appveyor.com/project/fulmicoton/tantivy/branch/main)
[![Crates.io](https://img.shields.io/crates/v/tantivy.svg)](https://crates.io/crates/tantivy)
![Tantivy](https://tantivy-search.github.io/logo/tantivy-logo.png)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/0)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/0)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/1)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/1)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/2)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/2)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/3)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/3)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/4)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/4)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/5)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/5)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/6)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/6)
[![](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/images/7)](https://sourcerer.io/fame/fulmicoton/tantivy-search/tantivy/links/7)
[![Become a patron](https://c5.patreon.com/external/logo/become_a_patron_button.png)](https://www.patreon.com/fulmicoton)
**Tantivy** is a **full text search engine library** written in Rust.
**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
an off-the-shelf search engine server, but rather a crate that can be used
@@ -28,19 +16,23 @@ to build such a search engine.
Tantivy is, in fact, strongly inspired by Lucene's design.
If you are looking for an alternative to Elasticsearch or Apache Solr, check out [Quickwit](https://github.com/quickwit-oss/quickwit), our search engine built on top of Tantivy.
# Benchmark
The following [benchmark](https://tantivy-search.github.io/bench/) break downs
performance for different type of queries / collection.
The following [benchmark](https://tantivy-search.github.io/bench/) breakdowns
performance for different types of queries/collections.
Your mileage WILL vary depending on the nature of queries and their load.
<img src="doc/assets/images/searchbenchmark.png">
# Features
- Full-text search
- Configurable tokenizer (stemming available for 17 Latin languages with third party support for Chinese ([tantivy-jieba](https://crates.io/crates/tantivy-jieba) and [cang-jie](https://crates.io/crates/cang-jie)), Japanese ([lindera](https://github.com/lindera-morphology/lindera-tantivy) and [tantivy-tokenizer-tiny-segmenter](https://crates.io/crates/tantivy-tokenizer-tiny-segmenter)) and Korean ([lindera](https://github.com/lindera-morphology/lindera-tantivy) + [lindera-ko-dic-builder](https://github.com/lindera-morphology/lindera-ko-dic-builder))
- Configurable tokenizer (stemming available for 17 Latin languages with third party support for Chinese ([tantivy-jieba](https://crates.io/crates/tantivy-jieba) and [cang-jie](https://crates.io/crates/cang-jie)), Japanese ([lindera](https://github.com/lindera-morphology/lindera-tantivy), [Vaporetto](https://crates.io/crates/vaporetto_tantivy), and [tantivy-tokenizer-tiny-segmenter](https://crates.io/crates/tantivy-tokenizer-tiny-segmenter)) and Korean ([lindera](https://github.com/lindera-morphology/lindera-tantivy) + [lindera-ko-dic-builder](https://github.com/lindera-morphology/lindera-ko-dic-builder))
- Fast (check out the :racehorse: :sparkles: [benchmark](https://tantivy-search.github.io/bench/) :sparkles: :racehorse:)
- Tiny startup time (<10ms), perfect for command line tools
- Tiny startup time (<10ms), perfect for command-line tools
- BM25 scoring (the same as Lucene)
- Natural query language (e.g. `(michael AND jackson) OR "king of pop"`)
- Phrase queries search (e.g. `"michael jackson"`)
@@ -55,36 +47,37 @@ Your mileage WILL vary depending on the nature of queries and their load.
- Range queries
- Faceted search
- Configurable indexing (optional term frequency and position indexing)
- JSON Field
- Aggregation Collector: range buckets, average, and stats metrics
- LogMergePolicy with deletes
- Searcher Warmer API
- Cheesy logo with a horse
## Non-features
- Distributed search is out of the scope of Tantivy. That being said, Tantivy is a
library upon which one could build a distributed search. Serializable/mergeable collector state for instance,
are within the scope of Tantivy.
Distributed search is out of the scope of Tantivy, but if you are looking for this feature, check out [Quickwit](https://github.com/quickwit-oss/quickwit/).
# Getting started
Tantivy works on stable Rust (>= 1.27) and supports Linux, MacOS, and Windows.
Tantivy works on stable Rust (>= 1.27) and supports Linux, macOS, and Windows.
- [Tantivy's simple search example](https://tantivy-search.github.io/examples/basic_search.html)
- [tantivy-cli and its tutorial](https://github.com/tantivy-search/tantivy-cli) - `tantivy-cli` is an actual command line interface that makes it easy for you to create a search engine,
- [tantivy-cli and its tutorial](https://github.com/quickwit-oss/tantivy-cli) - `tantivy-cli` is an actual command-line interface that makes it easy for you to create a search engine,
index documents, and search via the CLI or a small server with a REST API.
It walks you through getting a wikipedia search engine up and running in a few minutes.
It walks you through getting a Wikipedia search engine up and running in a few minutes.
- [Reference doc for the last released version](https://docs.rs/tantivy/)
# How can I support this project?
There are many ways to support this project.
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 +89,7 @@ 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-oss/tantivy.git
cd tantivy
cargo build
```
@@ -132,3 +125,31 @@ By default, `rustc` compiles everything in the `examples/` directory in debug mo
rust-gdb target/debug/examples/$EXAMPLE_NAME
$ gdb run
```
# Companies Using Tantivy
<p align="left">
<img align="center" src="doc/assets/images/Nuclia.png#gh-light-mode-only" alt="Nuclia" height="25" width="auto" /> &nbsp;
<img align="center" src="doc/assets/images/humanfirst.png#gh-light-mode-only" alt="Humanfirst.ai" height="30" width="auto" />
<img align="center" src="doc/assets/images/element.io.svg#gh-light-mode-only" alt="Element.io" height="25" width="auto" />
<img align="center" src="doc/assets/images/nuclia-dark-theme.png#gh-dark-mode-only" alt="Nuclia" height="35" width="auto" /> &nbsp;
<img align="center" src="doc/assets/images/humanfirst.ai-dark-theme.png#gh-dark-mode-only" alt="Humanfirst.ai" height="25" width="auto" />&nbsp; &nbsp;
<img align="center" src="doc/assets/images/element-dark-theme.png#gh-dark-mode-only" alt="Element.io" height="25" width="auto" />
</p>
# FAQ
### Can I use Tantivy in other languages?
- Python → [tantivy-py](https://github.com/quickwit-oss/tantivy-py)
- Ruby → [tantiny](https://github.com/baygeldin/tantiny)
You can also find other bindings on [GitHub](https://github.com/search?q=tantivy) but they may be less maintained.
### What are some examples of Tantivy use?
- [seshat](https://github.com/matrix-org/seshat/): A matrix message database/indexer
- [tantiny](https://github.com/baygeldin/tantiny): Tiny full-text search for Ruby
- [lnx](https://github.com/lnx-search/lnx): adaptable, typo tolerant search engine with a REST API
- and [more](https://github.com/search?q=tantivy)!
### On average, how much faster is Tantivy compared to Lucene?
- According to our [search latency benchmark](https://tantivy-search.github.io/bench/), Tantivy is approximately 2x faster than Lucene.

100000
benches/hdfs.json Normal file
View File

File diff suppressed because it is too large Load Diff

121
benches/index-bench.rs Normal file
View File

@@ -0,0 +1,121 @@
use criterion::{criterion_group, criterion_main, Criterion};
use pprof::criterion::{Output, PProfProfiler};
use tantivy::schema::{INDEXED, STORED, STRING, TEXT};
use tantivy::Index;
const HDFS_LOGS: &str = include_str!("hdfs.json");
const NUM_REPEATS: usize = 2;
pub fn hdfs_index_benchmark(c: &mut Criterion) {
let schema = {
let mut schema_builder = tantivy::schema::SchemaBuilder::new();
schema_builder.add_u64_field("timestamp", INDEXED);
schema_builder.add_text_field("body", TEXT);
schema_builder.add_text_field("severity", STRING);
schema_builder.build()
};
let schema_with_store = {
let mut schema_builder = tantivy::schema::SchemaBuilder::new();
schema_builder.add_u64_field("timestamp", INDEXED | STORED);
schema_builder.add_text_field("body", TEXT | STORED);
schema_builder.add_text_field("severity", STRING | STORED);
schema_builder.build()
};
let dynamic_schema = {
let mut schema_builder = tantivy::schema::SchemaBuilder::new();
schema_builder.add_json_field("json", TEXT);
schema_builder.build()
};
let mut group = c.benchmark_group("index-hdfs");
group.sample_size(20);
group.bench_function("index-hdfs-no-commit", |b| {
b.iter(|| {
let index = Index::create_in_ram(schema.clone());
let index_writer = index.writer_with_num_threads(1, 100_000_000).unwrap();
for _ in 0..NUM_REPEATS {
for doc_json in HDFS_LOGS.trim().split("\n") {
let doc = schema.parse_document(doc_json).unwrap();
index_writer.add_document(doc).unwrap();
}
}
})
});
group.bench_function("index-hdfs-with-commit", |b| {
b.iter(|| {
let index = Index::create_in_ram(schema.clone());
let mut index_writer = index.writer_with_num_threads(1, 100_000_000).unwrap();
for _ in 0..NUM_REPEATS {
for doc_json in HDFS_LOGS.trim().split("\n") {
let doc = schema.parse_document(doc_json).unwrap();
index_writer.add_document(doc).unwrap();
}
}
index_writer.commit().unwrap();
})
});
group.bench_function("index-hdfs-no-commit-with-docstore", |b| {
b.iter(|| {
let index = Index::create_in_ram(schema_with_store.clone());
let index_writer = index.writer_with_num_threads(1, 100_000_000).unwrap();
for _ in 0..NUM_REPEATS {
for doc_json in HDFS_LOGS.trim().split("\n") {
let doc = schema.parse_document(doc_json).unwrap();
index_writer.add_document(doc).unwrap();
}
}
})
});
group.bench_function("index-hdfs-with-commit-with-docstore", |b| {
b.iter(|| {
let index = Index::create_in_ram(schema_with_store.clone());
let mut index_writer = index.writer_with_num_threads(1, 100_000_000).unwrap();
for _ in 0..NUM_REPEATS {
for doc_json in HDFS_LOGS.trim().split("\n") {
let doc = schema.parse_document(doc_json).unwrap();
index_writer.add_document(doc).unwrap();
}
}
index_writer.commit().unwrap();
})
});
group.bench_function("index-hdfs-no-commit-json-without-docstore", |b| {
b.iter(|| {
let index = Index::create_in_ram(dynamic_schema.clone());
let json_field = dynamic_schema.get_field("json").unwrap();
let mut index_writer = index.writer_with_num_threads(1, 100_000_000).unwrap();
for _ in 0..NUM_REPEATS {
for doc_json in HDFS_LOGS.trim().split("\n") {
let json_val: serde_json::Map<String, serde_json::Value> =
serde_json::from_str(doc_json).unwrap();
let doc = tantivy::doc!(json_field=>json_val);
index_writer.add_document(doc).unwrap();
}
}
index_writer.commit().unwrap();
})
});
group.bench_function("index-hdfs-with-commit-json-without-docstore", |b| {
b.iter(|| {
let index = Index::create_in_ram(dynamic_schema.clone());
let json_field = dynamic_schema.get_field("json").unwrap();
let mut index_writer = index.writer_with_num_threads(1, 100_000_000).unwrap();
for _ in 0..NUM_REPEATS {
for doc_json in HDFS_LOGS.trim().split("\n") {
let json_val: serde_json::Map<String, serde_json::Value> =
serde_json::from_str(doc_json).unwrap();
let doc = tantivy::doc!(json_field=>json_val);
index_writer.add_document(doc).unwrap();
}
}
index_writer.commit().unwrap();
})
});
}
criterion_group! {
name = benches;
config = Criterion::default().with_profiler(PProfProfiler::new(100, Output::Flamegraph(None)));
targets = hdfs_index_benchmark
}
criterion_main!(benches);

4
bitpacker/Cargo.toml
View File

@@ -1,12 +1,12 @@
[package]
name = "tantivy-bitpacker"
version = "0.1.0"
version = "0.2.0"
edition = "2018"
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-oss/tantivy"
keywords = []

2
bitpacker/benches/bench.rs
View File

@@ -6,6 +6,7 @@ extern crate test;
mod tests {
use tantivy_bitpacker::BlockedBitpacker;
use test::Bencher;
#[bench]
fn bench_blockedbitp_read(b: &mut Bencher) {
let mut blocked_bitpacker = BlockedBitpacker::new();
@@ -20,6 +21,7 @@ mod tests {
out
});
}
#[bench]
fn bench_blockedbitp_create(b: &mut Bencher) {
b.iter(|| {

3
bitpacker/src/bitpacker.rs
View File

@@ -1,4 +1,5 @@
use std::{convert::TryInto, io};
use std::convert::TryInto;
use std::io;
pub struct BitPacker {
mini_buffer: u64,

5
bitpacker/src/blocked_bitpacker.rs
View File

@@ -1,12 +1,11 @@
use super::bitpacker::BitPacker;
use super::compute_num_bits;
use crate::{minmax, BitUnpacker};
use super::{bitpacker::BitPacker, compute_num_bits};
const BLOCK_SIZE: usize = 128;
/// `BlockedBitpacker` compresses data in blocks of
/// 128 elements, while keeping an index on it
///
#[derive(Debug, Clone)]
pub struct BlockedBitpacker {
// bitpacked blocks

32
bitpacker/src/lib.rs
View File

@@ -1,8 +1,7 @@
mod bitpacker;
mod blocked_bitpacker;
pub use crate::bitpacker::BitPacker;
pub use crate::bitpacker::BitUnpacker;
pub use crate::bitpacker::{BitPacker, BitUnpacker};
pub use crate::blocked_bitpacker::BlockedBitpacker;
/// Computes the number of bits that will be used for bitpacking.
@@ -50,3 +49,32 @@ where
}
None
}
#[test]
fn test_compute_num_bits() {
assert_eq!(compute_num_bits(1), 1u8);
assert_eq!(compute_num_bits(0), 0u8);
assert_eq!(compute_num_bits(2), 2u8);
assert_eq!(compute_num_bits(3), 2u8);
assert_eq!(compute_num_bits(4), 3u8);
assert_eq!(compute_num_bits(255), 8u8);
assert_eq!(compute_num_bits(256), 9u8);
assert_eq!(compute_num_bits(5_000_000_000), 33u8);
}
#[test]
fn test_minmax_empty() {
let vals: Vec<u32> = vec![];
assert_eq!(minmax(vals.into_iter()), None);
}
#[test]
fn test_minmax_one() {
assert_eq!(minmax(vec![1].into_iter()), Some((1, 1)));
}
#[test]
fn test_minmax_two() {
assert_eq!(minmax(vec![1, 2].into_iter()), Some((1, 2)));
assert_eq!(minmax(vec![2, 1].into_iter()), Some((1, 2)));
}

9
common/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "common"
version = "0.1.0"
name = "tantivy-common"
version = "0.3.0"
authors = ["Paul Masurel <paul@quickwit.io>", "Pascal Seitz <pascal@quickwit.io>"]
license = "MIT"
edition = "2018"
@@ -10,3 +10,8 @@ description = "common traits and utility functions used by multiple tantivy subc
[dependencies]
byteorder = "1.4.3"
ownedbytes = { version="0.3", path="../ownedbytes" }
[dev-dependencies]
proptest = "1.0.0"
rand = "0.8.4"

745
common/src/bitset.rs Normal file
View File

@@ -0,0 +1,745 @@
use std::convert::TryInto;
use std::io::Write;
use std::{fmt, io, u64};
use ownedbytes::OwnedBytes;
#[derive(Clone, Copy, Eq, PartialEq)]
pub struct TinySet(u64);
impl fmt::Debug for TinySet {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
self.into_iter().collect::<Vec<u32>>().fmt(f)
}
}
pub struct TinySetIterator(TinySet);
impl Iterator for TinySetIterator {
type Item = u32;
#[inline]
fn next(&mut self) -> Option<Self::Item> {
self.0.pop_lowest()
}
}
impl IntoIterator for TinySet {
type Item = u32;
type IntoIter = TinySetIterator;
fn into_iter(self) -> Self::IntoIter {
TinySetIterator(self)
}
}
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;
}
/// 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.
#[inline]
fn complement(self) -> TinySet {
TinySet(!self.0)
}
/// Returns true iff the `TinySet` contains the element `el`.
#[inline]
pub fn contains(self, el: u32) -> bool {
!self.intersect(TinySet::singleton(el)).is_empty()
}
/// Returns the number of elements in the TinySet.
#[inline]
pub fn len(self) -> u32 {
self.0.count_ones()
}
/// Returns the intersection of `self` and `other`
#[inline]
#[must_use]
pub fn intersect(self, other: TinySet) -> TinySet {
TinySet(self.0 & other.0)
}
/// Creates a new `TinySet` containing only one element
/// within `[0; 64[`
#[inline]
pub fn singleton(el: u32) -> TinySet {
TinySet(1u64 << u64::from(el))
}
/// Insert a new element within [0..64)
#[inline]
#[must_use]
pub fn insert(self, el: u32) -> TinySet {
self.union(TinySet::singleton(el))
}
/// Removes an element within [0..64)
#[inline]
#[must_use]
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;
*self = old.insert(el);
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]
#[must_use]
pub fn union(self, other: TinySet) -> TinySet {
TinySet(self.0 | other.0)
}
/// Returns true iff the `TinySet` is empty.
#[inline]
pub fn is_empty(self) -> bool {
self.0 == 0u64
}
/// Returns the lowest element in the `TinySet`
/// and removes it.
#[inline]
pub fn pop_lowest(&mut self) -> Option<u32> {
if self.is_empty() {
None
} else {
let lowest = self.0.trailing_zeros() as u32;
self.0 ^= TinySet::singleton(lowest).0;
Some(lowest)
}
}
/// Returns a `TinySet` than contains all values up
/// to limit excluded.
///
/// The limit is assumed to be strictly lower than 64.
pub fn range_lower(upper_bound: u32) -> TinySet {
TinySet((1u64 << u64::from(upper_bound % 64u32)) - 1u64)
}
/// Returns a `TinySet` that contains all values greater
/// or equal to the given limit, included. (and up to 63)
///
/// The limit is assumed to be strictly lower than 64.
pub fn range_greater_or_equal(from_included: u32) -> TinySet {
TinySet::range_lower(from_included).complement()
}
}
#[derive(Clone)]
pub struct BitSet {
tinysets: Box<[TinySet]>,
len: u64,
max_value: u32,
}
fn num_buckets(max_val: u32) -> u32 {
(max_val + 63u32) / 64u32
}
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)`.
pub fn with_max_value(max_value: u32) -> BitSet {
let num_buckets = num_buckets(max_value);
let tinybitsets = vec![TinySet::empty(); num_buckets as usize].into_boxed_slice();
BitSet {
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() {
*tinyset = TinySet::empty();
}
}
/// 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 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;
let lower = el % 64u32;
self.len += if self.tinysets[higher as usize].insert_mut(lower) {
1
} else {
0
};
}
/// 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)
}
/// Returns the first non-empty `TinySet` associated to a bucket lower
/// or greater than bucket.
///
/// Reminder: the tiny set with the bucket `bucket`, represents the
/// elements from `bucket * 64` to `(bucket+1) * 64`.
pub fn first_non_empty_bucket(&self, bucket: u32) -> Option<u32> {
self.tinysets[bucket as usize..]
.iter()
.cloned()
.position(|tinyset| !tinyset.is_empty())
.map(|delta_bucket| bucket + delta_bucket as u32)
}
#[inline]
pub fn max_value(&self) -> u32 {
self.max_value
}
/// Returns the tiny bitset representing the
/// the set restricted to the number range from
/// `bucket * 64` to `(bucket + 1) * 64`.
pub fn tinyset(&self, bucket: u32) -> TinySet {
self.tinysets[bucket as usize]
}
}
/// 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 std::collections::HashSet;
use ownedbytes::OwnedBytes;
use rand::distributions::Bernoulli;
use rand::rngs::StdRng;
use rand::{Rng, SeedableRng};
use super::{BitSet, ReadOnlyBitSet, TinySet};
#[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);
}
#[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());
{
let mut u = TinySet::empty().insert(1u32);
assert_eq!(u.pop_lowest(), Some(1u32));
assert!(u.pop_lowest().is_none())
}
{
let mut u = TinySet::empty().insert(1u32).insert(1u32);
assert_eq!(u.pop_lowest(), Some(1u32));
assert!(u.pop_lowest().is_none())
}
{
let mut u = TinySet::empty().insert(2u32);
assert_eq!(u.pop_lowest(), Some(2u32));
u.insert_mut(1u32);
assert_eq!(u.pop_lowest(), Some(1u32));
assert!(u.pop_lowest().is_none());
}
{
let mut u = TinySet::empty().insert(63u32);
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]
fn test_bitset() {
let test_against_hashset = |els: &[u32], max_value: u32| {
let mut hashset: HashSet<u32> = HashSet::new();
let mut bitset = BitSet::with_max_value(max_value);
for &el in els {
assert!(el < max_value);
hashset.insert(el);
bitset.insert(el);
}
for el in 0..max_value {
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);
test_against_hashset(&[], 1);
test_against_hashset(&[0u32], 1);
test_against_hashset(&[0u32], 100);
test_against_hashset(&[1u32, 2u32], 4);
test_against_hashset(&[99u32], 100);
test_against_hashset(&[63u32], 64);
test_against_hashset(&[62u32, 63u32], 64);
}
#[test]
fn test_bitset_num_buckets() {
use super::num_buckets;
assert_eq!(num_buckets(0u32), 0);
assert_eq!(num_buckets(1u32), 1);
assert_eq!(num_buckets(64u32), 1);
assert_eq!(num_buckets(65u32), 2);
assert_eq!(num_buckets(128u32), 2);
assert_eq!(num_buckets(129u32), 3);
}
#[test]
fn test_tinyset_range() {
assert_eq!(
TinySet::range_lower(3).into_iter().collect::<Vec<u32>>(),
[0, 1, 2]
);
assert!(TinySet::range_lower(0).is_empty());
assert_eq!(
TinySet::range_lower(63).into_iter().collect::<Vec<u32>>(),
(0u32..63u32).collect::<Vec<_>>()
);
assert_eq!(
TinySet::range_lower(1).into_iter().collect::<Vec<u32>>(),
[0]
);
assert_eq!(
TinySet::range_lower(2).into_iter().collect::<Vec<u32>>(),
[0, 1]
);
assert_eq!(
TinySet::range_greater_or_equal(3)
.into_iter()
.collect::<Vec<u32>>(),
(3u32..64u32).collect::<Vec<_>>()
);
}
#[test]
fn test_bitset_len() {
let mut bitset = BitSet::with_max_value(1_000);
assert_eq!(bitset.len(), 0);
bitset.insert(3u32);
assert_eq!(bitset.len(), 1);
bitset.insert(103u32);
assert_eq!(bitset.len(), 2);
bitset.insert(3u32);
assert_eq!(bitset.len(), 2);
bitset.insert(103u32);
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> {
StdRng::from_seed([seed_val; 32])
.sample_iter(&Bernoulli::new(ratio).unwrap())
.take(n as usize)
.enumerate()
.filter_map(|(val, keep)| if keep { Some(val as u32) } else { None })
.collect()
}
pub fn sample(n: u32, ratio: f64) -> Vec<u32> {
sample_with_seed(n, ratio, 4)
}
#[test]
fn test_bitset_clear() {
let mut bitset = BitSet::with_max_value(1_000);
let els = sample(1_000, 0.01f64);
for &el in &els {
bitset.insert(el);
}
assert!(els.iter().all(|el| bitset.contains(*el)));
bitset.clear();
for el in 0u32..1000u32 {
assert!(!bitset.contains(el));
}
}
}
#[cfg(all(test, feature = "unstable"))]
mod bench {
use test;
use super::{BitSet, TinySet};
#[bench]
fn bench_tinyset_pop(b: &mut test::Bencher) {
b.iter(|| {
let mut tinyset = TinySet::singleton(test::black_box(31u32));
tinyset.pop_lowest();
tinyset.pop_lowest();
tinyset.pop_lowest();
tinyset.pop_lowest();
tinyset.pop_lowest();
tinyset.pop_lowest();
});
}
#[bench]
fn bench_tinyset_sum(b: &mut test::Bencher) {
let tiny_set = TinySet::empty().insert(10u32).insert(14u32).insert(21u32);
b.iter(|| {
assert_eq!(test::black_box(tiny_set).into_iter().sum::<u32>(), 45u32);
});
}
#[bench]
fn bench_tinyarr_sum(b: &mut test::Bencher) {
let v = [10u32, 14u32, 21u32];
b.iter(|| test::black_box(v).iter().cloned().sum::<u32>());
}
#[bench]
fn bench_bitset_initialize(b: &mut test::Bencher) {
b.iter(|| BitSet::with_max_value(1_000_000));
}
}

161
common/src/lib.rs
View File

@@ -1,9 +1,170 @@
#![allow(clippy::len_without_is_empty)]
use std::ops::Deref;
pub use byteorder::LittleEndian as Endianness;
mod bitset;
mod serialize;
mod vint;
mod writer;
pub use bitset::*;
pub use serialize::{BinarySerializable, DeserializeFrom, FixedSize};
pub use vint::{read_u32_vint, read_u32_vint_no_advance, serialize_vint_u32, write_u32_vint, VInt};
pub use writer::{AntiCallToken, CountingWriter, TerminatingWrite};
/// Has length trait
pub trait HasLen {
/// Return length
fn len(&self) -> usize;
/// Returns true iff empty.
fn is_empty(&self) -> bool {
self.len() == 0
}
}
impl<T: Deref<Target = [u8]>> HasLen for T {
fn len(&self) -> usize {
self.deref().len()
}
}
const HIGHEST_BIT: u64 = 1 << 63;
/// Maps a `i64` to `u64`
///
/// For simplicity, tantivy internally handles `i64` as `u64`.
/// The mapping is defined by this function.
///
/// Maps `i64` to `u64` so that
/// `-2^63 .. 2^63-1` is mapped
/// to
/// `0 .. 2^64-1`
/// in that order.
///
/// This is more suited than simply casting (`val as u64`)
/// because of bitpacking.
///
/// Imagine a list of `i64` ranging from -10 to 10.
/// When casting negative values, the negative values are projected
/// to values over 2^63, and all values end up requiring 64 bits.
///
/// # See also
/// The [reverse mapping is `u64_to_i64`](./fn.u64_to_i64.html).
#[inline]
pub fn i64_to_u64(val: i64) -> u64 {
(val as u64) ^ HIGHEST_BIT
}
/// Reverse the mapping given by [`i64_to_u64`](./fn.i64_to_u64.html).
#[inline]
pub fn u64_to_i64(val: u64) -> i64 {
(val ^ HIGHEST_BIT) as i64
}
/// Maps a `f64` to `u64`
///
/// For simplicity, tantivy internally handles `f64` as `u64`.
/// The mapping is defined by this function.
///
/// Maps `f64` to `u64` in a monotonic manner, so that bytes lexical order is preserved.
///
/// This is more suited than simply casting (`val as u64`)
/// which would truncate the result
///
/// # Reference
///
/// Daniel Lemire's [blog post](https://lemire.me/blog/2020/12/14/converting-floating-point-numbers-to-integers-while-preserving-order/)
/// explains the mapping in a clear manner.
///
/// # See also
/// The [reverse mapping is `u64_to_f64`](./fn.u64_to_f64.html).
#[inline]
pub fn f64_to_u64(val: f64) -> u64 {
let bits = val.to_bits();
if val.is_sign_positive() {
bits ^ HIGHEST_BIT
} else {
!bits
}
}
/// Reverse the mapping given by [`i64_to_u64`](./fn.i64_to_u64.html).
#[inline]
pub fn u64_to_f64(val: u64) -> f64 {
f64::from_bits(if val & HIGHEST_BIT != 0 {
val ^ HIGHEST_BIT
} else {
!val
})
}
#[cfg(test)]
pub mod test {
use std::f64;
use proptest::prelude::*;
use super::{f64_to_u64, i64_to_u64, u64_to_f64, u64_to_i64, BinarySerializable, FixedSize};
fn test_i64_converter_helper(val: i64) {
assert_eq!(u64_to_i64(i64_to_u64(val)), val);
}
fn test_f64_converter_helper(val: f64) {
assert_eq!(u64_to_f64(f64_to_u64(val)), val);
}
pub fn fixed_size_test<O: BinarySerializable + FixedSize + Default>() {
let mut buffer = Vec::new();
O::default().serialize(&mut buffer).unwrap();
assert_eq!(buffer.len(), O::SIZE_IN_BYTES);
}
proptest! {
#[test]
fn test_f64_converter_monotonicity_proptest((left, right) in (proptest::num::f64::NORMAL, proptest::num::f64::NORMAL)) {
let left_u64 = f64_to_u64(left);
let right_u64 = f64_to_u64(right);
assert_eq!(left_u64 < right_u64, left < right);
}
}
#[test]
fn test_i64_converter() {
assert_eq!(i64_to_u64(i64::min_value()), u64::min_value());
assert_eq!(i64_to_u64(i64::max_value()), u64::max_value());
test_i64_converter_helper(0i64);
test_i64_converter_helper(i64::min_value());
test_i64_converter_helper(i64::max_value());
for i in -1000i64..1000i64 {
test_i64_converter_helper(i);
}
}
#[test]
fn test_f64_converter() {
test_f64_converter_helper(f64::INFINITY);
test_f64_converter_helper(f64::NEG_INFINITY);
test_f64_converter_helper(0.0);
test_f64_converter_helper(-0.0);
test_f64_converter_helper(1.0);
test_f64_converter_helper(-1.0);
}
#[test]
fn test_f64_order() {
assert!(!(f64_to_u64(f64::NEG_INFINITY)..f64_to_u64(f64::INFINITY))
.contains(&f64_to_u64(f64::NAN))); // nan is not a number
assert!(f64_to_u64(1.5) > f64_to_u64(1.0)); // same exponent, different mantissa
assert!(f64_to_u64(2.0) > f64_to_u64(1.0)); // same mantissa, different exponent
assert!(f64_to_u64(2.0) > f64_to_u64(1.5)); // different exponent and mantissa
assert!(f64_to_u64(1.0) > f64_to_u64(-1.0)); // pos > neg
assert!(f64_to_u64(-1.5) < f64_to_u64(-1.0));
assert!(f64_to_u64(-2.0) < f64_to_u64(1.0));
assert!(f64_to_u64(-2.0) < f64_to_u64(-1.5));
}
}

14
common/src/serialize.rs
View File

@@ -1,10 +1,9 @@
use crate::Endianness;
use crate::VInt;
use std::io::{Read, Write};
use std::{fmt, io};
use byteorder::{ReadBytesExt, WriteBytesExt};
use std::fmt;
use std::io;
use std::io::Read;
use std::io::Write;
use crate::{Endianness, VInt};
/// Trait for a simple binary serialization.
pub trait BinarySerializable: fmt::Debug + Sized {
@@ -202,8 +201,7 @@ impl BinarySerializable for String {
#[cfg(test)]
pub mod test {
use super::VInt;
use super::*;
use super::{VInt, *};
use crate::serialize::BinarySerializable;
pub fn fixed_size_test<O: BinarySerializable + FixedSize + Default>() {
let mut buffer = Vec::new();

13
common/src/vint.rs
View File

@@ -1,8 +1,9 @@
use super::BinarySerializable;
use byteorder::{ByteOrder, LittleEndian};
use std::io;
use std::io::Read;
use std::io::Write;
use std::io::{Read, Write};
use byteorder::{ByteOrder, LittleEndian};
use super::BinarySerializable;
/// Wrapper over a `u64` that serializes as a variable int.
#[derive(Clone, Copy, Debug, Eq, PartialEq)]
@@ -174,9 +175,7 @@ impl BinarySerializable for VInt {
#[cfg(test)]
mod tests {
use super::serialize_vint_u32;
use super::BinarySerializable;
use super::VInt;
use super::{serialize_vint_u32, BinarySerializable, VInt};
fn aux_test_vint(val: u64) {
let mut v = [14u8; 10];

10
common/src/writer.rs
View File

@@ -54,7 +54,8 @@ impl<W: TerminatingWrite> TerminatingWrite for CountingWriter<W> {
}
}
/// Struct used to prevent from calling [`terminate_ref`](trait.TerminatingWrite#method.terminate_ref) directly
/// Struct used to prevent from calling
/// [`terminate_ref`](trait.TerminatingWrite.html#tymethod.terminate_ref) directly
///
/// The point is that while the type is public, it cannot be built by anyone
/// outside of this module.
@@ -64,9 +65,7 @@ pub struct AntiCallToken(());
pub trait TerminatingWrite: Write {
/// Indicate that the writer will no longer be used. Internally call terminate_ref.
fn terminate(mut self) -> io::Result<()>
where
Self: Sized,
{
where Self: Sized {
self.terminate_ref(AntiCallToken(()))
}
@@ -97,9 +96,10 @@ impl<'a> TerminatingWrite for &'a mut Vec<u8> {
#[cfg(test)]
mod test {
use super::CountingWriter;
use std::io::Write;
use super::CountingWriter;
#[test]
fn test_counting_writer() {
let buffer: Vec<u8> = vec![];

BIN
doc/assets/images/Nuclia.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 3.1 KiB

BIN
doc/assets/images/element-dark-theme.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 56 KiB

8
doc/assets/images/element.io.svg Normal file
View File

@@ -0,0 +1,8 @@
<svg width="518" height="112" viewBox="0 0 518 112" fill="none" xmlns="http://www.w3.org/2000/svg">
<path fill-rule="evenodd" clip-rule="evenodd" d="M56 112C86.9279 112 112 86.9279 112 56C112 25.0721 86.9279 0 56 0C25.0721 0 0 25.0721 0 56C0 86.9279 25.0721 112 56 112Z" fill="#0DBD8B"/>
<path fill-rule="evenodd" clip-rule="evenodd" d="M45.7615 26.093C45.7615 23.8325 47.5977 22.0001 49.8629 22.0001C65.2154 22.0001 77.6611 34.4199 77.6611 49.7406C77.6611 52.001 75.8248 53.8335 73.5597 53.8335C71.2945 53.8335 69.4583 52.001 69.4583 49.7406C69.4583 38.9408 60.6851 30.1859 49.8629 30.1859C47.5977 30.1859 45.7615 28.3534 45.7615 26.093Z" fill="white"/>
<path fill-rule="evenodd" clip-rule="evenodd" d="M85.8986 45.6477C88.1637 45.6477 89.9999 47.4801 89.9999 49.7406C89.9999 65.0612 77.5543 77.4811 62.2017 77.4811C59.9366 77.4811 58.1003 75.6486 58.1003 73.3882C58.1003 71.1277 59.9366 69.2953 62.2017 69.2953C73.024 69.2953 81.7972 60.5403 81.7972 49.7406C81.7972 47.4801 83.6334 45.6477 85.8986 45.6477Z" fill="white"/>
<path fill-rule="evenodd" clip-rule="evenodd" d="M66.3031 85.907C66.3031 88.1675 64.4668 89.9999 62.2017 89.9999C46.8492 89.9999 34.4035 77.58 34.4035 62.2594C34.4035 59.9989 36.2398 58.1665 38.5049 58.1665C40.77 58.1665 42.6063 59.9989 42.6063 62.2594C42.6063 73.0592 51.3795 81.8141 62.2017 81.8141C64.4668 81.8141 66.3031 83.6466 66.3031 85.907Z" fill="white"/>
<path fill-rule="evenodd" clip-rule="evenodd" d="M26.1014 66.3523C23.8363 66.3523 22.0001 64.5199 22.0001 62.2594C22 46.9388 34.4457 34.5189 49.7983 34.5189C52.0634 34.5189 53.8997 36.3514 53.8997 38.6118C53.8997 40.8723 52.0634 42.7047 49.7983 42.7047C38.976 42.7047 30.2028 51.4597 30.2028 62.2594C30.2028 64.5199 28.3666 66.3523 26.1014 66.3523Z" fill="white"/>
<path d="M197 63.5H157.5C157.967 67.6333 159.467 70.9333 162 73.4C164.533 75.8 167.867 77 172 77C174.733 77 177.2 76.3333 179.4 75C181.6 73.6667 183.167 71.8667 184.1 69.6H196.1C194.5 74.8667 191.5 79.1333 187.1 82.4C182.767 85.6 177.633 87.2 171.7 87.2C163.967 87.2 157.7 84.6333 152.9 79.5C148.167 74.3667 145.8 67.8667 145.8 60C145.8 52.3333 148.2 45.9 153 40.7C157.8 35.5 164 32.9 171.6 32.9C179.2 32.9 185.333 35.4667 190 40.6C194.733 45.6667 197.1 52.0667 197.1 59.8L197 63.5ZM171.6 42.6C167.867 42.6 164.767 43.7 162.3 45.9C159.833 48.1 158.3 51.0333 157.7 54.7H185.3C184.767 51.0333 183.3 48.1 180.9 45.9C178.5 43.7 175.4 42.6 171.6 42.6ZM205.289 70.5V11H217.189V70.7C217.189 73.3667 218.656 74.7 221.589 74.7L223.689 74.6V85.9C222.556 86.1 221.356 86.2 220.089 86.2C214.956 86.2 211.189 84.9 208.789 82.3C206.456 79.7 205.289 75.7667 205.289 70.5ZM279.109 63.5H239.609C240.076 67.6333 241.576 70.9333 244.109 73.4C246.643 75.8 249.976 77 254.109 77C256.843 77 259.309 76.3333 261.509 75C263.709 73.6667 265.276 71.8667 266.209 69.6H278.209C276.609 74.8667 273.609 79.1333 269.209 82.4C264.876 85.6 259.743 87.2 253.809 87.2C246.076 87.2 239.809 84.6333 235.009 79.5C230.276 74.3667 227.909 67.8667 227.909 60C227.909 52.3333 230.309 45.9 235.109 40.7C239.909 35.5 246.109 32.9 253.709 32.9C261.309 32.9 267.443 35.4667 272.109 40.6C276.843 45.6667 279.209 52.0667 279.209 59.8L279.109 63.5ZM253.709 42.6C249.976 42.6 246.876 43.7 244.409 45.9C241.943 48.1 240.409 51.0333 239.809 54.7H267.409C266.876 51.0333 265.409 48.1 263.009 45.9C260.609 43.7 257.509 42.6 253.709 42.6ZM332.798 56.2V86H320.898V54.9C320.898 47.0333 317.632 43.1 311.098 43.1C307.565 43.1 304.732 44.2333 302.598 46.5C300.532 48.7667 299.498 51.8667 299.498 55.8V86H287.598V34.1H298.598V41C299.865 38.6667 301.798 36.7333 304.398 35.2C306.998 33.6667 310.232 32.9 314.098 32.9C321.298 32.9 326.498 35.6333 329.698 41.1C334.098 35.6333 339.965 32.9 347.298 32.9C353.365 32.9 358.032 34.8 361.298 38.6C364.565 42.3333 366.198 47.2667 366.198 53.4V86H354.298V54.9C354.298 47.0333 351.032 43.1 344.498 43.1C340.898 43.1 338.032 44.2667 335.898 46.6C333.832 48.8667 332.798 52.0667 332.798 56.2ZM425.379 63.5H385.879C386.346 67.6333 387.846 70.9333 390.379 73.4C392.912 75.8 396.246 77 400.379 77C403.112 77 405.579 76.3333 407.779 75C409.979 73.6667 411.546 71.8667 412.479 69.6H424.479C422.879 74.8667 419.879 79.1333 415.479 82.4C411.146 85.6 406.012 87.2 400.079 87.2C392.346 87.2 386.079 84.6333 381.279 79.5C376.546 74.3667 374.179 67.8667 374.179 60C374.179 52.3333 376.579 45.9 381.379 40.7C386.179 35.5 392.379 32.9 399.979 32.9C407.579 32.9 413.712 35.4667 418.379 40.6C423.112 45.6667 425.479 52.0667 425.479 59.8L425.379 63.5ZM399.979 42.6C396.246 42.6 393.146 43.7 390.679 45.9C388.212 48.1 386.679 51.0333 386.079 54.7H413.679C413.146 51.0333 411.679 48.1 409.279 45.9C406.879 43.7 403.779 42.6 399.979 42.6ZM444.868 34.1V41C446.068 38.7333 448.035 36.8333 450.768 35.3C453.568 33.7 456.935 32.9 460.868 32.9C467.001 32.9 471.735 34.7667 475.068 38.5C478.468 42.2333 480.168 47.2 480.168 53.4V86H468.268V54.9C468.268 51.2333 467.401 48.3667 465.668 46.3C464.001 44.1667 461.435 43.1 457.968 43.1C454.168 43.1 451.168 44.2333 448.968 46.5C446.835 48.7667 445.768 51.9 445.768 55.9V86H433.868V34.1H444.868ZM514.922 75.4V85.7C513.455 86.1 511.389 86.3 508.722 86.3C498.589 86.3 493.522 81.2 493.522 71V43.6H485.622V34.1H493.522V20.6H505.422V34.1H515.122V43.6H505.422V69.8C505.422 73.8667 507.355 75.9 511.222 75.9L514.922 75.4Z" fill="black"/>
</svg>
Side by Side

After

Width:  |  Height:  |  Size: 5.2 KiB

BIN
doc/assets/images/humanfirst.ai-dark-theme.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 23 KiB

BIN
doc/assets/images/humanfirst.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 102 KiB

BIN
doc/assets/images/nuclia-dark-theme.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 7.8 KiB

BIN
doc/assets/images/searchbenchmark.png Normal file
View File

Binary file not shown.
Side by Side

After

Width:  |  Height:  |  Size: 653 KiB

1
doc/src/SUMMARY.md
View File

@@ -7,6 +7,7 @@
- [Segments](./basis.md)
- [Defining your schema](./schema.md)
- [Facetting](./facetting.md)
- [Index Sorting](./index_sorting.md)
- [Innerworkings](./innerworkings.md)
- [Inverted index](./inverted_index.md)
- [Best practise](./inverted_index.md)

61
doc/src/index_sorting.md Normal file
View File

@@ -0,0 +1,61 @@
- [Index Sorting](#index-sorting)
+ [Why Sorting](#why-sorting)
* [Compression](#compression)
* [Top-N Optimization](#top-n-optimization)
* [Pruning](#pruning)
* [Other](#other)
+ [Usage](#usage)
# Index Sorting
Tantivy allows you to sort the index according to a property.
## Why Sorting
Presorting an index has several advantages:
###### Compression
When data is sorted it is easier to compress the data. E.g. the numbers sequence [5, 2, 3, 1, 4] would be sorted to [1, 2, 3, 4, 5].
If we apply delta encoding this list would be unsorted [5, -3, 1, -2, 3] vs. [1, 1, 1, 1, 1].
Compression ratio is mainly affected on the fast field of the sorted property, every thing else is likely unaffected.
###### Top-N Optimization
When data is presorted by a field and search queries request sorting by the same field, we can leverage the natural order of the documents.
E.g. if the data is sorted by timestamp and want the top n newest docs containing a term, we can simply leveraging the order of the docids.
Note: Tantivy 0.16 does not do this optimization yet.
###### Pruning
Let's say we want all documents and want to apply the filter `>= 2010-08-11`. When the data is sorted, we could make a lookup in the fast field to find the docid range and use this as the filter.
Note: Tantivy 0.16 does not do this optimization yet.
###### Other?
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/quickwit-oss/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 {
sort_by_field: Some(IndexSortByField {
field: "intval".to_string(),
order: Order::Desc,
}),
..Default::default()
};
let mut index_builder = Index::builder().schema(schema);
index_builder = index_builder.settings(settings);
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/quickwit-oss/tantivy/blob/000d76b11a139a84b16b9b95060a1c93e8b9851c/src/indexer/segment_writer.rs#L338) and [merging multiple segments](https://github.com/quickwit-oss/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).

128
doc/src/json.md Normal file
View File

@@ -0,0 +1,128 @@
# Json
As of tantivy 0.17, tantivy supports a json object type.
This type can be used to allow for a schema-less search index.
When indexing a json object, we "flatten" the JSON. This operation emits terms that represent a triplet `(json_path, value_type, value)`
For instance, if user is a json field, the following document:
```json
{
"user": {
"name": "Paul Masurel",
"address": {
"city": "Tokyo",
"country": "Japan"
},
"created_at": "2018-11-12T23:20:50.52Z"
}
}
```
emits the following tokens:
- ("name", Text, "Paul")
- ("name", Text, "Masurel")
- ("address.city", Text, "Tokyo")
- ("address.country", Text, "Japan")
- ("created_at", Date, 15420648505)
# Bytes-encoding and lexicographical sort.
Like any other terms, these triplets are encoded into a binary format as follows.
- `json_path`: the json path is a sequence of "segments". In the example above, `address.city`
is just a debug representation of the json path `["address", "city"]`.
Its representation is done by separating segments by a unicode char `\x01`, and ending the path by `\x00`.
- `value type`: One byte represents the `Value` type.
- `value`: The value representation is just the regular Value representation.
This representation is designed to align the natural sort of Terms with the lexicographical sort
of their binary representation (Tantivy's dictionary (whether fst or sstable) is sorted and does prefix encoding).
In the example above, the terms will be sorted as
- ("address.city", Text, "Tokyo")
- ("address.country", Text, "Japan")
- ("name", Text, "Masurel")
- ("name", Text, "Paul")
- ("created_at", Date, 15420648505)
As seen in "pitfalls", we may end up having to search for a value for a same path in several different fields. Putting the field code after the path makes it maximizes compression opportunities but also increases the chances for the two terms to end up in the actual same term dictionary block.
# Pitfalls, limitation and corner cases.
Json gives very little information about the type of the literals it stores.
All numeric types end up mapped as a "Number" and there are no types for dates.
At indexing, tantivy will try to interpret number and strings as different type with a
priority order.
Numbers will be interpreted as u64, i64 and f64 in that order.
Strings will be interpreted as rfc3999 dates or simple strings.
The first working type is picked and is the only term that is emitted for indexing.
Note this interpretation happens on a per-document basis, and there is no effort to try to sniff
a consistent field type at the scale of a segment.
On the query parser side on the other hand, we may end up emitting more than one type.
For instance, we do not even know if the type is a number or string based.
So the query
```
my_path.my_segment:233
```
Will be interpreted as
`(my_path.my_segment, String, 233) or (my_path.my_segment, u64, 233)`
Likewise, we need to emit two tokens if the query contains an rfc3999 date.
Indeed the date could have been actually a single token inside the text of a document at ingestion time. Generally speaking, we will always at least emit a string token in query parsing, and sometimes more.
If one more json field is defined, things get even more complicated.
## Default json field
If the schema contains a text field called "text" and a json field that is set as a default field:
`text:hello` could be reasonably interpreted as targetting the text field or as targetting the json field called `json_dynamic` with the json_path "text".
If there is such an ambiguity, we decide to only search in the "text" field: `text:hello`.
In other words, the parser will not search in default json fields if there is a schema hit.
This is a product decision.
The user can still target the JSON field by specifying its name explicitly:
`json_dynamic.text:hello`.
## Range queries are not supported.
Json field do not support range queries.
## Arrays do not work like nested object.
If json object contains an array, a search query might return more documents
than what might be expected.
Let's take an example.
```json
{
"cart_id": 3234234 ,
"cart": [
{"product_type": "sneakers", "attributes": {"color": "white"} },
{"product_type": "t-shirt", "attributes": {"color": "red"}},
]
}
```
Despite the array structure, a document in tantivy is a bag of terms.
The query:
```
cart.product_type:sneakers AND cart.attributes.color:red
```
Actually match the document above.

129
examples/aggregation.rs Normal file
View File

@@ -0,0 +1,129 @@
// # Aggregation example
//
// This example shows how you can use built-in aggregations.
// We will use range buckets and compute the average in each bucket.
//
use serde_json::Value;
use tantivy::aggregation::agg_req::{
Aggregation, Aggregations, BucketAggregation, BucketAggregationType, MetricAggregation,
RangeAggregation,
};
use tantivy::aggregation::agg_result::AggregationResults;
use tantivy::aggregation::metric::AverageAggregation;
use tantivy::aggregation::AggregationCollector;
use tantivy::query::TermQuery;
use tantivy::schema::{self, Cardinality, IndexRecordOption, Schema, TextFieldIndexing};
use tantivy::{doc, Index, Term};
fn main() -> tantivy::Result<()> {
let mut schema_builder = Schema::builder();
let text_fieldtype = schema::TextOptions::default()
.set_indexing_options(
TextFieldIndexing::default().set_index_option(IndexRecordOption::WithFreqs),
)
.set_stored();
let text_field = schema_builder.add_text_field("text", text_fieldtype);
let score_fieldtype =
crate::schema::NumericOptions::default().set_fast(Cardinality::SingleValue);
let highscore_field = schema_builder.add_f64_field("highscore", score_fieldtype.clone());
let price_field = schema_builder.add_f64_field("price", score_fieldtype.clone());
let schema = schema_builder.build();
// # Indexing documents
//
// Lets index a bunch of documents for this example.
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer(50_000_000)?;
// writing the segment
index_writer.add_document(doc!(
text_field => "cool",
highscore_field => 1f64,
price_field => 0f64,
))?;
index_writer.add_document(doc!(
text_field => "cool",
highscore_field => 3f64,
price_field => 1f64,
))?;
index_writer.add_document(doc!(
text_field => "cool",
highscore_field => 5f64,
price_field => 1f64,
))?;
index_writer.add_document(doc!(
text_field => "nohit",
highscore_field => 6f64,
price_field => 2f64,
))?;
index_writer.add_document(doc!(
text_field => "cool",
highscore_field => 7f64,
price_field => 2f64,
))?;
index_writer.commit()?;
index_writer.add_document(doc!(
text_field => "cool",
highscore_field => 11f64,
price_field => 10f64,
))?;
index_writer.add_document(doc!(
text_field => "cool",
highscore_field => 14f64,
price_field => 15f64,
))?;
index_writer.add_document(doc!(
text_field => "cool",
highscore_field => 15f64,
price_field => 20f64,
))?;
index_writer.commit()?;
let reader = index.reader()?;
let text_field = reader.searcher().schema().get_field("text").unwrap();
let term_query = TermQuery::new(
Term::from_field_text(text_field, "cool"),
IndexRecordOption::Basic,
);
let sub_agg_req_1: Aggregations = vec![(
"average_price".to_string(),
Aggregation::Metric(MetricAggregation::Average(
AverageAggregation::from_field_name("price".to_string()),
)),
)]
.into_iter()
.collect();
let agg_req_1: Aggregations = vec![(
"score_ranges".to_string(),
Aggregation::Bucket(BucketAggregation {
bucket_agg: BucketAggregationType::Range(RangeAggregation {
field: "highscore".to_string(),
ranges: vec![
(-1f64..9f64).into(),
(9f64..14f64).into(),
(14f64..20f64).into(),
],
}),
sub_aggregation: sub_agg_req_1.clone(),
}),
)]
.into_iter()
.collect();
let collector = AggregationCollector::from_aggs(agg_req_1);
let searcher = reader.searcher();
let agg_res: AggregationResults = searcher.search(&term_query, &collector).unwrap();
let res: Value = serde_json::to_value(&agg_res)?;
println!("{}", serde_json::to_string_pretty(&res)?);
Ok(())
}

12
examples/basic_search.rs
View File

@@ -73,7 +73,7 @@ fn main() -> tantivy::Result<()> {
// multithreaded.
//
// Here we give tantivy a budget of `50MB`.
// Using a bigger heap for the indexer may increase
// Using a bigger memory_arena for the indexer may increase
// throughput, but 50 MB is already plenty.
let mut index_writer = index.writer(50_000_000)?;
@@ -91,12 +91,12 @@ fn main() -> tantivy::Result<()> {
old_man_doc.add_text(title, "The Old Man and the Sea");
old_man_doc.add_text(
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.",
"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.",
);
// ... 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

21
examples/custom_collector.rs
View File

@@ -12,8 +12,7 @@
use tantivy::collector::{Collector, SegmentCollector};
use tantivy::fastfield::{DynamicFastFieldReader, FastFieldReader};
use tantivy::query::QueryParser;
use tantivy::schema::Field;
use tantivy::schema::{Schema, FAST, INDEXED, TEXT};
use tantivy::schema::{Field, Schema, FAST, INDEXED, TEXT};
use tantivy::{doc, Index, Score, SegmentReader};
#[derive(Default)]
@@ -86,12 +85,10 @@ impl Collector for StatsCollector {
fn merge_fruits(&self, segment_stats: Vec<Option<Stats>>) -> tantivy::Result<Option<Stats>> {
let mut stats = Stats::default();
for segment_stats_opt in segment_stats {
if let Some(segment_stats) = segment_stats_opt {
stats.count += segment_stats.count;
stats.sum += segment_stats.sum;
stats.squared_sum += segment_stats.squared_sum;
}
for segment_stats in segment_stats.into_iter().flatten() {
stats.count += segment_stats.count;
stats.sum += segment_stats.sum;
stats.squared_sum += segment_stats.squared_sum;
}
Ok(stats.non_zero_count())
}
@@ -147,23 +144,23 @@ 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()?;

8
examples/custom_tokenizer.rs
View File

@@ -62,13 +62,13 @@ fn main() -> tantivy::Result<()> {
// multithreaded.
//
// Here we use a buffer of 50MB per thread. Using a bigger
// heap for the indexer can increase its throughput.
// memory arena for the indexer can increase its throughput.
let mut index_writer = index.writer(50_000_000)?;
index_writer.add_document(doc!(
title => "The Old Man and the Sea",
body => "He was an old man who fished alone in a skiff in the Gulf Stream and \
he had gone eighty-four days now without taking a fish."
));
))?;
index_writer.add_document(doc!(
title => "Of Mice and Men",
body => r#"A few miles south of Soledad, the Salinas River drops in close to the hillside
@@ -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()?;

13
examples/deleting_updating_documents.rs
View File

@@ -56,8 +56,9 @@ fn main() -> tantivy::Result<()> {
// If it is `text`, let's make sure to keep it `raw` and let's avoid
// running any text processing on it.
// This is done by associating this field to the tokenizer named `raw`.
// Rather than building our [`TextOptions`](//docs.rs/tantivy/~0/tantivy/schema/struct.TextOptions.html) manually,
// We use the `STRING` shortcut. `STRING` stands for indexed (without term frequency or positions)
// Rather than building our
// [`TextOptions`](//docs.rs/tantivy/~0/tantivy/schema/struct.TextOptions.html) manually, We
// use the `STRING` shortcut. `STRING` stands for indexed (without term frequency or positions)
// and untokenized.
//
// Because we also want to be able to see this `id` in our returned documents,
@@ -76,15 +77,15 @@ 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 +123,7 @@ 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.

18
examples/faceted_search.rs
View File

@@ -23,7 +23,7 @@ fn main() -> tantivy::Result<()> {
let name = schema_builder.add_text_field("felin_name", TEXT | STORED);
// this is our faceted field: its scientific classification
let classification = schema_builder.add_facet_field("classification", INDEXED);
let classification = schema_builder.add_facet_field("classification", FacetOptions::default());
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
@@ -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()?;

14
examples/faceted_search_with_tweaked_score.rs
View File

@@ -1,15 +1,15 @@
use std::collections::HashSet;
use tantivy::collector::TopDocs;
use tantivy::doc;
use tantivy::query::BooleanQuery;
use tantivy::schema::*;
use tantivy::{DocId, Index, Score, SegmentReader};
use tantivy::{doc, DocId, Index, Score, SegmentReader};
fn main() -> tantivy::Result<()> {
let mut schema_builder = Schema::builder();
let title = schema_builder.add_text_field("title", STORED);
let ingredient = schema_builder.add_facet_field("ingredient", INDEXED);
let ingredient = schema_builder.add_facet_field("ingredient", FacetOptions::default());
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
@@ -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()?;
@@ -87,7 +87,7 @@ fn main() -> tantivy::Result<()> {
.unwrap()
.get_first(title)
.unwrap()
.text()
.as_text()
.unwrap()
.to_owned()
})

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()
}

22
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()?;
@@ -52,11 +52,11 @@ fn main() -> tantivy::Result<()> {
let term_the = Term::from_field_text(title, "the");
// This segment posting object is like a cursor over the documents matching the term.
// The `IndexRecordOption` arguments tells tantivy we will be interested in both term frequencies
// and positions.
// The `IndexRecordOption` arguments tells tantivy we will be interested in both term
// frequencies and positions.
//
// If you don't need all this information, you may get better performance by decompressing less
// information.
// If you don't need all this information, you may get better performance by decompressing
// less information.
if let Some(mut segment_postings) =
inverted_index.read_postings(&term_the, IndexRecordOption::WithFreqsAndPositions)?
{
@@ -109,11 +109,11 @@ fn main() -> tantivy::Result<()> {
let inverted_index = segment_reader.inverted_index(title)?;
// This segment posting object is like a cursor over the documents matching the term.
// The `IndexRecordOption` arguments tells tantivy we will be interested in both term frequencies
// and positions.
// The `IndexRecordOption` arguments tells tantivy we will be interested in both term
// frequencies and positions.
//
// If you don't need all this information, you may get better performance by decompressing less
// information.
// If you don't need all this information, you may get better performance by decompressing
// less information.
if let Some(mut block_segment_postings) =
inverted_index.read_block_postings(&term_the, IndexRecordOption::Basic)?
{

105
examples/json_field.rs Normal file
View File

@@ -0,0 +1,105 @@
// # Json field example
//
// This example shows how the json field can be used
// to make tantivy partially schemaless by setting it as
// default query parser field.
use tantivy::collector::{Count, TopDocs};
use tantivy::query::QueryParser;
use tantivy::schema::{Schema, FAST, STORED, STRING, TEXT};
use tantivy::Index;
fn main() -> tantivy::Result<()> {
// # Defining the schema
let mut schema_builder = Schema::builder();
schema_builder.add_date_field("timestamp", FAST | STORED);
let event_type = schema_builder.add_text_field("event_type", STRING | STORED);
let attributes = schema_builder.add_json_field("attributes", STORED | TEXT);
let schema = schema_builder.build();
// # Indexing documents
let index = Index::create_in_ram(schema.clone());
let mut index_writer = index.writer(50_000_000)?;
let doc = schema.parse_document(
r#"{
"timestamp": "2022-02-22T23:20:50.53Z",
"event_type": "click",
"attributes": {
"target": "submit-button",
"cart": {"product_id": 103},
"description": "the best vacuum cleaner ever"
}
}"#,
)?;
index_writer.add_document(doc)?;
let doc = schema.parse_document(
r#"{
"timestamp": "2022-02-22T23:20:51.53Z",
"event_type": "click",
"attributes": {
"target": "submit-button",
"cart": {"product_id": 133},
"description": "das keyboard",
"event_type": "holiday-sale"
}
}"#,
)?;
index_writer.add_document(doc)?;
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
// # Default fields: event_type and attributes
// By setting attributes as a default field it allows omitting attributes itself, e.g. "target",
// instead of "attributes.target"
let query_parser = QueryParser::for_index(&index, vec![event_type, attributes]);
{
let query = query_parser.parse_query("target:submit-button")?;
let count_docs = searcher.search(&*query, &TopDocs::with_limit(2))?;
assert_eq!(count_docs.len(), 2);
}
{
let query = query_parser.parse_query("target:submit")?;
let count_docs = searcher.search(&*query, &TopDocs::with_limit(2))?;
assert_eq!(count_docs.len(), 2);
}
{
let query = query_parser.parse_query("cart.product_id:103")?;
let count_docs = searcher.search(&*query, &Count)?;
assert_eq!(count_docs, 1);
}
{
let query = query_parser.parse_query("click AND cart.product_id:133")?;
let hits = searcher.search(&*query, &TopDocs::with_limit(2))?;
assert_eq!(hits.len(), 1);
}
{
// The sub-fields in the json field marked as default field still need to be explicitly
// addressed
let query = query_parser.parse_query("click AND 133")?;
let hits = searcher.search(&*query, &TopDocs::with_limit(2))?;
assert_eq!(hits.len(), 0);
}
{
// Default json fields are ignored if they collide with the schema
let query = query_parser.parse_query("event_type:holiday-sale")?;
let hits = searcher.search(&*query, &TopDocs::with_limit(2))?;
assert_eq!(hits.len(), 0);
}
// # Query via full attribute path
{
// This only searches in our schema's `event_type` field
let query = query_parser.parse_query("event_type:click")?;
let hits = searcher.search(&*query, &TopDocs::with_limit(2))?;
assert_eq!(hits.len(), 2);
}
{
// Default json fields can still be accessed by full path
let query = query_parser.parse_query("attributes.event_type:holiday-sale")?;
let hits = searcher.search(&*query, &TopDocs::with_limit(2))?;
assert_eq!(hits.len(), 1);
}
Ok(())
}

14
examples/multiple_producer.rs
View File

@@ -28,8 +28,9 @@
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 +60,11 @@ 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.
@@ -78,19 +80,21 @@ 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.
for _ in 0..10 {
let opstamp: Opstamp = {
// Committing or rollbacking on the other hand requires write lock. This will block other threads.
// 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);

13
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()?;
@@ -57,7 +57,10 @@ fn main() -> tantivy::Result<()> {
let doc = searcher.doc(doc_address)?;
let snippet = snippet_generator.snippet_from_doc(&doc);
println!("Document score {}:", score);
println!("title: {}", doc.get_first(title).unwrap().text().unwrap());
println!(
"title: {}",
doc.get_first(title).unwrap().as_text().unwrap()
);
println!("snippet: {}", snippet.to_html());
println!("custom highlighting: {}", highlight(snippet));
}
@@ -70,13 +73,13 @@ fn highlight(snippet: Snippet) -> String {
let mut start_from = 0;
for fragment_range in snippet.highlighted() {
result.push_str(&snippet.fragments()[start_from..fragment_range.start]);
result.push_str(&snippet.fragment()[start_from..fragment_range.start]);
result.push_str(" --> ");
result.push_str(&snippet.fragments()[fragment_range.clone()]);
result.push_str(&snippet.fragment()[fragment_range.clone()]);
result.push_str(" <-- ");
start_from = fragment_range.end;
}
result.push_str(&snippet.fragments()[start_from..]);
result.push_str(&snippet.fragment()[start_from..]);
result
}

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()?;

224
examples/warmer.rs Normal file
View File

@@ -0,0 +1,224 @@
use std::cmp::Reverse;
use std::collections::{HashMap, HashSet};
use std::sync::{Arc, RwLock, Weak};
use tantivy::collector::TopDocs;
use tantivy::fastfield::FastFieldReader;
use tantivy::query::QueryParser;
use tantivy::schema::{Field, Schema, FAST, TEXT};
use tantivy::{
doc, DocAddress, DocId, Index, IndexReader, Opstamp, Searcher, SearcherGeneration, SegmentId,
SegmentReader, Warmer,
};
// This example shows how warmers can be used to
// load a values from an external sources using the Warmer API.
//
// In this example, we assume an e-commerce search engine.
type ProductId = u64;
/// Price
type Price = u32;
pub trait PriceFetcher: Send + Sync + 'static {
fn fetch_prices(&self, product_ids: &[ProductId]) -> Vec<Price>;
}
struct DynamicPriceColumn {
field: Field,
price_cache: RwLock<HashMap<(SegmentId, Option<Opstamp>), Arc<Vec<Price>>>>,
price_fetcher: Box<dyn PriceFetcher>,
}
impl DynamicPriceColumn {
pub fn with_product_id_field<T: PriceFetcher>(field: Field, price_fetcher: T) -> Self {
DynamicPriceColumn {
field,
price_cache: Default::default(),
price_fetcher: Box::new(price_fetcher),
}
}
pub fn price_for_segment(&self, segment_reader: &SegmentReader) -> Option<Arc<Vec<Price>>> {
let segment_key = (segment_reader.segment_id(), segment_reader.delete_opstamp());
self.price_cache.read().unwrap().get(&segment_key).cloned()
}
}
impl Warmer for DynamicPriceColumn {
fn warm(&self, searcher: &Searcher) -> tantivy::Result<()> {
for segment in searcher.segment_readers() {
let key = (segment.segment_id(), segment.delete_opstamp());
let product_id_reader = segment.fast_fields().u64(self.field)?;
let product_ids: Vec<ProductId> = segment
.doc_ids_alive()
.map(|doc| product_id_reader.get(doc))
.collect();
let mut prices_it = self.price_fetcher.fetch_prices(&product_ids).into_iter();
let mut price_vals: Vec<Price> = Vec::new();
for doc in 0..segment.max_doc() {
if segment.is_deleted(doc) {
price_vals.push(0);
} else {
price_vals.push(prices_it.next().unwrap())
}
}
self.price_cache
.write()
.unwrap()
.insert(key, Arc::new(price_vals));
}
Ok(())
}
fn garbage_collect(&self, live_generations: &[&SearcherGeneration]) {
let live_segment_id_and_delete_ops: HashSet<(SegmentId, Option<Opstamp>)> =
live_generations
.iter()
.flat_map(|gen| gen.segments())
.map(|(&segment_id, &opstamp)| (segment_id, opstamp))
.collect();
let mut price_cache_wrt = self.price_cache.write().unwrap();
// let price_cache = std::mem::take(&mut *price_cache_wrt);
// Drain would be nicer here.
*price_cache_wrt = std::mem::take(&mut *price_cache_wrt)
.into_iter()
.filter(|(seg_id_and_op, _)| !live_segment_id_and_delete_ops.contains(seg_id_and_op))
.collect();
}
}
/// For the sake of this example, the table is just an editable HashMap behind a RwLock.
/// This map represents a map (ProductId -> Price)
///
/// In practise, it could be fetching things from an external service, like a SQL table.
#[derive(Default, Clone)]
pub struct ExternalPriceTable {
prices: Arc<RwLock<HashMap<ProductId, Price>>>,
}
impl ExternalPriceTable {
pub fn update_price(&self, product_id: ProductId, price: Price) {
let mut prices_wrt = self.prices.write().unwrap();
prices_wrt.insert(product_id, price);
}
}
impl PriceFetcher for ExternalPriceTable {
fn fetch_prices(&self, product_ids: &[ProductId]) -> Vec<Price> {
let prices_read = self.prices.read().unwrap();
product_ids
.iter()
.map(|product_id| prices_read.get(product_id).cloned().unwrap_or(0))
.collect()
}
}
fn main() -> tantivy::Result<()> {
// Declaring our schema.
let mut schema_builder = Schema::builder();
// The product id is assumed to be a primary id for our external price source.
let product_id = schema_builder.add_u64_field("product_id", FAST);
let text = schema_builder.add_text_field("text", TEXT);
let schema: Schema = schema_builder.build();
let price_table = ExternalPriceTable::default();
let price_dynamic_column = Arc::new(DynamicPriceColumn::with_product_id_field(
product_id,
price_table.clone(),
));
price_table.update_price(OLIVE_OIL, 12);
price_table.update_price(GLOVES, 13);
price_table.update_price(SNEAKERS, 80);
const OLIVE_OIL: ProductId = 323423;
const GLOVES: ProductId = 3966623;
const SNEAKERS: ProductId = 23222;
let index = Index::create_in_ram(schema);
let mut writer = index.writer_with_num_threads(1, 10_000_000)?;
writer.add_document(doc!(product_id=>OLIVE_OIL, text=>"cooking olive oil from greece"))?;
writer.add_document(doc!(product_id=>GLOVES, text=>"kitchen gloves, perfect for cooking"))?;
writer.add_document(doc!(product_id=>SNEAKERS, text=>"uber sweet sneakers"))?;
writer.commit()?;
let warmers: Vec<Weak<dyn Warmer>> = vec![Arc::downgrade(
&(price_dynamic_column.clone() as Arc<dyn Warmer>),
)];
let reader: IndexReader = index
.reader_builder()
.warmers(warmers)
.num_searchers(1)
.try_into()?;
reader.reload()?;
let query_parser = QueryParser::for_index(&index, vec![text]);
let query = query_parser.parse_query("cooking")?;
let searcher = reader.searcher();
let score_by_price = move |segment_reader: &SegmentReader| {
let price = price_dynamic_column
.price_for_segment(segment_reader)
.unwrap();
move |doc_id: DocId| Reverse(price[doc_id as usize])
};
let most_expensive_first = TopDocs::with_limit(10).custom_score(score_by_price);
let hits = searcher.search(&query, &most_expensive_first)?;
assert_eq!(
&hits,
&[
(
Reverse(12u32),
DocAddress {
segment_ord: 0,
doc_id: 0u32
}
),
(
Reverse(13u32),
DocAddress {
segment_ord: 0,
doc_id: 1u32
}
),
]
);
// Olive oil just got more expensive!
price_table.update_price(OLIVE_OIL, 15);
// The price update are directly reflected on `reload`.
//
// Be careful here though!...
// You may have spotted that we are still using the same `Searcher`.
//
// It is up to the `Warmer` implementer to decide how
// to control this behavior.
reader.reload()?;
let hits_with_new_prices = searcher.search(&query, &most_expensive_first)?;
assert_eq!(
&hits_with_new_prices,
&[
(
Reverse(13u32),
DocAddress {
segment_ord: 0,
doc_id: 1u32
}
),
(
Reverse(15u32),
DocAddress {
segment_ord: 0,
doc_id: 0u32
}
),
]
);
Ok(())
}

6
fastfield_codecs/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "fastfield_codecs"
version = "0.1.0"
version = "0.2.0"
authors = ["Pascal Seitz <pascal@quickwit.io>"]
license = "MIT"
edition = "2018"
@@ -9,8 +9,8 @@ description = "Fast field codecs used by tantivy"
# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
[dependencies]
common = { path = "../common/" }
tantivy-bitpacker = { path = "../bitpacker/" }
common = { version = "0.3", path = "../common/", package = "tantivy-common" }
tantivy-bitpacker = { version="0.2", path = "../bitpacker/" }
prettytable-rs = {version="0.8.0", optional= true}
rand = {version="0.8.3", optional= true}

14
fastfield_codecs/benches/bench.rs
View File

@@ -4,14 +4,14 @@ extern crate test;
#[cfg(test)]
mod tests {
use fastfield_codecs::{
bitpacked::{BitpackedFastFieldReader, BitpackedFastFieldSerializer},
linearinterpol::{LinearInterpolFastFieldReader, LinearInterpolFastFieldSerializer},
multilinearinterpol::{
MultiLinearInterpolFastFieldReader, MultiLinearInterpolFastFieldSerializer,
},
*,
use fastfield_codecs::bitpacked::{BitpackedFastFieldReader, BitpackedFastFieldSerializer};
use fastfield_codecs::linearinterpol::{
LinearInterpolFastFieldReader, LinearInterpolFastFieldSerializer,
};
use fastfield_codecs::multilinearinterpol::{
MultiLinearInterpolFastFieldReader, MultiLinearInterpolFastFieldSerializer,
};
use fastfield_codecs::*;
fn get_data() -> Vec<u64> {
let mut data: Vec<_> = (100..55000_u64)

12
fastfield_codecs/src/bitpacked.rs
View File

@@ -1,13 +1,9 @@
use crate::FastFieldCodecReader;
use crate::FastFieldCodecSerializer;
use crate::FastFieldDataAccess;
use crate::FastFieldStats;
use common::BinarySerializable;
use std::io::{self, Write};
use tantivy_bitpacker::compute_num_bits;
use tantivy_bitpacker::BitPacker;
use tantivy_bitpacker::BitUnpacker;
use common::BinarySerializable;
use tantivy_bitpacker::{compute_num_bits, BitPacker, BitUnpacker};
use crate::{FastFieldCodecReader, FastFieldCodecSerializer, FastFieldDataAccess, FastFieldStats};
/// Depending on the field type, a different
/// fast field is required.

16
fastfield_codecs/src/lib.rs
View File

@@ -53,7 +53,8 @@ pub trait FastFieldCodecSerializer {
pub trait FastFieldDataAccess {
/// Return the value associated to the given position.
///
/// Whenever possible use the Iterator passed to the fastfield creation instead, for performance reasons.
/// Whenever possible use the Iterator passed to the fastfield creation instead, for performance
/// reasons.
///
/// # Panics
///
@@ -62,6 +63,7 @@ pub trait FastFieldDataAccess {
}
#[derive(Debug, Clone)]
/// Statistics are used in codec detection and stored in the fast field footer.
pub struct FastFieldStats {
pub min_value: u64,
pub max_value: u64,
@@ -82,12 +84,10 @@ impl FastFieldDataAccess for Vec<u64> {
#[cfg(test)]
mod tests {
use crate::{
bitpacked::{BitpackedFastFieldReader, BitpackedFastFieldSerializer},
linearinterpol::{LinearInterpolFastFieldReader, LinearInterpolFastFieldSerializer},
multilinearinterpol::{
MultiLinearInterpolFastFieldReader, MultiLinearInterpolFastFieldSerializer,
},
use crate::bitpacked::{BitpackedFastFieldReader, BitpackedFastFieldSerializer};
use crate::linearinterpol::{LinearInterpolFastFieldReader, LinearInterpolFastFieldSerializer};
use crate::multilinearinterpol::{
MultiLinearInterpolFastFieldReader, MultiLinearInterpolFastFieldSerializer,
};
pub fn create_and_validate<S: FastFieldCodecSerializer, R: FastFieldCodecReader>(
@@ -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)> {

35
fastfield_codecs/src/linearinterpol.rs
View File

@@ -1,15 +1,10 @@
use crate::FastFieldCodecReader;
use crate::FastFieldCodecSerializer;
use crate::FastFieldDataAccess;
use crate::FastFieldStats;
use std::io::{self, Read, Write};
use std::ops::Sub;
use tantivy_bitpacker::compute_num_bits;
use tantivy_bitpacker::BitPacker;
use common::BinarySerializable;
use common::FixedSize;
use tantivy_bitpacker::BitUnpacker;
use common::{BinarySerializable, FixedSize};
use tantivy_bitpacker::{compute_num_bits, BitPacker, BitUnpacker};
use crate::{FastFieldCodecReader, FastFieldCodecSerializer, FastFieldDataAccess, FastFieldStats};
/// Depending on the field type, a different
/// fast field is required.
@@ -137,7 +132,7 @@ impl FastFieldCodecSerializer for LinearInterpolFastFieldSerializer {
// will be offset to 0
offset = offset.max(calculated_value - actual_value);
} else {
//positive value no offset reuqired
// positive value no offset reuqired
rel_positive_max = rel_positive_max.max(actual_value - calculated_value);
}
}
@@ -171,7 +166,7 @@ impl FastFieldCodecSerializer for LinearInterpolFastFieldSerializer {
stats: FastFieldStats,
) -> bool {
if stats.num_vals < 3 {
return false; //disable compressor for this case
return false; // disable compressor for this case
}
// On serialisation the offset is added to the actual value.
// We need to make sure this won't run into overflow calculation issues.
@@ -211,8 +206,8 @@ impl FastFieldCodecSerializer for LinearInterpolFastFieldSerializer {
.max()
.unwrap_or(0);
// the theory would be that we don't have the actual max_distance, but we are close within 50%
// threshold.
// the theory would be that we don't have the actual max_distance, but we are close within
// 50% threshold.
// It is multiplied by 2 because in a log case scenario the line would be as much above as
// below. So the offset would = max_distance
//
@@ -239,11 +234,21 @@ 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]

15
fastfield_codecs/src/main.rs
View File

@@ -1,10 +1,8 @@
#[macro_use]
extern crate prettytable;
use fastfield_codecs::{
linearinterpol::LinearInterpolFastFieldSerializer,
multilinearinterpol::MultiLinearInterpolFastFieldSerializer, FastFieldCodecSerializer,
FastFieldStats,
};
use fastfield_codecs::linearinterpol::LinearInterpolFastFieldSerializer;
use fastfield_codecs::multilinearinterpol::MultiLinearInterpolFastFieldSerializer;
use fastfield_codecs::{FastFieldCodecSerializer, FastFieldStats};
use prettytable::{Cell, Row, Table};
fn main() {
@@ -24,7 +22,7 @@ fn main() {
);
results.push(res);
//let best_estimation_codec = results
// let best_estimation_codec = results
//.iter()
//.min_by(|res1, res2| res1.partial_cmp(&res2).unwrap())
//.unwrap();
@@ -41,7 +39,6 @@ fn main() {
} else {
(est.to_string(), comp.to_string())
};
#[allow(clippy::all)]
let style = if comp == best_compression_ratio_codec.1 {
"Fb"
} else {
@@ -49,7 +46,7 @@ fn main() {
};
table.add_row(Row::new(vec![
Cell::new(&name.to_string()).style_spec("bFg"),
Cell::new(name).style_spec("bFg"),
Cell::new(&ratio_cell).style_spec(style),
Cell::new(&est_cell).style_spec(""),
]));
@@ -73,7 +70,7 @@ pub fn get_codec_test_data_sets() -> Vec<(Vec<u64>, &'static str)> {
current_cumulative
})
.collect::<Vec<_>>();
//let data = (1..=200000_u64).map(|num| num + num).collect::<Vec<_>>();
// let data = (1..=200000_u64).map(|num| num + num).collect::<Vec<_>>();
data_and_names.push((data, "Monotonically increasing concave"));
let mut current_cumulative = 0;

58
fastfield_codecs/src/multilinearinterpol.rs
View File

@@ -1,16 +1,22 @@
use crate::FastFieldCodecReader;
use crate::FastFieldCodecSerializer;
use crate::FastFieldDataAccess;
use crate::FastFieldStats;
use common::CountingWriter;
//! MultiLinearInterpol compressor uses linear interpolation to guess a values and stores the
//! offset, but in blocks of 512.
//!
//! With a CHUNK_SIZE of 512 and 29 byte metadata per block, we get a overhead for metadata of 232 /
//! 512 = 0,45 bits per element. The additional space required per element in a block is the the
//! maximum deviation of the linear interpolation estimation function.
//!
//! E.g. if the maximum deviation of an element is 12, all elements cost 4bits.
//!
//! Size per block:
//! Num Elements * Maximum Deviation from Interpolation + 29 Byte Metadata
use std::io::{self, Read, Write};
use std::ops::Sub;
use tantivy_bitpacker::compute_num_bits;
use tantivy_bitpacker::BitPacker;
use common::BinarySerializable;
use common::DeserializeFrom;
use tantivy_bitpacker::BitUnpacker;
use common::{BinarySerializable, CountingWriter, DeserializeFrom};
use tantivy_bitpacker::{compute_num_bits, BitPacker, BitUnpacker};
use crate::{FastFieldCodecReader, FastFieldCodecSerializer, FastFieldDataAccess, FastFieldStats};
const CHUNK_SIZE: u64 = 512;
@@ -43,7 +49,7 @@ 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 {
@@ -238,11 +244,11 @@ impl FastFieldCodecSerializer for MultiLinearInterpolFastFieldSerializer {
);
if calculated_value > actual_value {
// negative value we need to apply an offset
// we ignore negative values in the max value calculation, because negative values
// will be offset to 0
// we ignore negative values in the max value calculation, because negative
// values will be offset to 0
offset = offset.max(calculated_value - actual_value);
} else {
//positive value no offset reuqired
// positive value no offset reuqired
rel_positive_max = rel_positive_max.max(actual_value - calculated_value);
}
}
@@ -336,8 +342,8 @@ impl FastFieldCodecSerializer for MultiLinearInterpolFastFieldSerializer {
.unwrap();
// Estimate one block and extrapolate the cost to all blocks.
// the theory would be that we don't have the actual max_distance, but we are close within 50%
// threshold.
// the theory would be that we don't have the actual max_distance, but we are close within
// 50% threshold.
// It is multiplied by 2 because in a log case scenario the line would be as much above as
// below. So the offset would = max_distance
//
@@ -364,11 +370,22 @@ 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]
@@ -400,10 +417,9 @@ 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 _ = create_and_validate(&data, "random");
data.reverse();
create_and_validate(&data, "random");
}

3
ownedbytes/Cargo.toml
View File

@@ -1,9 +1,10 @@
[package]
authors = ["Paul Masurel <paul@quickwit.io>", "Pascal Seitz <pascal@quickwit.io>"]
name = "ownedbytes"
version = "0.1.0"
version = "0.3.0"
edition = "2018"
description = "Expose data as static slice"
license = "MIT"
# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
[dependencies]

63
ownedbytes/src/lib.rs
View File

@@ -1,9 +1,9 @@
use stable_deref_trait::StableDeref;
use std::convert::TryInto;
use std::mem;
use std::ops::{Deref, Range};
use std::sync::Arc;
use std::{fmt, io};
use std::{fmt, io, mem};
use stable_deref_trait::StableDeref;
/// An OwnedBytes simply wraps an object that owns a slice of data and exposes
/// this data as a static slice.
@@ -35,6 +35,8 @@ impl OwnedBytes {
}
/// creates a fileslice that is just a view over a slice of the data.
#[must_use]
#[inline]
pub fn slice(&self, range: Range<usize>) -> Self {
OwnedBytes {
data: &self.data[range],
@@ -63,6 +65,8 @@ impl OwnedBytes {
/// On the other hand, both `left` and `right` retain a handle over
/// the entire slice of memory. In other words, the memory will only
/// be released when both left and right are dropped.
#[inline]
#[must_use]
pub fn split(self, split_len: usize) -> (OwnedBytes, OwnedBytes) {
let right_box_stable_deref = self.box_stable_deref.clone();
let left = OwnedBytes {
@@ -76,6 +80,19 @@ 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 {
@@ -83,8 +100,6 @@ impl OwnedBytes {
}
/// Drops the left most `advance_len` bytes.
///
/// See also [.clip(clip_len: usize))](#method.clip).
#[inline]
pub fn advance(&mut self, advance_len: usize) {
self.data = &self.data[advance_len..]
@@ -124,6 +139,34 @@ 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 +330,14 @@ 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");
}
}

7
query-grammar/Cargo.toml
View File

@@ -1,13 +1,12 @@
[package]
name = "tantivy-query-grammar"
version = "0.15.0"
version = "0.18.0"
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"
homepage = "https://github.com/quickwit-oss/tantivy"
repository = "https://github.com/quickwit-oss/tantivy"
readme = "README.md"
keywords = ["search", "information", "retrieval"]
edition = "2018"

48
query-grammar/src/query_grammar.rs
View File

@@ -1,21 +1,24 @@
use super::user_input_ast::{UserInputAst, UserInputBound, UserInputLeaf, UserInputLiteral};
use crate::Occur;
use combine::error::StringStreamError;
use combine::parser::char::{char, digit, space, spaces, string};
use combine::parser::combinator::recognize;
use combine::parser::range::{take_while, take_while1};
use combine::parser::repeat::escaped;
use combine::parser::Parser;
use combine::{
attempt, choice, eof, many, many1, one_of, optional, parser, satisfy, skip_many1, value,
};
use combine::{error::StringStreamError, parser::combinator::recognize};
use once_cell::sync::Lazy;
use regex::Regex;
// Note: '-' char is only forbidden at the beginning of a field name, would be clearer to add it to special characters.
use super::user_input_ast::{UserInputAst, UserInputBound, UserInputLeaf, UserInputLiteral};
use crate::Occur;
// Note: '-' char is only forbidden at the beginning of a field name, would be clearer to add it to
// special characters.
const SPECIAL_CHARS: &[char] = &[
'+', '^', '`', ':', '{', '}', '"', '[', ']', '(', ')', '~', '!', '\\', '*', ' ',
];
const ESCAPED_SPECIAL_CHARS_PATTERN: &str = r#"\\(\+|\^|`|:|\{|\}|"|\[|\]|\(|\)|\~|!|\\|\*| )"#;
const ESCAPED_SPECIAL_CHARS_PATTERN: &str = r#"\\(\+|\^|`|:|\{|\}|"|\[|\]|\(|\)|\~|!|\\|\*|\s)"#;
/// Parses a field_name
/// A field name must have at least one character and be followed by a colon.
@@ -31,7 +34,8 @@ fn field_name<'a>() -> impl Parser<&'a str, Output = String> {
take_while(|c| !SPECIAL_CHARS.contains(&c)),
),
'\\',
satisfy(|c| SPECIAL_CHARS.contains(&c)),
satisfy(|_| true), /* if the next character is not a special char, the \ will be treated
* as the \ character. */
))
.skip(char(':'))
.map(|s| ESCAPED_SPECIAL_CHARS_RE.replace_all(&s, "$1").to_string())
@@ -64,7 +68,7 @@ fn word<'a>() -> impl Parser<&'a str, Output = String> {
///
/// NOTE: also accepts 999999-99-99T99:99:99.266051969+99:99
/// We delegate rejecting such invalid dates to the logical AST compuation code
/// which invokes chrono::DateTime::parse_from_rfc3339 on the value to actually parse
/// which invokes time::OffsetDateTime::parse(..., &Rfc3339) on the value to actually parse
/// it (instead of merely extracting the datetime value as string as done here).
fn date_time<'a>() -> impl Parser<&'a str, Output = String> {
let two_digits = || recognize::<String, _, _>((digit(), digit()));
@@ -363,9 +367,10 @@ mod test {
type TestParseResult = Result<(), StringStreamError>;
use super::*;
use combine::parser::Parser;
use super::*;
pub fn nearly_equals(a: f64, b: f64) -> bool {
(a - b).abs() < 0.0005 * (a + b).abs()
}
@@ -512,15 +517,27 @@ mod test {
}
#[test]
fn test_field_name() -> TestParseResult {
fn test_field_name() {
assert_eq!(
super::field_name().parse(".my.field.name:a"),
Ok((".my.field.name".to_string(), "a"))
);
assert_eq!(
super::field_name().parse(r#"my\　field:a"#),
Ok(("my　field".to_string(), "a"))
);
assert_eq!(
super::field_name().parse(r#"にんじん:a"#),
Ok(("にんじん".to_string(), "a"))
);
assert_eq!(
super::field_name().parse("my\\ field\\ name:a"),
Ok(("my field name".to_string(), "a"))
);
assert_eq!(
super::field_name().parse(r#"my\field:a"#),
Ok((r#"my\field"#.to_string(), "a"))
);
assert!(super::field_name().parse("my field:a").is_err());
assert_eq!(
super::field_name().parse("\\(1\\+1\\):2"),
@@ -530,14 +547,21 @@ mod test {
super::field_name().parse("my_field_name:a"),
Ok(("my_field_name".to_string(), "a"))
);
assert_eq!(
super::field_name().parse("myfield.b:hello").unwrap(),
("myfield.b".to_string(), "hello")
);
assert_eq!(
super::field_name().parse(r#"myfield\.b:hello"#).unwrap(),
(r#"myfield\.b"#.to_string(), "hello")
);
assert!(super::field_name().parse("my_field_name").is_err());
assert!(super::field_name().parse(":a").is_err());
assert!(super::field_name().parse("-my_field:a").is_err());
assert_eq!(
super::field_name().parse("_my_field:a")?,
("_my_field".to_string(), "a")
super::field_name().parse("_my_field:a"),
Ok(("_my_field".to_string(), "a"))
);
Ok(())
}
#[test]

5
query-grammar/src/user_input_ast.rs
View File

@@ -59,7 +59,7 @@ pub enum UserInputBound {
}
impl UserInputBound {
fn display_lower(&self, formatter: &mut fmt::Formatter<'_>) -> Result<(), fmt::Error> {
fn display_lower(&self, formatter: &mut fmt::Formatter) -> Result<(), fmt::Error> {
match *self {
UserInputBound::Inclusive(ref word) => write!(formatter, "[\"{}\"", word),
UserInputBound::Exclusive(ref word) => write!(formatter, "{{\"{}\"", word),
@@ -67,7 +67,7 @@ impl UserInputBound {
}
}
fn display_upper(&self, formatter: &mut fmt::Formatter<'_>) -> Result<(), fmt::Error> {
fn display_upper(&self, formatter: &mut fmt::Formatter) -> Result<(), fmt::Error> {
match *self {
UserInputBound::Inclusive(ref word) => write!(formatter, "\"{}\"]", word),
UserInputBound::Exclusive(ref word) => write!(formatter, "\"{}\"}}", word),
@@ -91,6 +91,7 @@ pub enum UserInputAst {
}
impl UserInputAst {
#[must_use]
pub fn unary(self, occur: Occur) -> UserInputAst {
UserInputAst::Clause(vec![(Some(occur), self)])
}

8
rustfmt.toml
View File

@@ -1 +1,7 @@
use_try_shorthand = true
comment_width = 120
format_strings = true
group_imports = "StdExternalCrate"
imports_granularity = "Module"
normalize_comments = true
where_single_line = true
wrap_comments = true

36
src/aggregation/README.md Normal file
View File

@@ -0,0 +1,36 @@
# Contributing
When adding new bucket aggregation make sure to extend the "test_aggregation_flushing" test for at least 2 levels.
# Code Organization
Tantivy's aggregations have been designed to mimic the
[aggregations of elasticsearch](https://www.elastic.co/guide/en/elasticsearch/reference/current/search-aggregations.html).
The code is organized in submodules:
## bucket
Contains all bucket aggregations, like range aggregation. These bucket aggregations group documents into buckets and can contain sub-aggegations.
## metric
Contains all metric aggregations, like average aggregation. Metric aggregations do not have sub aggregations.
#### agg_req
agg_req contains the users aggregation request. Deserialization from json is compatible with elasticsearch aggregation requests.
#### agg_req_with_accessor
agg_req_with_accessor contains the users aggregation request enriched with fast field accessors etc, which are
used during collection.
#### segment_agg_result
segment_agg_result contains the aggregation result tree, which is used for collection of a segment.
The tree from agg_req_with_accessor is passed during collection.
#### intermediate_agg_result
intermediate_agg_result contains the aggregation tree for merging with other trees.
#### agg_result
agg_result contains the final aggregation tree.

358
src/aggregation/agg_req.rs Normal file
View File

@@ -0,0 +1,358 @@
//! Contains the aggregation request tree. Used to build an
//! [AggregationCollector](super::AggregationCollector).
//!
//! [Aggregations] is the top level entry point to create a request, which is a `HashMap<String,
//! Aggregation>`.
//!
//! Requests are compatible with the json format of elasticsearch.
//!
//! # Example
//!
//! ```
//! use tantivy::aggregation::bucket::RangeAggregation;
//! use tantivy::aggregation::agg_req::BucketAggregationType;
//! use tantivy::aggregation::agg_req::{Aggregation, Aggregations};
//! use tantivy::aggregation::agg_req::BucketAggregation;
//! let agg_req1: Aggregations = vec![
//! (
//! "range".to_string(),
//! Aggregation::Bucket(BucketAggregation {
//! bucket_agg: BucketAggregationType::Range(RangeAggregation{
//! field: "score".to_string(),
//! ranges: vec![(3f64..7f64).into(), (7f64..20f64).into()],
//! }),
//! sub_aggregation: Default::default(),
//! }),
//! ),
//! ]
//! .into_iter()
//! .collect();
//!
//! let elasticsearch_compatible_json_req = r#"
//! {
//! "range": {
//! "range": {
//! "field": "score",
//! "ranges": [
//! { "from": 3.0, "to": 7.0 },
//! { "from": 7.0, "to": 20.0 }
//! ]
//! }
//! }
//! }"#;
//! let agg_req2: Aggregations = serde_json::from_str(elasticsearch_compatible_json_req).unwrap();
//! assert_eq!(agg_req1, agg_req2);
//! ```
use std::collections::{HashMap, HashSet};
use serde::{Deserialize, Serialize};
pub use super::bucket::RangeAggregation;
use super::bucket::{HistogramAggregation, TermsAggregation};
use super::metric::{AverageAggregation, StatsAggregation};
use super::VecWithNames;
/// The top-level aggregation request structure, which contains [Aggregation] and their user defined
/// names. It is also used in [buckets](BucketAggregation) to define sub-aggregations.
///
/// The key is the user defined name of the aggregation.
pub type Aggregations = HashMap<String, Aggregation>;
/// Like Aggregations, but optimized to work with the aggregation result
#[derive(Clone, Debug)]
pub(crate) struct AggregationsInternal {
pub(crate) metrics: VecWithNames<MetricAggregation>,
pub(crate) buckets: VecWithNames<BucketAggregationInternal>,
}
impl From<Aggregations> for AggregationsInternal {
fn from(aggs: Aggregations) -> Self {
let mut metrics = vec![];
let mut buckets = vec![];
for (key, agg) in aggs {
match agg {
Aggregation::Bucket(bucket) => buckets.push((
key,
BucketAggregationInternal {
bucket_agg: bucket.bucket_agg,
sub_aggregation: bucket.sub_aggregation.into(),
},
)),
Aggregation::Metric(metric) => metrics.push((key, metric)),
}
}
Self {
metrics: VecWithNames::from_entries(metrics),
buckets: VecWithNames::from_entries(buckets),
}
}
}
#[derive(Clone, Debug)]
// Like BucketAggregation, but optimized to work with the result
pub(crate) struct BucketAggregationInternal {
/// Bucket aggregation strategy to group documents.
pub bucket_agg: BucketAggregationType,
/// The sub_aggregations in the buckets. Each bucket will aggregate on the document set in the
/// bucket.
pub sub_aggregation: AggregationsInternal,
}
impl BucketAggregationInternal {
pub(crate) fn as_histogram(&self) -> Option<&HistogramAggregation> {
match &self.bucket_agg {
BucketAggregationType::Histogram(histogram) => Some(histogram),
_ => None,
}
}
pub(crate) fn as_term(&self) -> Option<&TermsAggregation> {
match &self.bucket_agg {
BucketAggregationType::Terms(terms) => Some(terms),
_ => None,
}
}
}
/// Extract all fields, where the term directory is used in the tree.
pub fn get_term_dict_field_names(aggs: &Aggregations) -> HashSet<String> {
let mut term_dict_field_names = Default::default();
for el in aggs.values() {
el.get_term_dict_field_names(&mut term_dict_field_names)
}
term_dict_field_names
}
/// Extract all fast field names used in the tree.
pub fn get_fast_field_names(aggs: &Aggregations) -> HashSet<String> {
let mut fast_field_names = Default::default();
for el in aggs.values() {
el.get_fast_field_names(&mut fast_field_names)
}
fast_field_names
}
/// Aggregation request of [BucketAggregation] or [MetricAggregation].
///
/// An aggregation is either a bucket or a metric.
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
#[serde(untagged)]
pub enum Aggregation {
/// Bucket aggregation, see [BucketAggregation] for details.
Bucket(BucketAggregation),
/// Metric aggregation, see [MetricAggregation] for details.
Metric(MetricAggregation),
}
impl Aggregation {
fn get_term_dict_field_names(&self, term_field_names: &mut HashSet<String>) {
if let Aggregation::Bucket(bucket) = self {
bucket.get_term_dict_field_names(term_field_names)
}
}
fn get_fast_field_names(&self, fast_field_names: &mut HashSet<String>) {
match self {
Aggregation::Bucket(bucket) => bucket.get_fast_field_names(fast_field_names),
Aggregation::Metric(metric) => metric.get_fast_field_names(fast_field_names),
}
}
}
/// BucketAggregations create buckets of documents. Each bucket is associated with a rule which
/// determines whether or not a document in the falls into it. In other words, the buckets
/// effectively define document sets. Buckets are not necessarily disjunct, therefore a document can
/// fall into multiple buckets. In addition to the buckets themselves, the bucket aggregations also
/// compute and return the number of documents for each bucket. Bucket aggregations, as opposed to
/// metric aggregations, can hold sub-aggregations. These sub-aggregations will be aggregated for
/// the buckets created by their "parent" bucket aggregation. There are different bucket
/// aggregators, each with a different "bucketing" strategy. Some define a single bucket, some
/// define fixed number of multiple buckets, and others dynamically create the buckets during the
/// aggregation process.
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct BucketAggregation {
/// Bucket aggregation strategy to group documents.
#[serde(flatten)]
pub bucket_agg: BucketAggregationType,
/// The sub_aggregations in the buckets. Each bucket will aggregate on the document set in the
/// bucket.
#[serde(rename = "aggs")]
#[serde(default)]
#[serde(skip_serializing_if = "Aggregations::is_empty")]
pub sub_aggregation: Aggregations,
}
impl BucketAggregation {
fn get_term_dict_field_names(&self, term_dict_field_names: &mut HashSet<String>) {
if let BucketAggregationType::Terms(terms) = &self.bucket_agg {
term_dict_field_names.insert(terms.field.to_string());
}
term_dict_field_names.extend(get_term_dict_field_names(&self.sub_aggregation));
}
fn get_fast_field_names(&self, fast_field_names: &mut HashSet<String>) {
self.bucket_agg.get_fast_field_names(fast_field_names);
fast_field_names.extend(get_fast_field_names(&self.sub_aggregation));
}
}
/// The bucket aggregation types.
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub enum BucketAggregationType {
/// Put data into buckets of user-defined ranges.
#[serde(rename = "range")]
Range(RangeAggregation),
/// Put data into buckets of user-defined ranges.
#[serde(rename = "histogram")]
Histogram(HistogramAggregation),
/// Put data into buckets of terms.
#[serde(rename = "terms")]
Terms(TermsAggregation),
}
impl BucketAggregationType {
fn get_fast_field_names(&self, fast_field_names: &mut HashSet<String>) {
match self {
BucketAggregationType::Terms(terms) => fast_field_names.insert(terms.field.to_string()),
BucketAggregationType::Range(range) => fast_field_names.insert(range.field.to_string()),
BucketAggregationType::Histogram(histogram) => {
fast_field_names.insert(histogram.field.to_string())
}
};
}
}
/// The aggregations in this family compute metrics based on values extracted
/// from the documents that are being aggregated. Values are extracted from the fast field of
/// the document.
/// Some aggregations output a single numeric metric (e.g. Average) and are called
/// single-value numeric metrics aggregation, others generate multiple metrics (e.g. Stats) and are
/// called multi-value numeric metrics aggregation.
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub enum MetricAggregation {
/// Calculates the average.
#[serde(rename = "avg")]
Average(AverageAggregation),
/// Calculates stats sum, average, min, max, standard_deviation on a field.
#[serde(rename = "stats")]
Stats(StatsAggregation),
}
impl MetricAggregation {
fn get_fast_field_names(&self, fast_field_names: &mut HashSet<String>) {
match self {
MetricAggregation::Average(avg) => fast_field_names.insert(avg.field.to_string()),
MetricAggregation::Stats(stats) => fast_field_names.insert(stats.field.to_string()),
};
}
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn serialize_to_json_test() {
let agg_req1: Aggregations = vec![(
"range".to_string(),
Aggregation::Bucket(BucketAggregation {
bucket_agg: BucketAggregationType::Range(RangeAggregation {
field: "score".to_string(),
ranges: vec![
(f64::MIN..3f64).into(),
(3f64..7f64).into(),
(7f64..20f64).into(),
(20f64..f64::MAX).into(),
],
}),
sub_aggregation: Default::default(),
}),
)]
.into_iter()
.collect();
let elasticsearch_compatible_json_req = r#"{
"range": {
"range": {
"field": "score",
"ranges": [
{
"to": 3.0
},
{
"from": 3.0,
"to": 7.0
},
{
"from": 7.0,
"to": 20.0
},
{
"from": 20.0
}
]
}
}
}"#;
let agg_req2: String = serde_json::to_string_pretty(&agg_req1).unwrap();
assert_eq!(agg_req2, elasticsearch_compatible_json_req);
}
#[test]
fn test_get_fast_field_names() {
let agg_req2: Aggregations = vec![
(
"range".to_string(),
Aggregation::Bucket(BucketAggregation {
bucket_agg: BucketAggregationType::Range(RangeAggregation {
field: "score2".to_string(),
ranges: vec![
(f64::MIN..3f64).into(),
(3f64..7f64).into(),
(7f64..20f64).into(),
(20f64..f64::MAX).into(),
],
}),
sub_aggregation: Default::default(),
}),
),
(
"metric".to_string(),
Aggregation::Metric(MetricAggregation::Average(
AverageAggregation::from_field_name("field123".to_string()),
)),
),
]
.into_iter()
.collect();
let agg_req1: Aggregations = vec![(
"range".to_string(),
Aggregation::Bucket(BucketAggregation {
bucket_agg: BucketAggregationType::Range(RangeAggregation {
field: "score".to_string(),
ranges: vec![
(f64::MIN..3f64).into(),
(3f64..7f64).into(),
(7f64..20f64).into(),
(20f64..f64::MAX).into(),
],
}),
sub_aggregation: agg_req2,
}),
)]
.into_iter()
.collect();
assert_eq!(
get_fast_field_names(&agg_req1),
vec![
"score".to_string(),
"score2".to_string(),
"field123".to_string()
]
.into_iter()
.collect()
)
}
}

203
src/aggregation/agg_req_with_accessor.rs Normal file
View File

@@ -0,0 +1,203 @@
//! This will enhance the request tree with access to the fastfield and metadata.
use std::sync::Arc;
use super::agg_req::{Aggregation, Aggregations, BucketAggregationType, MetricAggregation};
use super::bucket::{HistogramAggregation, RangeAggregation, TermsAggregation};
use super::metric::{AverageAggregation, StatsAggregation};
use super::VecWithNames;
use crate::fastfield::{
type_and_cardinality, DynamicFastFieldReader, FastType, MultiValuedFastFieldReader,
};
use crate::schema::{Cardinality, Type};
use crate::{InvertedIndexReader, SegmentReader, TantivyError};
#[derive(Clone, Default)]
pub(crate) struct AggregationsWithAccessor {
pub metrics: VecWithNames<MetricAggregationWithAccessor>,
pub buckets: VecWithNames<BucketAggregationWithAccessor>,
}
impl AggregationsWithAccessor {
fn from_data(
metrics: VecWithNames<MetricAggregationWithAccessor>,
buckets: VecWithNames<BucketAggregationWithAccessor>,
) -> Self {
Self { metrics, buckets }
}
pub fn is_empty(&self) -> bool {
self.metrics.is_empty() && self.buckets.is_empty()
}
}
#[derive(Clone)]
pub(crate) enum FastFieldAccessor {
Multi(MultiValuedFastFieldReader<u64>),
Single(DynamicFastFieldReader<u64>),
}
impl FastFieldAccessor {
pub fn as_single(&self) -> Option<&DynamicFastFieldReader<u64>> {
match self {
FastFieldAccessor::Multi(_) => None,
FastFieldAccessor::Single(reader) => Some(reader),
}
}
pub fn as_multi(&self) -> Option<&MultiValuedFastFieldReader<u64>> {
match self {
FastFieldAccessor::Multi(reader) => Some(reader),
FastFieldAccessor::Single(_) => None,
}
}
}
#[derive(Clone)]
pub struct BucketAggregationWithAccessor {
/// In general there can be buckets without fast field access, e.g. buckets that are created
/// based on search terms. So eventually this needs to be Option or moved.
pub(crate) accessor: FastFieldAccessor,
pub(crate) inverted_index: Option<Arc<InvertedIndexReader>>,
pub(crate) field_type: Type,
pub(crate) bucket_agg: BucketAggregationType,
pub(crate) sub_aggregation: AggregationsWithAccessor,
}
impl BucketAggregationWithAccessor {
fn try_from_bucket(
bucket: &BucketAggregationType,
sub_aggregation: &Aggregations,
reader: &SegmentReader,
) -> crate::Result<BucketAggregationWithAccessor> {
let mut inverted_index = None;
let (accessor, field_type) = match &bucket {
BucketAggregationType::Range(RangeAggregation {
field: field_name,
ranges: _,
}) => get_ff_reader_and_validate(reader, field_name, Cardinality::SingleValue)?,
BucketAggregationType::Histogram(HistogramAggregation {
field: field_name, ..
}) => get_ff_reader_and_validate(reader, field_name, Cardinality::SingleValue)?,
BucketAggregationType::Terms(TermsAggregation {
field: field_name, ..
}) => {
let field = reader
.schema()
.get_field(field_name)
.ok_or_else(|| TantivyError::FieldNotFound(field_name.to_string()))?;
inverted_index = Some(reader.inverted_index(field)?);
get_ff_reader_and_validate(reader, field_name, Cardinality::MultiValues)?
}
};
let sub_aggregation = sub_aggregation.clone();
Ok(BucketAggregationWithAccessor {
accessor,
field_type,
sub_aggregation: get_aggs_with_accessor_and_validate(&sub_aggregation, reader)?,
bucket_agg: bucket.clone(),
inverted_index,
})
}
}
/// Contains the metric request and the fast field accessor.
#[derive(Clone)]
pub struct MetricAggregationWithAccessor {
pub metric: MetricAggregation,
pub field_type: Type,
pub accessor: DynamicFastFieldReader<u64>,
}
impl MetricAggregationWithAccessor {
fn try_from_metric(
metric: &MetricAggregation,
reader: &SegmentReader,
) -> crate::Result<MetricAggregationWithAccessor> {
match &metric {
MetricAggregation::Average(AverageAggregation { field: field_name })
| MetricAggregation::Stats(StatsAggregation { field: field_name }) => {
let (accessor, field_type) =
get_ff_reader_and_validate(reader, field_name, Cardinality::SingleValue)?;
Ok(MetricAggregationWithAccessor {
accessor: accessor
.as_single()
.expect("unexpected fast field cardinality")
.clone(),
field_type,
metric: metric.clone(),
})
}
}
}
}
pub(crate) fn get_aggs_with_accessor_and_validate(
aggs: &Aggregations,
reader: &SegmentReader,
) -> crate::Result<AggregationsWithAccessor> {
let mut metrics = vec![];
let mut buckets = vec![];
for (key, agg) in aggs.iter() {
match agg {
Aggregation::Bucket(bucket) => buckets.push((
key.to_string(),
BucketAggregationWithAccessor::try_from_bucket(
&bucket.bucket_agg,
&bucket.sub_aggregation,
reader,
)?,
)),
Aggregation::Metric(metric) => metrics.push((
key.to_string(),
MetricAggregationWithAccessor::try_from_metric(metric, reader)?,
)),
}
}
Ok(AggregationsWithAccessor::from_data(
VecWithNames::from_entries(metrics),
VecWithNames::from_entries(buckets),
))
}
/// Get fast field reader with given cardinatility.
fn get_ff_reader_and_validate(
reader: &SegmentReader,
field_name: &str,
cardinality: Cardinality,
) -> crate::Result<(FastFieldAccessor, Type)> {
let field = reader
.schema()
.get_field(field_name)
.ok_or_else(|| TantivyError::FieldNotFound(field_name.to_string()))?;
let field_type = reader.schema().get_field_entry(field).field_type();
if let Some((ff_type, field_cardinality)) = type_and_cardinality(field_type) {
if ff_type == FastType::Date {
return Err(TantivyError::InvalidArgument(
"Unsupported field type date in aggregation".to_string(),
));
}
if cardinality != field_cardinality {
return Err(TantivyError::InvalidArgument(format!(
"Invalid field cardinality on field {} expected {:?}, but got {:?}",
field_name, cardinality, field_cardinality
)));
}
} else {
return Err(TantivyError::InvalidArgument(format!(
"Only fast fields of type f64, u64, i64 are supported, but got {:?} ",
field_type.value_type()
)));
};
let ff_fields = reader.fast_fields();
match cardinality {
Cardinality::SingleValue => ff_fields
.u64_lenient(field)
.map(|field| (FastFieldAccessor::Single(field), field_type.value_type())),
Cardinality::MultiValues => ff_fields
.u64s_lenient(field)
.map(|field| (FastFieldAccessor::Multi(field), field_type.value_type())),
}
}

404
src/aggregation/agg_result.rs Normal file
View File

@@ -0,0 +1,404 @@
//! Contains the final aggregation tree.
//! This tree can be converted via the `into()` method from `IntermediateAggregationResults`.
//! This conversion computes the final result. For example: The intermediate result contains
//! intermediate average results, which is the sum and the number of values. The actual average is
//! calculated on the step from intermediate to final aggregation result tree.
use std::cmp::Ordering;
use std::collections::HashMap;
use serde::{Deserialize, Serialize};
use super::agg_req::{
Aggregations, AggregationsInternal, BucketAggregationInternal, MetricAggregation,
};
use super::bucket::{intermediate_buckets_to_final_buckets, GetDocCount};
use super::intermediate_agg_result::{
IntermediateAggregationResults, IntermediateBucketResult, IntermediateHistogramBucketEntry,
IntermediateMetricResult, IntermediateRangeBucketEntry,
};
use super::metric::{SingleMetricResult, Stats};
use super::{Key, VecWithNames};
use crate::TantivyError;
#[derive(Clone, Default, Debug, PartialEq, Serialize, Deserialize)]
/// The final aggegation result.
pub struct AggregationResults(pub HashMap<String, AggregationResult>);
impl AggregationResults {
pub(crate) fn get_value_from_aggregation(
&self,
name: &str,
agg_property: &str,
) -> crate::Result<Option<f64>> {
if let Some(agg) = self.0.get(name) {
agg.get_value_from_aggregation(name, agg_property)
} else {
// Validation is be done during request parsing, so we can't reach this state.
Err(TantivyError::InternalError(format!(
"Can't find aggregation {:?} in sub_aggregations",
name
)))
}
}
/// Convert and intermediate result and its aggregation request to the final result
pub fn from_intermediate_and_req(
results: IntermediateAggregationResults,
agg: Aggregations,
) -> crate::Result<Self> {
AggregationResults::from_intermediate_and_req_internal(results, &(agg.into()))
}
/// Convert and intermediate result and its aggregation request to the final result
///
/// Internal function, CollectorAggregations is used instead Aggregations, which is optimized
/// for internal processing, by splitting metric and buckets into seperate groups.
pub(crate) fn from_intermediate_and_req_internal(
intermediate_results: IntermediateAggregationResults,
req: &AggregationsInternal,
) -> crate::Result<Self> {
// Important assumption:
// When the tree contains buckets/metric, we expect it to have all buckets/metrics from the
// request
let mut results: HashMap<String, AggregationResult> = HashMap::new();
if let Some(buckets) = intermediate_results.buckets {
add_coverted_final_buckets_to_result(&mut results, buckets, &req.buckets)?
} else {
// When there are no buckets, we create empty buckets, so that the serialized json
// format is constant
add_empty_final_buckets_to_result(&mut results, &req.buckets)?
};
if let Some(metrics) = intermediate_results.metrics {
add_converted_final_metrics_to_result(&mut results, metrics);
} else {
// When there are no metrics, we create empty metric results, so that the serialized
// json format is constant
add_empty_final_metrics_to_result(&mut results, &req.metrics)?;
}
Ok(Self(results))
}
}
fn add_converted_final_metrics_to_result(
results: &mut HashMap<String, AggregationResult>,
metrics: VecWithNames<IntermediateMetricResult>,
) {
results.extend(
metrics
.into_iter()
.map(|(key, metric)| (key, AggregationResult::MetricResult(metric.into()))),
);
}
fn add_empty_final_metrics_to_result(
results: &mut HashMap<String, AggregationResult>,
req_metrics: &VecWithNames<MetricAggregation>,
) -> crate::Result<()> {
results.extend(req_metrics.iter().map(|(key, req)| {
let empty_bucket = IntermediateMetricResult::empty_from_req(req);
(
key.to_string(),
AggregationResult::MetricResult(empty_bucket.into()),
)
}));
Ok(())
}
fn add_empty_final_buckets_to_result(
results: &mut HashMap<String, AggregationResult>,
req_buckets: &VecWithNames<BucketAggregationInternal>,
) -> crate::Result<()> {
let requested_buckets = req_buckets.iter();
for (key, req) in requested_buckets {
let empty_bucket = AggregationResult::BucketResult(BucketResult::empty_from_req(req)?);
results.insert(key.to_string(), empty_bucket);
}
Ok(())
}
fn add_coverted_final_buckets_to_result(
results: &mut HashMap<String, AggregationResult>,
buckets: VecWithNames<IntermediateBucketResult>,
req_buckets: &VecWithNames<BucketAggregationInternal>,
) -> crate::Result<()> {
assert_eq!(buckets.len(), req_buckets.len());
let buckets_with_request = buckets.into_iter().zip(req_buckets.values());
for ((key, bucket), req) in buckets_with_request {
let result =
AggregationResult::BucketResult(BucketResult::from_intermediate_and_req(bucket, req)?);
results.insert(key, result);
}
Ok(())
}
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
#[serde(untagged)]
/// An aggregation is either a bucket or a metric.
pub enum AggregationResult {
/// Bucket result variant.
BucketResult(BucketResult),
/// Metric result variant.
MetricResult(MetricResult),
}
impl AggregationResult {
pub(crate) fn get_value_from_aggregation(
&self,
_name: &str,
agg_property: &str,
) -> crate::Result<Option<f64>> {
match self {
AggregationResult::BucketResult(_bucket) => Err(TantivyError::InternalError(
"Tried to retrieve value from bucket aggregation. This is not supported and \
should not happen during collection, but should be catched during validation"
.to_string(),
)),
AggregationResult::MetricResult(metric) => metric.get_value(agg_property),
}
}
}
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
#[serde(untagged)]
/// MetricResult
pub enum MetricResult {
/// Average metric result.
Average(SingleMetricResult),
/// Stats metric result.
Stats(Stats),
}
impl MetricResult {
fn get_value(&self, agg_property: &str) -> crate::Result<Option<f64>> {
match self {
MetricResult::Average(avg) => Ok(avg.value),
MetricResult::Stats(stats) => stats.get_value(agg_property),
}
}
}
impl From<IntermediateMetricResult> for MetricResult {
fn from(metric: IntermediateMetricResult) -> Self {
match metric {
IntermediateMetricResult::Average(avg_data) => {
MetricResult::Average(avg_data.finalize().into())
}
IntermediateMetricResult::Stats(intermediate_stats) => {
MetricResult::Stats(intermediate_stats.finalize())
}
}
}
}
/// BucketEntry holds bucket aggregation result types.
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
#[serde(untagged)]
pub enum BucketResult {
/// This is the range entry for a bucket, which contains a key, count, from, to, and optionally
/// sub_aggregations.
Range {
/// The range buckets sorted by range.
buckets: Vec<RangeBucketEntry>,
},
/// This is the histogram entry for a bucket, which contains a key, count, and optionally
/// sub_aggregations.
Histogram {
/// The buckets.
///
/// If there are holes depends on the request, if min_doc_count is 0, then there are no
/// holes between the first and last bucket.
/// See [HistogramAggregation](super::bucket::HistogramAggregation)
buckets: Vec<BucketEntry>,
},
/// This is the term result
Terms {
/// The buckets.
///
/// See [TermsAggregation](super::bucket::TermsAggregation)
buckets: Vec<BucketEntry>,
/// The number of documents that didn’t make it into to TOP N due to shard_size or size
sum_other_doc_count: u64,
#[serde(skip_serializing_if = "Option::is_none")]
/// The upper bound error for the doc count of each term.
doc_count_error_upper_bound: Option<u64>,
},
}
impl BucketResult {
pub(crate) fn empty_from_req(req: &BucketAggregationInternal) -> crate::Result<Self> {
let empty_bucket = IntermediateBucketResult::empty_from_req(&req.bucket_agg);
BucketResult::from_intermediate_and_req(empty_bucket, req)
}
fn from_intermediate_and_req(
bucket_result: IntermediateBucketResult,
req: &BucketAggregationInternal,
) -> crate::Result<Self> {
match bucket_result {
IntermediateBucketResult::Range(range_res) => {
let mut buckets: Vec<RangeBucketEntry> = range_res
.buckets
.into_iter()
.map(|(_, bucket)| {
RangeBucketEntry::from_intermediate_and_req(bucket, &req.sub_aggregation)
})
.collect::<crate::Result<Vec<_>>>()?;
buckets.sort_by(|left, right| {
// TODO use total_cmp next stable rust release
left.from
.unwrap_or(f64::MIN)
.partial_cmp(&right.from.unwrap_or(f64::MIN))
.unwrap_or(Ordering::Equal)
});
Ok(BucketResult::Range { buckets })
}
IntermediateBucketResult::Histogram { buckets } => {
let buckets = intermediate_buckets_to_final_buckets(
buckets,
req.as_histogram()
.expect("unexpected aggregation, expected histogram aggregation"),
&req.sub_aggregation,
)?;
Ok(BucketResult::Histogram { buckets })
}
IntermediateBucketResult::Terms(terms) => terms.into_final_result(
req.as_term()
.expect("unexpected aggregation, expected term aggregation"),
&req.sub_aggregation,
),
}
}
}
/// This is the default entry for a bucket, which contains a key, count, and optionally
/// sub_aggregations.
///
/// # JSON Format
/// ```json
/// {
/// ...
/// "my_histogram": {
/// "buckets": [
/// {
/// "key": "2.0",
/// "doc_count": 5
/// },
/// {
/// "key": "4.0",
/// "doc_count": 2
/// },
/// {
/// "key": "6.0",
/// "doc_count": 3
/// }
/// ]
/// }
/// ...
/// }
/// ```
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct BucketEntry {
/// The identifier of the bucket.
pub key: Key,
/// Number of documents in the bucket.
pub doc_count: u64,
#[serde(flatten)]
/// Sub-aggregations in this bucket.
pub sub_aggregation: AggregationResults,
}
impl BucketEntry {
pub(crate) fn from_intermediate_and_req(
entry: IntermediateHistogramBucketEntry,
req: &AggregationsInternal,
) -> crate::Result<Self> {
Ok(BucketEntry {
key: Key::F64(entry.key),
doc_count: entry.doc_count,
sub_aggregation: AggregationResults::from_intermediate_and_req_internal(
entry.sub_aggregation,
req,
)?,
})
}
}
impl GetDocCount for &BucketEntry {
fn doc_count(&self) -> u64 {
self.doc_count
}
}
impl GetDocCount for BucketEntry {
fn doc_count(&self) -> u64 {
self.doc_count
}
}
/// This is the range entry for a bucket, which contains a key, count, and optionally
/// sub_aggregations.
///
/// # JSON Format
/// ```json
/// {
/// ...
/// "my_ranges": {
/// "buckets": [
/// {
/// "key": "*-10",
/// "to": 10,
/// "doc_count": 5
/// },
/// {
/// "key": "10-20",
/// "from": 10,
/// "to": 20,
/// "doc_count": 2
/// },
/// {
/// "key": "20-*",
/// "from": 20,
/// "doc_count": 3
/// }
/// ]
/// }
/// ...
/// }
/// ```
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct RangeBucketEntry {
/// The identifier of the bucket.
pub key: Key,
/// Number of documents in the bucket.
pub doc_count: u64,
#[serde(flatten)]
/// sub-aggregations in this bucket.
pub sub_aggregation: AggregationResults,
/// The from range of the bucket. Equals f64::MIN when None.
#[serde(skip_serializing_if = "Option::is_none")]
pub from: Option<f64>,
/// The to range of the bucket. Equals f64::MAX when None.
#[serde(skip_serializing_if = "Option::is_none")]
pub to: Option<f64>,
}
impl RangeBucketEntry {
fn from_intermediate_and_req(
entry: IntermediateRangeBucketEntry,
req: &AggregationsInternal,
) -> crate::Result<Self> {
Ok(RangeBucketEntry {
key: entry.key,
doc_count: entry.doc_count,
sub_aggregation: AggregationResults::from_intermediate_and_req_internal(
entry.sub_aggregation,
req,
)?,
to: entry.to,
from: entry.from,
})
}
}

1392
src/aggregation/bucket/histogram/histogram.rs Normal file
View File

File diff suppressed because it is too large Load Diff

2
src/aggregation/bucket/histogram/mod.rs Normal file
View File

@@ -0,0 +1,2 @@
mod histogram;
pub use histogram::*;

140
src/aggregation/bucket/mod.rs Normal file
View File

@@ -0,0 +1,140 @@
//! Module for all bucket aggregations.
//!
//! BucketAggregations create buckets of documents
//! [BucketAggregation](super::agg_req::BucketAggregation).
//!
//! Results of final buckets are [BucketResult](super::agg_result::BucketResult).
//! Results of intermediate buckets are
//! [IntermediateBucketResult](super::intermediate_agg_result::IntermediateBucketResult)
mod histogram;
mod range;
mod term_agg;
use std::collections::HashMap;
pub(crate) use histogram::SegmentHistogramCollector;
pub use histogram::*;
pub(crate) use range::SegmentRangeCollector;
pub use range::*;
use serde::{de, Deserialize, Deserializer, Serialize, Serializer};
pub use term_agg::*;
/// Order for buckets in a bucket aggregation.
#[derive(Clone, Copy, Debug, PartialEq, Serialize, Deserialize)]
pub enum Order {
/// Asc order
#[serde(rename = "asc")]
Asc,
/// Desc order
#[serde(rename = "desc")]
Desc,
}
impl Default for Order {
fn default() -> Self {
Order::Desc
}
}
#[derive(Clone, Debug, PartialEq)]
/// Order property by which to apply the order
pub enum OrderTarget {
/// The key of the bucket
Key,
/// The doc count of the bucket
Count,
/// Order by value of the sub aggregation metric with identified by given `String`.
///
/// Only single value metrics are supported currently
SubAggregation(String),
}
impl Default for OrderTarget {
fn default() -> Self {
OrderTarget::Count
}
}
impl From<&str> for OrderTarget {
fn from(val: &str) -> Self {
match val {
"_key" => OrderTarget::Key,
"_count" => OrderTarget::Count,
_ => OrderTarget::SubAggregation(val.to_string()),
}
}
}
impl ToString for OrderTarget {
fn to_string(&self) -> String {
match self {
OrderTarget::Key => "_key".to_string(),
OrderTarget::Count => "_count".to_string(),
OrderTarget::SubAggregation(agg) => agg.to_string(),
}
}
}
/// Set the order. target is either "_count", "_key", or the name of
/// a metric sub_aggregation.
///
/// De/Serializes to elasticsearch compatible JSON.
///
/// Examples in JSON format:
/// { "_count": "asc" }
/// { "_key": "asc" }
/// { "average_price": "asc" }
#[derive(Clone, Default, Debug, PartialEq)]
pub struct CustomOrder {
/// The target property by which to sort by
pub target: OrderTarget,
/// The order asc or desc
pub order: Order,
}
impl Serialize for CustomOrder {
fn serialize<S>(&self, serializer: S) -> Result<S::Ok, S::Error>
where S: Serializer {
let map: HashMap<String, Order> =
std::iter::once((self.target.to_string(), self.order)).collect();
map.serialize(serializer)
}
}
impl<'de> Deserialize<'de> for CustomOrder {
fn deserialize<D>(deserializer: D) -> Result<CustomOrder, D::Error>
where D: Deserializer<'de> {
HashMap::<String, Order>::deserialize(deserializer).and_then(|map| {
if let Some((key, value)) = map.into_iter().next() {
Ok(CustomOrder {
target: key.as_str().into(),
order: value,
})
} else {
Err(de::Error::custom(
"unexpected empty map in order".to_string(),
))
}
})
}
}
#[test]
fn custom_order_serde_test() {
let order = CustomOrder {
target: OrderTarget::Key,
order: Order::Desc,
};
let order_str = serde_json::to_string(&order).unwrap();
assert_eq!(order_str, "{\"_key\":\"desc\"}");
let order_deser = serde_json::from_str(&order_str).unwrap();
assert_eq!(order, order_deser);
let order_deser: serde_json::Result<CustomOrder> = serde_json::from_str("{}");
assert!(order_deser.is_err());
let order_deser: serde_json::Result<CustomOrder> = serde_json::from_str("[]");
assert!(order_deser.is_err());
}

620
src/aggregation/bucket/range.rs Normal file
View File

@@ -0,0 +1,620 @@
use std::fmt::Debug;
use std::ops::Range;
use serde::{Deserialize, Serialize};
use crate::aggregation::agg_req_with_accessor::{
AggregationsWithAccessor, BucketAggregationWithAccessor,
};
use crate::aggregation::intermediate_agg_result::{
IntermediateBucketResult, IntermediateRangeBucketEntry, IntermediateRangeBucketResult,
};
use crate::aggregation::segment_agg_result::SegmentAggregationResultsCollector;
use crate::aggregation::{f64_from_fastfield_u64, f64_to_fastfield_u64, Key};
use crate::fastfield::FastFieldReader;
use crate::schema::Type;
use crate::{DocId, TantivyError};
/// Provide user-defined buckets to aggregate on.
/// Two special buckets will automatically be created to cover the whole range of values.
/// The provided buckets have to be continous.
/// During the aggregation, the values extracted from the fast_field `field` will be checked
/// against each bucket range. Note that this aggregation includes the from value and excludes the
/// to value for each range.
///
/// Result type is [BucketResult](crate::aggregation::agg_result::BucketResult) with
/// [RangeBucketEntry](crate::aggregation::agg_result::RangeBucketEntry) on the
/// AggregationCollector.
///
/// Result type is
/// [crate::aggregation::intermediate_agg_result::IntermediateBucketResult] with
/// [crate::aggregation::intermediate_agg_result::IntermediateRangeBucketEntry] on the
/// DistributedAggregationCollector.
///
/// # Limitations/Compatibility
/// Overlapping ranges are not yet supported.
///
/// The keyed parameter (elasticsearch) is not yet supported.
///
/// # Request JSON Format
/// ```json
/// {
/// "my_ranges": {
/// "field": "score",
/// "ranges": [
/// { "to": 3.0 },
/// { "from": 3.0, "to": 7.0 },
/// { "from": 7.0, "to": 20.0 },
/// { "from": 20.0 }
/// ]
/// }
/// }
/// ```
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct RangeAggregation {
/// The field to aggregate on.
pub field: String,
/// Note that this aggregation includes the from value and excludes the to value for each
/// range. Extra buckets will be created until the first to, and last from, if necessary.
pub ranges: Vec<RangeAggregationRange>,
}
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
/// The range for one range bucket.
pub struct RangeAggregationRange {
/// The from range value, which is inclusive in the range.
/// None equals to an open ended interval.
#[serde(skip_serializing_if = "Option::is_none", default)]
pub from: Option<f64>,
/// The to range value, which is not inclusive in the range.
/// None equals to an open ended interval.
#[serde(skip_serializing_if = "Option::is_none", default)]
pub to: Option<f64>,
}
impl From<Range<f64>> for RangeAggregationRange {
fn from(range: Range<f64>) -> Self {
let from = if range.start == f64::MIN {
None
} else {
Some(range.start)
};
let to = if range.end == f64::MAX {
None
} else {
Some(range.end)
};
RangeAggregationRange { from, to }
}
}
#[derive(Clone, Debug, PartialEq)]
pub(crate) struct SegmentRangeAndBucketEntry {
range: Range<u64>,
bucket: SegmentRangeBucketEntry,
}
/// The collector puts values from the fast field into the correct buckets and does a conversion to
/// the correct datatype.
#[derive(Clone, Debug, PartialEq)]
pub struct SegmentRangeCollector {
/// The buckets containing the aggregation data.
buckets: Vec<SegmentRangeAndBucketEntry>,
field_type: Type,
}
#[derive(Clone, PartialEq)]
pub(crate) struct SegmentRangeBucketEntry {
pub key: Key,
pub doc_count: u64,
pub sub_aggregation: Option<SegmentAggregationResultsCollector>,
/// The from range of the bucket. Equals f64::MIN when None.
pub from: Option<f64>,
/// The to range of the bucket. Equals f64::MAX when None. Open interval, `to` is not
/// inclusive.
pub to: Option<f64>,
}
impl Debug for SegmentRangeBucketEntry {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
f.debug_struct("SegmentRangeBucketEntry")
.field("key", &self.key)
.field("doc_count", &self.doc_count)
.field("from", &self.from)
.field("to", &self.to)
.finish()
}
}
impl SegmentRangeBucketEntry {
pub(crate) fn into_intermediate_bucket_entry(
self,
agg_with_accessor: &AggregationsWithAccessor,
) -> crate::Result<IntermediateRangeBucketEntry> {
let sub_aggregation = if let Some(sub_aggregation) = self.sub_aggregation {
sub_aggregation.into_intermediate_aggregations_result(agg_with_accessor)?
} else {
Default::default()
};
Ok(IntermediateRangeBucketEntry {
key: self.key,
doc_count: self.doc_count,
sub_aggregation,
from: self.from,
to: self.to,
})
}
}
impl SegmentRangeCollector {
pub fn into_intermediate_bucket_result(
self,
agg_with_accessor: &BucketAggregationWithAccessor,
) -> crate::Result<IntermediateBucketResult> {
let field_type = self.field_type;
let buckets = self
.buckets
.into_iter()
.map(move |range_bucket| {
Ok((
range_to_string(&range_bucket.range, &field_type),
range_bucket
.bucket
.into_intermediate_bucket_entry(&agg_with_accessor.sub_aggregation)?,
))
})
.collect::<crate::Result<_>>()?;
Ok(IntermediateBucketResult::Range(
IntermediateRangeBucketResult { buckets },
))
}
pub(crate) fn from_req_and_validate(
req: &RangeAggregation,
sub_aggregation: &AggregationsWithAccessor,
field_type: Type,
) -> crate::Result<Self> {
// The range input on the request is f64.
// We need to convert to u64 ranges, because we read the values as u64.
// The mapping from the conversion is monotonic so ordering is preserved.
let buckets = extend_validate_ranges(&req.ranges, &field_type)?
.iter()
.map(|range| {
let to = if range.end == u64::MAX {
None
} else {
Some(f64_from_fastfield_u64(range.end, &field_type))
};
let from = if range.start == u64::MIN {
None
} else {
Some(f64_from_fastfield_u64(range.start, &field_type))
};
let sub_aggregation = if sub_aggregation.is_empty() {
None
} else {
Some(SegmentAggregationResultsCollector::from_req_and_validate(
sub_aggregation,
)?)
};
Ok(SegmentRangeAndBucketEntry {
range: range.clone(),
bucket: SegmentRangeBucketEntry {
key: range_to_key(range, &field_type),
doc_count: 0,
sub_aggregation,
from,
to,
},
})
})
.collect::<crate::Result<_>>()?;
Ok(SegmentRangeCollector {
buckets,
field_type,
})
}
#[inline]
pub(crate) fn collect_block(
&mut self,
doc: &[DocId],
bucket_with_accessor: &BucketAggregationWithAccessor,
force_flush: bool,
) {
let mut iter = doc.chunks_exact(4);
let accessor = bucket_with_accessor
.accessor
.as_single()
.expect("unexpected fast field cardinatility");
for docs in iter.by_ref() {
let val1 = accessor.get(docs[0]);
let val2 = accessor.get(docs[1]);
let val3 = accessor.get(docs[2]);
let val4 = accessor.get(docs[3]);
let bucket_pos1 = self.get_bucket_pos(val1);
let bucket_pos2 = self.get_bucket_pos(val2);
let bucket_pos3 = self.get_bucket_pos(val3);
let bucket_pos4 = self.get_bucket_pos(val4);
self.increment_bucket(bucket_pos1, docs[0], &bucket_with_accessor.sub_aggregation);
self.increment_bucket(bucket_pos2, docs[1], &bucket_with_accessor.sub_aggregation);
self.increment_bucket(bucket_pos3, docs[2], &bucket_with_accessor.sub_aggregation);
self.increment_bucket(bucket_pos4, docs[3], &bucket_with_accessor.sub_aggregation);
}
for doc in iter.remainder() {
let val = accessor.get(*doc);
let bucket_pos = self.get_bucket_pos(val);
self.increment_bucket(bucket_pos, *doc, &bucket_with_accessor.sub_aggregation);
}
if force_flush {
for bucket in &mut self.buckets {
if let Some(sub_aggregation) = &mut bucket.bucket.sub_aggregation {
sub_aggregation
.flush_staged_docs(&bucket_with_accessor.sub_aggregation, force_flush);
}
}
}
}
#[inline]
fn increment_bucket(
&mut self,
bucket_pos: usize,
doc: DocId,
bucket_with_accessor: &AggregationsWithAccessor,
) {
let bucket = &mut self.buckets[bucket_pos];
bucket.bucket.doc_count += 1;
if let Some(sub_aggregation) = &mut bucket.bucket.sub_aggregation {
sub_aggregation.collect(doc, bucket_with_accessor);
}
}
#[inline]
fn get_bucket_pos(&self, val: u64) -> usize {
let pos = self
.buckets
.binary_search_by_key(&val, |probe| probe.range.start)
.unwrap_or_else(|pos| pos - 1);
debug_assert!(self.buckets[pos].range.contains(&val));
pos
}
}
/// Converts the user provided f64 range value to fast field value space.
///
/// Internally fast field values are always stored as u64.
/// If the fast field has u64 [1,2,5], these values are stored as is in the fast field.
/// A fast field with f64 [1.0, 2.0, 5.0] is converted to u64 space, using a
/// monotonic mapping function, so the order is preserved.
///
/// Consequently, a f64 user range 1.0..3.0 needs to be converted to fast field value space using
/// the same monotonic mapping function, so that the provided ranges contain the u64 values in the
/// fast field.
/// The alternative would be that every value read would be converted to the f64 range, but that is
/// more computational expensive when many documents are hit.
fn to_u64_range(range: &RangeAggregationRange, field_type: &Type) -> crate::Result<Range<u64>> {
let start = if let Some(from) = range.from {
f64_to_fastfield_u64(from, field_type)
.ok_or_else(|| TantivyError::InvalidArgument("invalid field type".to_string()))?
} else {
u64::MIN
};
let end = if let Some(to) = range.to {
f64_to_fastfield_u64(to, field_type)
.ok_or_else(|| TantivyError::InvalidArgument("invalid field type".to_string()))?
} else {
u64::MAX
};
Ok(start..end)
}
/// Extends the provided buckets to contain the whole value range, by inserting buckets at the
/// beginning and end.
fn extend_validate_ranges(
buckets: &[RangeAggregationRange],
field_type: &Type,
) -> crate::Result<Vec<Range<u64>>> {
let mut converted_buckets = buckets
.iter()
.map(|range| to_u64_range(range, field_type))
.collect::<crate::Result<Vec<_>>>()?;
converted_buckets.sort_by_key(|bucket| bucket.start);
if converted_buckets[0].start != u64::MIN {
converted_buckets.insert(0, u64::MIN..converted_buckets[0].start);
}
if converted_buckets[converted_buckets.len() - 1].end != u64::MAX {
converted_buckets.push(converted_buckets[converted_buckets.len() - 1].end..u64::MAX);
}
// fill up holes in the ranges
let find_hole = |converted_buckets: &[Range<u64>]| {
for (pos, ranges) in converted_buckets.windows(2).enumerate() {
if ranges[0].end > ranges[1].start {
return Err(TantivyError::InvalidArgument(format!(
"Overlapping ranges not supported range {:?}, range+1 {:?}",
ranges[0], ranges[1]
)));
}
if ranges[0].end != ranges[1].start {
return Ok(Some(pos));
}
}
Ok(None)
};
while let Some(hole_pos) = find_hole(&converted_buckets)? {
let new_range = converted_buckets[hole_pos].end..converted_buckets[hole_pos + 1].start;
converted_buckets.insert(hole_pos + 1, new_range);
}
Ok(converted_buckets)
}
pub(crate) fn range_to_string(range: &Range<u64>, field_type: &Type) -> String {
// is_start is there for malformed requests, e.g. ig the user passes the range u64::MIN..0.0,
// it should be rendererd as "*-0" and not "*-*"
let to_str = |val: u64, is_start: bool| {
if (is_start && val == u64::MIN) || (!is_start && val == u64::MAX) {
"*".to_string()
} else {
f64_from_fastfield_u64(val, field_type).to_string()
}
};
format!("{}-{}", to_str(range.start, true), to_str(range.end, false))
}
pub(crate) fn range_to_key(range: &Range<u64>, field_type: &Type) -> Key {
Key::Str(range_to_string(range, field_type))
}
#[cfg(test)]
mod tests {
use serde_json::Value;
use super::*;
use crate::aggregation::agg_req::{
Aggregation, Aggregations, BucketAggregation, BucketAggregationType,
};
use crate::aggregation::tests::get_test_index_with_num_docs;
use crate::aggregation::AggregationCollector;
use crate::fastfield::FastValue;
use crate::query::AllQuery;
pub fn get_collector_from_ranges(
ranges: Vec<RangeAggregationRange>,
field_type: Type,
) -> SegmentRangeCollector {
let req = RangeAggregation {
field: "dummy".to_string(),
ranges,
};
SegmentRangeCollector::from_req_and_validate(&req, &Default::default(), field_type)
.expect("unexpected error")
}
#[test]
fn range_fraction_test() -> crate::Result<()> {
let index = get_test_index_with_num_docs(false, 100)?;
let agg_req: Aggregations = vec![(
"range".to_string(),
Aggregation::Bucket(BucketAggregation {
bucket_agg: BucketAggregationType::Range(RangeAggregation {
field: "fraction_f64".to_string(),
ranges: vec![(0f64..0.1f64).into(), (0.1f64..0.2f64).into()],
}),
sub_aggregation: Default::default(),
}),
)]
.into_iter()
.collect();
let collector = AggregationCollector::from_aggs(agg_req);
let reader = index.reader()?;
let searcher = reader.searcher();
let agg_res = searcher.search(&AllQuery, &collector).unwrap();
let res: Value = serde_json::from_str(&serde_json::to_string(&agg_res)?)?;
assert_eq!(res["range"]["buckets"][0]["key"], "*-0");
assert_eq!(res["range"]["buckets"][0]["doc_count"], 0);
assert_eq!(res["range"]["buckets"][1]["key"], "0-0.1");
assert_eq!(res["range"]["buckets"][1]["doc_count"], 10);
assert_eq!(res["range"]["buckets"][2]["key"], "0.1-0.2");
assert_eq!(res["range"]["buckets"][2]["doc_count"], 10);
assert_eq!(res["range"]["buckets"][3]["key"], "0.2-*");
assert_eq!(res["range"]["buckets"][3]["doc_count"], 80);
Ok(())
}
#[test]
fn bucket_test_extend_range_hole() {
let buckets = vec![(10f64..20f64).into(), (30f64..40f64).into()];
let collector = get_collector_from_ranges(buckets, Type::F64);
let buckets = collector.buckets;
assert_eq!(buckets[0].range.start, u64::MIN);
assert_eq!(buckets[0].range.end, 10f64.to_u64());
assert_eq!(buckets[1].range.start, 10f64.to_u64());
assert_eq!(buckets[1].range.end, 20f64.to_u64());
// Added bucket to fill hole
assert_eq!(buckets[2].range.start, 20f64.to_u64());
assert_eq!(buckets[2].range.end, 30f64.to_u64());
assert_eq!(buckets[3].range.start, 30f64.to_u64());
assert_eq!(buckets[3].range.end, 40f64.to_u64());
}
#[test]
fn bucket_test_range_conversion_special_case() {
// the monotonic conversion between f64 and u64, does not map f64::MIN.to_u64() ==
// u64::MIN, but the into trait converts f64::MIN/MAX to None
let buckets = vec![
(f64::MIN..10f64).into(),
(10f64..20f64).into(),
(20f64..f64::MAX).into(),
];
let collector = get_collector_from_ranges(buckets, Type::F64);
let buckets = collector.buckets;
assert_eq!(buckets[0].range.start, u64::MIN);
assert_eq!(buckets[0].range.end, 10f64.to_u64());
assert_eq!(buckets[1].range.start, 10f64.to_u64());
assert_eq!(buckets[1].range.end, 20f64.to_u64());
assert_eq!(buckets[2].range.start, 20f64.to_u64());
assert_eq!(buckets[2].range.end, u64::MAX);
assert_eq!(buckets.len(), 3);
}
#[test]
fn bucket_range_test_negative_vals() {
let buckets = vec![(-10f64..-1f64).into()];
let collector = get_collector_from_ranges(buckets, Type::F64);
let buckets = collector.buckets;
assert_eq!(&buckets[0].bucket.key.to_string(), "*--10");
assert_eq!(&buckets[buckets.len() - 1].bucket.key.to_string(), "-1-*");
}
#[test]
fn bucket_range_test_positive_vals() {
let buckets = vec![(0f64..10f64).into()];
let collector = get_collector_from_ranges(buckets, Type::F64);
let buckets = collector.buckets;
assert_eq!(&buckets[0].bucket.key.to_string(), "*-0");
assert_eq!(&buckets[buckets.len() - 1].bucket.key.to_string(), "10-*");
}
#[test]
fn range_binary_search_test_u64() {
let check_ranges = |ranges: Vec<RangeAggregationRange>| {
let collector = get_collector_from_ranges(ranges, Type::U64);
let search = |val: u64| collector.get_bucket_pos(val);
assert_eq!(search(u64::MIN), 0);
assert_eq!(search(9), 0);
assert_eq!(search(10), 1);
assert_eq!(search(11), 1);
assert_eq!(search(99), 1);
assert_eq!(search(100), 2);
assert_eq!(search(u64::MAX - 1), 2); // Since the end range is never included, the max
// value
};
let ranges = vec![(10.0..100.0).into()];
check_ranges(ranges);
let ranges = vec![
RangeAggregationRange {
to: Some(10.0),
from: None,
},
(10.0..100.0).into(),
];
check_ranges(ranges);
let ranges = vec![
RangeAggregationRange {
to: Some(10.0),
from: None,
},
(10.0..100.0).into(),
RangeAggregationRange {
to: None,
from: Some(100.0),
},
];
check_ranges(ranges);
}
#[test]
fn range_binary_search_test_f64() {
let ranges = vec![(10.0..100.0).into()];
let collector = get_collector_from_ranges(ranges, Type::F64);
let search = |val: u64| collector.get_bucket_pos(val);
assert_eq!(search(u64::MIN), 0);
assert_eq!(search(9f64.to_u64()), 0);
assert_eq!(search(10f64.to_u64()), 1);
assert_eq!(search(11f64.to_u64()), 1);
assert_eq!(search(99f64.to_u64()), 1);
assert_eq!(search(100f64.to_u64()), 2);
assert_eq!(search(u64::MAX - 1), 2); // Since the end range is never included,
// the max value
}
}
#[cfg(all(test, feature = "unstable"))]
mod bench {
use itertools::Itertools;
use rand::seq::SliceRandom;
use rand::thread_rng;
use super::*;
use crate::aggregation::bucket::range::tests::get_collector_from_ranges;
const TOTAL_DOCS: u64 = 1_000_000u64;
const NUM_DOCS: u64 = 50_000u64;
fn get_collector_with_buckets(num_buckets: u64, num_docs: u64) -> SegmentRangeCollector {
let bucket_size = num_docs / num_buckets;
let mut buckets: Vec<RangeAggregationRange> = vec![];
for i in 0..num_buckets {
let bucket_start = (i * bucket_size) as f64;
buckets.push((bucket_start..bucket_start + bucket_size as f64).into())
}
get_collector_from_ranges(buckets, Type::U64)
}
fn get_rand_docs(total_docs: u64, num_docs_returned: u64) -> Vec<u64> {
let mut rng = thread_rng();
let all_docs = (0..total_docs - 1).collect_vec();
let mut vals = all_docs
.as_slice()
.choose_multiple(&mut rng, num_docs_returned as usize)
.cloned()
.collect_vec();
vals.sort();
vals
}
fn bench_range_binary_search(b: &mut test::Bencher, num_buckets: u64) {
let collector = get_collector_with_buckets(num_buckets, TOTAL_DOCS);
let vals = get_rand_docs(TOTAL_DOCS, NUM_DOCS);
b.iter(|| {
let mut bucket_pos = 0;
for val in &vals {
bucket_pos = collector.get_bucket_pos(*val);
}
bucket_pos
})
}
#[bench]
fn bench_range_100_buckets(b: &mut test::Bencher) {
bench_range_binary_search(b, 100)
}
#[bench]
fn bench_range_10_buckets(b: &mut test::Bencher) {
bench_range_binary_search(b, 10)
}
}

1320
src/aggregation/bucket/term_agg.rs Normal file
View File

File diff suppressed because it is too large Load Diff

145
src/aggregation/collector.rs Normal file
View File

@@ -0,0 +1,145 @@
use super::agg_req::Aggregations;
use super::agg_req_with_accessor::AggregationsWithAccessor;
use super::agg_result::AggregationResults;
use super::intermediate_agg_result::IntermediateAggregationResults;
use super::segment_agg_result::SegmentAggregationResultsCollector;
use crate::aggregation::agg_req_with_accessor::get_aggs_with_accessor_and_validate;
use crate::collector::{Collector, SegmentCollector};
use crate::SegmentReader;
/// Collector for aggregations.
///
/// The collector collects all aggregations by the underlying aggregation request.
pub struct AggregationCollector {
agg: Aggregations,
}
impl AggregationCollector {
/// Create collector from aggregation request.
pub fn from_aggs(agg: Aggregations) -> Self {
Self { agg }
}
}
/// Collector for distributed aggregations.
///
/// The collector collects all aggregations by the underlying aggregation request.
///
/// # Purpose
/// AggregationCollector returns `IntermediateAggregationResults` and not the final
/// `AggregationResults`, so that results from differenct indices can be merged and then converted
/// into the final `AggregationResults` via the `into()` method.
pub struct DistributedAggregationCollector {
agg: Aggregations,
}
impl DistributedAggregationCollector {
/// Create collector from aggregation request.
pub fn from_aggs(agg: Aggregations) -> Self {
Self { agg }
}
}
impl Collector for DistributedAggregationCollector {
type Fruit = IntermediateAggregationResults;
type Child = AggregationSegmentCollector;
fn for_segment(
&self,
_segment_local_id: crate::SegmentOrdinal,
reader: &crate::SegmentReader,
) -> crate::Result<Self::Child> {
AggregationSegmentCollector::from_agg_req_and_reader(&self.agg, reader)
}
fn requires_scoring(&self) -> bool {
false
}
fn merge_fruits(
&self,
segment_fruits: Vec<<Self::Child as SegmentCollector>::Fruit>,
) -> crate::Result<Self::Fruit> {
merge_fruits(segment_fruits)
}
}
impl Collector for AggregationCollector {
type Fruit = AggregationResults;
type Child = AggregationSegmentCollector;
fn for_segment(
&self,
_segment_local_id: crate::SegmentOrdinal,
reader: &crate::SegmentReader,
) -> crate::Result<Self::Child> {
AggregationSegmentCollector::from_agg_req_and_reader(&self.agg, reader)
}
fn requires_scoring(&self) -> bool {
false
}
fn merge_fruits(
&self,
segment_fruits: Vec<<Self::Child as SegmentCollector>::Fruit>,
) -> crate::Result<Self::Fruit> {
let res = merge_fruits(segment_fruits)?;
AggregationResults::from_intermediate_and_req(res, self.agg.clone())
}
}
fn merge_fruits(
mut segment_fruits: Vec<crate::Result<IntermediateAggregationResults>>,
) -> crate::Result<IntermediateAggregationResults> {
if let Some(fruit) = segment_fruits.pop() {
let mut fruit = fruit?;
for next_fruit in segment_fruits {
fruit.merge_fruits(next_fruit?);
}
Ok(fruit)
} else {
Ok(IntermediateAggregationResults::default())
}
}
/// AggregationSegmentCollector does the aggregation collection on a segment.
pub struct AggregationSegmentCollector {
aggs_with_accessor: AggregationsWithAccessor,
result: SegmentAggregationResultsCollector,
}
impl AggregationSegmentCollector {
/// Creates an AggregationSegmentCollector from an [Aggregations] request and a segment reader.
/// Also includes validation, e.g. checking field types and existence.
pub fn from_agg_req_and_reader(
agg: &Aggregations,
reader: &SegmentReader,
) -> crate::Result<Self> {
let aggs_with_accessor = get_aggs_with_accessor_and_validate(agg, reader)?;
let result =
SegmentAggregationResultsCollector::from_req_and_validate(&aggs_with_accessor)?;
Ok(AggregationSegmentCollector {
aggs_with_accessor,
result,
})
}
}
impl SegmentCollector for AggregationSegmentCollector {
type Fruit = crate::Result<IntermediateAggregationResults>;
#[inline]
fn collect(&mut self, doc: crate::DocId, _score: crate::Score) {
self.result.collect(doc, &self.aggs_with_accessor);
}
fn harvest(mut self) -> Self::Fruit {
self.result
.flush_staged_docs(&self.aggs_with_accessor, true);
self.result
.into_intermediate_aggregations_result(&self.aggs_with_accessor)
}
}

555
src/aggregation/intermediate_agg_result.rs Normal file
View File

@@ -0,0 +1,555 @@
//! Contains the intermediate aggregation tree, that can be merged.
//! Intermediate aggregation results can be used to merge results between segments or between
//! indices.
use std::cmp::Ordering;
use fnv::FnvHashMap;
use itertools::Itertools;
use serde::{Deserialize, Serialize};
use super::agg_req::{AggregationsInternal, BucketAggregationType, MetricAggregation};
use super::agg_result::BucketResult;
use super::bucket::{
cut_off_buckets, get_agg_name_and_property, GetDocCount, Order, OrderTarget,
SegmentHistogramBucketEntry, TermsAggregation,
};
use super::metric::{IntermediateAverage, IntermediateStats};
use super::segment_agg_result::SegmentMetricResultCollector;
use super::{Key, SerializedKey, VecWithNames};
use crate::aggregation::agg_result::{AggregationResults, BucketEntry};
use crate::aggregation::bucket::TermsAggregationInternal;
/// Contains the intermediate aggregation result, which is optimized to be merged with other
/// intermediate results.
#[derive(Default, Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct IntermediateAggregationResults {
#[serde(skip_serializing_if = "Option::is_none")]
pub(crate) metrics: Option<VecWithNames<IntermediateMetricResult>>,
#[serde(skip_serializing_if = "Option::is_none")]
pub(crate) buckets: Option<VecWithNames<IntermediateBucketResult>>,
}
impl IntermediateAggregationResults {
pub(crate) fn empty_from_req(req: &AggregationsInternal) -> Self {
let metrics = if req.metrics.is_empty() {
None
} else {
let metrics = req
.metrics
.iter()
.map(|(key, req)| {
(
key.to_string(),
IntermediateMetricResult::empty_from_req(req),
)
})
.collect();
Some(VecWithNames::from_entries(metrics))
};
let buckets = if req.buckets.is_empty() {
None
} else {
let buckets = req
.buckets
.iter()
.map(|(key, req)| {
(
key.to_string(),
IntermediateBucketResult::empty_from_req(&req.bucket_agg),
)
})
.collect();
Some(VecWithNames::from_entries(buckets))
};
Self { metrics, buckets }
}
/// Merge an other intermediate aggregation result into this result.
///
/// The order of the values need to be the same on both results. This is ensured when the same
/// (key values) are present on the underlying VecWithNames struct.
pub fn merge_fruits(&mut self, other: IntermediateAggregationResults) {
if let (Some(buckets_left), Some(buckets_right)) = (&mut self.buckets, other.buckets) {
for (bucket_left, bucket_right) in
buckets_left.values_mut().zip(buckets_right.into_values())
{
bucket_left.merge_fruits(bucket_right);
}
}
if let (Some(metrics_left), Some(metrics_right)) = (&mut self.metrics, other.metrics) {
for (metric_left, metric_right) in
metrics_left.values_mut().zip(metrics_right.into_values())
{
metric_left.merge_fruits(metric_right);
}
}
}
}
/// An aggregation is either a bucket or a metric.
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub enum IntermediateAggregationResult {
/// Bucket variant
Bucket(IntermediateBucketResult),
/// Metric variant
Metric(IntermediateMetricResult),
}
/// Holds the intermediate data for metric results
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub enum IntermediateMetricResult {
/// Average containing intermediate average data result
Average(IntermediateAverage),
/// AverageData variant
Stats(IntermediateStats),
}
impl From<SegmentMetricResultCollector> for IntermediateMetricResult {
fn from(tree: SegmentMetricResultCollector) -> Self {
match tree {
SegmentMetricResultCollector::Average(collector) => {
IntermediateMetricResult::Average(IntermediateAverage::from_collector(collector))
}
SegmentMetricResultCollector::Stats(collector) => {
IntermediateMetricResult::Stats(collector.stats)
}
}
}
}
impl IntermediateMetricResult {
pub(crate) fn empty_from_req(req: &MetricAggregation) -> Self {
match req {
MetricAggregation::Average(_) => {
IntermediateMetricResult::Average(IntermediateAverage::default())
}
MetricAggregation::Stats(_) => {
IntermediateMetricResult::Stats(IntermediateStats::default())
}
}
}
fn merge_fruits(&mut self, other: IntermediateMetricResult) {
match (self, other) {
(
IntermediateMetricResult::Average(avg_data_left),
IntermediateMetricResult::Average(avg_data_right),
) => {
avg_data_left.merge_fruits(avg_data_right);
}
(
IntermediateMetricResult::Stats(stats_left),
IntermediateMetricResult::Stats(stats_right),
) => {
stats_left.merge_fruits(stats_right);
}
_ => {
panic!("incompatible fruit types in tree");
}
}
}
}
/// The intermediate bucket results. Internally they can be easily merged via the keys of the
/// buckets.
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub enum IntermediateBucketResult {
/// This is the range entry for a bucket, which contains a key, count, from, to, and optionally
/// sub_aggregations.
Range(IntermediateRangeBucketResult),
/// This is the histogram entry for a bucket, which contains a key, count, and optionally
/// sub_aggregations.
Histogram {
/// The buckets
buckets: Vec<IntermediateHistogramBucketEntry>,
},
/// Term aggregation
Terms(IntermediateTermBucketResult),
}
impl IntermediateBucketResult {
pub(crate) fn empty_from_req(req: &BucketAggregationType) -> Self {
match req {
BucketAggregationType::Terms(_) => IntermediateBucketResult::Terms(Default::default()),
BucketAggregationType::Range(_) => IntermediateBucketResult::Range(Default::default()),
BucketAggregationType::Histogram(_) => {
IntermediateBucketResult::Histogram { buckets: vec![] }
}
}
}
fn merge_fruits(&mut self, other: IntermediateBucketResult) {
match (self, other) {
(
IntermediateBucketResult::Terms(term_res_left),
IntermediateBucketResult::Terms(term_res_right),
) => {
merge_maps(&mut term_res_left.entries, term_res_right.entries);
term_res_left.sum_other_doc_count += term_res_right.sum_other_doc_count;
term_res_left.doc_count_error_upper_bound +=
term_res_right.doc_count_error_upper_bound;
}
(
IntermediateBucketResult::Range(range_res_left),
IntermediateBucketResult::Range(range_res_right),
) => {
merge_maps(&mut range_res_left.buckets, range_res_right.buckets);
}
(
IntermediateBucketResult::Histogram {
buckets: buckets_left,
..
},
IntermediateBucketResult::Histogram {
buckets: buckets_right,
..
},
) => {
let buckets = buckets_left
.drain(..)
.merge_join_by(buckets_right.into_iter(), |left, right| {
left.key.partial_cmp(&right.key).unwrap_or(Ordering::Equal)
})
.map(|either| match either {
itertools::EitherOrBoth::Both(mut left, right) => {
left.merge_fruits(right);
left
}
itertools::EitherOrBoth::Left(left) => left,
itertools::EitherOrBoth::Right(right) => right,
})
.collect();
*buckets_left = buckets;
}
(IntermediateBucketResult::Range(_), _) => {
panic!("try merge on different types")
}
(IntermediateBucketResult::Histogram { .. }, _) => {
panic!("try merge on different types")
}
(IntermediateBucketResult::Terms { .. }, _) => {
panic!("try merge on different types")
}
}
}
}
#[derive(Default, Clone, Debug, PartialEq, Serialize, Deserialize)]
/// Range aggregation including error counts
pub struct IntermediateRangeBucketResult {
pub(crate) buckets: FnvHashMap<SerializedKey, IntermediateRangeBucketEntry>,
}
#[derive(Default, Clone, Debug, PartialEq, Serialize, Deserialize)]
/// Term aggregation including error counts
pub struct IntermediateTermBucketResult {
pub(crate) entries: FnvHashMap<String, IntermediateTermBucketEntry>,
pub(crate) sum_other_doc_count: u64,
pub(crate) doc_count_error_upper_bound: u64,
}
impl IntermediateTermBucketResult {
pub(crate) fn into_final_result(
self,
req: &TermsAggregation,
sub_aggregation_req: &AggregationsInternal,
) -> crate::Result<BucketResult> {
let req = TermsAggregationInternal::from_req(req);
let mut buckets: Vec<BucketEntry> = self
.entries
.into_iter()
.filter(|bucket| bucket.1.doc_count >= req.min_doc_count)
.map(|(key, entry)| {
Ok(BucketEntry {
key: Key::Str(key),
doc_count: entry.doc_count,
sub_aggregation: AggregationResults::from_intermediate_and_req_internal(
entry.sub_aggregation,
sub_aggregation_req,
)?,
})
})
.collect::<crate::Result<_>>()?;
let order = req.order.order;
match req.order.target {
OrderTarget::Key => {
buckets.sort_by(|left, right| {
if req.order.order == Order::Desc {
left.key.partial_cmp(&right.key)
} else {
right.key.partial_cmp(&left.key)
}
.expect("expected type string, which is always sortable")
});
}
OrderTarget::Count => {
if req.order.order == Order::Desc {
buckets.sort_unstable_by_key(|bucket| std::cmp::Reverse(bucket.doc_count()));
} else {
buckets.sort_unstable_by_key(|bucket| bucket.doc_count());
}
}
OrderTarget::SubAggregation(name) => {
let (agg_name, agg_property) = get_agg_name_and_property(&name);
let mut buckets_with_val = buckets
.into_iter()
.map(|bucket| {
let val = bucket
.sub_aggregation
.get_value_from_aggregation(agg_name, agg_property)?
.unwrap_or(f64::NAN);
Ok((bucket, val))
})
.collect::<crate::Result<Vec<_>>>()?;
buckets_with_val.sort_by(|(_, val1), (_, val2)| {
// TODO use total_cmp in next rust stable release
match &order {
Order::Desc => val2.partial_cmp(val1).unwrap_or(std::cmp::Ordering::Equal),
Order::Asc => val1.partial_cmp(val2).unwrap_or(std::cmp::Ordering::Equal),
}
});
buckets = buckets_with_val
.into_iter()
.map(|(bucket, _val)| bucket)
.collect_vec();
}
}
// We ignore _term_doc_count_before_cutoff here, because it increases the upperbound error
// only for terms that didn't make it into the top N.
//
// This can be interesting, as a value of quality of the results, but not good to check the
// actual error count for the returned terms.
let (_term_doc_count_before_cutoff, sum_other_doc_count) =
cut_off_buckets(&mut buckets, req.size as usize);
let doc_count_error_upper_bound = if req.show_term_doc_count_error {
Some(self.doc_count_error_upper_bound)
} else {
None
};
Ok(BucketResult::Terms {
buckets,
sum_other_doc_count: self.sum_other_doc_count + sum_other_doc_count,
doc_count_error_upper_bound,
})
}
}
trait MergeFruits {
fn merge_fruits(&mut self, other: Self);
}
fn merge_maps<V: MergeFruits + Clone>(
entries_left: &mut FnvHashMap<SerializedKey, V>,
mut entries_right: FnvHashMap<SerializedKey, V>,
) {
for (name, entry_left) in entries_left.iter_mut() {
if let Some(entry_right) = entries_right.remove(name) {
entry_left.merge_fruits(entry_right);
}
}
for (key, res) in entries_right.into_iter() {
entries_left.entry(key).or_insert(res);
}
}
/// This is the histogram entry for a bucket, which contains a key, count, and optionally
/// sub_aggregations.
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct IntermediateHistogramBucketEntry {
/// The unique the bucket is identified.
pub key: f64,
/// The number of documents in the bucket.
pub doc_count: u64,
/// The sub_aggregation in this bucket.
pub sub_aggregation: IntermediateAggregationResults,
}
impl From<SegmentHistogramBucketEntry> for IntermediateHistogramBucketEntry {
fn from(entry: SegmentHistogramBucketEntry) -> Self {
IntermediateHistogramBucketEntry {
key: entry.key,
doc_count: entry.doc_count,
sub_aggregation: Default::default(),
}
}
}
/// This is the range entry for a bucket, which contains a key, count, and optionally
/// sub_aggregations.
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct IntermediateRangeBucketEntry {
/// The unique the bucket is identified.
pub key: Key,
/// The number of documents in the bucket.
pub doc_count: u64,
/// The sub_aggregation in this bucket.
pub sub_aggregation: IntermediateAggregationResults,
/// The from range of the bucket. Equals f64::MIN when None.
#[serde(skip_serializing_if = "Option::is_none")]
pub from: Option<f64>,
/// The to range of the bucket. Equals f64::MAX when None.
#[serde(skip_serializing_if = "Option::is_none")]
pub to: Option<f64>,
}
/// This is the term entry for a bucket, which contains a count, and optionally
/// sub_aggregations.
#[derive(Clone, Default, Debug, PartialEq, Serialize, Deserialize)]
pub struct IntermediateTermBucketEntry {
/// The number of documents in the bucket.
pub doc_count: u64,
/// The sub_aggregation in this bucket.
pub sub_aggregation: IntermediateAggregationResults,
}
impl MergeFruits for IntermediateTermBucketEntry {
fn merge_fruits(&mut self, other: IntermediateTermBucketEntry) {
self.doc_count += other.doc_count;
self.sub_aggregation.merge_fruits(other.sub_aggregation);
}
}
impl MergeFruits for IntermediateRangeBucketEntry {
fn merge_fruits(&mut self, other: IntermediateRangeBucketEntry) {
self.doc_count += other.doc_count;
self.sub_aggregation.merge_fruits(other.sub_aggregation);
}
}
impl MergeFruits for IntermediateHistogramBucketEntry {
fn merge_fruits(&mut self, other: IntermediateHistogramBucketEntry) {
self.doc_count += other.doc_count;
self.sub_aggregation.merge_fruits(other.sub_aggregation);
}
}
#[cfg(test)]
mod tests {
use std::collections::HashMap;
use pretty_assertions::assert_eq;
use super::*;
fn get_sub_test_tree(data: &[(String, u64)]) -> IntermediateAggregationResults {
let mut map = HashMap::new();
let mut buckets = FnvHashMap::default();
for (key, doc_count) in data {
buckets.insert(
key.to_string(),
IntermediateRangeBucketEntry {
key: Key::Str(key.to_string()),
doc_count: *doc_count,
sub_aggregation: Default::default(),
from: None,
to: None,
},
);
}
map.insert(
"my_agg_level2".to_string(),
IntermediateBucketResult::Range(IntermediateRangeBucketResult { buckets }),
);
IntermediateAggregationResults {
buckets: Some(VecWithNames::from_entries(map.into_iter().collect())),
metrics: Default::default(),
}
}
fn get_intermediat_tree_with_ranges(
data: &[(String, u64, String, u64)],
) -> IntermediateAggregationResults {
let mut map = HashMap::new();
let mut buckets: FnvHashMap<_, _> = Default::default();
for (key, doc_count, sub_aggregation_key, sub_aggregation_count) in data {
buckets.insert(
key.to_string(),
IntermediateRangeBucketEntry {
key: Key::Str(key.to_string()),
doc_count: *doc_count,
from: None,
to: None,
sub_aggregation: get_sub_test_tree(&[(
sub_aggregation_key.to_string(),
*sub_aggregation_count,
)]),
},
);
}
map.insert(
"my_agg_level1".to_string(),
IntermediateBucketResult::Range(IntermediateRangeBucketResult { buckets }),
);
IntermediateAggregationResults {
buckets: Some(VecWithNames::from_entries(map.into_iter().collect())),
metrics: Default::default(),
}
}
#[test]
fn test_merge_fruits_tree_1() {
let mut tree_left = get_intermediat_tree_with_ranges(&[
("red".to_string(), 50, "1900".to_string(), 25),
("blue".to_string(), 30, "1900".to_string(), 30),
]);
let tree_right = get_intermediat_tree_with_ranges(&[
("red".to_string(), 60, "1900".to_string(), 30),
("blue".to_string(), 25, "1900".to_string(), 50),
]);
tree_left.merge_fruits(tree_right);
let tree_expected = get_intermediat_tree_with_ranges(&[
("red".to_string(), 110, "1900".to_string(), 55),
("blue".to_string(), 55, "1900".to_string(), 80),
]);
assert_eq!(tree_left, tree_expected);
}
#[test]
fn test_merge_fruits_tree_2() {
let mut tree_left = get_intermediat_tree_with_ranges(&[
("red".to_string(), 50, "1900".to_string(), 25),
("blue".to_string(), 30, "1900".to_string(), 30),
]);
let tree_right = get_intermediat_tree_with_ranges(&[
("red".to_string(), 60, "1900".to_string(), 30),
("green".to_string(), 25, "1900".to_string(), 50),
]);
tree_left.merge_fruits(tree_right);
let tree_expected = get_intermediat_tree_with_ranges(&[
("red".to_string(), 110, "1900".to_string(), 55),
("blue".to_string(), 30, "1900".to_string(), 30),
("green".to_string(), 25, "1900".to_string(), 50),
]);
assert_eq!(tree_left, tree_expected);
}
#[test]
fn test_merge_fruits_tree_empty() {
let mut tree_left = get_intermediat_tree_with_ranges(&[
("red".to_string(), 50, "1900".to_string(), 25),
("blue".to_string(), 30, "1900".to_string(), 30),
]);
let orig = tree_left.clone();
tree_left.merge_fruits(IntermediateAggregationResults::default());
assert_eq!(tree_left, orig);
}
}

114
src/aggregation/metric/average.rs Normal file
View File

@@ -0,0 +1,114 @@
use std::fmt::Debug;
use serde::{Deserialize, Serialize};
use crate::aggregation::f64_from_fastfield_u64;
use crate::fastfield::{DynamicFastFieldReader, FastFieldReader};
use crate::schema::Type;
use crate::DocId;
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
/// A single-value metric aggregation that computes the average of numeric values that are
/// extracted from the aggregated documents.
/// Supported field types are u64, i64, and f64.
/// See [super::SingleMetricResult] for return value.
///
/// # JSON Format
/// ```json
/// {
/// "avg": {
/// "field": "score",
/// }
/// }
/// ```
pub struct AverageAggregation {
/// The field name to compute the stats on.
pub field: String,
}
impl AverageAggregation {
/// Create new AverageAggregation from a field.
pub fn from_field_name(field_name: String) -> Self {
AverageAggregation { field: field_name }
}
/// Return the field name.
pub fn field_name(&self) -> &str {
&self.field
}
}
#[derive(Clone, PartialEq)]
pub(crate) struct SegmentAverageCollector {
pub data: IntermediateAverage,
field_type: Type,
}
impl Debug for SegmentAverageCollector {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
f.debug_struct("AverageCollector")
.field("data", &self.data)
.finish()
}
}
impl SegmentAverageCollector {
pub fn from_req(field_type: Type) -> Self {
Self {
field_type,
data: Default::default(),
}
}
pub(crate) fn collect_block(&mut self, doc: &[DocId], field: &DynamicFastFieldReader<u64>) {
let mut iter = doc.chunks_exact(4);
for docs in iter.by_ref() {
let val1 = field.get(docs[0]);
let val2 = field.get(docs[1]);
let val3 = field.get(docs[2]);
let val4 = field.get(docs[3]);
let val1 = f64_from_fastfield_u64(val1, &self.field_type);
let val2 = f64_from_fastfield_u64(val2, &self.field_type);
let val3 = f64_from_fastfield_u64(val3, &self.field_type);
let val4 = f64_from_fastfield_u64(val4, &self.field_type);
self.data.collect(val1);
self.data.collect(val2);
self.data.collect(val3);
self.data.collect(val4);
}
for doc in iter.remainder() {
let val = field.get(*doc);
let val = f64_from_fastfield_u64(val, &self.field_type);
self.data.collect(val);
}
}
}
/// Contains mergeable version of average data.
#[derive(Default, Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct IntermediateAverage {
pub(crate) sum: f64,
pub(crate) doc_count: u64,
}
impl IntermediateAverage {
pub(crate) fn from_collector(collector: SegmentAverageCollector) -> Self {
collector.data
}
/// Merge average data into this instance.
pub fn merge_fruits(&mut self, other: IntermediateAverage) {
self.sum += other.sum;
self.doc_count += other.doc_count;
}
/// compute final result
pub fn finalize(&self) -> Option<f64> {
if self.doc_count == 0 {
None
} else {
Some(self.sum / self.doc_count as f64)
}
}
#[inline]
fn collect(&mut self, val: f64) {
self.doc_count += 1;
self.sum += val;
}
}

30
src/aggregation/metric/mod.rs Normal file
View File

@@ -0,0 +1,30 @@
//! Module for all metric aggregations.
//!
//! The aggregations in this family compute metrics, see [super::agg_req::MetricAggregation] for
//! details.
mod average;
mod stats;
pub use average::*;
use serde::{Deserialize, Serialize};
pub use stats::*;
/// Single-metric aggregations use this common result structure.
///
/// Main reason to wrap it in value is to match elasticsearch output structure.
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct SingleMetricResult {
/// The value of the single value metric.
pub value: Option<f64>,
}
impl From<f64> for SingleMetricResult {
fn from(value: f64) -> Self {
Self { value: Some(value) }
}
}
impl From<Option<f64>> for SingleMetricResult {
fn from(value: Option<f64>) -> Self {
Self { value }
}
}

370
src/aggregation/metric/stats.rs Normal file
View File

@@ -0,0 +1,370 @@
use serde::{Deserialize, Serialize};
use crate::aggregation::f64_from_fastfield_u64;
use crate::fastfield::{DynamicFastFieldReader, FastFieldReader};
use crate::schema::Type;
use crate::{DocId, TantivyError};
/// A multi-value metric aggregation that computes stats of numeric values that are
/// extracted from the aggregated documents.
/// Supported field types are u64, i64, and f64.
/// See [Stats] for returned statistics.
///
/// # JSON Format
/// ```json
/// {
/// "stats": {
/// "field": "score",
/// }
/// }
/// ```
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct StatsAggregation {
/// The field name to compute the stats on.
pub field: String,
}
impl StatsAggregation {
/// Create new StatsAggregation from a field.
pub fn from_field_name(field_name: String) -> Self {
StatsAggregation { field: field_name }
}
/// Return the field name.
pub fn field_name(&self) -> &str {
&self.field
}
}
/// Stats contains a collection of statistics.
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct Stats {
/// The number of documents.
pub count: usize,
/// The sum of the fast field values.
pub sum: f64,
/// The standard deviation of the fast field values. None for count == 0.
pub standard_deviation: Option<f64>,
/// The min value of the fast field values.
pub min: Option<f64>,
/// The max value of the fast field values.
pub max: Option<f64>,
/// The average of the values. None for count == 0.
pub avg: Option<f64>,
}
impl Stats {
pub(crate) fn get_value(&self, agg_property: &str) -> crate::Result<Option<f64>> {
match agg_property {
"count" => Ok(Some(self.count as f64)),
"sum" => Ok(Some(self.sum)),
"standard_deviation" => Ok(self.standard_deviation),
"min" => Ok(self.min),
"max" => Ok(self.max),
"avg" => Ok(self.avg),
_ => Err(TantivyError::InvalidArgument(format!(
"unknown property {} on stats metric aggregation",
agg_property
))),
}
}
}
/// IntermediateStats contains the mergeable version for stats.
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
pub struct IntermediateStats {
count: usize,
sum: f64,
squared_sum: f64,
min: f64,
max: f64,
}
impl Default for IntermediateStats {
fn default() -> Self {
Self {
count: 0,
sum: 0.0,
squared_sum: 0.0,
min: f64::MAX,
max: f64::MIN,
}
}
}
impl IntermediateStats {
pub(crate) fn avg(&self) -> Option<f64> {
if self.count == 0 {
None
} else {
Some(self.sum / (self.count as f64))
}
}
fn square_mean(&self) -> f64 {
self.squared_sum / (self.count as f64)
}
pub(crate) fn standard_deviation(&self) -> Option<f64> {
self.avg()
.map(|average| (self.square_mean() - average * average).sqrt())
}
/// Merge data from other stats into this instance.
pub fn merge_fruits(&mut self, other: IntermediateStats) {
self.count += other.count;
self.sum += other.sum;
self.squared_sum += other.squared_sum;
self.min = self.min.min(other.min);
self.max = self.max.max(other.max);
}
/// compute final resultimprove_docs
pub fn finalize(&self) -> Stats {
let min = if self.count == 0 {
None
} else {
Some(self.min)
};
let max = if self.count == 0 {
None
} else {
Some(self.max)
};
Stats {
count: self.count,
sum: self.sum,
standard_deviation: self.standard_deviation(),
min,
max,
avg: self.avg(),
}
}
#[inline]
fn collect(&mut self, value: f64) {
self.count += 1;
self.sum += value;
self.squared_sum += value * value;
self.min = self.min.min(value);
self.max = self.max.max(value);
}
}
#[derive(Clone, Debug, PartialEq)]
pub(crate) struct SegmentStatsCollector {
pub(crate) stats: IntermediateStats,
field_type: Type,
}
impl SegmentStatsCollector {
pub fn from_req(field_type: Type) -> Self {
Self {
field_type,
stats: IntermediateStats::default(),
}
}
pub(crate) fn collect_block(&mut self, doc: &[DocId], field: &DynamicFastFieldReader<u64>) {
let mut iter = doc.chunks_exact(4);
for docs in iter.by_ref() {
let val1 = field.get(docs[0]);
let val2 = field.get(docs[1]);
let val3 = field.get(docs[2]);
let val4 = field.get(docs[3]);
let val1 = f64_from_fastfield_u64(val1, &self.field_type);
let val2 = f64_from_fastfield_u64(val2, &self.field_type);
let val3 = f64_from_fastfield_u64(val3, &self.field_type);
let val4 = f64_from_fastfield_u64(val4, &self.field_type);
self.stats.collect(val1);
self.stats.collect(val2);
self.stats.collect(val3);
self.stats.collect(val4);
}
for doc in iter.remainder() {
let val = field.get(*doc);
let val = f64_from_fastfield_u64(val, &self.field_type);
self.stats.collect(val);
}
}
}
#[cfg(test)]
mod tests {
use std::iter;
use serde_json::Value;
use crate::aggregation::agg_req::{
Aggregation, Aggregations, BucketAggregation, BucketAggregationType, MetricAggregation,
RangeAggregation,
};
use crate::aggregation::agg_result::AggregationResults;
use crate::aggregation::metric::StatsAggregation;
use crate::aggregation::tests::{get_test_index_2_segments, get_test_index_from_values};
use crate::aggregation::AggregationCollector;
use crate::query::{AllQuery, TermQuery};
use crate::schema::IndexRecordOption;
use crate::Term;
#[test]
fn test_aggregation_stats_empty_index() -> crate::Result<()> {
// test index without segments
let values = vec![];
let index = get_test_index_from_values(false, &values)?;
let agg_req_1: Aggregations = vec![(
"stats".to_string(),
Aggregation::Metric(MetricAggregation::Stats(StatsAggregation::from_field_name(
"score".to_string(),
))),
)]
.into_iter()
.collect();
let collector = AggregationCollector::from_aggs(agg_req_1);
let reader = index.reader()?;
let searcher = reader.searcher();
let agg_res: AggregationResults = searcher.search(&AllQuery, &collector).unwrap();
let res: Value = serde_json::from_str(&serde_json::to_string(&agg_res)?)?;
assert_eq!(
res["stats"],
json!({
"avg": Value::Null,
"count": 0,
"max": Value::Null,
"min": Value::Null,
"standard_deviation": Value::Null,
"sum": 0.0
})
);
Ok(())
}
#[test]
fn test_aggregation_stats() -> crate::Result<()> {
let index = get_test_index_2_segments(false)?;
let reader = index.reader()?;
let text_field = reader.searcher().schema().get_field("text").unwrap();
let term_query = TermQuery::new(
Term::from_field_text(text_field, "cool"),
IndexRecordOption::Basic,
);
let agg_req_1: Aggregations = vec![
(
"stats_i64".to_string(),
Aggregation::Metric(MetricAggregation::Stats(StatsAggregation::from_field_name(
"score_i64".to_string(),
))),
),
(
"stats_f64".to_string(),
Aggregation::Metric(MetricAggregation::Stats(StatsAggregation::from_field_name(
"score_f64".to_string(),
))),
),
(
"stats".to_string(),
Aggregation::Metric(MetricAggregation::Stats(StatsAggregation::from_field_name(
"score".to_string(),
))),
),
(
"range".to_string(),
Aggregation::Bucket(BucketAggregation {
bucket_agg: BucketAggregationType::Range(RangeAggregation {
field: "score".to_string(),
ranges: vec![
(3f64..7f64).into(),
(7f64..19f64).into(),
(19f64..20f64).into(),
],
}),
sub_aggregation: iter::once((
"stats".to_string(),
Aggregation::Metric(MetricAggregation::Stats(
StatsAggregation::from_field_name("score".to_string()),
)),
))
.collect(),
}),
),
]
.into_iter()
.collect();
let collector = AggregationCollector::from_aggs(agg_req_1);
let searcher = reader.searcher();
let agg_res: AggregationResults = searcher.search(&term_query, &collector).unwrap();
let res: Value = serde_json::from_str(&serde_json::to_string(&agg_res)?)?;
assert_eq!(
res["stats"],
json!({
"avg": 12.142857142857142,
"count": 7,
"max": 44.0,
"min": 1.0,
"standard_deviation": 13.65313748796613,
"sum": 85.0
})
);
assert_eq!(
res["stats_i64"],
json!({
"avg": 12.142857142857142,
"count": 7,
"max": 44.0,
"min": 1.0,
"standard_deviation": 13.65313748796613,
"sum": 85.0
})
);
assert_eq!(
res["stats_f64"],
json!({
"avg": 12.214285714285714,
"count": 7,
"max": 44.5,
"min": 1.0,
"standard_deviation": 13.819905785437443,
"sum": 85.5
})
);
assert_eq!(
res["range"]["buckets"][2]["stats"],
json!({
"avg": 10.666666666666666,
"count": 3,
"max": 14.0,
"min": 7.0,
"standard_deviation": 2.867441755680877,
"sum": 32.0
})
);
assert_eq!(
res["range"]["buckets"][3]["stats"],
json!({
"avg": serde_json::Value::Null,
"count": 0,
"max": serde_json::Value::Null,
"min": serde_json::Value::Null,
"standard_deviation": serde_json::Value::Null,
"sum": 0.0,
})
);
Ok(())
}
}

1591
src/aggregation/mod.rs Normal file
View File

File diff suppressed because it is too large Load Diff

272
src/aggregation/segment_agg_result.rs Normal file
View File

@@ -0,0 +1,272 @@
//! Contains aggregation trees which is used during collection in a segment.
//! This tree contains datastructrues optimized for fast collection.
//! The tree can be converted to an intermediate tree, which contains datastructrues optimized for
//! merging.
use std::fmt::Debug;
use super::agg_req::MetricAggregation;
use super::agg_req_with_accessor::{
AggregationsWithAccessor, BucketAggregationWithAccessor, MetricAggregationWithAccessor,
};
use super::bucket::{SegmentHistogramCollector, SegmentRangeCollector, SegmentTermCollector};
use super::intermediate_agg_result::{IntermediateAggregationResults, IntermediateBucketResult};
use super::metric::{
AverageAggregation, SegmentAverageCollector, SegmentStatsCollector, StatsAggregation,
};
use super::VecWithNames;
use crate::aggregation::agg_req::BucketAggregationType;
use crate::DocId;
pub(crate) const DOC_BLOCK_SIZE: usize = 64;
pub(crate) type DocBlock = [DocId; DOC_BLOCK_SIZE];
#[derive(Clone, PartialEq)]
pub(crate) struct SegmentAggregationResultsCollector {
pub(crate) metrics: Option<VecWithNames<SegmentMetricResultCollector>>,
pub(crate) buckets: Option<VecWithNames<SegmentBucketResultCollector>>,
staged_docs: DocBlock,
num_staged_docs: usize,
}
impl Default for SegmentAggregationResultsCollector {
fn default() -> Self {
Self {
metrics: Default::default(),
buckets: Default::default(),
staged_docs: [0; DOC_BLOCK_SIZE],
num_staged_docs: Default::default(),
}
}
}
impl Debug for SegmentAggregationResultsCollector {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
f.debug_struct("SegmentAggregationResultsCollector")
.field("metrics", &self.metrics)
.field("buckets", &self.buckets)
.field("staged_docs", &&self.staged_docs[..self.num_staged_docs])
.field("num_staged_docs", &self.num_staged_docs)
.finish()
}
}
impl SegmentAggregationResultsCollector {
pub fn into_intermediate_aggregations_result(
self,
agg_with_accessor: &AggregationsWithAccessor,
) -> crate::Result<IntermediateAggregationResults> {
let buckets = if let Some(buckets) = self.buckets {
let entries = buckets
.into_iter()
.zip(agg_with_accessor.buckets.values())
.map(|((key, bucket), acc)| Ok((key, bucket.into_intermediate_bucket_result(acc)?)))
.collect::<crate::Result<Vec<(String, _)>>>()?;
Some(VecWithNames::from_entries(entries))
} else {
None
};
let metrics = self.metrics.map(VecWithNames::from_other);
Ok(IntermediateAggregationResults { metrics, buckets })
}
pub(crate) fn from_req_and_validate(req: &AggregationsWithAccessor) -> crate::Result<Self> {
let buckets = req
.buckets
.entries()
.map(|(key, req)| {
Ok((
key.to_string(),
SegmentBucketResultCollector::from_req_and_validate(req)?,
))
})
.collect::<crate::Result<Vec<(String, _)>>>()?;
let metrics = req
.metrics
.entries()
.map(|(key, req)| {
Ok((
key.to_string(),
SegmentMetricResultCollector::from_req_and_validate(req)?,
))
})
.collect::<crate::Result<Vec<(String, _)>>>()?;
let metrics = if metrics.is_empty() {
None
} else {
Some(VecWithNames::from_entries(metrics))
};
let buckets = if buckets.is_empty() {
None
} else {
Some(VecWithNames::from_entries(buckets))
};
Ok(SegmentAggregationResultsCollector {
metrics,
buckets,
staged_docs: [0; DOC_BLOCK_SIZE],
num_staged_docs: 0,
})
}
#[inline]
pub(crate) fn collect(
&mut self,
doc: crate::DocId,
agg_with_accessor: &AggregationsWithAccessor,
) {
self.staged_docs[self.num_staged_docs] = doc;
self.num_staged_docs += 1;
if self.num_staged_docs == self.staged_docs.len() {
self.flush_staged_docs(agg_with_accessor, false);
}
}
pub(crate) fn flush_staged_docs(
&mut self,
agg_with_accessor: &AggregationsWithAccessor,
force_flush: bool,
) {
if self.num_staged_docs == 0 {
return;
}
if let Some(metrics) = &mut self.metrics {
for (collector, agg_with_accessor) in
metrics.values_mut().zip(agg_with_accessor.metrics.values())
{
collector
.collect_block(&self.staged_docs[..self.num_staged_docs], agg_with_accessor);
}
}
if let Some(buckets) = &mut self.buckets {
for (collector, agg_with_accessor) in
buckets.values_mut().zip(agg_with_accessor.buckets.values())
{
collector.collect_block(
&self.staged_docs[..self.num_staged_docs],
agg_with_accessor,
force_flush,
);
}
}
self.num_staged_docs = 0;
}
}
#[derive(Clone, Debug, PartialEq)]
pub(crate) enum SegmentMetricResultCollector {
Average(SegmentAverageCollector),
Stats(SegmentStatsCollector),
}
impl SegmentMetricResultCollector {
pub fn from_req_and_validate(req: &MetricAggregationWithAccessor) -> crate::Result<Self> {
match &req.metric {
MetricAggregation::Average(AverageAggregation { field: _ }) => {
Ok(SegmentMetricResultCollector::Average(
SegmentAverageCollector::from_req(req.field_type),
))
}
MetricAggregation::Stats(StatsAggregation { field: _ }) => {
Ok(SegmentMetricResultCollector::Stats(
SegmentStatsCollector::from_req(req.field_type),
))
}
}
}
pub(crate) fn collect_block(&mut self, doc: &[DocId], metric: &MetricAggregationWithAccessor) {
match self {
SegmentMetricResultCollector::Average(avg_collector) => {
avg_collector.collect_block(doc, &metric.accessor);
}
SegmentMetricResultCollector::Stats(stats_collector) => {
stats_collector.collect_block(doc, &metric.accessor);
}
}
}
}
/// SegmentBucketAggregationResultCollectors will have specialized buckets for collection inside
/// segments.
/// The typical structure of Map<Key, Bucket> is not suitable during collection for performance
/// reasons.
#[derive(Clone, Debug, PartialEq)]
pub(crate) enum SegmentBucketResultCollector {
Range(SegmentRangeCollector),
Histogram(Box<SegmentHistogramCollector>),
Terms(Box<SegmentTermCollector>),
}
impl SegmentBucketResultCollector {
pub fn into_intermediate_bucket_result(
self,
agg_with_accessor: &BucketAggregationWithAccessor,
) -> crate::Result<IntermediateBucketResult> {
match self {
SegmentBucketResultCollector::Terms(terms) => {
terms.into_intermediate_bucket_result(agg_with_accessor)
}
SegmentBucketResultCollector::Range(range) => {
range.into_intermediate_bucket_result(agg_with_accessor)
}
SegmentBucketResultCollector::Histogram(histogram) => {
histogram.into_intermediate_bucket_result(agg_with_accessor)
}
}
}
pub fn from_req_and_validate(req: &BucketAggregationWithAccessor) -> crate::Result<Self> {
match &req.bucket_agg {
BucketAggregationType::Terms(terms_req) => Ok(Self::Terms(Box::new(
SegmentTermCollector::from_req_and_validate(
terms_req,
&req.sub_aggregation,
req.field_type,
req.accessor
.as_multi()
.expect("unexpected fast field cardinality"),
)?,
))),
BucketAggregationType::Range(range_req) => {
Ok(Self::Range(SegmentRangeCollector::from_req_and_validate(
range_req,
&req.sub_aggregation,
req.field_type,
)?))
}
BucketAggregationType::Histogram(histogram) => Ok(Self::Histogram(Box::new(
SegmentHistogramCollector::from_req_and_validate(
histogram,
&req.sub_aggregation,
req.field_type,
req.accessor
.as_single()
.expect("unexpected fast field cardinality"),
)?,
))),
}
}
#[inline]
pub(crate) fn collect_block(
&mut self,
doc: &[DocId],
bucket_with_accessor: &BucketAggregationWithAccessor,
force_flush: bool,
) {
match self {
SegmentBucketResultCollector::Range(range) => {
range.collect_block(doc, bucket_with_accessor, force_flush);
}
SegmentBucketResultCollector::Histogram(histogram) => {
histogram.collect_block(doc, bucket_with_accessor, force_flush)
}
SegmentBucketResultCollector::Terms(terms) => {
terms.collect_block(doc, bucket_with_accessor, force_flush)
}
}
}
}

16
src/collector/count_collector.rs
View File

@@ -1,9 +1,6 @@
use super::Collector;
use crate::collector::SegmentCollector;
use crate::DocId;
use crate::Score;
use crate::SegmentOrdinal;
use crate::SegmentReader;
use crate::{DocId, Score, SegmentOrdinal, SegmentReader};
/// `CountCollector` collector only counts how many
/// documents match the query.
@@ -20,10 +17,10 @@ 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();
@@ -80,8 +77,7 @@ impl SegmentCollector for SegmentCountCollector {
#[cfg(test)]
mod tests {
use super::{Count, SegmentCountCollector};
use crate::collector::Collector;
use crate::collector::SegmentCollector;
use crate::collector::{Collector, SegmentCollector};
#[test]
fn test_count_collect_does_not_requires_scoring() {

6
src/collector/custom_score_top_collector.rs
View File

@@ -8,8 +8,7 @@ pub(crate) struct CustomScoreTopCollector<TCustomScorer, TScore = Score> {
}
impl<TCustomScorer, TScore> CustomScoreTopCollector<TCustomScorer, TScore>
where
TScore: Clone + PartialOrd,
where TScore: Clone + PartialOrd
{
pub(crate) fn new(
custom_scorer: TCustomScorer,
@@ -114,8 +113,7 @@ where
}
impl<F, TScore> CustomSegmentScorer<TScore> for F
where
F: 'static + FnMut(DocId) -> TScore,
where F: 'static + FnMut(DocId) -> TScore
{
fn score(&mut self, doc: DocId) -> TScore {
(self)(doc)

3
src/collector/docset_collector.rs
View File

@@ -1,8 +1,7 @@
use std::collections::HashSet;
use crate::{DocAddress, DocId, Score};
use super::{Collector, SegmentCollector};
use crate::{DocAddress, DocId, Score};
/// Collectors that returns the set of DocAddress that matches the query.
///

121
src/collector/facet_collector.rs
View File

@@ -1,21 +1,14 @@
use crate::collector::Collector;
use crate::collector::SegmentCollector;
use crate::fastfield::FacetReader;
use crate::schema::Facet;
use crate::schema::Field;
use crate::DocId;
use crate::Score;
use crate::SegmentOrdinal;
use crate::SegmentReader;
use std::cmp::Ordering;
use std::collections::btree_map;
use std::collections::BTreeMap;
use std::collections::BTreeSet;
use std::collections::BinaryHeap;
use std::collections::{btree_map, BTreeMap, BTreeSet, BinaryHeap};
use std::iter::Peekable;
use std::ops::Bound;
use std::{u64, usize};
use crate::collector::{Collector, SegmentCollector};
use crate::fastfield::FacetReader;
use crate::schema::{Facet, Field};
use crate::{DocId, Score, SegmentOrdinal, SegmentReader};
struct Hit<'a> {
count: u64,
facet: &'a Facet,
@@ -83,7 +76,7 @@ fn facet_depth(facet_bytes: &[u8]) -> usize {
/// ```rust
/// use tantivy::collector::FacetCollector;
/// use tantivy::query::AllQuery;
/// use tantivy::schema::{Facet, Schema, INDEXED, TEXT};
/// use tantivy::schema::{Facet, Schema, FacetOptions, TEXT};
/// use tantivy::{doc, Index};
///
/// fn example() -> tantivy::Result<()> {
@@ -92,7 +85,7 @@ fn facet_depth(facet_bytes: &[u8]) -> usize {
/// // Facet have their own specific type.
/// // It is not a bad practise to put all of your
/// // facet information in the same field.
/// let facet = schema_builder.add_facet_field("facet", INDEXED);
/// let facet = schema_builder.add_facet_field("facet", FacetOptions::default());
/// let title = schema_builder.add_text_field("title", TEXT);
/// let schema = schema_builder.build();
/// let index = Index::create_in_ram(schema);
@@ -103,23 +96,23 @@ 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()?;
@@ -240,9 +233,7 @@ impl FacetCollector {
/// If you need the correct number of unique documents for two such facets,
/// just add them in separate `FacetCollector`.
pub fn add_facet<T>(&mut self, facet_from: T)
where
Facet: From<T>,
{
where Facet: From<T> {
let facet = Facet::from(facet_from);
for old_facet in &self.facets {
assert!(
@@ -400,11 +391,9 @@ impl<'a> Iterator for FacetChildIterator<'a> {
impl FacetCounts {
/// Returns an iterator over all of the facet count pairs inside this result.
/// See the documentation for `FacetCollector` for a usage example.
/// See the documentation for [FacetCollector] for a usage example.
pub fn get<T>(&self, facet_from: T) -> FacetChildIterator<'_>
where
Facet: From<T>,
{
where Facet: From<T> {
let facet = Facet::from(facet_from);
let left_bound = Bound::Excluded(facet.clone());
let right_bound = if facet.is_root() {
@@ -421,11 +410,9 @@ impl FacetCounts {
}
/// Returns a vector of top `k` facets with their counts, sorted highest-to-lowest by counts.
/// See the documentation for `FacetCollector` for a usage example.
/// See the documentation for [FacetCollector] for a usage example.
pub fn top_k<T>(&self, facet: T, k: usize) -> Vec<(&Facet, u64)>
where
Facet: From<T>,
{
where Facet: From<T> {
let mut heap = BinaryHeap::with_capacity(k);
let mut it = self.get(facet);
@@ -458,25 +445,27 @@ impl FacetCounts {
#[cfg(test)]
mod tests {
use std::iter;
use rand::distributions::Uniform;
use rand::prelude::SliceRandom;
use rand::{thread_rng, Rng};
use super::{FacetCollector, FacetCounts};
use crate::collector::Count;
use crate::core::Index;
use crate::query::{AllQuery, QueryParser, TermQuery};
use crate::schema::{Document, Facet, Field, IndexRecordOption, Schema, INDEXED};
use crate::schema::{Document, Facet, FacetOptions, Field, IndexRecordOption, Schema};
use crate::Term;
use rand::distributions::Uniform;
use rand::prelude::SliceRandom;
use rand::{thread_rng, Rng};
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 facet_field = schema_builder.add_facet_field("facet", FacetOptions::default());
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 +480,14 @@ 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
@@ -518,11 +507,13 @@ mod tests {
.collect::<Vec<_>>()
);
}
Ok(())
}
#[test]
#[should_panic(expected = "Tried to add a facet which is a descendant of \
an already added facet.")]
#[should_panic(
expected = "Tried to add a facet which is a descendant of an already added facet."
)]
fn test_misused_facet_collector() {
let mut facet_collector = FacetCollector::for_field(Field::from_field_id(0));
facet_collector.add_facet(Facet::from("/country"));
@@ -530,48 +521,49 @@ 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 facet_field = schema_builder.add_facet_field("facets", FacetOptions::default());
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]
fn test_doc_search_by_facet() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let facet_field = schema_builder.add_facet_field("facet", INDEXED);
let facet_field = schema_builder.add_facet_field("facet", FacetOptions::default());
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/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();
@@ -613,7 +605,7 @@ mod tests {
#[test]
fn test_facet_collector_topk() {
let mut schema_builder = Schema::builder();
let facet_field = schema_builder.add_facet_field("facet", INDEXED);
let facet_field = schema_builder.add_facet_field("facet", FacetOptions::default());
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
@@ -637,7 +629,7 @@ 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();
@@ -662,7 +654,7 @@ mod tests {
#[test]
fn test_facet_collector_topk_tie_break() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let facet_field = schema_builder.add_facet_field("facet", INDEXED);
let facet_field = schema_builder.add_facet_field("facet", FacetOptions::default());
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
@@ -677,7 +669,7 @@ 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()?;
@@ -698,13 +690,14 @@ mod tests {
#[cfg(all(test, feature = "unstable"))]
mod bench {
use rand::seq::SliceRandom;
use rand::thread_rng;
use test::Bencher;
use crate::collector::FacetCollector;
use crate::query::AllQuery;
use crate::schema::{Facet, Schema, INDEXED};
use crate::Index;
use rand::seq::SliceRandom;
use rand::thread_rng;
use test::Bencher;
#[bench]
fn bench_facet_collector(b: &mut Bencher) {
@@ -725,7 +718,7 @@ mod bench {
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 reader = index.reader().unwrap();

31
src/collector/filter_collector_wrapper.rs
View File

@@ -16,8 +16,9 @@ use crate::fastfield::{DynamicFastFieldReader, FastFieldReader, FastValue};
use crate::schema::Field;
use crate::{Score, SegmentReader, TantivyError};
/// The `FilterCollector` collector filters docs using a fast field value and a predicate.
/// Only the documents for which the predicate returned "true" will be passed on to the next collector.
/// The `FilterCollector` filters docs using a fast field value and a predicate.
/// Only the documents for which the predicate returned "true" will be passed on to the next
/// collector.
///
/// ```rust
/// use tantivy::collector::{TopDocs, FilterCollector};
@@ -25,38 +26,40 @@ 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
TPredicate: 'static + Clone,
where TPredicate: 'static + Clone
{
field: Field,
collector: TCollector,

46
src/collector/histogram_collector.rs
View File

@@ -1,8 +1,9 @@
use fastdivide::DividerU64;
use crate::collector::{Collector, SegmentCollector};
use crate::fastfield::{DynamicFastFieldReader, FastFieldReader, FastValue};
use crate::schema::{Field, Type};
use crate::{DocId, Score};
use fastdivide::DividerU64;
/// Histogram builds an histogram of the values of a fastfield for the
/// collected DocSet.
@@ -18,7 +19,7 @@ use fastdivide::DividerU64;
///
/// # Warning
///
/// f64 field. are not supported.
/// f64 fields are not supported.
#[derive(Clone)]
pub struct HistogramCollector {
min_value: u64,
@@ -36,8 +37,8 @@ impl HistogramCollector {
/// - `bucket_width`: the length of the interval that is associated to each buckets.
/// - `num_buckets`: The overall number of buckets.
///
/// Together, this parameters define a partition of `[min_value, min_value + num_buckets * bucket_width)`
/// into `num_buckets` intervals of width bucket that we call `bucket`.
/// Together, this parameters define a partition of `[min_value, min_value + num_buckets *
/// bucket_width)` into `num_buckets` intervals of width bucket that we call `bucket`.
///
/// # Disclaimer
/// This function panics if the field given is of type f64.
@@ -147,13 +148,14 @@ fn add_vecs(mut vals_list: Vec<Vec<u64>>, len: usize) -> Vec<u64> {
#[cfg(test)]
mod tests {
use super::{add_vecs, HistogramCollector, HistogramComputer};
use crate::chrono::{TimeZone, Utc};
use crate::schema::{Schema, FAST};
use crate::{doc, query, Index};
use fastdivide::DividerU64;
use query::AllQuery;
use super::{add_vecs, HistogramCollector, HistogramComputer};
use crate::schema::{Schema, FAST};
use crate::time::{Date, Month};
use crate::{doc, query, DateTime, Index};
#[test]
fn test_add_histograms_simple() {
assert_eq!(
@@ -226,10 +228,10 @@ 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 +249,13 @@ 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();
@@ -271,16 +273,20 @@ 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=>DateTime::from_primitive(Date::from_calendar_date(1982, Month::September, 17)?.with_hms(0, 0, 0)?)))?;
writer.add_document(
doc!(date_field=>DateTime::from_primitive(Date::from_calendar_date(1986, Month::March, 9)?.with_hms(0, 0, 0)?)),
)?;
writer.add_document(doc!(date_field=>DateTime::from_primitive(Date::from_calendar_date(1983, Month::September, 27)?.with_hms(0, 0, 0)?)))?;
writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let all_query = AllQuery;
let week_histogram_collector = HistogramCollector::new(
date_field,
Utc.ymd(1980, 1, 1).and_hms(0, 0, 0),
DateTime::from_primitive(
Date::from_calendar_date(1980, Month::January, 1)?.with_hms(0, 0, 0)?,
),
3600 * 24 * 365, // it is just for a unit test... sorry leap years.
10,
);

182
src/collector/mod.rs
View File

@@ -1,95 +1,90 @@
/*!
//! # Collectors
//!
//! Collectors define the information you want to extract from the documents matching the queries.
//! In tantivy jargon, we call this information your search "fruit".
//!
//! Your fruit could for instance be :
//! - [the count of matching documents](./struct.Count.html)
//! - [the top 10 documents, by relevancy or by a fast field](./struct.TopDocs.html)
//! - [facet counts](./struct.FacetCollector.html)
//!
//! At one point in your code, you will trigger the actual search operation by calling
//! [the `search(...)` method of your `Searcher` object](../struct.Searcher.html#method.search).
//! This call will look like this.
//!
//! ```verbatim
//! let fruit = searcher.search(&query, &collector)?;
//! ```
//!
//! Here the type of fruit is actually determined as an associated type of the collector
//! (`Collector::Fruit`).
//!
//!
//! # Combining several collectors
//!
//! A rich search experience often requires to run several collectors on your search query.
//! For instance,
//! - selecting the top-K products matching your query
//! - counting the matching documents
//! - computing several facets
//! - computing statistics about the matching product prices
//!
//! A simple and efficient way to do that is to pass your collectors as one tuple.
//! The resulting `Fruit` will then be a typed tuple with each collector's original fruits
//! in their respective position.
//!
//! ```rust
//! # use tantivy::schema::*;
//! # use tantivy::*;
//! # use tantivy::query::*;
//! use tantivy::collector::{Count, TopDocs};
//! #
//! # fn main() -> tantivy::Result<()> {
//! # let mut schema_builder = Schema::builder();
//! # let title = schema_builder.add_text_field("title", TEXT);
//! # let schema = schema_builder.build();
//! # let index = Index::create_in_ram(schema);
//! # let mut index_writer = index.writer(3_000_000)?;
//! # index_writer.add_document(doc!(
//! # title => "The Name of the Wind",
//! # ))?;
//! # index_writer.add_document(doc!(
//! # title => "The Diary of Muadib",
//! # ))?;
//! # index_writer.commit()?;
//! # let reader = index.reader()?;
//! # let searcher = reader.searcher();
//! # let query_parser = QueryParser::for_index(&index, vec![title]);
//! # let query = query_parser.parse_query("diary")?;
//! let (doc_count, top_docs): (usize, Vec<(Score, DocAddress)>) =
//! searcher.search(&query, &(Count, TopDocs::with_limit(2)))?;
//! # Ok(())
//! # }
//! ```
//!
//! The `Collector` trait is implemented for up to 4 collectors.
//! If you have more than 4 collectors, you can either group them into
//! tuples of tuples `(a,(b,(c,d)))`, or rely on [`MultiCollector`](./struct.MultiCollector.html).
//!
//! # Combining several collectors dynamically
//!
//! Combining collectors into a tuple is a zero-cost abstraction: everything
//! happens as if you had manually implemented a single collector
//! combining all of our features.
//!
//! Unfortunately it requires you to know at compile time your collector types.
//! If on the other hand, the collectors depend on some query parameter,
//! you can rely on `MultiCollector`'s.
//!
//!
//! # Implementing your own collectors.
//!
//! See the `custom_collector` example.
# Collectors
Collectors define the information you want to extract from the documents matching the queries.
In tantivy jargon, we call this information your search "fruit".
Your fruit could for instance be :
- [the count of matching documents](./struct.Count.html)
- [the top 10 documents, by relevancy or by a fast field](./struct.TopDocs.html)
- [facet counts](./struct.FacetCollector.html)
At one point in your code, you will trigger the actual search operation by calling
[the `search(...)` method of your `Searcher` object](../struct.Searcher.html#method.search).
This call will look like this.
```verbatim
let fruit = searcher.search(&query, &collector)?;
```
Here the type of fruit is actually determined as an associated type of the collector (`Collector::Fruit`).
# Combining several collectors
A rich search experience often requires to run several collectors on your search query.
For instance,
- selecting the top-K products matching your query
- counting the matching documents
- computing several facets
- computing statistics about the matching product prices
A simple and efficient way to do that is to pass your collectors as one tuple.
The resulting `Fruit` will then be a typed tuple with each collector's original fruits
in their respective position.
```rust
# use tantivy::schema::*;
# use tantivy::*;
# use tantivy::query::*;
use tantivy::collector::{Count, TopDocs};
#
# fn main() -> tantivy::Result<()> {
# let mut schema_builder = Schema::builder();
# let title = schema_builder.add_text_field("title", TEXT);
# let schema = schema_builder.build();
# let index = Index::create_in_ram(schema);
# let mut index_writer = index.writer(3_000_000)?;
# index_writer.add_document(doc!(
# title => "The Name of the Wind",
# ));
# index_writer.add_document(doc!(
# title => "The Diary of Muadib",
# ));
# index_writer.commit()?;
# let reader = index.reader()?;
# let searcher = reader.searcher();
# let query_parser = QueryParser::for_index(&index, vec![title]);
# let query = query_parser.parse_query("diary")?;
let (doc_count, top_docs): (usize, Vec<(Score, DocAddress)>) =
searcher.search(&query, &(Count, TopDocs::with_limit(2)))?;
# Ok(())
# }
```
The `Collector` trait is implemented for up to 4 collectors.
If you have more than 4 collectors, you can either group them into
tuples of tuples `(a,(b,(c,d)))`, or rely on [`MultiCollector`](./struct.MultiCollector.html).
# Combining several collectors dynamically
Combining collectors into a tuple is a zero-cost abstraction: everything
happens as if you had manually implemented a single collector
combining all of our features.
Unfortunately it requires you to know at compile time your collector types.
If on the other hand, the collectors depend on some query parameter,
you can rely on `MultiCollector`'s.
# Implementing your own collectors.
See the `custom_collector` example.
*/
use crate::DocId;
use crate::Score;
use crate::SegmentOrdinal;
use crate::SegmentReader;
use downcast_rs::impl_downcast;
use crate::{DocId, Score, SegmentOrdinal, SegmentReader};
mod count_collector;
pub use self::count_collector::Count;
@@ -97,7 +92,7 @@ mod histogram_collector;
pub use histogram_collector::HistogramCollector;
mod multi_collector;
pub use self::multi_collector::MultiCollector;
pub use self::multi_collector::{FruitHandle, MultiCollector, MultiFruit};
mod top_collector;
@@ -111,8 +106,7 @@ mod tweak_score_top_collector;
pub use self::tweak_score_top_collector::{ScoreSegmentTweaker, ScoreTweaker};
mod facet_collector;
pub use self::facet_collector::FacetCollector;
pub use self::facet_collector::FacetCounts;
pub use self::facet_collector::{FacetCollector, FacetCounts};
use crate::query::Weight;
mod docset_collector;
@@ -178,9 +172,9 @@ 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);
}
})?;

68
src/collector/multi_collector.rs
View File

@@ -1,14 +1,11 @@
use super::Collector;
use super::SegmentCollector;
use crate::collector::Fruit;
use crate::DocId;
use crate::Score;
use crate::SegmentOrdinal;
use crate::SegmentReader;
use crate::TantivyError;
use std::marker::PhantomData;
use std::ops::Deref;
use super::{Collector, SegmentCollector};
use crate::collector::Fruit;
use crate::{DocId, Score, SegmentOrdinal, SegmentReader, TantivyError};
/// MultiFruit keeps Fruits from every nested Collector
pub struct MultiFruit {
sub_fruits: Vec<Option<Box<dyn Fruit>>>,
}
@@ -83,12 +80,17 @@ impl<TSegmentCollector: SegmentCollector> BoxableSegmentCollector
}
}
/// FruitHandle stores reference to the corresponding collector inside MultiCollector
pub struct FruitHandle<TFruit: Fruit> {
pos: usize,
_phantom: PhantomData<TFruit>,
}
impl<TFruit: Fruit> FruitHandle<TFruit> {
/// Extract a typed fruit off a multifruit.
///
/// This function involves downcasting and can panic if the multifruit was
/// created using faulty code.
pub fn extract(self, fruits: &mut MultiFruit) -> TFruit {
let boxed_fruit = fruits.sub_fruits[self.pos].take().expect("");
*boxed_fruit
@@ -104,7 +106,8 @@ impl<TFruit: Fruit> FruitHandle<TFruit> {
///
/// If the type of the collectors is known, you can just group yours collectors
/// in a tuple. See the
/// [Combining several collectors section of the collector documentation](./index.html#combining-several-collectors).
/// [Combining several collectors section of the collector
/// documentation](./index.html#combining-several-collectors).
///
/// ```rust
/// use tantivy::collector::{Count, TopDocs, MultiCollector};
@@ -112,19 +115,19 @@ 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 +142,8 @@ impl<TFruit: Fruit> FruitHandle<TFruit> {
///
/// assert_eq!(count, 2);
/// assert_eq!(top_docs.len(), 2);
/// # Ok(())
/// # }
/// ```
#[allow(clippy::type_complexity)]
#[derive(Default)]
@@ -246,30 +251,28 @@ mod tests {
use super::*;
use crate::collector::{Count, TopDocs};
use crate::query::TermQuery;
use crate::schema::IndexRecordOption;
use crate::schema::{Schema, TEXT};
use crate::Index;
use crate::Term;
use crate::schema::{IndexRecordOption, Schema, TEXT};
use crate::{Index, 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 +283,6 @@ mod tests {
assert_eq!(count_handler.extract(&mut multifruits), 5);
assert_eq!(topdocs_handler.extract(&mut multifruits).len(), 2);
Ok(())
}
}

53
src/collector/tests.rs
View File

@@ -1,20 +1,12 @@
use super::*;
use crate::core::SegmentReader;
use crate::fastfield::BytesFastFieldReader;
use crate::fastfield::DynamicFastFieldReader;
use crate::fastfield::FastFieldReader;
use crate::schema::Field;
use crate::DocId;
use crate::Score;
use crate::SegmentOrdinal;
use crate::{DocAddress, Document, Searcher};
use crate::collector::{Count, FilterCollector, TopDocs};
use crate::core::SegmentReader;
use crate::fastfield::{BytesFastFieldReader, DynamicFastFieldReader, FastFieldReader};
use crate::query::{AllQuery, QueryParser};
use crate::schema::{Schema, FAST, TEXT};
use crate::DateTime;
use crate::{doc, Index};
use std::str::FromStr;
use crate::schema::{Field, Schema, FAST, TEXT};
use crate::time::format_description::well_known::Rfc3339;
use crate::time::OffsetDateTime;
use crate::{doc, DateTime, DocAddress, DocId, Document, Index, Score, Searcher, SegmentOrdinal};
pub const TEST_COLLECTOR_WITH_SCORE: TestCollector = TestCollector {
compute_score: true,
@@ -25,7 +17,7 @@ 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 +25,25 @@ 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_utc(OffsetDateTime::parse("1898-04-09T00:00:00+00:00", &Rfc3339).unwrap())))?;
index_writer.add_document(doc!(title => "The Diary of Muadib", price => 29_240u64, date => DateTime::from_utc(OffsetDateTime::parse("2020-04-09T00:00:00+00:00", &Rfc3339).unwrap())))?;
index_writer.add_document(doc!(title => "The Diary of Anne Frank", price => 18_240u64, date => DateTime::from_utc(OffsetDateTime::parse("2019-04-20T00:00:00+00:00", &Rfc3339).unwrap())))?;
index_writer.add_document(doc!(title => "A Dairy Cow", price => 21_240u64, date => DateTime::from_utc(OffsetDateTime::parse("2019-04-09T00:00:00+00:00", &Rfc3339).unwrap())))?;
index_writer.add_document(doc!(title => "The Diary of a Young Girl", price => 20_120u64, date => DateTime::from_utc(OffsetDateTime::parse("2018-04-09T00:00:00+00:00", &Rfc3339).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));
@@ -63,13 +55,16 @@ pub fn test_filter_collector() {
assert_eq!(filtered_top_docs.len(), 0);
fn date_filter(value: DateTime) -> bool {
(value - DateTime::from_str("2019-04-09T00:00:00+00:00").unwrap()).num_weeks() > 0
(value.into_utc() - OffsetDateTime::parse("2019-04-09T00:00:00+00:00", &Rfc3339).unwrap())
.whole_weeks()
> 0
}
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 +269,8 @@ 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())
}

21
src/collector/top_collector.rs
View File

@@ -1,11 +1,9 @@
use crate::DocAddress;
use crate::DocId;
use crate::SegmentOrdinal;
use crate::SegmentReader;
use std::cmp::Ordering;
use std::collections::BinaryHeap;
use std::marker::PhantomData;
use crate::{DocAddress, DocId, SegmentOrdinal, SegmentReader};
/// Contains a feature (field, score, etc.) of a document along with the document address.
///
/// It has a custom implementation of `PartialOrd` that reverses the order. This is because the
@@ -62,17 +60,14 @@ pub(crate) struct TopCollector<T> {
}
impl<T> TopCollector<T>
where
T: PartialOrd + Clone,
where T: PartialOrd + Clone
{
/// Creates a top collector, with a number of documents equal to "limit".
///
/// # Panics
/// The method panics if limit is 0
pub fn with_limit(limit: usize) -> TopCollector<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,
@@ -178,8 +173,7 @@ impl<T: PartialOrd + Clone> TopSegmentCollector<T> {
.collect()
}
/// Return true iff at least K documents have gone through
/// the collector.
/// Return true if more documents have been collected than the limit.
#[inline]
pub(crate) fn at_capacity(&self) -> bool {
self.heap.len() >= self.limit
@@ -255,7 +249,7 @@ mod tests {
// when harvesting we have to guarantee stable sorting in case of a tie
// on the score
let doc_ids_collection = [4, 5, 6];
let score = 3.14;
let score = 3.3f32;
let mut top_collector_limit_2 = TopSegmentCollector::new(0, 2);
for id in &doc_ids_collection {
@@ -324,9 +318,10 @@ mod tests {
#[cfg(all(test, feature = "unstable"))]
mod bench {
use super::TopSegmentCollector;
use test::Bencher;
use super::TopSegmentCollector;
#[bench]
fn bench_top_segment_collector_collect_not_at_capacity(b: &mut Bencher) {
let mut top_collector = TopSegmentCollector::new(0, 400);

259
src/collector/top_score_collector.rs
View File

@@ -1,21 +1,18 @@
use std::collections::BinaryHeap;
use std::fmt;
use std::marker::PhantomData;
use super::Collector;
use crate::collector::top_collector::{ComparableDoc, TopCollector};
use crate::collector::custom_score_top_collector::CustomScoreTopCollector;
use crate::collector::top_collector::{ComparableDoc, TopCollector, TopSegmentCollector};
use crate::collector::tweak_score_top_collector::TweakedScoreTopCollector;
use crate::collector::{
CustomScorer, CustomSegmentScorer, ScoreSegmentTweaker, ScoreTweaker, SegmentCollector,
};
use crate::fastfield::{DynamicFastFieldReader, FastFieldReader};
use crate::fastfield::{DynamicFastFieldReader, FastFieldReader, FastValue};
use crate::query::Weight;
use crate::schema::Field;
use crate::DocAddress;
use crate::DocId;
use crate::Score;
use crate::SegmentOrdinal;
use crate::SegmentReader;
use crate::{collector::custom_score_top_collector::CustomScoreTopCollector, fastfield::FastValue};
use crate::{collector::top_collector::TopSegmentCollector, TantivyError};
use std::fmt;
use std::{collections::BinaryHeap, marker::PhantomData};
use crate::{DocAddress, DocId, Score, SegmentOrdinal, SegmentReader, TantivyError};
struct FastFieldConvertCollector<
TCollector: Collector<Fruit = Vec<(u64, DocAddress)>>,
@@ -94,27 +91,30 @@ 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>);
@@ -180,41 +180,46 @@ 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(())
/// # }
/// ```
#[must_use]
pub fn and_offset(self, offset: usize) -> TopDocs {
TopDocs(self.0.and_offset(offset))
}
/// Set top-K to rank documents by a given fast field.
///
/// If the field is not a fast or does not exist, this method returns successfully (it is not aware of any schema).
/// An error will be returned at the moment of search.
/// If the field is not a fast or does not exist, this method returns successfully (it is not
/// aware of any schema). An error will be returned at the moment of search.
///
/// If the field is a FAST field but not a u64 field, search will return successfully but it will return
/// returns a monotonic u64-representation (ie. the order is still correct) of the requested field type.
/// If the field is a FAST field but not a u64 field, search will return successfully but it
/// will return returns a monotonic u64-representation (ie. the order is still correct) of
/// the requested field type.
///
/// # Example
///
@@ -234,11 +239,11 @@ 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)?;
@@ -289,14 +294,15 @@ impl TopDocs {
/// Set top-K to rank documents by a given fast field.
///
/// If the field is not a fast field, or its field type does not match the generic type, this method does not panic,
/// but an explicit error will be returned at the moment of collection.
/// If the field is not a fast field, or its field type does not match the generic type, this
/// method does not panic, but an explicit error will be returned at the moment of
/// collection.
///
/// Note that this method is a generic. The requested fast field type will be often
/// inferred in your code by the rust compiler.
///
/// Implementation-wise, for performance reason, tantivy will manipulate the u64 representation of your fast
/// field until the last moment.
/// Implementation-wise, for performance reason, tantivy will manipulate the u64 representation
/// of your fast field until the last moment.
///
/// # Example
///
@@ -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 {
@@ -708,25 +714,22 @@ mod tests {
use crate::collector::Collector;
use crate::query::{AllQuery, Query, QueryParser};
use crate::schema::{Field, Schema, FAST, STORED, TEXT};
use crate::Index;
use crate::IndexWriter;
use crate::Score;
use crate::{DocAddress, DocId, SegmentReader};
use crate::time::format_description::well_known::Rfc3339;
use crate::time::OffsetDateTime;
use crate::{DateTime, DocAddress, DocId, Index, IndexWriter, Score, 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 +740,15 @@ 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 +757,12 @@ 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 +777,7 @@ 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 +798,7 @@ 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 +819,7 @@ 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 +850,35 @@ 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[..],
&[
@@ -879,32 +887,37 @@ mod tests {
(12, DocAddress::new(0, 0))
]
);
Ok(())
}
#[test]
fn test_top_field_collector_datetime() -> crate::Result<()> {
use std::str::FromStr;
let mut schema_builder = Schema::builder();
let name = schema_builder.add_text_field("name", TEXT);
let birthday = schema_builder.add_date_field("birthday", FAST);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_for_tests()?;
let pr_birthday = crate::DateTime::from_str("1898-04-09T00:00:00+00:00")?;
let pr_birthday = DateTime::from_utc(OffsetDateTime::parse(
"1898-04-09T00:00:00+00:00",
&Rfc3339,
)?);
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")?;
birthday => pr_birthday,
))?;
let mr_birthday = DateTime::from_utc(OffsetDateTime::parse(
"1947-11-08T00:00:00+00:00",
&Rfc3339,
)?);
index_writer.add_document(doc!(
name => "Minnie Riperton",
birthday => mr_birthday
));
birthday => mr_birthday,
))?;
index_writer.commit()?;
let searcher = index.reader()?.searcher();
let top_collector = TopDocs::with_limit(3).order_by_fast_field(birthday);
let top_docs: Vec<(crate::DateTime, DocAddress)> =
searcher.search(&AllQuery, &top_collector)?;
let top_docs: Vec<(DateTime, DocAddress)> = searcher.search(&AllQuery, &top_collector)?;
assert_eq!(
&top_docs[..],
&[
@@ -926,11 +939,11 @@ 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 +969,11 @@ 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 +996,12 @@ 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 +1018,7 @@ 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 +1035,7 @@ 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 +1048,26 @@ 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();

9
src/collector/tweak_score_top_collector.rs
View File

@@ -1,7 +1,6 @@
use crate::collector::top_collector::{TopCollector, TopSegmentCollector};
use crate::collector::{Collector, SegmentCollector};
use crate::DocAddress;
use crate::{DocId, Result, Score, SegmentReader};
use crate::{DocAddress, DocId, Result, Score, SegmentReader};
pub(crate) struct TweakedScoreTopCollector<TScoreTweaker, TScore = Score> {
score_tweaker: TScoreTweaker,
@@ -9,8 +8,7 @@ pub(crate) struct TweakedScoreTopCollector<TScoreTweaker, TScore = Score> {
}
impl<TScoreTweaker, TScore> TweakedScoreTopCollector<TScoreTweaker, TScore>
where
TScore: Clone + PartialOrd,
where TScore: Clone + PartialOrd
{
pub fn new(
score_tweaker: TScoreTweaker,
@@ -118,8 +116,7 @@ where
}
impl<F, TScore> ScoreSegmentTweaker<TScore> for F
where
F: 'static + FnMut(DocId, Score) -> TScore,
where F: 'static + FnMut(DocId, Score) -> TScore
{
fn score(&mut self, doc: DocId, score: Score) -> TScore {
(self)(doc, score)

396
src/common/bitset.rs
View File

@@ -1,396 +0,0 @@
use std::fmt;
use std::u64;
#[derive(Clone, Copy, Eq, PartialEq)]
pub(crate) struct TinySet(u64);
impl fmt::Debug for TinySet {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
self.into_iter().collect::<Vec<u32>>().fmt(f)
}
}
pub struct TinySetIterator(TinySet);
impl Iterator for TinySetIterator {
type Item = u32;
fn next(&mut self) -> Option<Self::Item> {
self.0.pop_lowest()
}
}
impl IntoIterator for TinySet {
type Item = u32;
type IntoIter = TinySetIterator;
fn into_iter(self) -> Self::IntoIter {
TinySetIterator(self)
}
}
impl TinySet {
/// Returns an empty `TinySet`.
pub fn empty() -> TinySet {
TinySet(0u64)
}
pub fn clear(&mut self) {
self.0 = 0u64;
}
/// Returns the complement of the set in `[0, 64[`.
fn complement(self) -> TinySet {
TinySet(!self.0)
}
/// Returns true iff the `TinySet` contains the element `el`.
pub fn contains(self, el: u32) -> bool {
!self.intersect(TinySet::singleton(el)).is_empty()
}
/// Returns the number of elements in the TinySet.
pub fn len(self) -> u32 {
self.0.count_ones()
}
/// Returns the intersection of `self` and `other`
pub fn intersect(self, other: TinySet) -> TinySet {
TinySet(self.0 & other.0)
}
/// Creates a new `TinySet` containing only one element
/// within `[0; 64[`
#[inline]
pub fn singleton(el: u32) -> TinySet {
TinySet(1u64 << u64::from(el))
}
/// 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[
#[inline]
pub fn insert_mut(&mut self, el: u32) -> bool {
let old = *self;
*self = old.insert(el);
old != *self
}
/// Returns the union of two tinysets
#[inline]
pub fn union(self, other: TinySet) -> TinySet {
TinySet(self.0 | other.0)
}
/// Returns true iff the `TinySet` is empty.
#[inline]
pub fn is_empty(self) -> bool {
self.0 == 0u64
}
/// Returns the lowest element in the `TinySet`
/// and removes it.
#[inline]
pub fn pop_lowest(&mut self) -> Option<u32> {
if self.is_empty() {
None
} else {
let lowest = self.0.trailing_zeros() as u32;
self.0 ^= TinySet::singleton(lowest).0;
Some(lowest)
}
}
/// Returns a `TinySet` than contains all values up
/// to limit excluded.
///
/// The limit is assumed to be strictly lower than 64.
pub fn range_lower(upper_bound: u32) -> TinySet {
TinySet((1u64 << u64::from(upper_bound % 64u32)) - 1u64)
}
/// Returns a `TinySet` that contains all values greater
/// or equal to the given limit, included. (and up to 63)
///
/// The limit is assumed to be strictly lower than 64.
pub fn range_greater_or_equal(from_included: u32) -> TinySet {
TinySet::range_lower(from_included).complement()
}
}
#[derive(Clone)]
pub struct BitSet {
tinysets: Box<[TinySet]>,
len: usize,
max_value: u32,
}
fn num_buckets(max_val: u32) -> u32 {
(max_val + 63u32) / 64u32
}
impl BitSet {
/// Create a new `BitSet` that may contain elements
/// 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();
BitSet {
tinysets: tinybisets,
len: 0,
max_value,
}
}
/// Removes all elements from the `BitSet`.
pub fn clear(&mut self) {
for tinyset in self.tinysets.iter_mut() {
*tinyset = TinySet::empty();
}
}
/// Returns the number of elements in the `BitSet`.
pub fn len(&self) -> usize {
self.len
}
/// Inserts an element in the `BitSet`
pub fn insert(&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].insert_mut(lower) {
1
} else {
0
};
}
/// Returns true iff the elements is in the `BitSet`.
pub fn contains(&self, el: u32) -> bool {
self.tinyset(el / 64u32).contains(el % 64)
}
/// Returns the first non-empty `TinySet` associated to a bucket lower
/// or greater than bucket.
///
/// Reminder: the tiny set with the bucket `bucket`, represents the
/// elements from `bucket * 64` to `(bucket+1) * 64`.
pub(crate) fn first_non_empty_bucket(&self, bucket: u32) -> Option<u32> {
self.tinysets[bucket as usize..]
.iter()
.cloned()
.position(|tinyset| !tinyset.is_empty())
.map(|delta_bucket| bucket + delta_bucket as u32)
}
pub fn max_value(&self) -> u32 {
self.max_value
}
/// Returns the tiny bitset representing the
/// the set restricted to the number range from
/// `bucket * 64` to `(bucket + 1) * 64`.
pub(crate) fn tinyset(&self, bucket: u32) -> TinySet {
self.tinysets[bucket as usize]
}
}
#[cfg(test)]
mod tests {
use super::BitSet;
use super::TinySet;
use crate::docset::{DocSet, TERMINATED};
use crate::query::BitSetDocSet;
use crate::tests;
use crate::tests::generate_nonunique_unsorted;
use std::collections::BTreeSet;
use std::collections::HashSet;
#[test]
fn test_tiny_set() {
assert!(TinySet::empty().is_empty());
{
let mut u = TinySet::empty().insert(1u32);
assert_eq!(u.pop_lowest(), Some(1u32));
assert!(u.pop_lowest().is_none())
}
{
let mut u = TinySet::empty().insert(1u32).insert(1u32);
assert_eq!(u.pop_lowest(), Some(1u32));
assert!(u.pop_lowest().is_none())
}
{
let mut u = TinySet::empty().insert(2u32);
assert_eq!(u.pop_lowest(), Some(2u32));
u.insert_mut(1u32);
assert_eq!(u.pop_lowest(), Some(1u32));
assert!(u.pop_lowest().is_none());
}
{
let mut u = TinySet::empty().insert(63u32);
assert_eq!(u.pop_lowest(), Some(63u32));
assert!(u.pop_lowest().is_none());
}
}
#[test]
fn test_bitset() {
let test_against_hashset = |els: &[u32], max_value: u32| {
let mut hashset: HashSet<u32> = HashSet::new();
let mut bitset = BitSet::with_max_value(max_value);
for &el in els {
assert!(el < max_value);
hashset.insert(el);
bitset.insert(el);
}
for el in 0..max_value {
assert_eq!(hashset.contains(&el), bitset.contains(el));
}
assert_eq!(bitset.max_value(), max_value);
};
test_against_hashset(&[], 0);
test_against_hashset(&[], 1);
test_against_hashset(&[0u32], 1);
test_against_hashset(&[0u32], 100);
test_against_hashset(&[1u32, 2u32], 4);
test_against_hashset(&[99u32], 100);
test_against_hashset(&[63u32], 64);
test_against_hashset(&[62u32, 63u32], 64);
}
#[test]
fn test_bitset_large() {
let arr = generate_nonunique_unsorted(100_000, 5_000);
let mut btreeset: BTreeSet<u32> = BTreeSet::new();
let mut bitset = BitSet::with_max_value(100_000);
for el in arr {
btreeset.insert(el);
bitset.insert(el);
}
for i in 0..100_000 {
assert_eq!(btreeset.contains(&i), bitset.contains(i));
}
assert_eq!(btreeset.len(), bitset.len());
let mut bitset_docset = BitSetDocSet::from(bitset);
let mut remaining = true;
for el in btreeset.into_iter() {
assert!(remaining);
assert_eq!(bitset_docset.doc(), el);
remaining = bitset_docset.advance() != TERMINATED;
}
assert!(!remaining);
}
#[test]
fn test_bitset_num_buckets() {
use super::num_buckets;
assert_eq!(num_buckets(0u32), 0);
assert_eq!(num_buckets(1u32), 1);
assert_eq!(num_buckets(64u32), 1);
assert_eq!(num_buckets(65u32), 2);
assert_eq!(num_buckets(128u32), 2);
assert_eq!(num_buckets(129u32), 3);
}
#[test]
fn test_tinyset_range() {
assert_eq!(
TinySet::range_lower(3).into_iter().collect::<Vec<u32>>(),
[0, 1, 2]
);
assert!(TinySet::range_lower(0).is_empty());
assert_eq!(
TinySet::range_lower(63).into_iter().collect::<Vec<u32>>(),
(0u32..63u32).collect::<Vec<_>>()
);
assert_eq!(
TinySet::range_lower(1).into_iter().collect::<Vec<u32>>(),
[0]
);
assert_eq!(
TinySet::range_lower(2).into_iter().collect::<Vec<u32>>(),
[0, 1]
);
assert_eq!(
TinySet::range_greater_or_equal(3)
.into_iter()
.collect::<Vec<u32>>(),
(3u32..64u32).collect::<Vec<_>>()
);
}
#[test]
fn test_bitset_len() {
let mut bitset = BitSet::with_max_value(1_000);
assert_eq!(bitset.len(), 0);
bitset.insert(3u32);
assert_eq!(bitset.len(), 1);
bitset.insert(103u32);
assert_eq!(bitset.len(), 2);
bitset.insert(3u32);
assert_eq!(bitset.len(), 2);
bitset.insert(103u32);
assert_eq!(bitset.len(), 2);
bitset.insert(104u32);
assert_eq!(bitset.len(), 3);
}
#[test]
fn test_bitset_clear() {
let mut bitset = BitSet::with_max_value(1_000);
let els = tests::sample(1_000, 0.01f64);
for &el in &els {
bitset.insert(el);
}
assert!(els.iter().all(|el| bitset.contains(*el)));
bitset.clear();
for el in 0u32..1000u32 {
assert!(!bitset.contains(el));
}
}
}
#[cfg(all(test, feature = "unstable"))]
mod bench {
use super::BitSet;
use super::TinySet;
use test;
#[bench]
fn bench_tinyset_pop(b: &mut test::Bencher) {
b.iter(|| {
let mut tinyset = TinySet::singleton(test::black_box(31u32));
tinyset.pop_lowest();
tinyset.pop_lowest();
tinyset.pop_lowest();
tinyset.pop_lowest();
tinyset.pop_lowest();
tinyset.pop_lowest();
});
}
#[bench]
fn bench_tinyset_sum(b: &mut test::Bencher) {
let tiny_set = TinySet::empty().insert(10u32).insert(14u32).insert(21u32);
b.iter(|| {
assert_eq!(test::black_box(tiny_set).into_iter().sum::<u32>(), 45u32);
});
}
#[bench]
fn bench_tinyarr_sum(b: &mut test::Bencher) {
let v = [10u32, 14u32, 21u32];
b.iter(|| test::black_box(v).iter().cloned().sum::<u32>());
}
#[bench]
fn bench_bitset_initialize(b: &mut test::Bencher) {
b.iter(|| BitSet::with_max_value(1_000_000));
}
}

203
src/common/mod.rs
View File

@@ -1,203 +0,0 @@
mod bitset;
mod composite_file;
pub use self::bitset::BitSet;
pub(crate) use self::bitset::TinySet;
pub(crate) use self::composite_file::{CompositeFile, CompositeWrite};
pub use byteorder::LittleEndian as Endianness;
pub use common::CountingWriter;
pub use common::{
read_u32_vint, read_u32_vint_no_advance, serialize_vint_u32, write_u32_vint, VInt,
};
pub use common::{BinarySerializable, DeserializeFrom, FixedSize};
/// Segment's max doc must be `< MAX_DOC_LIMIT`.
///
/// We do not allow segments with more than
pub const MAX_DOC_LIMIT: u32 = 1 << 31;
/// Has length trait
pub trait HasLen {
/// Return length
fn len(&self) -> usize;
/// Returns true iff empty.
fn is_empty(&self) -> bool {
self.len() == 0
}
}
const HIGHEST_BIT: u64 = 1 << 63;
/// Maps a `i64` to `u64`
///
/// For simplicity, tantivy internally handles `i64` as `u64`.
/// The mapping is defined by this function.
///
/// Maps `i64` to `u64` so that
/// `-2^63 .. 2^63-1` is mapped
/// to
/// `0 .. 2^64-1`
/// in that order.
///
/// This is more suited than simply casting (`val as u64`)
/// because of bitpacking.
///
/// Imagine a list of `i64` ranging from -10 to 10.
/// When casting negative values, the negative values are projected
/// to values over 2^63, and all values end up requiring 64 bits.
///
/// # See also
/// The [reverse mapping is `u64_to_i64`](./fn.u64_to_i64.html).
#[inline]
pub fn i64_to_u64(val: i64) -> u64 {
(val as u64) ^ HIGHEST_BIT
}
/// Reverse the mapping given by [`i64_to_u64`](./fn.i64_to_u64.html).
#[inline]
pub fn u64_to_i64(val: u64) -> i64 {
(val ^ HIGHEST_BIT) as i64
}
/// Maps a `f64` to `u64`
///
/// For simplicity, tantivy internally handles `f64` as `u64`.
/// The mapping is defined by this function.
///
/// Maps `f64` to `u64` in a monotonic manner, so that bytes lexical order is preserved.
///
/// This is more suited than simply casting (`val as u64`)
/// which would truncate the result
///
/// # Reference
///
/// Daniel Lemire's [blog post](https://lemire.me/blog/2020/12/14/converting-floating-point-numbers-to-integers-while-preserving-order/)
/// explains the mapping in a clear manner.
///
/// # See also
/// The [reverse mapping is `u64_to_f64`](./fn.u64_to_f64.html).
#[inline]
pub fn f64_to_u64(val: f64) -> u64 {
let bits = val.to_bits();
if val.is_sign_positive() {
bits ^ HIGHEST_BIT
} else {
!bits
}
}
/// Reverse the mapping given by [`i64_to_u64`](./fn.i64_to_u64.html).
#[inline]
pub fn u64_to_f64(val: u64) -> f64 {
f64::from_bits(if val & HIGHEST_BIT != 0 {
val ^ HIGHEST_BIT
} else {
!val
})
}
#[cfg(test)]
pub(crate) mod test {
use super::{f64_to_u64, i64_to_u64, u64_to_f64, u64_to_i64};
use common::{BinarySerializable, FixedSize};
use proptest::prelude::*;
use std::f64;
use tantivy_bitpacker::compute_num_bits;
pub use tantivy_bitpacker::minmax;
fn test_i64_converter_helper(val: i64) {
assert_eq!(u64_to_i64(i64_to_u64(val)), val);
}
fn test_f64_converter_helper(val: f64) {
assert_eq!(u64_to_f64(f64_to_u64(val)), val);
}
pub fn fixed_size_test<O: BinarySerializable + FixedSize + Default>() {
let mut buffer = Vec::new();
O::default().serialize(&mut buffer).unwrap();
assert_eq!(buffer.len(), O::SIZE_IN_BYTES);
}
proptest! {
#[test]
fn test_f64_converter_monotonicity_proptest((left, right) in (proptest::num::f64::NORMAL, proptest::num::f64::NORMAL)) {
let left_u64 = f64_to_u64(left);
let right_u64 = f64_to_u64(right);
assert_eq!(left_u64 < right_u64, left < right);
}
}
#[test]
fn test_i64_converter() {
assert_eq!(i64_to_u64(i64::min_value()), u64::min_value());
assert_eq!(i64_to_u64(i64::max_value()), u64::max_value());
test_i64_converter_helper(0i64);
test_i64_converter_helper(i64::min_value());
test_i64_converter_helper(i64::max_value());
for i in -1000i64..1000i64 {
test_i64_converter_helper(i);
}
}
#[test]
fn test_f64_converter() {
test_f64_converter_helper(f64::INFINITY);
test_f64_converter_helper(f64::NEG_INFINITY);
test_f64_converter_helper(0.0);
test_f64_converter_helper(-0.0);
test_f64_converter_helper(1.0);
test_f64_converter_helper(-1.0);
}
#[test]
fn test_f64_order() {
assert!(!(f64_to_u64(f64::NEG_INFINITY)..f64_to_u64(f64::INFINITY))
.contains(&f64_to_u64(f64::NAN))); //nan is not a number
assert!(f64_to_u64(1.5) > f64_to_u64(1.0)); //same exponent, different mantissa
assert!(f64_to_u64(2.0) > f64_to_u64(1.0)); //same mantissa, different exponent
assert!(f64_to_u64(2.0) > f64_to_u64(1.5)); //different exponent and mantissa
assert!(f64_to_u64(1.0) > f64_to_u64(-1.0)); // pos > neg
assert!(f64_to_u64(-1.5) < f64_to_u64(-1.0));
assert!(f64_to_u64(-2.0) < f64_to_u64(1.0));
assert!(f64_to_u64(-2.0) < f64_to_u64(-1.5));
}
#[test]
fn test_compute_num_bits() {
assert_eq!(compute_num_bits(1), 1u8);
assert_eq!(compute_num_bits(0), 0u8);
assert_eq!(compute_num_bits(2), 2u8);
assert_eq!(compute_num_bits(3), 2u8);
assert_eq!(compute_num_bits(4), 3u8);
assert_eq!(compute_num_bits(255), 8u8);
assert_eq!(compute_num_bits(256), 9u8);
assert_eq!(compute_num_bits(5_000_000_000), 33u8);
}
#[test]
fn test_max_doc() {
// this is the first time I write a unit test for a constant.
assert!(((super::MAX_DOC_LIMIT - 1) as i32) >= 0);
assert!((super::MAX_DOC_LIMIT as i32) < 0);
}
#[test]
fn test_minmax_empty() {
let vals: Vec<u32> = vec![];
assert_eq!(minmax(vals.into_iter()), None);
}
#[test]
fn test_minmax_one() {
assert_eq!(minmax(vec![1].into_iter()), Some((1, 1)));
}
#[test]
fn test_minmax_two() {
assert_eq!(minmax(vec![1, 2].into_iter()), Some((1, 2)));
assert_eq!(minmax(vec![2, 1].into_iter()), Some((1, 2)));
}
}

47
src/core/executor.rs
View File

@@ -1,6 +1,7 @@
use crossbeam::channel;
use rayon::{ThreadPool, ThreadPoolBuilder};
use crate::TantivyError;
/// Search executor whether search request are single thread or multithread.
///
/// We don't expose Rayon thread pool directly here for several reasons.
@@ -47,17 +48,24 @@ impl Executor {
match self {
Executor::SingleThread => args.map(f).collect::<crate::Result<_>>(),
Executor::ThreadPool(pool) => {
let args_with_indices: Vec<(usize, A)> = args.enumerate().collect();
let num_fruits = args_with_indices.len();
let args: Vec<A> = args.collect();
let num_fruits = args.len();
let fruit_receiver = {
let (fruit_sender, fruit_receiver) = channel::unbounded();
let (fruit_sender, fruit_receiver) = crossbeam_channel::unbounded();
pool.scope(|scope| {
for arg_with_idx in args_with_indices {
scope.spawn(|_| {
let (idx, arg) = arg_with_idx;
let fruit = f(arg);
if let Err(err) = fruit_sender.send((idx, fruit)) {
error!("Failed to send search task. It probably means all search threads have panicked. {:?}", err);
for (idx, arg) in args.into_iter().enumerate() {
// We name references for f and fruit_sender_ref because we do not
// want these two to be moved into the closure.
let f_ref = &f;
let fruit_sender_ref = &fruit_sender;
scope.spawn(move |_| {
let fruit = f_ref(arg);
if let Err(err) = fruit_sender_ref.send((idx, fruit)) {
error!(
"Failed to send search task. It probably means all search \
threads have panicked. {:?}",
err
);
}
});
}
@@ -67,18 +75,19 @@ impl Executor {
// This is important as it makes it possible for the fruit_receiver iteration to
// terminate.
};
// This is lame, but safe.
let mut results_with_position = Vec::with_capacity(num_fruits);
let mut result_placeholders: Vec<Option<R>> =
std::iter::repeat_with(|| None).take(num_fruits).collect();
for (pos, fruit_res) in fruit_receiver {
let fruit = fruit_res?;
results_with_position.push((pos, fruit));
result_placeholders[pos] = Some(fruit);
}
results_with_position.sort_by_key(|(pos, _)| *pos);
assert_eq!(results_with_position.len(), num_fruits);
Ok(results_with_position
.into_iter()
.map(|(_, fruit)| fruit)
.collect::<Vec<_>>())
let results: Vec<R> = result_placeholders.into_iter().flatten().collect();
if results.len() != num_fruits {
return Err(TantivyError::InternalError(
"One of the mapped execution failed.".to_string(),
));
}
Ok(results)
}
}
}

315
src/core/index.rs
View File

@@ -1,35 +1,27 @@
use super::{segment::Segment, IndexSettings};
use crate::core::Executor;
use crate::core::IndexMeta;
use crate::core::SegmentId;
use crate::core::SegmentMeta;
use crate::core::SegmentMetaInventory;
use crate::core::META_FILEPATH;
use crate::directory::error::OpenReadError;
use crate::directory::ManagedDirectory;
#[cfg(feature = "mmap")]
use crate::directory::MmapDirectory;
use crate::directory::INDEX_WRITER_LOCK;
use crate::directory::{Directory, RamDirectory};
use crate::error::DataCorruption;
use crate::error::TantivyError;
use crate::indexer::index_writer::{HEAP_SIZE_MIN, MAX_NUM_THREAD};
use crate::indexer::segment_updater::save_new_metas;
use crate::reader::IndexReader;
use crate::reader::IndexReaderBuilder;
use crate::schema::Field;
use crate::schema::FieldType;
use crate::schema::Schema;
use crate::tokenizer::{TextAnalyzer, TokenizerManager};
use crate::IndexWriter;
use std::collections::HashSet;
use std::fmt;
#[cfg(feature = "mmap")]
use std::path::Path;
use std::path::PathBuf;
use std::sync::Arc;
use super::segment::Segment;
use super::IndexSettings;
use crate::core::{
Executor, IndexMeta, SegmentId, SegmentMeta, SegmentMetaInventory, META_FILEPATH,
};
use crate::directory::error::OpenReadError;
#[cfg(feature = "mmap")]
use crate::directory::MmapDirectory;
use crate::directory::{Directory, ManagedDirectory, RamDirectory, INDEX_WRITER_LOCK};
use crate::error::{DataCorruption, TantivyError};
use crate::indexer::index_writer::{MAX_NUM_THREAD, MEMORY_ARENA_NUM_BYTES_MIN};
use crate::indexer::segment_updater::save_new_metas;
use crate::reader::{IndexReader, IndexReaderBuilder};
use crate::schema::{Field, FieldType, Schema};
use crate::tokenizer::{TextAnalyzer, TokenizerManager};
use crate::IndexWriter;
fn load_metas(
directory: &dyn Directory,
inventory: &SegmentMetaInventory,
@@ -72,17 +64,17 @@ fn load_metas(
/// let body_field = schema_builder.add_text_field("body", TEXT);
/// let number_field = schema_builder.add_u64_field(
/// "number",
/// IntOptions::default().set_fast(Cardinality::SingleValue),
/// NumericOptions::default().set_fast(Cardinality::SingleValue),
/// );
///
/// let schema = schema_builder.build();
/// let settings = IndexSettings{sort_by_field: Some(IndexSortByField{field:"number".to_string(), order:Order::Asc}), ..Default::default()};
/// let index = Index::builder().schema(schema).settings(settings).create_in_ram();
///
/// ```
pub struct IndexBuilder {
schema: Option<Schema>,
index_settings: IndexSettings,
tokenizer_manager: TokenizerManager,
}
impl Default for IndexBuilder {
fn default() -> Self {
@@ -95,18 +87,30 @@ impl IndexBuilder {
Self {
schema: None,
index_settings: IndexSettings::default(),
tokenizer_manager: TokenizerManager::default(),
}
}
/// Set the settings
#[must_use]
pub fn settings(mut self, settings: IndexSettings) -> Self {
self.index_settings = settings;
self
}
/// Set the schema
#[must_use]
pub fn schema(mut self, schema: Schema) -> Self {
self.schema = Some(schema);
self
}
/// Set the tokenizers .
pub fn tokenizers(mut self, tokenizers: TokenizerManager) -> Self {
self.tokenizer_manager = tokenizers;
self
}
/// Creates a new index using the `RAMDirectory`.
///
/// The index will be allocated in anonymous memory.
@@ -117,18 +121,20 @@ impl IndexBuilder {
.create(ram_directory)
.expect("Creating a RAMDirectory should never fail"))
}
/// Creates a new index in a given filepath.
/// The index will use the `MMapDirectory`.
///
/// If a previous index was in this directory, then its meta file will be destroyed.
/// 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)
}
/// Creates a new index in a temp directory.
///
/// The index will use the `MMapDirectory` in a newly created directory.
@@ -139,21 +145,25 @@ 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> {
self.schema
.as_ref()
.cloned()
.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)?;
let mut index = Index::open(dir)?;
index.set_tokenizers(self.tokenizer_manager.clone());
if index.schema() == self.get_expect_schema()? {
Ok(index)
} else {
@@ -165,7 +175,8 @@ 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()?,
@@ -174,7 +185,8 @@ impl IndexBuilder {
)?;
let mut metas = IndexMeta::with_schema(self.get_expect_schema()?);
metas.index_settings = self.index_settings;
let index = Index::open_from_metas(directory, &metas, SegmentMetaInventory::default());
let mut index = Index::open_from_metas(directory, &metas, SegmentMetaInventory::default());
index.set_tokenizers(self.tokenizer_manager);
Ok(index)
}
}
@@ -198,7 +210,7 @@ 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)
}
@@ -215,7 +227,7 @@ impl Index {
/// Replace the default single thread search executor pool
/// by a thread pool with a given number of threads.
pub fn set_multithread_executor(&mut self, num_threads: usize) -> crate::Result<()> {
self.executor = Arc::new(Executor::multi_thread(num_threads, "thrd-tantivy-search-")?);
self.executor = Arc::new(Executor::multi_thread(num_threads, "tantivy-search-")?);
Ok(())
}
@@ -229,7 +241,8 @@ impl Index {
/// Creates a new index using the `RamDirectory`.
///
/// The index will be allocated in anonymous memory.
/// This should only be used for unit tests.
/// This is useful for indexing small set of documents
/// for instances like unit test or temporary in memory index.
pub fn create_in_ram(schema: Schema) -> Index {
IndexBuilder::new().schema(schema).create_in_ram().unwrap()
}
@@ -237,7 +250,7 @@ impl Index {
/// Creates a new index in a given filepath.
/// The index will use the `MMapDirectory`.
///
/// If a previous index was in this directory, then its meta file will be destroyed.
/// If a previous index was in this directory, then it returns an `IndexAlreadyExists` error.
#[cfg(feature = "mmap")]
pub fn create_in_dir<P: AsRef<Path>>(
directory_path: P,
@@ -249,7 +262,11 @@ 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)
}
@@ -269,11 +286,12 @@ 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)
@@ -296,6 +314,11 @@ impl Index {
}
}
/// Setter for the tokenizer manager.
pub fn set_tokenizers(&mut self, tokenizers: TokenizerManager) {
self.tokenizers = tokenizers;
}
/// Accessor for the tokenizer manager.
pub fn tokenizers(&self) -> &TokenizerManager {
&self.tokenizers
@@ -306,20 +329,31 @@ impl Index {
let field_entry = self.schema.get_field_entry(field);
let field_type = field_entry.field_type();
let tokenizer_manager: &TokenizerManager = self.tokenizers();
let tokenizer_name_opt: Option<TextAnalyzer> = match field_type {
FieldType::Str(text_options) => text_options
.get_indexing_options()
.map(|text_indexing_options| text_indexing_options.tokenizer().to_string())
.and_then(|tokenizer_name| tokenizer_manager.get(&tokenizer_name)),
_ => None,
let indexing_options_opt = match field_type {
FieldType::JsonObject(options) => options.get_text_indexing_options(),
FieldType::Str(options) => options.get_indexing_options(),
_ => {
return Err(TantivyError::SchemaError(format!(
"{:?} is not a text field.",
field_entry.name()
)))
}
};
match tokenizer_name_opt {
Some(tokenizer) => Ok(tokenizer),
None => Err(TantivyError::SchemaError(format!(
"{:?} is not a text field.",
field_entry.name()
))),
}
let indexing_options = indexing_options_opt.ok_or_else(|| {
TantivyError::InvalidArgument(format!(
"No indexing options set for field {:?}",
field_entry
))
})?;
tokenizer_manager
.get(indexing_options.tokenizer())
.ok_or_else(|| {
TantivyError::InvalidArgument(format!(
"No Tokenizer found for field {:?}",
field_entry
))
})
}
/// Create a default `IndexReader` for the given index.
@@ -364,7 +398,8 @@ 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)?;
@@ -388,19 +423,18 @@ impl Index {
/// - `num_threads` defines the number of indexing workers that
/// should work at the same time.
///
/// - `overall_heap_size_in_bytes` sets the amount of memory
/// - `overall_memory_arena_in_bytes` sets the amount of memory
/// allocated for all indexing thread.
/// Each thread will receive a budget of `overall_heap_size_in_bytes / num_threads`.
/// Each thread will receive a budget of `overall_memory_arena_in_bytes / num_threads`.
///
/// # 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 memory arena per thread is too small or too big, returns
/// `TantivyError::InvalidArgument`
pub fn writer_with_num_threads(
&self,
num_threads: usize,
overall_heap_size_in_bytes: usize,
overall_memory_arena_in_bytes: usize,
) -> crate::Result<IndexWriter> {
let directory_lock = self
.directory
@@ -409,26 +443,25 @@ impl Index {
TantivyError::LockFailure(
err,
Some(
"Failed to acquire index lock. If you are using \
a regular directory, this means there is already an \
`IndexWriter` working on this `Directory`, in this process \
or in a different process."
"Failed to acquire index lock. If you are using a regular directory, this \
means there is already an `IndexWriter` working on this `Directory`, in \
this process or in a different process."
.to_string(),
),
)
})?;
let heap_size_in_bytes_per_thread = overall_heap_size_in_bytes / num_threads;
let memory_arena_in_bytes_per_thread = overall_memory_arena_in_bytes / num_threads;
IndexWriter::new(
self,
num_threads,
heap_size_in_bytes_per_thread,
memory_arena_in_bytes_per_thread,
directory_lock,
)
}
/// Helper to create an index writer for tests.
///
/// That index writer only simply has a single thread and a heap of 10 MB.
/// That index writer only simply has a single thread and a memory arena of 10 MB.
/// Using a single thread gives us a deterministic allocation of DocId.
#[cfg(test)]
pub fn writer_for_tests(&self) -> crate::Result<IndexWriter> {
@@ -438,31 +471,29 @@ impl Index {
/// Creates a multithreaded writer
///
/// Tantivy will automatically define the number of threads to use, but
/// no more than [`MAX_NUM_THREAD`] threads.
/// `overall_heap_size_in_bytes` is the total target memory usage that will be split
/// no more than 8 threads.
/// `overall_memory_arena_in_bytes` is the total target memory usage that will be split
/// between a given number of threads.
///
/// # Errors
/// If the lockfile already exists, returns `Error::FileAlreadyExists`.
/// # Panics
/// If the heap size per thread is too small, panics.
pub fn writer(&self, overall_heap_size_in_bytes: usize) -> crate::Result<IndexWriter> {
/// If the memory arena per thread is too small or too big, returns
/// `TantivyError::InvalidArgument`
pub fn writer(&self, memory_arena_num_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;
if heap_size_in_bytes_per_thread < HEAP_SIZE_MIN {
num_threads = (overall_heap_size_in_bytes / HEAP_SIZE_MIN).max(1);
let memory_arena_num_bytes_per_thread = memory_arena_num_bytes / num_threads;
if memory_arena_num_bytes_per_thread < MEMORY_ARENA_NUM_BYTES_MIN {
num_threads = (memory_arena_num_bytes / MEMORY_ARENA_NUM_BYTES_MIN).max(1);
}
self.writer_with_num_threads(num_threads, overall_heap_size_in_bytes)
self.writer_with_num_threads(num_threads, memory_arena_num_bytes)
}
/// Accessor to the index settings
///
pub fn settings(&self) -> &IndexSettings {
&self.settings
}
/// Accessor to the index settings
///
pub fn settings_mut(&mut self) -> &mut IndexSettings {
&mut self.settings
}
@@ -523,7 +554,22 @@ impl Index {
/// Returns the set of corrupted files
pub fn validate_checksum(&self) -> crate::Result<HashSet<PathBuf>> {
self.directory.list_damaged().map_err(Into::into)
let managed_files = self.directory.list_managed_files();
let active_segments_files: HashSet<PathBuf> = self
.searchable_segment_metas()?
.iter()
.flat_map(|segment_meta| segment_meta.list_files())
.collect();
let active_existing_files: HashSet<&PathBuf> =
active_segments_files.intersection(&managed_files).collect();
let mut damaged_files = HashSet::new();
for path in active_existing_files {
if !self.directory.validate_checksum(path)? {
damaged_files.insert((*path).clone());
}
}
Ok(damaged_files)
}
}
@@ -535,15 +581,10 @@ impl fmt::Debug for Index {
#[cfg(test)]
mod tests {
use crate::schema::Field;
use crate::schema::{Schema, INDEXED, TEXT};
use crate::IndexReader;
use crate::ReloadPolicy;
use crate::{
directory::{RamDirectory, WatchCallback},
IndexSettings,
};
use crate::{Directory, Index};
use crate::directory::{RamDirectory, WatchCallback};
use crate::schema::{Field, Schema, INDEXED, TEXT};
use crate::tokenizer::TokenizerManager;
use crate::{Directory, Index, IndexBuilder, IndexReader, IndexSettings, ReloadPolicy};
#[test]
fn test_indexer_for_field() {
@@ -559,17 +600,32 @@ mod tests {
);
}
#[test]
fn test_set_tokenizer_manager() {
let mut schema_builder = Schema::builder();
schema_builder.add_u64_field("num_likes", INDEXED);
schema_builder.add_text_field("body", TEXT);
let schema = schema_builder.build();
let index = IndexBuilder::new()
// set empty tokenizer manager
.tokenizers(TokenizerManager::new())
.schema(schema)
.create_in_ram()
.unwrap();
assert!(index.tokenizers().get("raw").is_none());
}
#[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]
@@ -582,27 +638,27 @@ 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(),
@@ -636,7 +692,7 @@ 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);
@@ -646,19 +702,21 @@ 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")]
mod mmap_specific {
use super::*;
use crate::Directory;
use std::path::PathBuf;
use tempfile::TempDir;
use super::*;
use crate::Directory;
#[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();
@@ -670,7 +728,7 @@ 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]
@@ -685,8 +743,8 @@ mod tests {
.reload_policy(ReloadPolicy::Manual)
.try_into()?;
assert_eq!(reader.searcher().num_docs(), 0);
writer.add_document(doc!(field=>1u64));
let (sender, receiver) = crossbeam::channel::unbounded();
writer.add_document(doc!(field=>1u64))?;
let (sender, receiver) = crossbeam_channel::unbounded();
let _handle = index.directory_mut().watch(WatchCallback::new(move || {
let _ = sender.send(());
}));
@@ -699,7 +757,7 @@ 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();
@@ -712,20 +770,24 @@ 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 (sender, receiver) = crossbeam_channel::unbounded();
let _watch_handle = reader_index
.directory_mut()
.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.
@@ -735,7 +797,7 @@ mod tests {
break;
}
}
writer.add_document(doc!(field=>2u64));
writer.add_document(doc!(field=>2u64))?;
writer.commit().unwrap();
// ... Same as above
loop {
@@ -744,41 +806,41 @@ 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();
assert_eq!(reader.searcher().segment_readers().len(), 8);
writer.wait_merging_threads()?;
let mem_right_after_merge_finished = directory.total_mem_usage();
reader.reload().unwrap();
let searcher = reader.searcher();
assert_eq!(searcher.segment_readers().len(), 1);
assert_eq!(searcher.num_docs(), 8_000);
assert!(
mem_right_after_merge_finished < mem_right_after_commit,
@@ -786,5 +848,6 @@ mod tests {
mem_right_after_merge_finished,
mem_right_after_commit
);
Ok(())
}
}

68
src/core/index_meta.rs
View File

@@ -1,12 +1,16 @@
use super::SegmentComponent;
use crate::schema::Schema;
use crate::Opstamp;
use crate::{core::SegmentId, store::Compressor};
use census::{Inventory, TrackedObject};
use serde::{Deserialize, Serialize};
use std::collections::HashSet;
use std::fmt;
use std::path::PathBuf;
use std::{collections::HashSet, sync::atomic::AtomicBool};
use std::{fmt, sync::Arc};
use std::sync::atomic::AtomicBool;
use std::sync::Arc;
use serde::{Deserialize, Serialize};
use super::SegmentComponent;
use crate::core::SegmentId;
use crate::schema::Schema;
use crate::store::Compressor;
use crate::{Inventory, Opstamp, TrackedObject};
#[derive(Clone, Debug, Serialize, Deserialize)]
struct DeleteMeta {
@@ -101,6 +105,7 @@ impl SegmentMeta {
/// Returns the list of files that
/// are required for the segment meta.
/// Note: Some of the returned files may not exist depending on the state of the segment.
///
/// This is useful as the way tantivy removes files
/// is by removing all files that have been created by tantivy
@@ -187,7 +192,12 @@ impl SegmentMeta {
}
#[doc(hidden)]
#[must_use]
pub fn with_delete_meta(self, num_deleted_docs: u32, opstamp: Opstamp) -> SegmentMeta {
assert!(
num_deleted_docs <= self.max_doc(),
"There cannot be more deleted docs than there are docs."
);
let delete_meta = DeleteMeta {
num_deleted_docs,
opstamp,
@@ -229,7 +239,7 @@ impl InnerSegmentMeta {
///
/// Contains settings which are applied on the whole
/// index, like presort documents.
#[derive(Clone, Default, Serialize, Deserialize, Eq, PartialEq)]
#[derive(Clone, Debug, Serialize, Deserialize, Eq, PartialEq)]
pub struct IndexSettings {
/// Sorts the documents by information
/// provided in `IndexSortByField`
@@ -238,13 +248,32 @@ pub struct IndexSettings {
/// The `Compressor` used to compress the doc store.
#[serde(default)]
pub docstore_compression: Compressor,
#[serde(default = "default_docstore_blocksize")]
/// The size of each block that will be compressed and written to disk
pub docstore_blocksize: usize,
}
/// Must be a function to be compatible with serde defaults
fn default_docstore_blocksize() -> usize {
16_384
}
impl Default for IndexSettings {
fn default() -> Self {
Self {
sort_by_field: None,
docstore_compression: Compressor::default(),
docstore_blocksize: default_docstore_blocksize(),
}
}
}
/// Settings to presort the documents in an index
///
/// Presorting documents can greatly performance
/// in some scenarios, by applying top n
/// optimizations.
#[derive(Clone, Serialize, Deserialize, Eq, PartialEq)]
#[derive(Clone, Debug, Serialize, Deserialize, Eq, PartialEq)]
pub struct IndexSortByField {
/// The field to sort the documents by
pub field: String,
@@ -252,7 +281,7 @@ pub struct IndexSortByField {
pub order: Order,
}
/// The order to sort by
#[derive(Clone, Serialize, Deserialize, Eq, PartialEq)]
#[derive(Clone, Debug, Serialize, Deserialize, Eq, PartialEq)]
pub enum Order {
/// Ascending Order
Asc,
@@ -277,7 +306,6 @@ impl Order {
/// * the searchable segments,
/// * the index `docstamp`
/// * the schema
///
#[derive(Clone, Serialize)]
pub struct IndexMeta {
/// `IndexSettings` to configure index options.
@@ -289,12 +317,12 @@ pub struct IndexMeta {
pub schema: Schema,
/// Opstamp associated to the last `commit` operation.
pub opstamp: Opstamp,
#[serde(skip_serializing_if = "Option::is_none")]
/// Payload associated to the last commit.
///
/// Upon commit, clients can optionally add a small `String` payload to their commit
/// to help identify this commit.
/// This payload is entirely unused by tantivy.
#[serde(skip_serializing_if = "Option::is_none")]
pub payload: Option<String>,
}
@@ -365,10 +393,9 @@ impl fmt::Debug for IndexMeta {
mod tests {
use super::IndexMeta;
use crate::{
schema::{Schema, TEXT},
IndexSettings, IndexSortByField, Order,
};
use crate::core::index_meta::UntrackedIndexMeta;
use crate::schema::{Schema, TEXT};
use crate::{IndexSettings, IndexSortByField, Order};
#[test]
fn test_serialize_metas() {
@@ -393,7 +420,12 @@ mod tests {
let json = serde_json::ser::to_string(&index_metas).expect("serialization failed");
assert_eq!(
json,
r#"{"index_settings":{"sort_by_field":{"field":"text","order":"Asc"},"docstore_compression":"lz4"},"segments":[],"schema":[{"name":"text","type":"text","options":{"indexing":{"record":"position","tokenizer":"default"},"stored":false}}],"opstamp":0}"#
r#"{"index_settings":{"sort_by_field":{"field":"text","order":"Asc"},"docstore_compression":"lz4","docstore_blocksize":16384},"segments":[],"schema":[{"name":"text","type":"text","options":{"indexing":{"record":"position","fieldnorms":true,"tokenizer":"default"},"stored":false,"fast":false}}],"opstamp":0}"#
);
let deser_meta: UntrackedIndexMeta = serde_json::from_str(&json).unwrap();
assert_eq!(index_metas.index_settings, deser_meta.index_settings);
assert_eq!(index_metas.schema, deser_meta.schema);
assert_eq!(index_metas.opstamp, deser_meta.opstamp);
}
}

45
src/core/inverted_index_reader.rs
View File

@@ -1,12 +1,11 @@
use std::io;
use crate::common::BinarySerializable;
use common::BinarySerializable;
use crate::directory::FileSlice;
use crate::positions::PositionReader;
use crate::postings::TermInfo;
use crate::postings::{BlockSegmentPostings, SegmentPostings};
use crate::schema::IndexRecordOption;
use crate::schema::Term;
use crate::postings::{BlockSegmentPostings, SegmentPostings, TermInfo};
use crate::schema::{IndexRecordOption, Term};
use crate::termdict::TermDictionary;
/// The inverted index reader is in charge of accessing
@@ -89,7 +88,8 @@ impl InvertedIndexReader {
let postings_slice = self
.postings_file_slice
.slice(term_info.postings_range.clone());
block_postings.reset(term_info.doc_freq, postings_slice.read_bytes()?);
let postings_bytes = postings_slice.read_bytes()?;
block_postings.reset(term_info.doc_freq, postings_bytes)?;
Ok(())
}
@@ -198,3 +198,36 @@ impl InvertedIndexReader {
.unwrap_or(0u32))
}
}
#[cfg(feature = "quickwit")]
impl InvertedIndexReader {
pub(crate) async fn get_term_info_async(
&self,
term: &Term,
) -> crate::AsyncIoResult<Option<TermInfo>> {
self.termdict.get_async(term.value_bytes()).await
}
/// Returns a block postings given a `Term`.
/// This method is for an advanced usage only.
///
/// Most user should prefer using `read_postings` instead.
pub async fn warm_postings(
&self,
term: &Term,
with_positions: bool,
) -> crate::AsyncIoResult<()> {
let term_info_opt = self.get_term_info_async(term).await?;
if let Some(term_info) = term_info_opt {
self.postings_file_slice
.read_bytes_slice_async(term_info.postings_range.clone())
.await?;
if with_positions {
self.positions_file_slice
.read_bytes_slice_async(term_info.positions_range.clone())
.await?;
}
}
Ok(())
}
}

9
src/core/mod.rs
View File

@@ -8,21 +8,22 @@ mod segment_component;
mod segment_id;
mod segment_reader;
use std::path::Path;
use once_cell::sync::Lazy;
pub use self::executor::Executor;
pub use self::index::{Index, IndexBuilder};
pub use self::index_meta::{
IndexMeta, IndexSettings, IndexSortByField, Order, SegmentMeta, SegmentMetaInventory,
};
pub use self::inverted_index_reader::InvertedIndexReader;
pub use self::searcher::Searcher;
pub use self::searcher::{Searcher, SearcherGeneration};
pub use self::segment::Segment;
pub use self::segment_component::SegmentComponent;
pub use self::segment_id::SegmentId;
pub use self::segment_reader::SegmentReader;
use once_cell::sync::Lazy;
use std::path::Path;
/// The meta file contains all the information about the list of segments and the schema
/// of the index.
pub static META_FILEPATH: Lazy<&'static Path> = Lazy::new(|| Path::new("meta.json"));

81
src/core/searcher.rs
View File

@@ -1,28 +1,73 @@
use crate::collector::Collector;
use crate::core::Executor;
use std::collections::BTreeMap;
use std::{fmt, io};
use crate::core::SegmentReader;
use crate::collector::Collector;
use crate::core::{Executor, SegmentReader};
use crate::query::Query;
use crate::schema::Document;
use crate::schema::Schema;
use crate::schema::Term;
use crate::schema::{Document, Schema, Term};
use crate::space_usage::SearcherSpaceUsage;
use crate::store::StoreReader;
use crate::DocAddress;
use crate::Index;
use crate::{DocAddress, Index, Opstamp, SegmentId, TrackedObject};
use std::{fmt, io};
/// Identifies the searcher generation accessed by a [Searcher].
///
/// While this might seem redundant, a [SearcherGeneration] contains
/// both a `generation_id` AND a list of `(SegmentId, DeleteOpstamp)`.
///
/// This is on purpose. This object is used by the `Warmer` API.
/// Having both information makes it possible to identify which
/// artifact should be refreshed or garbage collected.
///
/// Depending on the use case, `Warmer`'s implementers can decide to
/// produce artifacts per:
/// - `generation_id` (e.g. some searcher level aggregates)
/// - `(segment_id, delete_opstamp)` (e.g. segment level aggregates)
/// - `segment_id` (e.g. for immutable document level information)
/// - `(generation_id, segment_id)` (e.g. for consistent dynamic column)
/// - ...
#[derive(Debug, Clone, PartialEq, Eq, PartialOrd, Ord, Hash)]
pub struct SearcherGeneration {
segments: BTreeMap<SegmentId, Option<Opstamp>>,
generation_id: u64,
}
impl SearcherGeneration {
pub(crate) fn from_segment_readers(
segment_readers: &[SegmentReader],
generation_id: u64,
) -> Self {
let mut segment_id_to_del_opstamp = BTreeMap::new();
for segment_reader in segment_readers {
segment_id_to_del_opstamp
.insert(segment_reader.segment_id(), segment_reader.delete_opstamp());
}
Self {
segments: segment_id_to_del_opstamp,
generation_id,
}
}
/// Returns the searcher generation id.
pub fn generation_id(&self) -> u64 {
self.generation_id
}
/// Return a `(SegmentId -> DeleteOpstamp)` mapping.
pub fn segments(&self) -> &BTreeMap<SegmentId, Option<Opstamp>> {
&self.segments
}
}
/// Holds a list of `SegmentReader`s ready for search.
///
/// It guarantees that the `Segment` will not be removed before
/// the destruction of the `Searcher`.
///
pub struct Searcher {
schema: Schema,
index: Index,
segment_readers: Vec<SegmentReader>,
store_readers: Vec<StoreReader>,
generation: TrackedObject<SearcherGeneration>,
}
impl Searcher {
@@ -31,6 +76,7 @@ impl Searcher {
schema: Schema,
index: Index,
segment_readers: Vec<SegmentReader>,
generation: TrackedObject<SearcherGeneration>,
) -> io::Result<Searcher> {
let store_readers: Vec<StoreReader> = segment_readers
.iter()
@@ -41,6 +87,7 @@ impl Searcher {
index,
segment_readers,
store_readers,
generation,
})
}
@@ -49,6 +96,11 @@ impl Searcher {
&self.index
}
/// [SearcherGeneration] which identifies the version of the snapshot held by this `Searcher`.
pub fn generation(&self) -> &SearcherGeneration {
self.generation.as_ref()
}
/// Fetches a document from tantivy's store given a `DocAddress`.
///
/// The searcher uses the segment ordinal to route the
@@ -58,6 +110,13 @@ impl Searcher {
store_reader.get(doc_address.doc_id)
}
/// Fetches a document in an asynchronous manner.
#[cfg(feature = "quickwit")]
pub async fn doc_async(&self, doc_address: DocAddress) -> crate::Result<Document> {
let store_reader = &self.store_readers[doc_address.segment_ord as usize];
store_reader.get_async(doc_address.doc_id).await
}
/// Access the schema associated to the index of this searcher.
pub fn schema(&self) -> &Schema {
&self.schema
@@ -88,7 +147,7 @@ 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]
}

17
src/core/segment.rs
View File

@@ -1,15 +1,13 @@
use super::SegmentComponent;
use crate::core::Index;
use crate::core::SegmentId;
use crate::core::SegmentMeta;
use crate::directory::error::{OpenReadError, OpenWriteError};
use crate::directory::Directory;
use crate::directory::{FileSlice, WritePtr};
use crate::schema::Schema;
use crate::Opstamp;
use std::fmt;
use std::path::PathBuf;
use super::SegmentComponent;
use crate::core::{Index, SegmentId, SegmentMeta};
use crate::directory::error::{OpenReadError, OpenWriteError};
use crate::directory::{Directory, FileSlice, WritePtr};
use crate::schema::Schema;
use crate::Opstamp;
/// A segment is a piece of the index.
#[derive(Clone)]
pub struct Segment {
@@ -56,6 +54,7 @@ impl Segment {
}
#[doc(hidden)]
#[must_use]
pub fn with_delete_meta(self, num_deleted_docs: u32, opstamp: Opstamp) -> Segment {
Segment {
index: self.index,

16
src/core/segment_id.rs
View File

@@ -1,14 +1,14 @@
use std::cmp::{Ord, Ordering};
use std::error::Error;
use std::fmt;
use uuid::Uuid;
use std::str::FromStr;
#[cfg(test)]
use std::sync::atomic;
#[cfg(test)]
use once_cell::sync::Lazy;
use serde::{Deserialize, Serialize};
use std::error::Error;
use std::str::FromStr;
#[cfg(test)]
use std::sync::atomic;
use uuid::Uuid;
/// Uuid identifying a segment.
///
@@ -35,7 +35,7 @@ const ZERO_ARRAY: [u8; 8] = [0u8; 8];
#[cfg(test)]
fn create_uuid() -> Uuid {
let new_auto_inc_id = (*AUTO_INC_COUNTER).fetch_add(1, atomic::Ordering::SeqCst);
Uuid::from_fields(new_auto_inc_id as u32, 0, 0, &ZERO_ARRAY).unwrap()
Uuid::from_fields(new_auto_inc_id as u32, 0, 0, &ZERO_ARRAY)
}
#[cfg(not(test))]
@@ -57,7 +57,7 @@ impl SegmentId {
/// Picking the first 8 chars is ok to identify
/// segments in a display message (e.g. a5c4dfcb).
pub fn short_uuid_string(&self) -> String {
(&self.0.to_simple_ref().to_string()[..8]).to_string()
(&self.0.as_simple().to_string()[..8]).to_string()
}
/// Returns a segment uuid string.
@@ -65,7 +65,7 @@ impl SegmentId {
/// It consists in 32 lowercase hexadecimal chars
/// (e.g. a5c4dfcbdfe645089129e308e26d5523)
pub fn uuid_string(&self) -> String {
self.0.to_simple_ref().to_string()
self.0.as_simple().to_string()
}
/// Build a `SegmentId` string from the full uuid string.

175
src/core/segment_reader.rs
View File

@@ -1,25 +1,19 @@
use crate::core::InvertedIndexReader;
use crate::core::Segment;
use crate::core::SegmentComponent;
use crate::core::SegmentId;
use crate::directory::FileSlice;
use crate::fastfield::DeleteBitSet;
use crate::fastfield::FacetReader;
use crate::fastfield::FastFieldReaders;
use std::collections::HashMap;
use std::sync::{Arc, RwLock};
use std::{fmt, io};
use fail::fail_point;
use crate::core::{InvertedIndexReader, Segment, SegmentComponent, SegmentId};
use crate::directory::{CompositeFile, FileSlice};
use crate::error::DataCorruption;
use crate::fastfield::{intersect_alive_bitsets, AliveBitSet, FacetReader, FastFieldReaders};
use crate::fieldnorm::{FieldNormReader, FieldNormReaders};
use crate::schema::FieldType;
use crate::schema::Schema;
use crate::schema::{Field, IndexRecordOption};
use crate::schema::{Field, FieldType, IndexRecordOption, Schema};
use crate::space_usage::SegmentSpaceUsage;
use crate::store::StoreReader;
use crate::termdict::TermDictionary;
use crate::DocId;
use crate::{common::CompositeFile, error::DataCorruption};
use fail::fail_point;
use std::fmt;
use std::sync::Arc;
use std::sync::RwLock;
use std::{collections::HashMap, io};
use crate::{DocId, Opstamp};
/// Entry point to access all of the datastructures of the `Segment`
///
@@ -36,6 +30,8 @@ pub struct SegmentReader {
inv_idx_reader_cache: Arc<RwLock<HashMap<Field, Arc<InvertedIndexReader>>>>,
segment_id: SegmentId,
delete_opstamp: Option<Opstamp>,
max_doc: DocId,
num_docs: DocId,
@@ -46,7 +42,7 @@ pub struct SegmentReader {
fieldnorm_readers: FieldNormReaders,
store_file: FileSlice,
delete_bitset_opt: Option<DeleteBitSet>,
alive_bitset_opt: Option<AliveBitSet>,
schema: Schema,
}
@@ -71,14 +67,12 @@ 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.
/// Returns true if 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.
@@ -100,7 +94,7 @@ impl SegmentReader {
let field_entry = self.schema.get_field_entry(field);
match field_entry.field_type() {
FieldType::HierarchicalFacet(_) => {
FieldType::Facet(_) => {
let term_ords_reader = self.fast_fields().u64s(field)?;
let termdict = self
.termdict_composite
@@ -127,13 +121,17 @@ impl SegmentReader {
self.fieldnorm_readers.get_field(field)?.ok_or_else(|| {
let field_name = self.schema.get_field_name(field);
let err_msg = format!(
"Field norm not found for field {:?}. Was it marked as indexed during indexing?",
field_name
"Field norm not found for field {field_name:?}. Was the field set to record norm \
during indexing?"
);
crate::TantivyError::SchemaError(err_msg)
})
}
pub(crate) fn fieldnorms_readers(&self) -> &FieldNormReaders {
&self.fieldnorm_readers
}
/// Accessor to the segment's `StoreReader`.
pub fn get_store_reader(&self) -> io::Result<StoreReader> {
StoreReader::open(self.store_file.clone())
@@ -141,6 +139,14 @@ 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)?;
@@ -163,31 +169,39 @@ impl SegmentReader {
let fast_fields_data = segment.open_read(SegmentComponent::FastFields)?;
let fast_fields_composite = CompositeFile::open(&fast_fields_data)?;
let fast_field_readers =
let fast_fields_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,
fast_fields_readers,
fieldnorm_readers,
segment_id: segment.id(),
delete_opstamp: segment.meta().delete_opstamp(),
store_file,
delete_bitset_opt,
alive_bitset_opt,
positions_composite,
schema,
})
@@ -236,19 +250,24 @@ impl SegmentReader {
let record_option = record_option_opt.unwrap();
let postings_file = postings_file_opt.unwrap();
let termdict_file: FileSlice = self.termdict_composite.open_read(field)
.ok_or_else(||
DataCorruption::comment_only(format!("Failed to open field {:?}'s term dictionary in the composite file. Has the schema been modified?", field_entry.name()))
)?;
let positions_file = self
.positions_composite
.open_read(field)
.ok_or_else(|| {
let error_msg = format!("Failed to open field {:?}'s positions in the composite file. Has the schema been modified?", field_entry.name());
DataCorruption::comment_only(error_msg)
let termdict_file: FileSlice =
self.termdict_composite.open_read(field).ok_or_else(|| {
DataCorruption::comment_only(format!(
"Failed to open field {:?}'s term dictionary in the composite file. Has the \
schema been modified?",
field_entry.name()
))
})?;
let positions_file = self.positions_composite.open_read(field).ok_or_else(|| {
let error_msg = format!(
"Failed to open field {:?}'s positions in the composite file. Has the schema been \
modified?",
field_entry.name()
);
DataCorruption::comment_only(error_msg)
})?;
let inv_idx_reader = Arc::new(InvertedIndexReader::new(
TermDictionary::open(termdict_file)?,
postings_file,
@@ -271,23 +290,32 @@ impl SegmentReader {
self.segment_id
}
/// Returns the bitset representing
/// the documents that have been deleted.
pub fn delete_bitset(&self) -> Option<&DeleteBitSet> {
self.delete_bitset_opt.as_ref()
/// Returns the delete opstamp
pub fn delete_opstamp(&self) -> Option<Opstamp> {
self.delete_opstamp
}
/// Returns true iff the `doc` is marked
/// Returns the bitset representing
/// the documents that have been deleted.
pub fn alive_bitset(&self) -> Option<&AliveBitSet> {
self.alive_bitset_opt.as_ref()
}
/// Returns true if 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.
@@ -300,14 +328,29 @@ 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)
@@ -330,10 +373,10 @@ 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"));
@@ -356,10 +399,10 @@ 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()?;
}

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