rust/tantivy
1
0
Fork 0
mirror of https://github.com/quickwit-oss/tantivy.git synced 2026-06-17 16:00:42 +00:00
Code Issues Packages Projects Releases Wiki Activity

Compare commits

base: rust:storage_abstraction
Branches Tags
rust:main rust:seek_danger_range_doc_set rust:moshiki-ready rust:faster_union rust:dependabot/github_actions/github/codeql-action-4.36.2 rust:postings-writer-enum-code-june2026 rust:bench/alyze-vs-unicode-seg-tokenizer rust:larger-collect-block-buffer rust:paul.masurel/upstream-datafusion rust:PSeitz-patch-1 rust:dependabot/cargo/fs4-1.1.0 rust:dependabot/cargo/lru-0.18.0 rust:paul.masurel/bench-nested rust:paul.masurel/2892-nested-term-agg-opt rust:paul.masurel/nested-aggregation-slow rust:dependabot/cargo/rand_distr-0.6 rust:storage_abstraction rust:bench_0.25_compare rust:dependabot/cargo/rand-0.10 rust:dependabot/cargo/oneshot-0.2.1 rust:composite-agg-rebase rust:composite-agg rust:postings-writer-enum-codec5 rust:congxie/supportAgg rust:postings-writer-enum-codec4 rust:segmentreader_trait rust:stuhood.intersection-termination-upstream rust:postings-writer-enum-codec3 rust:postings-writer-enum-codec2 rust:postings-writer-enum rust:paul.masurel/skip-first-doc-on-intersection rust:stuhood.lazy-scorers-blocks-upstream rust:paul.masurel/seek_into_the_danger_zone_result rust:paul.masurel/search_into_danger_zone_fastfield rust:stuhood.lazy-scorers-blocks rust:flatheadmill-geo rust:paradedb-paradedb/fix-overflow-issue rust:low_card_optimization rust:paul.masurel/clippy rust:qw-airmail-20250916 rust:unit-test-aggregation-sum-other rust:clippy-and-cleanups rust:trinity.pointard/opti-sstable-sorted-ords-to-term rust:release_0.24 rust:tagged-user-input-ast rust:1686a/parse-agg-res rust:issue/2411-adding-panic-handler-to-rayon-merge-thread-pool rust:dependabot/cargo/nom-8 rust:upgrade_deps rust:test_parser 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:0.26.1 rust:0.26.0 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:seek_danger_range_doc_set
Branches Tags
rust:main rust:seek_danger_range_doc_set rust:moshiki-ready rust:faster_union rust:dependabot/github_actions/github/codeql-action-4.36.2 rust:postings-writer-enum-code-june2026 rust:bench/alyze-vs-unicode-seg-tokenizer rust:larger-collect-block-buffer rust:paul.masurel/upstream-datafusion rust:PSeitz-patch-1 rust:dependabot/cargo/fs4-1.1.0 rust:dependabot/cargo/lru-0.18.0 rust:paul.masurel/bench-nested rust:paul.masurel/2892-nested-term-agg-opt rust:paul.masurel/nested-aggregation-slow rust:dependabot/cargo/rand_distr-0.6 rust:storage_abstraction rust:bench_0.25_compare rust:dependabot/cargo/rand-0.10 rust:dependabot/cargo/oneshot-0.2.1 rust:composite-agg-rebase rust:composite-agg rust:postings-writer-enum-codec5 rust:congxie/supportAgg rust:postings-writer-enum-codec4 rust:segmentreader_trait rust:stuhood.intersection-termination-upstream rust:postings-writer-enum-codec3 rust:postings-writer-enum-codec2 rust:postings-writer-enum rust:paul.masurel/skip-first-doc-on-intersection rust:stuhood.lazy-scorers-blocks-upstream rust:paul.masurel/seek_into_the_danger_zone_result rust:paul.masurel/search_into_danger_zone_fastfield rust:stuhood.lazy-scorers-blocks rust:flatheadmill-geo rust:paradedb-paradedb/fix-overflow-issue rust:low_card_optimization rust:paul.masurel/clippy rust:qw-airmail-20250916 rust:unit-test-aggregation-sum-other rust:clippy-and-cleanups rust:trinity.pointard/opti-sstable-sorted-ords-to-term rust:release_0.24 rust:tagged-user-input-ast rust:1686a/parse-agg-res rust:issue/2411-adding-panic-handler-to-rayon-merge-thread-pool rust:dependabot/cargo/nom-8 rust:upgrade_deps rust:test_parser 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:0.26.1 rust:0.26.0 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

92 Commits
storage_ab ... seek_dange

Author SHA1 Message Date
Pascal Seitz
7577a0bdda Speed up range-query intersections via seek_danger on RangeDocSet (up to ~50x faster) 
A regular seek on RangeDocSet is costly: on a miss it fetches blocks and
scans the column forward to materialize the next matching doc. As a
non-leading docset in an intersection that work is wasted — the driver only
asks "does this candidate match?". seek_danger answers that with a cheap
point lookup via Column::values_for_doc, returning a lower bound on a miss
and leaving forward progress to the caller.

Forward seek_danger through ConstScorer.

Benchmarks (bool_queries_with_range, _all_results / DocSetCollector):

```
dense and 0.1% a
a_AND_num_rand:[0_TO_9]_all_results                                 Avg: 0.0827ms (-4.60%)     Median: 0.0825ms (-4.82%)     [0.0809ms .. 0.0891ms]    Output: 43
a_AND_num_asc:[0_TO_9]_all_results                                  Avg: 0.1937ms (-3.70%)     Median: 0.1930ms (-3.59%)     [0.1806ms .. 0.2044ms]    Output: 100
a_AND_num_rand_fast:[0_TO_9]_all_results                            Avg: 0.0367ms (-92.67%)    Median: 0.0365ms (-92.65%)    [0.0340ms .. 0.0398ms]    Output: 43
a_AND_num_asc_fast:[0_TO_9]_all_results                             Avg: 0.1052ms (-98.05%)    Median: 0.1050ms (-97.98%)    [0.1009ms .. 0.1117ms]    Output: 100
num_rand_fast:[0_TO_9]_AND_num_asc_fast:[0_TO_9]_all_results        Avg: 2.7147ms (-51.42%)    Median: 2.7075ms (-49.58%)    [2.6806ms .. 2.7799ms]    Output: 968
dense and 1% a
a_AND_num_rand:[0_TO_9]_all_results                                 Avg: 0.4373ms (-9.71%)     Median: 0.4357ms (-10.12%)    [0.4117ms .. 0.4711ms]    Output: 463
a_AND_num_asc:[0_TO_9]_all_results                                  Avg: 0.2342ms (-2.50%)     Median: 0.2338ms (-2.56%)     [0.2247ms .. 0.2452ms]    Output: 1_054
a_AND_num_rand_fast:[0_TO_9]_all_results                            Avg: 0.3956ms (-82.86%)    Median: 0.3943ms (-82.90%)    [0.3815ms .. 0.4119ms]    Output: 463
a_AND_num_asc_fast:[0_TO_9]_all_results                             Avg: 0.4896ms (-91.16%)    Median: 0.4862ms (-90.81%)    [0.4797ms .. 0.5084ms]    Output: 1_054
num_rand_fast:[0_TO_9]_AND_num_asc_fast:[0_TO_9]_all_results        Avg: 2.7108ms (-50.81%)    Median: 2.6925ms (-49.51%)    [2.6688ms .. 2.7868ms]    Output: 968
dense and 10% a
a_AND_num_rand:[0_TO_9]_all_results                                 Avg: 0.9869ms (-3.71%)     Median: 0.9833ms (-3.83%)     [0.9518ms .. 1.1218ms]    Output: 4_914
a_AND_num_asc:[0_TO_9]_all_results                                  Avg: 0.6352ms (-3.74%)     Median: 0.6363ms (-3.32%)     [0.6158ms .. 0.6488ms]    Output: 10_152
a_AND_num_rand_fast:[0_TO_9]_all_results                            Avg: 3.1264ms (+0.39%)     Median: 3.1466ms (+1.34%)     [3.0261ms .. 3.2051ms]    Output: 4_914
a_AND_num_asc_fast:[0_TO_9]_all_results                             Avg: 4.1547ms (-31.12%)    Median: 4.0933ms (-28.55%)    [3.7648ms .. 4.7600ms]    Output: 10_152
num_rand_fast:[0_TO_9]_AND_num_asc_fast:[0_TO_9]_all_results        Avg: 2.6973ms (-52.30%)    Median: 2.6901ms (-49.86%)    [2.6689ms .. 2.7677ms]    Output: 968
```

Gains are largest when the range query is the non-leading docset of a low-cardinality intersection.
 2026-06-17 12:03:45 +02:00
Pascal Seitz
c096b2ad89 aggregation/terms: charge fused term_counts to the memory limit 
term_counts (one u32/term) was allocated but not charged to
AggregationLimitsGuard, so a memory limit could be exceeded silently.
Charge it, skip allocating it when unbounded, and add a regression test.
 2026-06-16 21:23:23 +08:00
Pascal Seitz
ac7a3d347c add comment, hoist variables 2026-06-16 21:23:23 +08:00
Pascal Seitz
03520a0719 add top level comment 2026-06-16 21:23:23 +08:00
Pascal Seitz
86a4c47bed merge loops, histo with bounds may benefit from single vec opt 2026-06-16 21:23:23 +08:00
Pascal Seitz
fb23e8908f add histogram with bounds 2026-06-16 21:23:23 +08:00
Pascal Seitz
3ca510dff0 aggregation/terms: tidy fused term×histogram grid construction 
Rename the value threaded through build_segment_term_collector and
maybe_build_collector from max_term_id to col_max_val/max_column_val — it
is the column's max value, only later reused as the max term id. Make the
grid-size arithmetic overflow-/zero-safe (saturating_add, checked_div).
 2026-06-16 21:23:23 +08:00
Pascal Seitz
3cb400c300 clarify counts/term_counts field docs 
Spell out that `counts` is the flattened per-term × time-bucket grid (each
term's own contiguous slice) and that `term_counts` is only needed when the
per-term total can't be derived from that grid (i.e. with hard bounds).
 2026-06-16 21:23:23 +08:00
Pascal Seitz
ef13489d63 skip hard_bounds that can't exclude any value 
When a histogram's hard_bounds are wider than the column's value range, the
per-doc `bounds.contains` check can never fail. Collapse such bounds to the
unbounded sentinel in `normalize_histogram_req`, so both the general histogram
hot loop and the fused term×histogram path skip the check — the latter then
derives per-term counts from the grid (the ~17% win) instead of falling back to
per-doc counting just because `bounds != [MIN, MAX]`.

Only the collect-time filter is affected: empty-bucket emission reads
`req.hard_bounds` directly, and hard_bounds only ever clips that range, so a
wider-than-data bound leaves results unchanged. Covered by new tests on the
general and fused paths, including mid-interval (bucket-splitting) bounds.

Also tighten the fused-path u32-overflow guard to bound on `num_vals()` (the
per-value increment count) rather than `num_docs()`, and document why the fused
collector's hot-loop fields are hoisted into locals (re-reading them from memory
each iteration measured ~15% slower).
 2026-06-16 21:23:23 +08:00
Pascal Seitz
9f7aea4765 derive term counts 2026-06-16 21:23:23 +08:00
Pascal Seitz
2c8536ab11 add specialized TermHistogram 2026-06-16 21:23:23 +08:00
Pascal Seitz
05f4c02ac5 add dense histogram, optional sub-buckets 2026-06-16 21:23:23 +08:00
Pascal Seitz
d137779219 add no sub-gg fastpath 2026-06-16 21:23:23 +08:00
Pascal Seitz
8f9846ac80 use get_range when possible 2026-06-16 21:23:23 +08:00
Pascal Seitz
52e24a9757 add status -> date histogram bench 2026-06-16 21:23:23 +08:00
trinity-1686a
00714326af Merge pull request #2960 from Darkheir/fix/query_grammar_boost_and_escape 
fix(query-grammar): Fix issues on boosted and regex queries
 2026-06-16 12:03:23 +02:00
Mohammad Dashti
799f7b4646 Built SUM final result in each branch directly. 
Keeps the empty-bucket coercion visible at the boundary instead of a
shared binding, following the reviewer's suggested shape.
 2026-06-16 03:10:30 +08:00
Mohammad Dashti
fc88d80726 docs: drop downstream-specific name from none_if_no_match doc 
The flag's purpose is described well enough by "SQL-style consumers";
no need to call out a specific downstream.
 2026-06-16 03:10:30 +08:00
Mohammad Dashti
6a684e7c38 feat: opt-in none_if_no_match flag on SumAggregation for SQL-style null 
Switch the default serialized output of `sum` on empty / all-missing
buckets back to `"value": 0` to match Elasticsearch, and gate the
SQL-style `"value": null` behavior behind a new
`none_if_no_match: Option<bool>` flag on `SumAggregation`.

`IntermediateSum::finalize` still returns `Option<f64>` internally so
the Rust API stays parallel to min/max/avg, but the ES-vs-SQL choice is
made at the boundary in `IntermediateMetricResult::into_final_metric_result`:
`None` is coerced to `Some(0.0)` unless `none_if_no_match` is set on the
aggregation request.

Adds `AggregationVariants::as_sum()` accessor for that boundary check
and two end-to-end tests covering both the default ES behavior and the
opt-in null behavior on an empty index.
 2026-06-16 03:10:30 +08:00
Mohammad Dashti
94fe52cc67 docs: clarify SUM finalize returning None diverges from Elasticsearch 
Surface the trade-off in the doc comment so future reviewers see why
this differs from ES (which returns "value": 0 for sum over
empty/all-missing buckets) and what consumers (ParadeDB SQL NULL) the
None variant is meant to serve.
 2026-06-16 03:10:30 +08:00
Mohammad Dashti
2ff39f6f7f fix: return None from SUM when no values were collected 
IntermediateSum::finalize() returned Some(0.0) even when count==0
(all documents had missing/NULL values). This differs from MIN, MAX,
and AVG which all return None for count==0.

The 0.0 came from IntermediateStats' default sum initialization.
Consumers (like ParadeDB) that map None to SQL NULL were incorrectly
getting 0 for SUM on all-NULL groups.

Fixes paradedb/paradedb#4621
 2026-06-16 03:10:30 +08:00
Windforce17
1d06328cb3 Add BlockSegmentPostings::rank() for skip-list-based positional counting 
Add a public rank(target) method on BlockSegmentPostings that returns the
number of docs with a doc id strictly smaller than target. It jumps to the
candidate block through the skip list and decodes a single block, so the cost
is O(skip-list entries) + one block decode rather than O(doc_freq).

This is a useful primitive for range counting over a posting list (e.g. number
of matches in a [lo, hi) doc-id window) without iterating every matched doc.

To support it, expose SkipReader::remaining_docs() (pub(crate)). Like seek(),
rank() advances the cursor forward only and must be called with non-decreasing,
valid (<= TERMINATED) targets. Adds a unit test covering multi-block lists and
the below-first / above-last / empty edge cases.
 2026-06-15 18:56:49 +08:00
Darkheir
7fd1dbe9f5 fix(query-grammar): Fix issues on boosted and regex queries 
Signed-off-by: Darkheir <raphael.cohen@sekoia.io>
 2026-06-15 10:50:07 +02:00
Pascal Seitz
b19f0ddc77 fix clippy 2026-06-09 23:14:12 +08:00
Pascal Seitz
b4acfcf881 cleanup AggregationsSegmentCtx 
The metric/cardinality/histogram _mut getters had no callers needing
mutation; their two uses already pass the resulting reference as &T.

simplify req_data ownership: clone into collectors, Rc only for filter BitSet

Replace Vec<Option<Box<T>>> + take/put-back round-trip with Vec<T> +
direct clone into collector. Collectors now own their per-segment
request data outright, removing the borrow-checker dance that the
take/put-back pattern existed to satisfy.

The structural clones are cheap (Column<u64> is Arc-internal) except
for the filter aggregation, whose DocumentQueryEvaluator carries a
precomputed per-segment BitSet sized by max_doc. Wrap that in
Rc<DocumentQueryEvaluator> so FilterAggReqData::clone() bumps a
refcount instead of duplicating the BitSet. Move SegmentFilterCollector's
matching_docs_buffer out of FilterAggReqData so its pre-allocated
capacity is preserved per collector instead of being lost on every clone.
 2026-06-09 23:14:12 +08:00
dependabot[bot]
3a8240b123 Bump codecov/codecov-action from 6.0.0 to 7.0.0 
Bumps [codecov/codecov-action](https://github.com/codecov/codecov-action) from 6.0.0 to 7.0.0.
- [Release notes](https://github.com/codecov/codecov-action/releases)
- [Changelog](https://github.com/codecov/codecov-action/blob/main/CHANGELOG.md)
- [Commits](57e3a136b7...fb8b3582c8)

---
updated-dependencies:
- dependency-name: codecov/codecov-action
  dependency-version: 7.0.0
  dependency-type: direct:production
  update-type: version-update:semver-major
...

Signed-off-by: dependabot[bot] <support@github.com>
 2026-06-09 14:48:17 +08:00
dependabot[bot]
fd9713e1ca Bump actions/checkout from 6.0.2 to 6.0.3 (#2949) 
Bumps [actions/checkout](https://github.com/actions/checkout) from 6.0.2 to 6.0.3.
- [Release notes](https://github.com/actions/checkout/releases)
- [Changelog](https://github.com/actions/checkout/blob/main/CHANGELOG.md)
- [Commits](de0fac2e45...df4cb1c069)

---
updated-dependencies:
- dependency-name: actions/checkout
  dependency-version: 6.0.3
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2026-06-08 10:55:54 +02:00
dependabot[bot]
96f3784f79 Bump github/codeql-action from 4.35.2 to 4.36.1 (#2948) 
Bumps [github/codeql-action](https://github.com/github/codeql-action) from 4.35.2 to 4.36.1.
- [Release notes](https://github.com/github/codeql-action/releases)
- [Changelog](https://github.com/github/codeql-action/blob/main/CHANGELOG.md)
- [Commits](95e58e9a2c...87557b9c84)

---
updated-dependencies:
- dependency-name: github/codeql-action
  dependency-version: 4.36.1
  dependency-type: direct:production
  update-type: version-update:semver-minor
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2026-06-08 10:49:04 +02:00
dependabot[bot]
87a6679a79 Bump actions/upload-artifact from 7.0.0 to 7.0.1 (#2917) 
Bumps [actions/upload-artifact](https://github.com/actions/upload-artifact) from 7.0.0 to 7.0.1.
- [Release notes](https://github.com/actions/upload-artifact/releases)
- [Commits](bbbca2ddaa...043fb46d1a)

---
updated-dependencies:
- dependency-name: actions/upload-artifact
  dependency-version: 7.0.1
  dependency-type: direct:production
  update-type: version-update:semver-patch
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2026-06-08 10:48:48 +02:00
dependabot[bot]
864a6aa72c Update murmurhash32 requirement from 0.3 to 0.4 (#2894) 
Updates the requirements on [murmurhash32](https://github.com/quickwit-inc/murmurhash32) to permit the latest version.
- [Commits](https://github.com/quickwit-inc/murmurhash32/commits)

---
updated-dependencies:
- dependency-name: murmurhash32
  dependency-version: 0.4.0
  dependency-type: direct:production
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2026-06-08 10:48:32 +02:00
Paul Masurel
abcf6754a2 CR comments from https://github.com/quickwit-oss/tantivy/pull/2940 (#2952) 
Co-authored-by: Paul Masurel <paul.masurel@datadoghq.com>
 2026-06-08 10:47:58 +02:00
Kanishk Sachan
70a8e56ee5 test(postings): add unit tests for TermFrequencyRecorder 
Closes #2285

The TermFrequencyRecorder was completely untested. Add five focused tests:

- term_frequency_recorder_has_term_freq: verifies the recorder
  correctly advertises term-frequency support via has_term_freq()
- term_frequency_recorder_zero_docs: term_doc_freq() returns Some(0)
  before any documents are recorded
- term_frequency_recorder_term_doc_freq_single_doc: one document with
  two occurrences yields term_doc_freq() == Some(1)
- term_frequency_recorder_term_doc_freq_multiple_docs: three documents
  with varying term frequencies yield term_doc_freq() == Some(3),
  confirming the count tracks documents, not occurrences
- term_frequency_recorder_single_occurrence_per_doc: each of three
  documents has exactly one occurrence
- term_frequency_recorder_high_frequency_doc: a single document with
  1000 occurrences still yields term_doc_freq() == Some(1)
 2026-06-06 14:44:51 +08:00
Paul Masurel
62705526e8 Add sve + neon filter vec implementation as spotted by Adam (#2940) 
* Add filter_vec benchmarks (dense, sparse, full coverage)

Uses get_ids_for_value_range to exercise both the bitpacking decode and
the filter_vec SIMD path together under realistic cache conditions.

* Add NEON and SVE implementations for filter_vec

Adds aarch64-specific SIMD paths (NEON always available on aarch64;
SVE gated on nightly + non-Apple target) with routing logic in mod.rs
that selects the best available instruction set at runtime.

* Using asm! to workaround the lack of stabilized SVE intrinsics

* showing instruction set

* improved proptesting

* removing build.rs

---------

Co-authored-by: Paul Masurel <paul.masurel@datadoghq.com>
 2026-06-04 17:51:26 +02:00
Paul Masurel
a27c64998f Cargo clippy fix (#2943) 
Co-authored-by: Paul Masurel <paul.masurel@datadoghq.com>
 2026-06-01 14:39:44 +02:00
Paul Masurel
46b3fb9ed3 Relying on upstream version of datasketch and stop using HLL 4. (#2936) 
We were relying on a fork for:

a bugfix in LIST serialization
a better API exposing a new Coupon type, required for caching coupons.
We also stop using HLL8 in hope to fix
https://datadoghq.atlassian.net/browse/CLOUDPREM-625

Co-authored-by: Paul Masurel <paul.masurel@datadoghq.com>
 2026-05-19 13:29:35 +02:00
trinity-1686a
fbe620b9b4 Merge pull request #2933 from quickwit-oss/1686a/sstable-opt 
optimise sstable index access pattern
 2026-05-19 11:43:17 +02:00
trinity-1686a
95d8a3989a cr 2026-05-19 11:38:48 +02:00
trinity-1686a
ea61a68db4 skip sstable index binary search when ordinal is in same block 2026-05-16 11:35:38 +02:00
trinity-1686a
c367df37c1 refactor sstable index 2026-05-16 11:30:02 +02:00
Mohammad Dashti
d99a5d4e91 Rename validate_aggregation_fields to validate_aggregation_fields_exist 
Applies @PSeitz's review suggestion to make the function name more
descriptive of what it checks. Also adds a doc note clarifying why
validation is opt-in rather than enforced by default.
 2026-05-16 15:45:20 +08:00
Mohammad Dashti
2de6f075ce Fixed the example 2026-05-16 15:45:20 +08:00
Mohammad Dashti
18080067c7 Applied PR comment: 
I would move it outside of the aggregation. You can fetch the fields from the aggregation request and do a validation in a helper function
 2026-05-16 15:45:20 +08:00
Mohammad Dashti
95db7d2e5c Revert "Revert all impl." 
This reverts commit d5e0991549a05bf80f19f853f7689ad69f96e7e5.
 2026-05-16 15:45:20 +08:00
Mohammad Dashti
fc017c4c74 Applied PR comments. 2026-05-16 15:45:20 +08:00
Mohammad Dashti
141c91d028 Added a flag: strict_validation 2026-05-16 15:45:20 +08:00
Mohammad Dashti
36a83e7c1a Fixed agg validation 2026-05-16 15:45:20 +08:00
jinhelin
be11f8a6a1 Fix opening positions file error 2026-05-14 15:55:59 +08:00
dependabot[bot]
4305e4029e Update binggan requirement from 0.16.1 to 0.17.0 
Updates the requirements on [binggan](https://github.com/pseitz/binggan) to permit the latest version.
- [Changelog](https://github.com/PSeitz/binggan/blob/main/CHANGELOG.md)
- [Commits](https://github.com/pseitz/binggan/commits)

---
updated-dependencies:
- dependency-name: binggan
  dependency-version: 0.17.0
  dependency-type: direct:production
...

Signed-off-by: dependabot[bot] <support@github.com>
 2026-05-12 15:10:20 +08:00
Pascal Seitz
edfb02b47e switch to enum, fix mixed types for cardinality agg 2026-05-05 16:39:51 +08:00
Pascal Seitz
d0fad88bac use bitsets for card agg 2026-05-05 16:39:51 +08:00
Pascal Seitz
351280c0b4 add card bench for high card 2026-05-05 16:39:51 +08:00
James Sewell
4480cf0a98 Enable BMW for single-scorer boolean queries by removing early return in scorer_union (#2915) 
The early return for `scorers.len() == 1` in `scorer_union` short-circuits a single TermScorer into `SpecializedScorer::Other`, bypassing the `TermUnion` path that enables block-max WAND (BMW) in `for_each_pruning`.

This was originally addressed in PR #2898 (backed out), which added a special case in `BooleanWeight::for_each_pruning`. PR #2912 (merged as d27ca164a) added a single-scorer fast path inside `block_wand` itself, but did not remove this early return — so a single SHOULD TermScorer still never reaches the BMW path.

Removing the early return lets a single TermScorer with freq reading flow through to `SpecializedScorer::TermUnion`, where `block_wand` → `block_wand_single_scorer` handles it efficiently.
 2026-04-28 14:49:53 -07:00
Pascal Seitz
d47abdf104 early cut off for order by sub agg in term agg 2026-04-28 16:59:59 +02:00
Pascal Seitz
c11952eb7c add order by agg benchmark 2026-04-28 16:59:59 +02:00
trinity-1686a
09667ee9c8 Merge pull request #2909 from osyniakov/claude/add-ossf-scorecard-1z6Vn 
Add OpenSSF Scorecard workflow
 2026-04-28 11:57:36 +02:00
trinity-1686a
333ccf5300 Merge pull request #2896 from osyniakov/claude/fix-issues-5945-5937-eQm1Q 
ci: pin GitHub Actions to full commit SHAs and restrict token permissions
 2026-04-28 11:57:18 +02:00
Oleksii Syniakov
60a39a4689 Merge branch 'main' into claude/fix-issues-5945-5937-eQm1Q 2026-04-28 10:28:23 +02:00
Oleksii Syniakov
f8f3e4277f remove not neeeded permissions for the public repo 2026-04-28 10:09:30 +02:00
Oleksii Syniakov
ff1433713a bump upload-sarif -> 4.35.2 
Co-authored-by: trinity-1686a <trinity.pointard@gmail.com>
 2026-04-28 10:07:45 +02:00
trinity-1686a
ca139d8eb1 Merge pull request #2910 from quickwit-oss/abdul.andha/composite-agg-after 
Composite aggregations: send after key on last page
 2026-04-27 23:38:52 +02:00
Abdul Andha
ac508108aa address pr comment 2026-04-27 12:39:38 -04:00
Paul Masurel
63da5a21b2 Optimizing top K using Adrien Grand's ideas (#2865) 
* Optimizing top K using Adrien Grand's ideas

https://jpountz.github.io/2025/08/28/compiled-vs-vectorized-search-engine-edition.html

* Suffix-sum pruning for multi-term intersection candidates

After scoring each secondary in Phase 2, check whether remaining
secondaries' block_max scores can still beat the threshold. Skip
to the next candidate early if impossible, avoiding expensive seeks
into later secondaries.

Improves three-term intersection by ~8% on the balanced benchmark
while keeping two-term performance neutral.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

* Claude CR comment

* Removed 16 term scorer limit.

---------

Co-authored-by: Paul Masurel <paul.masurel@datadoghq.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-26 12:14:40 +02:00
lif
54cd5bba98 fix: skip sentinel facet ords in harvest to prevent wrong root (#2867) 
When a document has the exact registered facet path (not a child),
compute_collapse_mapping_one maps it to a sentinel (u64::MAX, 0).
Without filtering, harvest() passes u64::MAX to ord_to_term which
resolves to the last dictionary entry, producing a spurious facet
from an unrelated branch.

Skip entries where facet_ord == u64::MAX in harvest().

Closes #2494

Signed-off-by: majiayu000 <1835304752@qq.com>
 2026-04-25 22:23:30 +02:00
Paul Masurel
d27ca164a9 block_wand: use single-scorer path when there is only one scorer 2026-04-25 16:35:00 +02:00
dependabot[bot]
2f5a48e8b1 Update criterion requirement from 0.5 to 0.8 (#2873) 
Updates the requirements on [criterion](https://github.com/criterion-rs/criterion.rs) to permit the latest version.
- [Release notes](https://github.com/criterion-rs/criterion.rs/releases)
- [Changelog](https://github.com/criterion-rs/criterion.rs/blob/master/CHANGELOG.md)
- [Commits](https://github.com/criterion-rs/criterion.rs/compare/0.5.0...criterion-v0.8.2)

---
updated-dependencies:
- dependency-name: criterion
  dependency-version: 0.8.2
  dependency-type: direct:production
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2026-04-25 14:15:53 +02:00
dependabot[bot]
ae0ab907fe Bump actions/checkout from 4 to 6 (#2875) 
Bumps [actions/checkout](https://github.com/actions/checkout) from 4 to 6.
- [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/v4...v6)

---
updated-dependencies:
- dependency-name: actions/checkout
  dependency-version: '6'
  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>
 2026-04-25 14:15:27 +02:00
dependabot[bot]
7d62e084e7 Bump codecov/codecov-action from 3 to 6 (#2876) 
Bumps [codecov/codecov-action](https://github.com/codecov/codecov-action) from 3 to 6.
- [Release notes](https://github.com/codecov/codecov-action/releases)
- [Changelog](https://github.com/codecov/codecov-action/blob/main/CHANGELOG.md)
- [Commits](https://github.com/codecov/codecov-action/compare/v3...v6)

---
updated-dependencies:
- dependency-name: codecov/codecov-action
  dependency-version: '6'
  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>
 2026-04-25 14:14:54 +02:00
James Sewell
322286ee16 Tighen Block-Max in single-scorer (#2897) 
In the Block-Max WAND single-scorer, it uses block_max_score() < threshold,
whereas the multi-term one uses  block_max_score_upperbound <= threshold.

As both of these are guarded later on with if score > threshold we can
use the more efficent form in single-scorer.

Single-scorer block skip (<, should be <=): https://github.com/quickwit-oss/tantivy/blob/main/src/query/boolean_query/block_wand.rs#L231
Multi-scorer block skip (already <=): https://github.com/quickwit-oss/tantivy/blob/main/src/query/boolean_query/block_wand.rs#L179
Single-scorer per-doc guard (>): https://github.com/quickwit-oss/tantivy/blob/main/src/query/boolean_query/block_wand.rs#L246
Multi-scorer per-doc guard (>): https://github.com/quickwit-oss/tantivy/blob/main/src/query/boolean_query/block_wand.rs#L206

This will improve performance when there are many identical scores.
 2026-04-25 14:13:07 +02:00
RJ Barman
73ad18fa1e fix: Add space for missing sentinel in allowed bitset when a missing key is provided (#119) (#2907) 
## Bug Overview
Under certain conditions, a `terms` aggregation request can cause a
bounds-check panic. Those conditions are:
- The queried field must be a text field
- There must be a segment where the number of distinct terms in it's
dictionary for the queried field is divisible by 64 (i.e.e where
`count(term_dict.keys) % 64 == 0`)
- That same segment must contain at least one document that does not
contain this field.
- The request contain a `missing` key that is a string.
- The request must contain an `include` or `exclude` filter.
For example:
```json
{
    "my_bool": {
        "terms": {
            "field": "title",
            "include": "foo",
            "missing": "__NULL__",
        }
    }
}
```
Check out the added tests in `src/aggregation/bucket/term_agg.rs` to see
this in action

## How the bug happens
### Preparation
While preparing the aggregation nodes:
1) When we've provided a `missing` key, we derive a missing sentinel.
For string keys this column's max value (which for string keys is always
the number of terms in this segment) + 1.
2) for string columns only, we optionally prep an "allowed" `BitSet` for
allowed term ids. (`build_allowed_term_ids_for_str` in
`src/aggregation/agg_data.rs`)
- If no `include` or `exclude` filter is provided, this just returns
`None`, causing this check to be skipped down the line
- Otherwise the bitset is initialized to be able to hold the exact
number of terms in the segments term dictionary, and the bits are set to
signify which terms are to be included in the results.

### Collection
If we have an "allowed" `BitSet`, filter documents against that. For
each document, we check if the `BitSet` contains the documents term id.
For documents without the field, this is the missing sentinel we derived
earlier, minus 1 (to account for zero-based indexing): `(num_terms + 1)
- 1`.However, the `BitSet`s size is only `num_terms`. Normally, this
slips by without a problem, but if `num_terms % 64 == 0`, this will
cause a panic.

### Why `BitSet` panics
`BitSet` is represented under the hood by a boxed slice of `u64`s. When
you go to check a bit using `BitSet::contains`, it must determine which
of those `u64`s the bit is in, and then the position within that `u64`
of the bit.

In cases where the number of terms is not divisible by 64, the `BitSet`
must waste some bits. When we then look up the missing sentinel's bit,
it happens to be one of those wasted bits, for which `BitSet` is happy
to return the value of. For example, if the number of terms was 63:
```rust
let bitset_init_size = 63; // so BitSet's boxed slice has a length of 1, capable of holding 64 bits, term id [0, 62]
let missing_sentinel = 63; // num_terms + 1 - 1;
let byte_pos = missing_sentinel / 64; // 0 - within the valid slice
let bit_pos = missing_sentinel % 64; // 63 - hits the 1 wasted bit
```

But if the number of terms is indeed divisible by 64, then the `BitSet`
is perfectly aligned to the byte boundary:
```rust
let bitset_init_size = 64; // so BitSet's boxed slice has a length of 1, capable of holding 64 bits, term ids [0, 63]
let missing_sentinel = 64; // num_terms + 1 - 1, 
let byte_pos = missing_sentinel / 64; // 1 - idx 1 >= slice length 1
let bit_pos = missing_sentinel % 64; // 0 
```
We try to access a byte outside of the bounds of the boxed slice,
causing a panic from the bounds check to failing.

## Fixing it
The fix is simple. If we need to account for the missing sentinel,
initialize the `BitSet` with capacity for one more bit.

## Tests
- Added a bunch of unit tests that hit these conditions. I ensured they
failed without the fix, and that they now pass.
- All unit tests pass with the fix in place

## Other
- The investigation that led to finding this bug began with
https://github.com/paradedb/paradedb/issues/4746.
 2026-04-25 14:11:47 +02:00
Abdul Andha
4fbae92187 send after key on last page 2026-04-24 15:33:26 -04:00
Cameron
89f0cef807 Fix O(2^n) query parser regression for deeply-nested queries (#2905) 
* Fix O(2^n) query parser regression for deeply-nested queries

The top-level `ast()` parser used `alt((boolean_expr, single_leaf))` at
every group level. When the group contained a single leaf with no
trailing operand, `boolean_expr` would parse `occur_leaf` (recursing
into the inner group), fail at `multispace1`, backtrack, and then
`single_leaf` would re-parse `occur_leaf` from scratch. Every nesting
level doubled the work, giving O(2^n) time for queries like
`(((((title:test)))))`.

Parse `occur_leaf` once and peek ahead for a trailing operand instead
of backtracking. This keeps parsing O(n) and also avoids the duplicate
parse for simple single-leaf queries.

Fixes #2498.

Measured on the issue reproducer (release build):

    depth   before     after
       20   0.87 s   <1 us
       25  28.23 s   <1 us
       60  (years)   ~5 us

Non-pathological queries are unaffected or slightly faster:

    query                     before     after
    hello                     650 ns     308 ns
    a AND b AND c            1380 ns    1364 ns
    title:rust AND (...)     3426 ns    3460 ns

All 53 existing grammar tests and 56 query_parser tests pass. Adds a
regression test at depth 60 that would not complete under the old
parser.

* Add ignored benchmark for nested query parsing at depth 20/21

Matches the depths from issue #2498 which reported 0.87 s / 1.72 s
under the regression. With the fix these parse in single-digit
microseconds. Runs via:

  cargo test -p tantivy-query-grammar --release bench_deeply_nested \
      -- --ignored --nocapture

* Propagate Err::Failure and Err::Incomplete from operand parser

`alt((boolean_expr, single_leaf))` only retried on `Err::Error` and
propagated `Err::Failure` and `Err::Incomplete`. The replacement was
catching all three with `Err(_)`, which would silently fall back to
a single leaf if any cut point were ever added to `operand_leaf` or
its descendants. Match specifically on `Err::Error` to preserve the
original `alt` semantics.

* Replace inline bench with binggan bench in benches/

Move the nested-query benchmark out of the query-grammar test module
and into a proper binggan benchmark at benches/query_parser_nested.rs,
registered as a harnessless bench in Cargo.toml. Keeps the correctness
regression test (depth 60) in place.

Run with: cargo bench --bench query_parser_nested

* Fix rustfmt import ordering in query_parser_nested bench
 2026-04-24 03:54:00 -04:00
Claude
a5d297c75f Add OpenSSF Scorecard workflow 
Runs weekly security analysis and uploads SARIF results to GitHub code
scanning. Third-party actions are pinned by commit SHA. Adds the Scorecard
badge to the README.

Based on quickwit-oss/quickwit#5969.
 2026-04-24 06:56:58 +00:00
Pascal Seitz
2e16243f9a fix memory consumption for histogram 2026-04-21 13:58:39 +02:00
Pascal Seitz
e015abab8e docs: add 0.26.1 changelog entry for aggregation perf fix 2026-04-21 11:12:37 +02:00
Pascal Seitz
73c711ec74 perf(agg): only measure active parent bucket in composite collect 
Same change as 26a589e for SegmentCompositeCollector: get_memory_consumption
summed across all parent_buckets on every block, scaling with outer bucket
cardinality. Pass parent_bucket_id and index the single bucket.
 2026-04-21 07:26:58 +02:00
Pascal Seitz
cb037c8079 add inline 2026-04-21 07:26:58 +02:00
Pascal Seitz
ed3453606b agg fix: compute memory consumption only for current bucket 2026-04-21 07:26:58 +02:00
Pascal Seitz
e9641f99c5 add nested term benchmark 2026-04-21 07:26:58 +02:00
Paul Masurel
13d74c3c20 Update binggan requirement from 0.16.0 to 0.16.1 (#2899) 2026-04-20 11:59:47 +02:00
Claude
3a6a3de8d7 ci: update pinned Action SHAs to current latest versions 
The previous commit pinned actions to commit SHAs but used stale
version tags (v4.2.2, v2.7.5, old nextest/cargo-llvm-cov refs).
Update to the actual current HEAD of each pinned tag:

  actions/checkout        v4.2.2 → v4.3.1  (34e114876b0b...)
  Swatinem/rust-cache     v2.7.5 → v2.9.1  (c19371144df3...)
  taiki-e/install-action  nextest           (56cc9adf3a3e...)
  taiki-e/install-action  cargo-llvm-cov    (e4b3a0453201...)

actions-rs/toolchain, actions-rs/clippy-check, and
codecov/codecov-action SHAs were already correct.

https://claude.ai/code/session_01VD7Bo8upj3cQwWDf9ni2Ln
 2026-04-16 06:49:47 +00:00
Claude
af3c6c0070 ci: pin GitHub Actions to full commit SHAs and restrict token permissions 
Fixes two supply chain / token security issues:

- Pin all third-party Actions to immutable full commit SHAs instead of
  mutable version tags (addresses unpinned-dependencies risk, analogous
  to quickwit-oss/quickwit#5937):
    actions/checkout v4.2.2
    actions-rs/toolchain v1.0.7
    Swatinem/rust-cache v2.7.5
    taiki-e/install-action nextest / cargo-llvm-cov
    actions-rs/clippy-check v1.0.7
    codecov/codecov-action v3.1.6

- Add explicit least-privilege `permissions` blocks at workflow and job
  level (addresses excessive GITHUB_TOKEN permissions, analogous to
  quickwit-oss/quickwit#5945):
    default: contents: read
    check job: also grants checks: write (required by clippy-check)

https://claude.ai/code/session_01VD7Bo8upj3cQwWDf9ni2Ln
 2026-04-15 20:55:43 +00:00
dependabot[bot]
058afff8b7 Update binggan requirement from 0.15.3 to 0.16.0 
Updates the requirements on [binggan](https://github.com/pseitz/binggan) to permit the latest version.
- [Changelog](https://github.com/PSeitz/binggan/blob/main/CHANGELOG.md)
- [Commits](https://github.com/pseitz/binggan/commits)

---
updated-dependencies:
- dependency-name: binggan
  dependency-version: 0.16.0
  dependency-type: direct:production
...

Signed-off-by: dependabot[bot] <support@github.com>
 2026-04-15 08:58:03 +02:00
Paul Masurel
58aa4b7074 Fix cardinality aggregation using invalid coupons (#2893) 
Previously, coupons were computed via murmurhash32 and fed as raw u32
to the HLL sketch, bypassing the sketch's internal hashing and producing
invalid (slot, value) pairs. Switch to Coupon::from_hash from the
datasketches crate which correctly derives coupons, and drop the
now-unused murmurhash32 dependency.

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
 2026-04-13 19:14:30 +02:00
Paul Masurel
04beab3b29 Performance improvement for nested cardinality aggregation 
When a string cardinality aggregation is nested it end up being applied to different buckets.
Dictionary encoding relies on a different dictionaries for each segment.

As a result, during segment collection, we only collect term ordinals in a HashSet, and decode them in the
term dictionary at the end of collection.

Before this PR, this decoding phase was done once for each bucket, causing the same work to be done over and over. This PR introduce a coupon cache. The HLL sketch relies on a hash of the string values.

We populate the cache before bucket collection, and get our values from it.

This PR also rename "caching" "buffering" in aggregation (it was never caching), and does several cleanups.
 2026-04-10 14:51:00 +02:00
alexanderbianchi
3cd9011f87 Make BucketEntries::iter, PercentileValuesVecEntry fields, and TopNComputer::threshold public (#2890) 
These items need to be accessible from the tantivy-datafusion crate:
- BucketEntries::iter() for iterating aggregation bucket results
- PercentileValuesVecEntry.key/.value for reading percentile results
- TopNComputer.threshold for Block-WAND score pruning in the inverted index provider

Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-authored-by: Paul Masurel <paul@quickwit.io>
 2026-04-09 13:32:31 +02:00
Paul Masurel
d2c1b8bc2c Optimized intersection count using a bitset when the first leg is dense 2026-04-06 12:01:52 -04:00
nuri
a65107135a Use BinaryHeap for score-based top-K collection (#2881) 
* Use BinaryHeap for score-based top-K collection

* Use peek_mut and add proptest for TopNHeap

---------

Co-authored-by: nryoo <nryoo@nryooui-MacBookPro.local>
 2026-04-04 19:49:05 +02:00
Pascal Seitz
5c344db1bf chore: Release 2026-03-31 17:15:34 +08:00
Pascal Seitz
dc0f31554d unbump for release and update Changelog.md 2026-03-31 17:15:34 +08:00
trinity-1686a
a28ce3ee54 Merge pull request #2869 from quickwit-oss/trinity.pointard/maint 
add dependabot cooldown
 2026-03-31 09:52:22 +02:00
dependabot[bot]
3abc137bfe Update binggan requirement from 0.14.2 to 0.15.3 (#2870) 
Updates the requirements on [binggan](https://github.com/pseitz/binggan) to permit the latest version.
- [Commits](https://github.com/pseitz/binggan/commits)

---
updated-dependencies:
- dependency-name: binggan
  dependency-version: 0.15.3
  dependency-type: direct:production
...

Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
 2026-03-31 07:59:02 +08:00
trinity Pointard
cf9800f981 add dependabot cooldown 2026-03-30 11:36:04 +02:00
174 changed files with 8419 additions and 4583 deletions
Expand all files Collapse all files

4
.github/dependabot.yml vendored
View File

@@ -6,6 +6,8 @@ updates:
interval: daily
time: "20:00"
open-pull-requests-limit: 10
cooldown:
default-days: 2
- package-ecosystem: "github-actions"
directory: "/"
@@ -13,3 +15,5 @@ updates:
interval: daily
time: "20:00"
open-pull-requests-limit: 10
cooldown:
default-days: 2

15
.github/workflows/coverage.yml vendored
View File

@@ -4,6 +4,9 @@ on:
push:
branches: [main]
permissions:
contents: read
# Ensures that we cancel running jobs for the same PR / same workflow.
concurrency:
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
@@ -12,16 +15,20 @@ concurrency:
jobs:
coverage:
runs-on: ubuntu-latest
permissions:
contents: read
steps:
- uses: actions/checkout@v4
- uses: actions/checkout@df4cb1c069e1874edd31b4311f1884172cec0e10 # v6.0.3
- name: Install Rust
run: rustup toolchain install nightly-2025-12-01 --profile minimal --component llvm-tools-preview
- uses: Swatinem/rust-cache@v2
- uses: taiki-e/install-action@cargo-llvm-cov
- uses: Swatinem/rust-cache@c19371144df3bb44fab255c43d04cbc2ab54d1c4 # v2.9.1
- uses: taiki-e/install-action@e4b3a0453201addddc06d3a72db90326aad87084 # cargo-llvm-cov
- name: Generate code coverage
run: cargo +nightly-2025-12-01 llvm-cov --all-features --workspace --doctests --lcov --output-path lcov.info
- name: Upload coverage to Codecov
uses: codecov/codecov-action@v3
uses: codecov/codecov-action@fb8b3582c8e4def4969c97caa2f19720cb33a72f # v7.0.0
continue-on-error: true
with:
token: ${{ secrets.CODECOV_TOKEN }} # not required for public repos

10
.github/workflows/long_running.yml vendored
View File

@@ -8,6 +8,9 @@ env:
CARGO_TERM_COLOR: always
NUM_FUNCTIONAL_TEST_ITERATIONS: 20000
permissions:
contents: read
# Ensures that we cancel running jobs for the same PR / same workflow.
concurrency:
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
@@ -18,10 +21,13 @@ jobs:
runs-on: ubuntu-latest
permissions:
contents: read
steps:
- uses: actions/checkout@v4
- uses: actions/checkout@df4cb1c069e1874edd31b4311f1884172cec0e10 # v6.0.3
- name: Install stable
uses: actions-rs/toolchain@v1
uses: actions-rs/toolchain@16499b5e05bf2e26879000db0c1d13f7e13fa3af # v1.0.7
with:
toolchain: stable
profile: minimal

49
.github/workflows/scorecard.yml vendored Normal file
View File

@@ -0,0 +1,49 @@
name: OpenSSF Scorecard
on:
schedule:
- cron: '0 0 * * 0'
push:
branches:
- main
permissions:
contents: read
jobs:
analysis:
name: Scorecards analysis
runs-on: ubuntu-latest
permissions:
# Needed to upload the results to code-scanning dashboard.
security-events: write
# Needed to publish results
id-token: write
steps:
- name: 'Checkout code'
uses: actions/checkout@df4cb1c069e1874edd31b4311f1884172cec0e10 # v6.0.3
with:
persist-credentials: false
- name: 'Run analysis'
uses: ossf/scorecard-action@4eaacf0543bb3f2c246792bd56e8cdeffafb205a # v2.4.3
with:
results_file: results.sarif
results_format: sarif
repo_token: ${{ secrets.GITHUB_TOKEN }}
publish_results: true
# Upload the results as artifacts.
- name: 'Upload artifact'
uses: actions/upload-artifact@043fb46d1a93c77aae656e7c1c64a875d1fc6a0a # v7.0.1
with:
name: SARIF file
path: results.sarif
retention-days: 5
# Upload the results to GitHub's code scanning dashboard.
- name: 'Upload to code-scanning'
uses: github/codeql-action/upload-sarif@87557b9c84dde89fdd9b10e88954ac2f4248e463 # v4.36.1
with:
sarif_file: results.sarif

28
.github/workflows/test.yml vendored
View File

@@ -9,6 +9,9 @@ on:
env:
CARGO_TERM_COLOR: always
permissions:
contents: read
# Ensures that we cancel running jobs for the same PR / same workflow.
concurrency:
group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
@@ -19,23 +22,27 @@ jobs:
runs-on: ubuntu-latest
permissions:
contents: read
checks: write
steps:
- uses: actions/checkout@v4
- uses: actions/checkout@df4cb1c069e1874edd31b4311f1884172cec0e10 # v6.0.3
- name: Install nightly
uses: actions-rs/toolchain@v1
uses: actions-rs/toolchain@16499b5e05bf2e26879000db0c1d13f7e13fa3af # v1.0.7
with:
toolchain: nightly
profile: minimal
components: rustfmt
- name: Install stable
uses: actions-rs/toolchain@v1
uses: actions-rs/toolchain@16499b5e05bf2e26879000db0c1d13f7e13fa3af # v1.0.7
with:
toolchain: stable
profile: minimal
components: clippy
- uses: Swatinem/rust-cache@v2
- uses: Swatinem/rust-cache@c19371144df3bb44fab255c43d04cbc2ab54d1c4 # v2.9.1
- name: Check Formatting
run: cargo +nightly fmt --all -- --check
@@ -47,7 +54,7 @@ jobs:
- name: Check Bench Compilation
run: cargo +nightly bench --no-run --profile=dev --all-features
- uses: actions-rs/clippy-check@v1
- uses: actions-rs/clippy-check@b5b5f21f4797c02da247df37026fcd0a5024aa4d # v1.0.7
with:
toolchain: stable
token: ${{ secrets.GITHUB_TOKEN }}
@@ -57,6 +64,9 @@ jobs:
runs-on: ubuntu-latest
permissions:
contents: read
strategy:
matrix:
features:
@@ -67,17 +77,17 @@ jobs:
name: test-${{ matrix.features.label}}
steps:
- uses: actions/checkout@v4
- uses: actions/checkout@df4cb1c069e1874edd31b4311f1884172cec0e10 # v6.0.3
- name: Install stable
uses: actions-rs/toolchain@v1
uses: actions-rs/toolchain@16499b5e05bf2e26879000db0c1d13f7e13fa3af # v1.0.7
with:
toolchain: stable
profile: minimal
override: true
- uses: taiki-e/install-action@nextest
- uses: Swatinem/rust-cache@v2
- uses: taiki-e/install-action@56cc9adf3a3e2c23eafb56e8acaf9d0373cb845a # nextest
- uses: Swatinem/rust-cache@c19371144df3bb44fab255c43d04cbc2ab54d1c4 # v2.9.1
- name: Run tests
run: |

7
CHANGELOG.md
View File

@@ -1,3 +1,9 @@
Tantivy 0.26.1
================================
## Performance
- Fix quadratic runtime in nested term and composite aggregations: memory accounting scanned all parent buckets on every collect instead of just the current parent (@PSeitz @fulmicoton)
Tantivy 0.26 (Unreleased)
================================
@@ -45,6 +51,7 @@ Tantivy 0.26 (Unreleased)
- Add `seek_danger` on `DocSet` for more efficient intersections [#2538](https://github.com/quickwit-oss/tantivy/pull/2538) [#2810](https://github.com/quickwit-oss/tantivy/pull/2810)(@PSeitz @stuhood @fulmicoton)
- Skip column traversal in `RangeDocSet` when query range does not overlap with column bounds [#2783](https://github.com/quickwit-oss/tantivy/pull/2783)(@ChangRui-Ryan)
- Speed up exclude queries by supporting multiple excluded `DocSet`s without intermediate union [#2825](https://github.com/quickwit-oss/tantivy/pull/2825)(@PSeitz)
- Improve union performance for non-score unions with `fill_buffer` and optimized `TinySet` [#2863](https://github.com/quickwit-oss/tantivy/pull/2863)(@PSeitz)
Tantivy 0.25
================================

25
Cargo.toml
View File

@@ -57,15 +57,15 @@ measure_time = "0.9.0"
arc-swap = "1.5.0"
bon = "3.3.1"
columnar = { version = "0.6", path = "./columnar", package = "tantivy-columnar" }
sstable = { version = "0.6", path = "./sstable", package = "tantivy-sstable", optional = true }
stacker = { version = "0.6", path = "./stacker", package = "tantivy-stacker" }
query-grammar = { version = "0.25.0", path = "./query-grammar", package = "tantivy-query-grammar" }
tantivy-bitpacker = { version = "0.9", path = "./bitpacker" }
common = { version = "0.10", path = "./common/", package = "tantivy-common" }
tokenizer-api = { version = "0.6", path = "./tokenizer-api", package = "tantivy-tokenizer-api" }
columnar = { version = "0.7", path = "./columnar", package = "tantivy-columnar" }
sstable = { version = "0.7", path = "./sstable", package = "tantivy-sstable", optional = true }
stacker = { version = "0.7", path = "./stacker", package = "tantivy-stacker" }
query-grammar = { version = "0.26.0", path = "./query-grammar", package = "tantivy-query-grammar" }
tantivy-bitpacker = { version = "0.10", path = "./bitpacker" }
common = { version = "0.11", path = "./common/", package = "tantivy-common" }
tokenizer-api = { version = "0.7", path = "./tokenizer-api", package = "tantivy-tokenizer-api" }
sketches-ddsketch = { version = "0.4", features = ["use_serde"] }
datasketches = "0.2.0"
datasketches = { version = "0.3.0", features = ["hll"] }
futures-util = { version = "0.3.28", optional = true }
futures-channel = { version = "0.3.28", optional = true }
fnv = "1.0.7"
@@ -75,7 +75,7 @@ typetag = "0.2.21"
winapi = "0.3.9"
[dev-dependencies]
binggan = "0.14.2"
binggan = "0.17.0"
rand = "0.9"
maplit = "1.0.2"
matches = "0.1.9"
@@ -92,7 +92,7 @@ postcard = { version = "1.0.4", features = [
], default-features = false }
[target.'cfg(not(windows))'.dev-dependencies]
criterion = { version = "0.5", default-features = false }
criterion = { version = "0.8", default-features = false }
[dev-dependencies.fail]
version = "0.5.0"
@@ -203,6 +203,9 @@ name = "regex_all_terms"
harness = false
[[bench]]
name = "fill_bitset"
name = "query_parser_nested"
harness = false
[[bench]]
name = "intersection_bench"
harness = false

1
README.md
View File

@@ -1,6 +1,7 @@
[![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)
[![OpenSSF Scorecard](https://api.scorecard.dev/projects/github.com/quickwit-oss/tantivy/badge)](https://scorecard.dev/viewer/?uri=github.com/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)
[![Crates.io](https://img.shields.io/crates/v/tantivy.svg)](https://crates.io/crates/tantivy)

194
benches/agg_bench.rs
View File

@@ -63,7 +63,12 @@ fn bench_agg(mut group: InputGroup<Index>) {
register!(group, terms_all_unique_with_avg_sub_agg);
register!(group, terms_many_with_avg_sub_agg);
register!(group, terms_status_with_avg_sub_agg);
register!(group, terms_status_with_terms_zipf_1000_sub_agg);
register!(group, terms_zipf_1000_with_terms_status_sub_agg);
register!(group, terms_status_with_histogram);
register!(group, terms_status_with_date_histogram);
register!(group, terms_status_with_date_histogram_hard_bounds);
register!(group, terms_status_with_date_histogram_and_sibling_terms);
register!(group, terms_zipf_1000);
register!(group, terms_zipf_1000_with_histogram);
register!(group, terms_zipf_1000_with_avg_sub_agg);
@@ -77,7 +82,12 @@ fn bench_agg(mut group: InputGroup<Index>) {
register!(group, composite_histogram_calendar);
register!(group, cardinality_agg);
register!(group, cardinality_agg_high_card);
register!(group, cardinality_agg_low_card);
register!(group, terms_status_with_cardinality_agg);
register!(group, terms_100_buckets_with_cardinality_agg);
register!(group, terms_many_with_single_term_order_by_card);
register!(group, terms_many_with_single_term_2_order_by_card);
register!(group, range_agg);
register!(group, range_agg_with_avg_sub_agg);
@@ -165,10 +175,52 @@ fn cardinality_agg(index: &Index) {
});
execute_agg(index, agg_req);
}
// Full-scan cardinality on a near-1M-cardinality string field.
// Hits the dense (PagedBitset) path: every doc has a unique term,
// so the bucket promotes from FxHashSet shortly into the scan.
fn cardinality_agg_high_card(index: &Index) {
let agg_req = json!({
"cardinality": {
"cardinality": {
"field": "text_all_unique_terms"
},
}
});
execute_agg(index, agg_req);
}
// Full-scan cardinality on a tiny-cardinality string field (7 distinct
// values). Stays on the FxHashSet path — the promotion threshold is
// never crossed. Validates no regression on the sparse path.
fn cardinality_agg_low_card(index: &Index) {
let agg_req = json!({
"cardinality": {
"cardinality": {
"field": "text_few_terms_status"
},
}
});
execute_agg(index, agg_req);
}
fn terms_status_with_cardinality_agg(index: &Index) {
let agg_req = json!({
"my_texts": {
"terms": { "field": "text_few_terms_status" },
"aggs": {
"cardinality": {
"cardinality": {
"field": "text_few_terms_status"
},
}
}
},
});
execute_agg(index, agg_req);
}
fn terms_100_buckets_with_cardinality_agg(index: &Index) {
let agg_req = json!({
"my_texts": {
"terms": { "field": "text_1000_terms_zipf", "size": 100 },
"aggs": {
"cardinality": {
"cardinality": {
@@ -181,6 +233,58 @@ fn terms_status_with_cardinality_agg(index: &Index) {
execute_agg(index, agg_req);
}
fn terms_many_with_single_term_order_by_card(index: &Index) {
let agg_req = json!({
"my_texts": {
"terms": { "field": "text_many_terms" },
"aggs": {
"nested_terms": {
"terms": {
"field": "single_term",
"order": { "cardinality": "desc" }
},
"aggs": {
"cardinality": {
"cardinality": { "field": "text_few_terms" }
}
}
}
}
},
});
execute_agg(index, agg_req);
}
// Two-level terms ordered by cardinality at each level: a high-card outer terms
// (text_many_terms) ordered by a cardinality sub-agg, with a nested low-card terms
// (text_few_terms_status) also ordered by a cardinality sub-agg, plus an avg.
fn terms_many_with_single_term_2_order_by_card(index: &Index) {
let agg_req = json!({
"by_ip": {
"terms": {
"field": "text_many_terms",
"order": { "card_few_terms": "desc" }
},
"aggs": {
"card_few_terms": {
"cardinality": { "field": "text_few_terms" }
},
"nested_terms": {
"terms": {
"field": " single_term",
"order": { "distinct_path2": "desc" }
},
"aggs": {
"avg_botscore": { "avg": { "field": "score" } },
"distinct_path2": { "cardinality": { "field": "text_few_terms" } }
}
}
}
}
});
execute_agg(index, agg_req);
}
fn terms_7(index: &Index) {
let agg_req = json!({
"my_texts": { "terms": { "field": "text_few_terms_status" } },
@@ -253,6 +357,30 @@ fn terms_all_unique_with_avg_sub_agg(index: &Index) {
});
execute_agg(index, agg_req);
}
fn terms_status_with_terms_zipf_1000_sub_agg(index: &Index) {
let agg_req = json!({
"my_texts": {
"terms": { "field": "text_few_terms_status" },
"aggs": {
"nested_terms": { "terms": { "field": "text_1000_terms_zipf" } }
}
}
});
execute_agg(index, agg_req);
}
fn terms_zipf_1000_with_terms_status_sub_agg(index: &Index) {
let agg_req = json!({
"my_texts": {
"terms": { "field": "text_1000_terms_zipf" },
"aggs": {
"nested_terms": { "terms": { "field": "text_few_terms_status" } }
}
}
});
execute_agg(index, agg_req);
}
fn terms_status_with_histogram(index: &Index) {
let agg_req = json!({
"my_texts": {
@@ -265,6 +393,57 @@ fn terms_status_with_histogram(index: &Index) {
execute_agg(index, agg_req);
}
fn terms_status_with_date_histogram(index: &Index) {
let agg_req = json!({
"my_texts": {
"terms": { "field": "text_few_terms_status" },
"aggs": {
"over_time": { "date_histogram": { "field": "timestamp", "fixed_interval": "1h" } }
}
}
});
execute_agg(index, agg_req);
}
/// Same fused terms × date_histogram, but with `hard_bounds`. The timestamps span 0..120h; the
/// bounds drop only the first and last hour (ms: 1h=3_600_000, 119h=428_400_000), so almost every
/// doc is in-bounds. This exercises the collector's hard-bounds path: `bounds.contains` runs per
/// doc (the `all_docs_in_bounds` short-circuit is off) and the rare out-of-bounds doc takes the
/// `term_counts` branch.
fn terms_status_with_date_histogram_hard_bounds(index: &Index) {
let agg_req = json!({
"my_texts": {
"terms": { "field": "text_few_terms_status" },
"aggs": {
"over_time": {
"date_histogram": {
"field": "timestamp",
"fixed_interval": "1h",
"hard_bounds": { "min": 3_600_000, "max": 428_400_000 }
}
}
}
}
});
execute_agg(index, agg_req);
}
/// Same fused terms × date_histogram, but with a sibling terms aggregation next to it. The fused
/// fast path should still trigger for `my_texts` (sibling aggregations are independent top-level
/// aggregations, so they don't change its eligibility).
fn terms_status_with_date_histogram_and_sibling_terms(index: &Index) {
let agg_req = json!({
"my_texts": {
"terms": { "field": "text_few_terms_status" },
"aggs": {
"over_time": { "date_histogram": { "field": "timestamp", "fixed_interval": "1h" } }
}
},
"other_texts": { "terms": { "field": "text_few_terms" } }
});
execute_agg(index, agg_req);
}
fn terms_zipf_1000_with_histogram(index: &Index) {
let agg_req = json!({
"my_texts": {
@@ -566,7 +745,8 @@ fn get_test_index_bench(cardinality: Cardinality) -> tantivy::Result<Index> {
TextFieldIndexing::default().set_index_option(IndexRecordOption::WithFreqs),
)
.set_stored();
let text_field = schema_builder.add_text_field("text", text_fieldtype);
let text_field = schema_builder.add_text_field("text", text_fieldtype.clone());
let single_term = schema_builder.add_text_field("single_term", FAST);
let json_field = schema_builder.add_json_field("json", FAST);
let text_field_all_unique_terms =
schema_builder.add_text_field("text_all_unique_terms", STRING | FAST);
@@ -630,6 +810,8 @@ fn get_test_index_bench(cardinality: Cardinality) -> tantivy::Result<Index> {
index_writer.add_document(doc!(
json_field => json!({"mixed_type": 10.0}),
json_field => json!({"mixed_type": 10.0}),
single_term => "single_term",
single_term => "single_term",
text_field => "cool",
text_field => "cool",
text_field_all_unique_terms => "cool",
@@ -655,7 +837,9 @@ fn get_test_index_bench(cardinality: Cardinality) -> tantivy::Result<Index> {
doc_with_value /= 20;
}
let _val_max = 1_000_000.0;
for _ in 0..doc_with_value {
const SPAN_MS: i64 = 120 * 3600 * 1000; // 120 hours in ms
const NOISE_MS: i64 = 2 * 3600 * 1000; // ±2h noise
for i in 0..doc_with_value {
let val: f64 = rng.random_range(0.0..1_000_000.0);
let json = if rng.random_bool(0.1) {
// 10% are numeric values
@@ -663,7 +847,11 @@ fn get_test_index_bench(cardinality: Cardinality) -> tantivy::Result<Index> {
} else {
json!({"mixed_type": many_terms_data.choose(&mut rng).unwrap().to_string()})
};
let base_ms = (i as i64 * SPAN_MS) / doc_with_value as i64;
let noise_ms = rng.random_range(-NOISE_MS..NOISE_MS);
let ts_ms = (base_ms + noise_ms).clamp(0, SPAN_MS);
index_writer.add_document(doc!(
single_term => "single_term",
text_field => "cool",
json_field => json,
text_field_all_unique_terms => format!("unique_term_{}", rng.random::<u64>()),
@@ -674,7 +862,7 @@ fn get_test_index_bench(cardinality: Cardinality) -> tantivy::Result<Index> {
score_field => val as u64,
score_field_f64 => lg_norm.sample(&mut rng),
score_field_i64 => val as i64,
date_field => DateTime::from_timestamp_millis((val * 1_000_000.) as i64),
date_field => DateTime::from_timestamp_millis(ts_ms),
))?;
if cardinality == Cardinality::OptionalSparse {
for _ in 0..20 {

38
benches/bool_queries_with_range.rs
View File

@@ -110,43 +110,31 @@ fn main() {
// Prepare corpora with varying scenarios
let scenarios = vec![
(
"dense and 99% a".to_string(),
10_000_000,
0.99,
"dense and 0.1% a".to_string(),
5_000_000,
0.001,
"dense",
0,
9,
),
("dense and 1% a".to_string(), 5_000_000, 0.01, "dense", 0, 9),
("dense and 10% a".to_string(), 5_000_000, 0.1, "dense", 0, 9),
(
"dense and 99% a".to_string(),
10_000_000,
0.99,
"dense",
990,
999,
),
(
"sparse and 99% a".to_string(),
10_000_000,
"sparse and 50% a".to_string(),
5_000_000,
0.99,
"sparse",
0,
9,
),
(
"sparse and 99% a".to_string(),
10_000_000,
0.99,
"sparse",
9_999_990,
9_999_999,
),
];
let mut runner = BenchRunner::new();
for (scenario_id, n, p_title_a, num_rand_distribution, range_low, range_high) in scenarios {
for (scenario_id, num_docs, p_title_a, num_rand_distribution, range_low, range_high) in
scenarios
{
// Build index for this scenario
let bench_index = build_shared_indices(n, p_title_a, num_rand_distribution);
let bench_index = build_shared_indices(num_docs, p_title_a, num_rand_distribution);
// Create benchmark group
let mut group = runner.new_group();
@@ -158,7 +146,7 @@ fn main() {
let field_names = ["num_rand", "num_asc", "num_rand_fast", "num_asc_fast"];
// Define the three terms we want to test with
let terms = ["a", "b", "z"];
let terms = ["a"];
// Generate all combinations of terms and field names
let mut queries = Vec::new();
@@ -203,7 +191,7 @@ fn run_benchmark_tasks(
bench_index,
query_str,
DocSetCollector,
"all results",
"all_results",
);
// Test top 100 by the field (if it's a FAST field)

112
benches/fill_bitset.rs
View File

@@ -1,112 +0,0 @@
use binggan::{black_box, BenchRunner, PeakMemAlloc, INSTRUMENTED_SYSTEM};
use common::BitSet;
use rand::rngs::StdRng;
use rand::{Rng, SeedableRng};
use tantivy::postings::BlockSegmentPostings;
use tantivy::schema::*;
use tantivy::{
doc, DocSet, Index, InvertedIndexReader, TantivyDocument, TantivyInvertedIndexReader,
};
#[global_allocator]
pub static GLOBAL: &PeakMemAlloc<std::alloc::System> = &INSTRUMENTED_SYSTEM;
fn main() {
let index = build_test_index();
let reader = index.reader().unwrap();
let searcher = reader.searcher();
let segment_reader = &searcher.segment_readers()[0];
let text_field = index.schema().get_field("text").unwrap();
let inverted_index = segment_reader.inverted_index(text_field).unwrap();
let max_doc = segment_reader.max_doc();
let term = Term::from_field_text(text_field, "hello");
let term_info = inverted_index.get_term_info(&term).unwrap().unwrap();
let mut runner = BenchRunner::new();
runner.set_name("fill_bitset");
let mut group = runner.new_group();
{
let inverted_index = &inverted_index;
let term_info = &term_info;
// This is the path used by queries (AutomatonWeight, RangeQuery, etc.)
// It dispatches via DynInvertedIndexReader::fill_bitset_from_terminfo.
group.register("fill_bitset_from_terminfo (via trait)", move |_| {
let mut bitset = BitSet::with_max_value(max_doc);
inverted_index
.fill_bitset_from_terminfo(term_info, &mut bitset)
.unwrap();
black_box(bitset);
});
}
{
let inverted_index = &inverted_index;
let term_info = &term_info;
// This constructs a SegmentPostings via read_docset_from_terminfo and calls fill_bitset.
group.register("read_docset + fill_bitset", move |_| {
let mut postings = inverted_index.read_docset_from_terminfo(term_info).unwrap();
let mut bitset = BitSet::with_max_value(max_doc);
postings.fill_bitset(&mut bitset);
black_box(bitset);
});
}
{
let inverted_index = &inverted_index;
let term_info = &term_info;
// This uses BlockSegmentPostings directly, bypassing SegmentPostings entirely.
let concrete_reader = inverted_index
.as_any()
.downcast_ref::<TantivyInvertedIndexReader>()
.expect("expected TantivyInvertedIndexReader");
group.register("BlockSegmentPostings direct", move |_| {
let raw = concrete_reader
.read_raw_postings_data(term_info, IndexRecordOption::Basic)
.unwrap();
let mut block_postings = BlockSegmentPostings::open(
term_info.doc_freq,
raw.postings_data,
raw.record_option,
raw.effective_option,
)
.unwrap();
let mut bitset = BitSet::with_max_value(max_doc);
loop {
let docs = block_postings.docs();
if docs.is_empty() {
break;
}
for &doc in docs {
bitset.insert(doc);
}
block_postings.advance();
}
black_box(bitset);
});
}
group.run();
}
fn build_test_index() -> Index {
let mut schema_builder = Schema::builder();
schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema.clone());
let text_field = schema.get_field("text").unwrap();
let mut writer = index.writer::<TantivyDocument>(250_000_000).unwrap();
let mut rng = StdRng::from_seed([42u8; 32]);
for _ in 0..100_000 {
if rng.random_bool(0.5) {
writer
.add_document(doc!(text_field => "hello world"))
.unwrap();
} else {
writer
.add_document(doc!(text_field => "goodbye world"))
.unwrap();
}
}
writer.commit().unwrap();
index
}

149
benches/intersection_bench.rs Normal file
View File

@@ -0,0 +1,149 @@
// Benchmarks top-K intersection of term scorers (block_wand_intersection).
//
// What's measured:
// - Conjunctive queries (+a +b, +a +b +c) with top-10 by score
// - Varying doc-frequency balance between terms (balanced, skewed, very skewed)
// - Realistic term frequencies (geometric distribution, mostly low)
// - 1M-doc single segment
//
// Run with: cargo bench --bench intersection_bench
use binggan::{black_box, BenchRunner};
use rand::prelude::*;
use rand::rngs::StdRng;
use rand::SeedableRng;
use tantivy::collector::TopDocs;
use tantivy::query::QueryParser;
use tantivy::schema::{Schema, TEXT};
use tantivy::{doc, Index, ReloadPolicy, Searcher};
const NUM_DOCS: usize = 1_000_000;
struct BenchIndex {
searcher: Searcher,
query_parser: QueryParser,
}
/// Generate term frequency from a geometric-like distribution.
/// Most values are 1, a few are 2-3, rarely higher.
/// p controls the decay: higher p → more weight on tf=1.
fn random_term_freq(rng: &mut StdRng, p: f64) -> u32 {
let mut tf = 1u32;
while tf < 10 && rng.random_bool(1.0 - p) {
tf += 1;
}
tf
}
/// Build an index with three terms (a, b, c) with given doc-frequency probabilities.
/// Each term occurrence has a realistic term frequency (geometric distribution).
/// Field length is padded with filler tokens to create varied fieldnorms.
fn build_index(p_a: f64, p_b: f64, p_c: f64) -> BenchIndex {
let mut schema_builder = Schema::builder();
let body = schema_builder.add_text_field("body", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut rng = StdRng::from_seed([42u8; 32]);
{
let mut writer = index.writer_with_num_threads(1, 500_000_000).unwrap();
for _ in 0..NUM_DOCS {
let mut tokens: Vec<String> = Vec::new();
if rng.random_bool(p_a) {
let tf = random_term_freq(&mut rng, 0.7);
for _ in 0..tf {
tokens.push("aaa".to_string());
}
}
if rng.random_bool(p_b) {
let tf = random_term_freq(&mut rng, 0.7);
for _ in 0..tf {
tokens.push("bbb".to_string());
}
}
if rng.random_bool(p_c) {
let tf = random_term_freq(&mut rng, 0.7);
for _ in 0..tf {
tokens.push("ccc".to_string());
}
}
// Pad with filler to create varied field lengths (5-30 tokens).
let filler_count = rng.random_range(5u32..30u32);
for _ in 0..filler_count {
tokens.push("filler".to_string());
}
let text = tokens.join(" ");
writer.add_document(doc!(body => text)).unwrap();
}
writer.commit().unwrap();
}
let reader = index
.reader_builder()
.reload_policy(ReloadPolicy::Manual)
.try_into()
.unwrap();
let searcher = reader.searcher();
let query_parser = QueryParser::for_index(&index, vec![body]);
BenchIndex {
searcher,
query_parser,
}
}
fn main() {
// Scenarios: (label, p_a, p_b, p_c)
//
// "balanced": all terms ~10% → intersection ~1% of docs
// "skewed": one common (50%), one rare (2%) → intersection ~1%
// "very_skewed": one very common (80%), one very rare (0.5%) → intersection ~0.4%
// "three_balanced": three terms ~20% each → intersection ~0.8%
// "three_skewed": 50% / 10% / 2% → intersection ~0.1%
let scenarios: Vec<(&str, f64, f64, f64)> = vec![
("balanced_10%_10%", 0.10, 0.10, 0.0),
("skewed_50%_2%", 0.50, 0.02, 0.0),
("very_skewed_80%_0.5%", 0.80, 0.005, 0.0),
("three_balanced_20%_20%_20%", 0.20, 0.20, 0.20),
("three_skewed_50%_10%_2%", 0.50, 0.10, 0.02),
];
let mut runner = BenchRunner::new();
for (label, p_a, p_b, p_c) in &scenarios {
let bench_index = build_index(*p_a, *p_b, *p_c);
let mut group = runner.new_group();
group.set_name(format!("intersection — {label}"));
// Two-term intersection
if *p_a > 0.0 && *p_b > 0.0 {
let query_str = "+aaa +bbb";
let query = bench_index.query_parser.parse_query(query_str).unwrap();
let searcher = bench_index.searcher.clone();
group.register(format!("{query_str} top10"), move |_| {
let collector = TopDocs::with_limit(10).order_by_score();
black_box(searcher.search(&query, &collector).unwrap());
1usize
});
}
// Three-term intersection
if *p_c > 0.0 {
let query_str = "+aaa +bbb +ccc";
let query = bench_index.query_parser.parse_query(query_str).unwrap();
let searcher = bench_index.searcher.clone();
group.register(format!("{query_str} top10"), move |_| {
let collector = TopDocs::with_limit(10).order_by_score();
black_box(searcher.search(&query, &collector).unwrap());
1usize
});
}
group.run();
}
}

35
benches/query_parser_nested.rs Normal file
View File

@@ -0,0 +1,35 @@
// Benchmark for the query grammar parsing deeply nested queries.
//
// Regression guard for https://github.com/quickwit-oss/tantivy/issues/2498:
// at depth 20/21 the old parser took 0.87 s / 1.72 s respectively because
// `ast()` retried `occur_leaf` on backtrack, giving O(2^n) time. With the
// fix parsing is linear and completes in microseconds.
//
// Run with: `cargo bench --bench query_parser_nested`.
use binggan::{black_box, BenchRunner};
use tantivy::query_grammar::parse_query;
fn nested_query(depth: usize, leading_plus: bool) -> String {
let leading = "(".repeat(depth);
let trailing = ")".repeat(depth);
let prefix = if leading_plus { "+" } else { "" };
format!("{prefix}{leading}title:test{trailing}")
}
fn main() {
let mut runner = BenchRunner::new();
for depth in [20, 21] {
for leading_plus in [false, true] {
let query = nested_query(depth, leading_plus);
let label = format!(
"parse_nested_depth_{depth}_{}",
if leading_plus { "plus" } else { "plain" },
);
runner.bench_function(&label, move |_| {
black_box(parse_query(black_box(&query)).unwrap());
});
}
}
}

3
benches/str_search_and_get.rs
View File

@@ -17,6 +17,7 @@ use rand::rngs::StdRng;
use rand::SeedableRng;
use tantivy::collector::{Count, DocSetCollector};
use tantivy::query::RangeQuery;
use tantivy::schema::document::TantivyDocument;
use tantivy::schema::{Schema, Value, FAST, STORED, STRING};
use tantivy::{doc, Index, ReloadPolicy, Searcher, Term};
@@ -405,7 +406,7 @@ impl FetchAllStringsFromDocTask {
for doc_address in docs {
// Get the document from the doc store (row store access)
if let Ok(doc) = self.searcher.doc(doc_address) {
if let Ok(doc) = self.searcher.doc::<TantivyDocument>(doc_address) {
// Extract string values from the stored field
if let Some(field_value) = doc.get_first(str_stored_field) {
if let Some(text) = field_value.as_value().as_str() {

7
bitpacker/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy-bitpacker"
version = "0.9.0"
version = "0.10.0"
edition = "2024"
authors = ["Paul Masurel <paul.masurel@gmail.com>"]
license = "MIT"
@@ -18,5 +18,10 @@ homepage = "https://github.com/quickwit-oss/tantivy"
bitpacking = { version = "0.9.2", default-features = false, features = ["bitpacker1x"] }
[dev-dependencies]
binggan = "0.17.0"
rand = "0.9"
proptest = "1"
[[bench]]
name = "bench"
harness = false

153
bitpacker/benches/bench.rs
View File

@@ -1,65 +1,110 @@
#![feature(test)]
use std::cell::RefCell;
extern crate test;
use binggan::{BenchRunner, black_box};
use rand::rng;
use rand::seq::IteratorRandom;
use tantivy_bitpacker::{BitPacker, BitUnpacker, BlockedBitpacker};
#[cfg(test)]
mod tests {
use rand::rng;
use rand::seq::IteratorRandom;
use tantivy_bitpacker::{BitPacker, BitUnpacker, BlockedBitpacker};
use test::Bencher;
fn create_bitpacked_data(bit_width: u8, num_els: u32) -> Vec<u8> {
let mut bitpacker = BitPacker::new();
let mut buffer = Vec::new();
for _ in 0..num_els {
bitpacker.write(0u64, bit_width, &mut buffer).unwrap();
bitpacker.flush(&mut buffer).unwrap();
}
buffer
}
#[inline(never)]
fn create_bitpacked_data(bit_width: u8, num_els: u32) -> Vec<u8> {
let mut bitpacker = BitPacker::new();
let mut buffer = Vec::new();
for _ in 0..num_els {
// the values do not matter.
bitpacker.write(0u64, bit_width, &mut buffer).unwrap();
bitpacker.flush(&mut buffer).unwrap();
const N: usize = 100_000;
const MAX_VAL: u64 = 1_000;
const BIT_WIDTH: u8 = 10; // 2^10 = 1024 > MAX_VAL
fn create_packed_data() -> (BitUnpacker, Vec<u8>) {
let mut bitpacker = BitPacker::new();
let mut data = Vec::new();
for i in 0..N as u64 {
let val = i * MAX_VAL / N as u64;
bitpacker.write(val, BIT_WIDTH, &mut data).unwrap();
}
bitpacker.close(&mut data).unwrap();
(BitUnpacker::new(BIT_WIDTH), data)
}
fn bench_bitpacking() {
let mut runner = BenchRunner::new();
let bit_width = 3;
let num_els = 1_000_000u32;
let bit_unpacker = BitUnpacker::new(bit_width);
let data = create_bitpacked_data(bit_width, num_els);
let idxs: Vec<u32> = (0..num_els).choose_multiple(&mut rng(), 100_000);
runner.bench_function("bitpacking_read", move |_| {
let mut out = 0u64;
for &idx in &idxs {
out = out.wrapping_add(bit_unpacker.get(idx, &data[..]));
}
buffer
}
black_box(out);
});
}
#[bench]
fn bench_bitpacking_read(b: &mut Bencher) {
let bit_width = 3;
let num_els = 1_000_000u32;
let bit_unpacker = BitUnpacker::new(bit_width);
let data = create_bitpacked_data(bit_width, num_els);
let idxs: Vec<u32> = (0..num_els).choose_multiple(&mut rng(), 100_000);
b.iter(|| {
let mut out = 0u64;
for &idx in &idxs {
out = out.wrapping_add(bit_unpacker.get(idx, &data[..]));
}
out
});
fn bench_blocked_bitpacker() {
let mut runner = BenchRunner::new();
let mut blocked_bitpacker = BlockedBitpacker::new();
for val in 0..=21500 {
blocked_bitpacker.add(val * val);
}
#[bench]
fn bench_blockedbitp_read(b: &mut Bencher) {
runner.bench_function("blockedbitp_read", move |_| {
let mut out = 0u64;
for val in 0..=21500 {
out = out.wrapping_add(blocked_bitpacker.get(val));
}
black_box(out);
});
runner.bench_function("blockedbitp_create", |_| {
let mut blocked_bitpacker = BlockedBitpacker::new();
for val in 0..=21500 {
blocked_bitpacker.add(val * val);
}
b.iter(|| {
let mut out = 0u64;
for val in 0..=21500 {
out = out.wrapping_add(blocked_bitpacker.get(val));
}
out
});
}
#[bench]
fn bench_blockedbitp_create(b: &mut Bencher) {
b.iter(|| {
let mut blocked_bitpacker = BlockedBitpacker::new();
for val in 0..=21500 {
blocked_bitpacker.add(val * val);
}
blocked_bitpacker
});
}
black_box(blocked_bitpacker);
});
}
fn bench_filter_vec() {
let mut runner = BenchRunner::new();
let (unpacker, data) = create_packed_data();
let positions = RefCell::new(Vec::with_capacity(N));
runner.bench_function("filter_vec_dense", move |_| {
unpacker.get_ids_for_value_range(
250..=750,
0..N as u32,
&data,
&mut positions.borrow_mut(),
);
black_box(positions.borrow().len());
});
let (unpacker, data) = create_packed_data();
let positions = RefCell::new(Vec::with_capacity(N));
runner.bench_function("filter_vec_sparse", move |_| {
unpacker.get_ids_for_value_range(0..=50, 0..N as u32, &data, &mut positions.borrow_mut());
black_box(positions.borrow().len());
});
let (unpacker, data) = create_packed_data();
let positions = RefCell::new(Vec::with_capacity(N));
runner.bench_function("filter_vec_full", move |_| {
unpacker.get_ids_for_value_range(
0..=MAX_VAL,
0..N as u32,
&data,
&mut positions.borrow_mut(),
);
black_box(positions.borrow().len());
});
}
fn main() {
bench_bitpacking();
bench_blocked_bitpacker();
bench_filter_vec();
}

160
bitpacker/src/filter_vec/mod.rs
View File

@@ -1,8 +1,17 @@
#[cfg(all(target_arch = "aarch64", not(target_vendor = "apple")))]
use std::arch::is_aarch64_feature_detected;
use std::ops::RangeInclusive;
#[cfg(target_arch = "x86_64")]
mod avx2;
#[cfg(target_arch = "aarch64")]
mod neon;
// SVE intrinsics are not exposed on aarch64-apple-darwin.
#[cfg(all(target_arch = "aarch64", not(target_vendor = "apple")))]
mod sve;
mod scalar;
#[derive(Clone, Copy, Eq, PartialEq, Debug)]
@@ -10,6 +19,10 @@ mod scalar;
enum FilterImplPerInstructionSet {
#[cfg(target_arch = "x86_64")]
AVX2 = 0u8,
#[cfg(all(target_arch = "aarch64", not(target_vendor = "apple")))]
SVE = 3u8,
#[cfg(target_arch = "aarch64")]
Neon = 2u8,
Scalar = 1u8,
}
@@ -19,29 +32,57 @@ impl FilterImplPerInstructionSet {
match *self {
#[cfg(target_arch = "x86_64")]
FilterImplPerInstructionSet::AVX2 => is_x86_feature_detected!("avx2"),
#[cfg(all(target_arch = "aarch64", not(target_vendor = "apple")))]
FilterImplPerInstructionSet::SVE => is_aarch64_feature_detected!("sve"),
// TIL Neon is required on aarch 64.
#[cfg(target_arch = "aarch64")]
FilterImplPerInstructionSet::Neon => true,
FilterImplPerInstructionSet::Scalar => true,
}
}
}
// List of available implementation in preferred order.
// List of available implementations in preferred order.
#[cfg(target_arch = "x86_64")]
const IMPLS: [FilterImplPerInstructionSet; 2] = [
FilterImplPerInstructionSet::AVX2,
FilterImplPerInstructionSet::Scalar,
];
#[cfg(not(target_arch = "x86_64"))]
// Non-Apple aarch64: try SVE, NEON, Scalar.
#[cfg(all(target_arch = "aarch64", not(target_vendor = "apple")))]
const IMPLS: [FilterImplPerInstructionSet; 3] = [
FilterImplPerInstructionSet::SVE,
FilterImplPerInstructionSet::Neon,
FilterImplPerInstructionSet::Scalar,
];
// Apple aarch64 (M-series): SVE not available; use NEON or Scalar.
#[cfg(all(target_arch = "aarch64", target_vendor = "apple"))]
const IMPLS: [FilterImplPerInstructionSet; 2] = [
FilterImplPerInstructionSet::Neon,
FilterImplPerInstructionSet::Scalar,
];
#[cfg(not(any(target_arch = "x86_64", target_arch = "aarch64")))]
const IMPLS: [FilterImplPerInstructionSet; 1] = [FilterImplPerInstructionSet::Scalar];
impl FilterImplPerInstructionSet {
#[inline]
#[allow(unused_variables)] // on non-x86_64, code is unused.
#[allow(unused_variables)]
fn from(code: u8) -> FilterImplPerInstructionSet {
#[cfg(target_arch = "x86_64")]
if code == FilterImplPerInstructionSet::AVX2 as u8 {
return FilterImplPerInstructionSet::AVX2;
}
#[cfg(all(target_arch = "aarch64", not(target_vendor = "apple")))]
if code == FilterImplPerInstructionSet::SVE as u8 {
return FilterImplPerInstructionSet::SVE;
}
#[cfg(target_arch = "aarch64")]
if code == FilterImplPerInstructionSet::Neon as u8 {
return FilterImplPerInstructionSet::Neon;
}
FilterImplPerInstructionSet::Scalar
}
@@ -50,6 +91,13 @@ impl FilterImplPerInstructionSet {
match self {
#[cfg(target_arch = "x86_64")]
FilterImplPerInstructionSet::AVX2 => avx2::filter_vec_in_place(range, offset, output),
#[cfg(all(target_arch = "aarch64", not(target_vendor = "apple")))]
// SAFETY: SVE availability was verified by is_available() before selecting this impl.
FilterImplPerInstructionSet::SVE => unsafe {
sve::filter_vec_in_place(range, offset, output)
},
#[cfg(target_arch = "aarch64")]
FilterImplPerInstructionSet::Neon => neon::filter_vec_in_place(range, offset, output),
FilterImplPerInstructionSet::Scalar => {
scalar::filter_vec_in_place(range, offset, output)
}
@@ -57,6 +105,12 @@ impl FilterImplPerInstructionSet {
}
}
fn available_impls() -> impl Iterator<Item = FilterImplPerInstructionSet> {
IMPLS
.into_iter()
.filter(FilterImplPerInstructionSet::is_available)
}
#[inline]
fn get_best_available_instruction_set() -> FilterImplPerInstructionSet {
use std::sync::atomic::{AtomicU8, Ordering};
@@ -64,10 +118,7 @@ fn get_best_available_instruction_set() -> FilterImplPerInstructionSet {
let instruction_set_byte: u8 = INSTRUCTION_SET_BYTE.load(Ordering::Relaxed);
if instruction_set_byte == u8::MAX {
// Let's initialize the instruction set and cache it.
let instruction_set = IMPLS
.into_iter()
.find(FilterImplPerInstructionSet::is_available)
.unwrap();
let instruction_set = available_impls().next().unwrap();
INSTRUCTION_SET_BYTE.store(instruction_set as u8, Ordering::Relaxed);
return instruction_set;
}
@@ -80,12 +131,12 @@ pub fn filter_vec_in_place(range: RangeInclusive<u32>, offset: u32, output: &mut
#[cfg(test)]
mod tests {
use proptest::strategy::Strategy;
use super::*;
#[test]
fn test_get_best_available_instruction_set() {
// This does not test much unfortunately.
// We just make sure the function returns without crashing and returns the same result.
let instruction_set = get_best_available_instruction_set();
assert_eq!(get_best_available_instruction_set(), instruction_set);
}
@@ -102,6 +153,31 @@ mod tests {
}
}
#[cfg(all(target_arch = "aarch64", not(target_vendor = "apple")))]
#[test]
fn test_instruction_set_to_code_from_code() {
for instruction_set in [
FilterImplPerInstructionSet::SVE,
FilterImplPerInstructionSet::Neon,
FilterImplPerInstructionSet::Scalar,
] {
let code = instruction_set as u8;
assert_eq!(instruction_set, FilterImplPerInstructionSet::from(code));
}
}
#[cfg(all(target_arch = "aarch64", target_vendor = "apple"))]
#[test]
fn test_instruction_set_to_code_from_code() {
for instruction_set in [
FilterImplPerInstructionSet::Neon,
FilterImplPerInstructionSet::Scalar,
] {
let code = instruction_set as u8;
assert_eq!(instruction_set, FilterImplPerInstructionSet::from(code));
}
}
fn test_filter_impl_empty_aux(filter_impl: FilterImplPerInstructionSet) {
let mut output = vec![];
filter_impl.filter_vec_in_place(0..=u32::MAX, 0, &mut output);
@@ -126,11 +202,20 @@ mod tests {
assert_eq!(&output, &[1, 3, 4, 5, 6, 7, 8]);
}
fn test_filter_impl_empty_range_aux(filter_impl: FilterImplPerInstructionSet) {
// start > end: RangeInclusive::contains always returns false; output must be empty.
// The SVE path's wrapping_sub would otherwise produce a huge range_width.
let mut output = vec![3, 2, 1, 5, 11, 2, 5, 10, 2];
filter_impl.filter_vec_in_place(10..=5, 0, &mut output);
assert_eq!(&output, &[]);
}
fn test_filter_impl_test_suite(filter_impl: FilterImplPerInstructionSet) {
test_filter_impl_empty_aux(filter_impl);
test_filter_impl_simple_aux(filter_impl);
test_filter_impl_simple_aux_shifted(filter_impl);
test_filter_impl_simple_outside_i32_range(filter_impl);
test_filter_impl_empty_range_aux(filter_impl);
}
#[test]
@@ -141,25 +226,60 @@ mod tests {
}
}
#[test]
#[cfg(all(target_arch = "aarch64", not(target_vendor = "apple")))]
fn test_filter_implementation_sve() {
if FilterImplPerInstructionSet::SVE.is_available() {
test_filter_impl_test_suite(FilterImplPerInstructionSet::SVE);
}
}
#[test]
#[cfg(target_arch = "aarch64")]
fn test_filter_implementation_neon() {
test_filter_impl_test_suite(FilterImplPerInstructionSet::Neon);
}
#[test]
fn test_filter_implementation_scalar() {
test_filter_impl_test_suite(FilterImplPerInstructionSet::Scalar);
}
#[cfg(target_arch = "x86_64")]
fn max_val_strategy() -> impl proptest::strategy::Strategy<Value = u32> {
proptest::prop_oneof![
0u32..10u32,
255u32..258u32,
proptest::prelude::Just(1u32 << 25),
proptest::prelude::Just(u32::MAX - 1),
proptest::prelude::Just(u32::MAX),
]
}
fn vals_strategy() -> impl proptest::strategy::Strategy<Value = Vec<u32>> {
proptest::prop_oneof![
proptest::collection::vec(proptest::prelude::any::<u32>(), 0..300),
max_val_strategy()
.prop_flat_map(|max_val| { proptest::collection::vec(0..=max_val, 0..300) })
]
}
proptest::proptest! {
#[test]
fn test_filter_compare_scalar_and_avx2_impl_proptest(
start in proptest::prelude::any::<u32>(),
end in proptest::prelude::any::<u32>(),
fn test_filter_compare_scalar_and_impls_impl_proptest(
start in 0u32..400u32,
end in 0u32..400u32,
offset in 0u32..2u32,
mut vals in proptest::collection::vec(0..u32::MAX, 0..30)) {
if FilterImplPerInstructionSet::AVX2.is_available() {
let mut vals_clone = vals.clone();
FilterImplPerInstructionSet::AVX2.filter_vec_in_place(start..=end, offset, &mut vals);
FilterImplPerInstructionSet::Scalar.filter_vec_in_place(start..=end, offset, &mut vals_clone);
assert_eq!(&vals, &vals_clone);
}
vals in vals_strategy()) {
for implementation in available_impls() {
if implementation == FilterImplPerInstructionSet::Scalar {
continue;
}
let mut impl_output = vals.clone();
let mut scalar_output = vals.clone();
implementation.filter_vec_in_place(start..=end, offset, &mut impl_output);
FilterImplPerInstructionSet::Scalar.filter_vec_in_place(start..=end, offset, &mut scalar_output);
assert_eq!(&impl_output, &scalar_output);
}
}
}
}

118
bitpacker/src/filter_vec/neon.rs Normal file
View File

@@ -0,0 +1,118 @@
use std::arch::aarch64::*;
use std::ops::RangeInclusive;
const NUM_LANES: usize = 4;
// Compacts matching lanes to the front using a byte-level shuffle.
// `mask` is a 4-bit value: bit k=1 means lane k should appear in the output.
#[inline]
#[target_feature(enable = "neon")]
unsafe fn compact(data: uint32x4_t, mask: u8) -> uint32x4_t {
unsafe {
// SAFETY: mask is always in [0, 15] by construction (max sum of [1,2,4,8]).
// BYTE_SHUFFLE_TABLE has 16 entries, so this is always in bounds.
let shuffle = BYTE_SHUFFLE_TABLE.get_unchecked(mask as usize);
let shuffle_vec = vld1q_u8(shuffle.as_ptr());
vreinterpretq_u32_u8(vqtbl1q_u8(vreinterpretq_u8_u32(data), shuffle_vec))
}
}
// Safe (not unsafe) because NEON is mandatory on aarch64: no runtime feature check needed.
#[inline(never)]
pub fn filter_vec_in_place(range: RangeInclusive<u32>, offset: u32, output: &mut Vec<u32>) {
let num_words = output.len() / NUM_LANES;
let mut output_len = unsafe {
filter_vec_neon_aux(
output.as_ptr(),
range.clone(),
output.as_mut_ptr(),
offset,
num_words,
)
};
let remainder_start = num_words * NUM_LANES;
for i in remainder_start..output.len() {
let val = output[i];
output[output_len] = offset + i as u32;
output_len += if range.contains(&val) { 1 } else { 0 };
}
output.truncate(output_len);
}
#[target_feature(enable = "neon")]
unsafe fn filter_vec_neon_aux(
input: *const u32,
range: RangeInclusive<u32>,
output: *mut u32,
offset: u32,
num_words: usize,
) -> usize {
unsafe {
let mut input = input;
let mut output_tail = output;
let range_start_simd = vdupq_n_u32(*range.start());
let range_end_simd = vdupq_n_u32(*range.end());
let mut ids = vld1q_u32([offset, offset + 1, offset + 2, offset + 3].as_ptr());
let shift = vdupq_n_u32(NUM_LANES as u32);
let bit_weights = vld1q_u32([1u32, 2, 4, 8].as_ptr());
for _ in 0..num_words {
let word = vld1q_u32(input);
// Unsigned compares: CMHS (compare higher or same) tests `word >= start`
// and `end >= word`. ANDing both gives the inside-range mask directly,
// which is cheaper than computing `outside` and then negating.
let ge_start = vcgeq_u32(word, range_start_simd);
let le_end = vcleq_u32(word, range_end_simd);
// inside[k] = 0xFFFFFFFF if val[k] is in range, 0 otherwise.
let inside = vandq_u32(ge_start, le_end);
// Build the 4-bit mask: AND bit_weights with the inside lane mask, so each
// inside lane contributes its bit_weight (1, 2, 4, or 8). Summing yields the
// 4-bit mask in one addv.
let inside_bits = vandq_u32(bit_weights, inside);
let mask = vaddvq_u32(inside_bits) as u8;
// mask is mathematically bounded: max value is 1+2+4+8=15 (all lanes match)
debug_assert!(mask <= 15, "mask must fit in 4 bits: {}", mask);
// Count of matching lanes = popcount(mask). Derives the count directly from
// the mask instead of running a parallel SIMD reduction over `outside`.
let added_len = mask.count_ones() as usize;
// Safe because mask is guaranteed to be in [0, 15]
let filtered_ids = compact(ids, mask);
vst1q_u32(output_tail, filtered_ids);
output_tail = output_tail.add(added_len);
ids = vaddq_u32(ids, shift);
input = input.add(NUM_LANES);
}
output_tail.offset_from(output) as usize
}
}
// Byte shuffle patterns to compact matching lanes to the front of the vector.
// Index is a 4-bit mask: bit k=1 means lane k (bytes 4k..4k+3) is in-range.
// The j-th set bit determines which input lane goes to output position j.
const BYTE_SHUFFLE_TABLE: [[u8; 16]; 16] = [
[
16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16,
], // 0b0000: none
[0, 1, 2, 3, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16], // 0b0001: lane 0
[4, 5, 6, 7, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16], // 0b0010: lane 1
[0, 1, 2, 3, 4, 5, 6, 7, 16, 16, 16, 16, 16, 16, 16, 16], // 0b0011: lanes 0,1
[8, 9, 10, 11, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16], // 0b0100: lane 2
[0, 1, 2, 3, 8, 9, 10, 11, 16, 16, 16, 16, 16, 16, 16, 16], // 0b0101: lanes 0,2
[4, 5, 6, 7, 8, 9, 10, 11, 16, 16, 16, 16, 16, 16, 16, 16], // 0b0110: lanes 1,2
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 16, 16, 16, 16], // 0b0111: lanes 0,1,2
[
12, 13, 14, 15, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16,
], // 0b1000: lane 3
[0, 1, 2, 3, 12, 13, 14, 15, 16, 16, 16, 16, 16, 16, 16, 16], // 0b1001: lanes 0,3
[4, 5, 6, 7, 12, 13, 14, 15, 16, 16, 16, 16, 16, 16, 16, 16], // 0b1010: lanes 1,3
[0, 1, 2, 3, 4, 5, 6, 7, 12, 13, 14, 15, 16, 16, 16, 16], // 0b1011: lanes 0,1,3
[8, 9, 10, 11, 12, 13, 14, 15, 16, 16, 16, 16, 16, 16, 16, 16], // 0b1100: lanes 2,3
[0, 1, 2, 3, 8, 9, 10, 11, 12, 13, 14, 15, 16, 16, 16, 16], // 0b1101: lanes 0,2,3
[4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 16, 16, 16], // 0b1110: lanes 1,2,3
[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15], // 0b1111: all lanes
];

260
bitpacker/src/filter_vec/sve.rs Normal file
View File

@@ -0,0 +1,260 @@
use std::ops::RangeInclusive;
// SVE vector length (in u32 lanes) is not a compile-time constant; query at runtime.
// Safe to call only when SVE is confirmed available via is_aarch64_feature_detected!("sve").
#[target_feature(enable = "sve")]
unsafe fn num_lanes() -> usize {
let vl: usize;
unsafe {
core::arch::asm!(
"cntw {vl}",
vl = out(reg) vl,
options(nostack, nomem, preserves_flags),
);
}
vl
}
// SAFETY: caller must ensure SVE is available (checked via is_aarch64_feature_detected!("sve")).
// Unlike NEON, SVE is optional on aarch64 and not guaranteed by the target architecture.
pub unsafe fn filter_vec_in_place(range: RangeInclusive<u32>, offset: u32, output: &mut Vec<u32>) {
if range.start() > range.end() {
output.clear();
return;
}
let vl = unsafe { num_lanes() };
let num_words = output.len() / vl;
let range_start = *range.start();
// Unsigned subtraction trick: val ∈ [lo, hi] ↔ (val - lo) ≤ᵤ (hi - lo).
// Values below lo wrap around to large u32, so the single unsigned ≤ excludes them.
let range_width = range.end().wrapping_sub(range_start);
let mut output_len = unsafe {
filter_vec_sve_aux(
output.as_ptr(),
range_start,
range_width,
output.as_mut_ptr(),
offset,
num_words,
vl,
)
};
let remainder_start = num_words * vl;
for i in remainder_start..output.len() {
let val = output[i];
output[output_len] = offset + i as u32;
output_len += if range.contains(&val) { 1 } else { 0 };
}
output.truncate(output_len);
}
// Register allocation for the asm! blocks:
// z0 ids_a (index vector for first half of each pair, advances by step2 each iter)
// z1 range_width broadcast
// z2 range_start broadcast
// z3 step2 broadcast (2 * vl)
// z4 ids_b (index vector for second half, = ids_a + step, advances by step2)
// z5 scratch: loaded word_a, then compacted_a
// z6 scratch: loaded word_b, then compacted_b
// p0 all-true predicate (ptrue p0.s)
// p1 in-range mask for word_a
// p2 in-range mask for word_b
#[target_feature(enable = "sve")]
unsafe fn filter_vec_sve_aux(
input: *const u32,
range_start: u32,
range_width: u32,
output: *mut u32,
offset: u32,
num_words: usize,
vl: usize,
) -> usize {
let num_pairs = num_words / 2;
let mut input_ptr = input;
let mut output_tail = output;
if num_pairs > 0 {
unsafe {
// We rely on asm! because the SVE intrinsics are not available in stable Rust.
// The code that follows was generated by Rustc nightly based on the intrinsics version
// at the bottom of this file.
core::arch::asm!(
// --- Setup ---
// All-true predicate for 32-bit lanes.
"ptrue p0.s",
// ids_a = [offset, offset+1, offset+2, ...]
"index z0.s, {offset:w}, #1",
// Broadcast scalars into SVE vectors.
"mov z1.s, {range_width:w}",
"mov z2.s, {range_start:w}",
// vl_gpr = number of 32-bit lanes (cntw).
"cntw {vl_gpr}",
// step2_bytes will first hold 2*vl (for the step2 vector), then 2*VL in bytes.
"lsl {step2_bytes}, {vl_gpr}, #1",
// z4 = step = [vl, vl, ...]; will become ids_b after the add below.
"mov z4.s, {vl_gpr:w}",
// z3 = step2 = [2*vl, 2*vl, ...], used to advance both id vectors each iter.
"mov z3.s, {step2_bytes:w}",
// Repurpose step2_bytes to hold the byte stride for advancing the input pointer
// by two full SVE vectors per iteration.
"rdvl {step2_bytes}, #2",
// ids_b = ids_a + step = [offset+vl, offset+vl+1, ...]
"add z4.s, z0.s, z4.s",
// --- Main loop: process two SVE vectors (ids_a and ids_b) per iteration ---
"0:",
// Load two consecutive SVE vectors from input.
"ld1w {{z5.s}}, p0/z, [{input}]",
"ld1w {{z6.s}}, p0/z, [{input}, #1, mul vl]",
// Advance input pointer by 2 * VL bytes.
"add {input}, {input}, {step2_bytes}",
// Unsigned shift: subtract range_start so in-range check becomes a single cmpu ≤.
"sub z5.s, z5.s, z2.s",
"sub z6.s, z6.s, z2.s",
// in_range: shifted value ≤ range_width (unsigned, so values below lo also fail).
"cmphs p1.s, p0/z, z1.s, z5.s",
"cmphs p2.s, p0/z, z1.s, z6.s",
// Count matching lanes; both cntp calls have independent inputs for OOO parallelism.
"cntp {cnt_a}, p0, p1.s",
"compact z5.s, p1, z0.s",
"compact z6.s, p2, z4.s",
"cntp {cnt_b}, p0, p2.s",
// Advance id vectors for the next iteration.
"add z0.s, z0.s, z3.s",
"add z4.s, z4.s, z3.s",
// Store compacted ids. Only the first cnt_a / cnt_b slots are valid; the rest
// will be overwritten by subsequent iterations before the final truncate.
"str z5, [{out}]",
"st1w {{z6.s}}, p0, [{out}, {cnt_a}, lsl #2]",
"add {out}, {out}, {cnt_a}, lsl #2",
"add {out}, {out}, {cnt_b}, lsl #2",
"subs {pairs}, {pairs}, #1",
"b.ne 0b",
// --- Operands ---
input = inout(reg) input_ptr,
out = inout(reg) output_tail,
pairs = inout(reg) num_pairs => _,
offset = in(reg) offset,
range_start = in(reg) range_start,
range_width = in(reg) range_width,
vl_gpr = out(reg) _,
step2_bytes = out(reg) _,
cnt_a = out(reg) _,
cnt_b = out(reg) _,
out("p0") _, out("p1") _, out("p2") _,
out("v0") _, out("v1") _, out("v2") _, out("v3") _,
out("v4") _, out("v5") _, out("v6") _,
options(nostack),
);
}
}
// Handle an odd trailing vector.
if num_words % 2 == 1 {
// ids_a for the odd word starts at offset + num_pairs * 2 * vl.
// input_ptr was advanced by the main loop and now points at the odd word.
let odd_offset =
offset.wrapping_add((num_pairs as u32).wrapping_mul(2).wrapping_mul(vl as u32));
unsafe {
core::arch::asm!(
"ptrue p0.s",
"index z0.s, {odd_offset:w}, #1",
"mov z1.s, {range_width:w}",
"mov z2.s, {range_start:w}",
"ld1w {{z3.s}}, p0/z, [{input}]",
"sub z3.s, z3.s, z2.s",
"cmphs p1.s, p0/z, z1.s, z3.s",
"cntp {cnt}, p0, p1.s",
"compact z0.s, p1, z0.s",
"str z0, [{out}]",
"add {out}, {out}, {cnt}, lsl #2",
odd_offset = in(reg) odd_offset,
range_width = in(reg) range_width,
range_start = in(reg) range_start,
input = in(reg) input_ptr,
out = inout(reg) output_tail,
cnt = out(reg) _,
out("p0") _, out("p1") _,
out("v0") _, out("v1") _, out("v2") _, out("v3") _,
options(nostack),
);
}
}
unsafe { output_tail.offset_from(output) as usize }
}
// SVE implements with intrinsics.
//
// #[target_feature(enable = "sve")]
// unsafe fn filter_vec_sve_aux(
// input: *const u32,
// range_start: u32,
// range_width: u32,
// output: *mut u32,
// offset: u32,
// num_words: usize,
// vl: usize,
// ) -> usize {
// unsafe {
// let all_true = svptrue_b32();
// let range_start_simd = svdup_n_u32(range_start);
// let range_width_simd = svdup_n_u32(range_width);
// // ids_a covers [offset .. offset+vl), ids_b covers the next vl ids.
// // Keeping them separate breaks the loop-carried dependency through ids so
// // both compact/cntp chains are fully independent within each unrolled body.
// let mut ids_a = svindex_u32(offset, 1);
// let step = svdup_n_u32(vl as u32);
// let step2 = svdup_n_u32(2 * vl as u32);
// let mut ids_b = svadd_u32_x(all_true, ids_a, step);
// let mut input = input;
// let mut output_tail = output;
// // Unrolled ×2: both cntp calls have independent inputs and execute in parallel.
// // The two output_tail updates are sequential but together cost 4+1+1=6 cy per
// // pair vs 5+5=10 cy for two scalar iterations, breaking the cntp latency chain.
// let num_pairs = num_words / 2;
// for _ in 0..num_pairs {
// let word_a = svld1_u32(all_true, input);
// let word_b = svld1_u32(all_true, input.add(vl));
// let shifted_a = svsub_u32_x(all_true, word_a, range_start_simd);
// let shifted_b = svsub_u32_x(all_true, word_b, range_start_simd);
// let in_range_a = svcmple_u32(all_true, shifted_a, range_width_simd);
// let in_range_b = svcmple_u32(all_true, shifted_b, range_width_simd);
// let compacted_a = svcompact_u32(in_range_a, ids_a);
// let compacted_b = svcompact_u32(in_range_b, ids_b);
// // cntp_a and cntp_b have independent inputs: OOO engine issues them in parallel.
// let added_len_a = svcntp_b32(all_true, in_range_a) as usize;
// let added_len_b = svcntp_b32(all_true, in_range_b) as usize;
// // Write the full vector — only the first added_len slots are valid.
// // Subsequent iterations overwrite the trailing zeros before truncate.
// svst1_u32(all_true, output_tail, compacted_a);
// output_tail = output_tail.add(added_len_a);
// svst1_u32(all_true, output_tail, compacted_b);
// output_tail = output_tail.add(added_len_b);
// ids_a = svadd_u32_x(all_true, ids_a, step2);
// ids_b = svadd_u32_x(all_true, ids_b, step2);
// input = input.add(2 * vl);
// }
// // Handle an odd trailing word.
// if num_words % 2 == 1 {
// let word = svld1_u32(all_true, input);
// let shifted = svsub_u32_x(all_true, word, range_start_simd);
// let in_range = svcmple_u32(all_true, shifted, range_width_simd);
// let added_len = svcntp_b32(all_true, in_range) as usize;
// let compacted_ids = svcompact_u32(in_range, ids_a);
// svst1_u32(all_true, output_tail, compacted_ids);
// output_tail = output_tail.add(added_len);
// }
// output_tail.offset_from(output) as usize
// }
// }

12
columnar/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy-columnar"
version = "0.6.0"
version = "0.7.0"
edition = "2024"
license = "MIT"
homepage = "https://github.com/quickwit-oss/tantivy"
@@ -12,10 +12,10 @@ categories = ["database-implementations", "data-structures", "compression"]
itertools = "0.14.0"
fastdivide = "0.4.0"
stacker = { version= "0.6", path = "../stacker", package="tantivy-stacker"}
sstable = { version= "0.6", path = "../sstable", package = "tantivy-sstable" }
common = { version= "0.10", path = "../common", package = "tantivy-common" }
tantivy-bitpacker = { version= "0.9", path = "../bitpacker/" }
stacker = { version= "0.7", path = "../stacker", package="tantivy-stacker"}
sstable = { version= "0.7", path = "../sstable", package = "tantivy-sstable" }
common = { version= "0.11", path = "../common", package = "tantivy-common" }
tantivy-bitpacker = { version= "0.10", path = "../bitpacker/" }
serde = "1.0.152"
downcast-rs = "2.0.1"
@@ -23,7 +23,7 @@ downcast-rs = "2.0.1"
proptest = "1"
more-asserts = "0.3.1"
rand = "0.9"
binggan = "0.14.0"
binggan = "0.17.0"
[[bench]]
name = "bench_merge"

97
columnar/src/block_accessor.rs
View File

@@ -15,9 +15,37 @@ impl<T: PartialOrd + Copy + std::fmt::Debug + Send + Sync + 'static + Default>
{
#[inline]
pub fn fetch_block<'a>(&'a mut self, docs: &'a [u32], accessor: &Column<T>) {
if accessor.index.get_cardinality().is_full() {
self.val_cache.resize(docs.len(), T::default());
accessor.values.get_vals(docs, &mut self.val_cache);
self.fetch_block_with_is_full(docs, accessor, accessor.index.get_cardinality().is_full());
}
/// Like [`Self::fetch_block`] but takes the column's fullness instead of querying
/// `accessor.index.get_cardinality()` each call — for callers that know it up front (e.g.
/// checked once at construction). `is_full` must equal
/// `accessor.index.get_cardinality().is_full()`.
#[inline]
pub fn fetch_block_with_is_full<'a>(
&'a mut self,
docs: &'a [u32],
accessor: &Column<T>,
is_full: bool,
) {
if is_full {
// Skip the resize when already the right length (common case: fixed-size blocks).
if self.val_cache.len() != docs.len() {
self.val_cache.resize(docs.len(), T::default());
}
// When the docs form a contiguous ascending run we can fetch the values
// as a single range. This lets codecs (e.g. bitpacked) bulk-decode the
// slice instead of gathering value-by-value, and avoids per-value dynamic
// dispatch. `docs` is always sorted ascending and free of duplicates here,
// so comparing the endpoints is enough to detect contiguity.
if is_contiguous(docs) {
accessor
.values
.get_range(docs[0] as u64, &mut self.val_cache);
} else {
accessor.values.get_vals(docs, &mut self.val_cache);
}
} else {
self.docid_cache.clear();
self.row_id_cache.clear();
@@ -33,14 +61,14 @@ impl<T: PartialOrd + Copy + std::fmt::Debug + Send + Sync + 'static + Default>
&mut self,
docs: &[u32],
accessor: &Column<T>,
missing: Option<T>,
missing_opt: Option<T>,
) {
self.fetch_block(docs, accessor);
// no missing values
if accessor.index.get_cardinality().is_full() {
return;
}
let Some(missing) = missing else {
let Some(missing) = missing_opt else {
return;
};
@@ -158,6 +186,22 @@ impl<T: PartialOrd + Copy + std::fmt::Debug + Send + Sync + 'static + Default>
}
}
/// Returns true if `docs` is a contiguous ascending run `[d, d + 1, ..., d + n - 1]`.
///
/// Assumes `docs` is sorted ascending and free of duplicates (the invariant for the
/// doc blocks passed to `fetch_block`), so comparing the endpoints is sufficient.
#[inline]
fn is_contiguous(docs: &[u32]) -> bool {
let (Some(&first), Some(&last)) = (docs.first(), docs.last()) else {
return false;
};
debug_assert!(
docs.windows(2).all(|w| w[0] < w[1]),
"fetch_block requires docs sorted ascending without duplicates"
);
(last - first) as usize + 1 == docs.len()
}
/// Given two sorted lists of docids `docs` and `hits`, hits is a subset of `docs`.
/// Return all docs that are not in `hits`.
fn find_missing_docs<F>(docs: &[u32], hits: &[u32], mut callback: F)
@@ -191,6 +235,7 @@ where F: FnMut(u32) {
}
#[cfg(test)]
#[allow(clippy::field_reassign_with_default)]
mod tests {
use super::*;
@@ -287,4 +332,46 @@ mod tests {
assert_eq!(accessor.docid_cache, vec![0]);
assert_eq!(accessor.val_cache, vec![1]);
}
#[test]
fn test_is_contiguous() {
assert!(!is_contiguous(&[]));
assert!(is_contiguous(&[5]));
assert!(is_contiguous(&[5, 6, 7, 8]));
assert!(is_contiguous(&[0, 1, 2]));
assert!(!is_contiguous(&[5, 7, 8]));
assert!(!is_contiguous(&[0, 1, 3]));
}
#[test]
fn test_fetch_block_contiguous_and_gather_match() {
use crate::column_index::ColumnIndex;
use crate::column_values::{
ALL_U64_CODEC_TYPES, serialize_and_load_u64_based_column_values,
};
let vals: Vec<u64> = (0..200u64).map(|i| i * 7 + 3).collect();
let values =
serialize_and_load_u64_based_column_values::<u64>(&&vals[..], &ALL_U64_CODEC_TYPES);
let column = Column {
index: ColumnIndex::Full,
values,
};
let check = |accessor: &mut ColumnBlockAccessor<u64>, docs: &[u32]| {
accessor.fetch_block(docs, &column);
let got: Vec<(u32, u64)> = accessor.iter_docid_vals(docs, &column).collect();
let expected: Vec<(u32, u64)> = docs.iter().map(|&d| (d, vals[d as usize])).collect();
assert_eq!(got, expected);
};
let mut accessor = ColumnBlockAccessor::<u64>::default();
// Contiguous block -> get_range fast path.
check(&mut accessor, &(10..74).collect::<Vec<u32>>());
// Non-contiguous block -> get_vals gather path.
check(&mut accessor, &[0, 5, 9, 100, 199]);
// Single doc and full span.
check(&mut accessor, &[42]);
check(&mut accessor, &(0..200).collect::<Vec<u32>>());
}
}

12
columnar/src/column_values/mod.rs
View File

@@ -119,8 +119,18 @@ pub trait ColumnValues<T: PartialOrd = u64>: Send + Sync + DowncastSync {
/// the segment's `maxdoc`.
#[inline(always)]
fn get_range(&self, start: u64, output: &mut [T]) {
for (out, idx) in output.iter_mut().zip(start..) {
let mut out_chunks = output.chunks_exact_mut(4);
let mut idx = start;
for out_x4 in out_chunks.by_ref() {
out_x4[0] = self.get_val(idx as u32);
out_x4[1] = self.get_val((idx + 1) as u32);
out_x4[2] = self.get_val((idx + 2) as u32);
out_x4[3] = self.get_val((idx + 3) as u32);
idx += 4;
}
for out in out_chunks.into_remainder() {
*out = self.get_val(idx as u32);
idx += 1;
}
}

16
columnar/src/column_values/u64_based/tests.rs
View File

@@ -121,6 +121,22 @@ pub(crate) fn create_and_validate<TColumnCodec: ColumnCodec>(
reader.get_vals(&all_docs, &mut buffer);
assert_eq!(vals, buffer);
// Validate `get_range` over the full column and a sub-range. The sub-range starts
// at a non-zero offset to exercise the entrance-ramp alignment of the batch decode.
buffer.resize(all_docs.len(), 0);
reader.get_range(0, &mut buffer);
assert_eq!(vals, buffer, "get_range (full) mismatch in data set {name}");
if vals.len() >= 2 {
let start = 1usize;
buffer.resize(vals.len() - start, 0);
reader.get_range(start as u64, &mut buffer);
assert_eq!(
&vals[start..],
&buffer[..],
"get_range (sub-range) mismatch in data set {name}"
);
}
if !vals.is_empty() {
let test_rand_idx = rand::rng().random_range(0..=vals.len() - 1);
let expected_positions: Vec<u32> = vals

4
common/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy-common"
version = "0.10.0"
version = "0.11.0"
authors = ["Paul Masurel <paul@quickwit.io>", "Pascal Seitz <pascal@quickwit.io>"]
license = "MIT"
edition = "2024"
@@ -19,6 +19,6 @@ time = { version = "0.3.47", features = ["serde-well-known"] }
serde = { version = "1.0.136", features = ["derive"] }
[dev-dependencies]
binggan = "0.14.0"
binggan = "0.17.0"
proptest = "1.0.0"
rand = "0.9"

14
common/src/bitset.rs
View File

@@ -47,6 +47,9 @@ impl TinySet {
TinySet(val)
}
/// An empty `TinySet` constant.
pub const EMPTY: TinySet = TinySet(0u64);
/// Returns an empty `TinySet`.
#[inline]
pub fn empty() -> TinySet {
@@ -193,8 +196,6 @@ impl TinySet {
#[derive(Clone)]
pub struct BitSet {
tinysets: Box<[TinySet]>,
// Tracking `len` on every insert/remove adds overhead even when `len()` is never called.
// Consider removing if `len()` usage is rare or not on a hot path.
len: u64,
max_value: u32,
}
@@ -254,7 +255,6 @@ impl BitSet {
/// Removes all elements from the `BitSet`.
pub fn clear(&mut self) {
self.len = 0;
for tinyset in self.tinysets.iter_mut() {
*tinyset = TinySet::empty();
}
@@ -274,11 +274,6 @@ impl BitSet {
}
}
/// Estimate the heap memory consumption of this `BitSet` in bytes.
pub fn get_memory_consumption(&self) -> usize {
self.tinysets.len() * std::mem::size_of::<TinySet>()
}
/// Returns the number of elements in the `BitSet`.
#[inline]
pub fn len(&self) -> usize {
@@ -322,9 +317,6 @@ impl BitSet {
.map(|delta_bucket| bucket + delta_bucket as u32)
}
/// Returns the maximum number of elements in the bitset.
///
/// Warning: The largest element the bitset can contain is `max_value - 1`.
#[inline]
pub fn max_value(&self) -> u32 {
self.max_value

2
common/src/file_slice.rs
View File

@@ -121,7 +121,7 @@ pub struct FileSlice {
impl fmt::Debug for FileSlice {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
write!(f, "FileSlice({:?}, {:?})", &self.data, self.range)
write!(f, "FileSlice({:?}, {:?})", self.data, self.range)
}
}

1
doc/src/SUMMARY.md
View File

@@ -8,7 +8,6 @@
- [Index Sorting](./index_sorting.md)
- [Innerworkings](./innerworkings.md)
- [Inverted index](./inverted_index.md)
- [Storage Abstraction](./storage_abstraction.md)
- [Best practise](./inverted_index.md)
[Frequently Asked Questions](./faq.md)

76
doc/src/storage_abstraction.md
View File

@@ -1,76 +0,0 @@
# Storage Abstraction — Design Notes
## Problem
tantivy's query engine needs to work with pluggable `SegmentReader` implementations while preserving the monomorphized fast path that avoids `Box<dyn Postings>` vtable
overhead in tight scoring loops (`advance()`, `doc()`, `score()`) or similar cases.
## Requirements
- **Pluggable `SegmentReader`.** External crates can provide their own `SegmentReader` implementation (with their own `InvertedIndexReader`, postings types, etc.) and tantivy's query engine works with it.
- **No performance regression.** tantivy's default path (`SegmentPostings` → `TermScorer<SegmentPostings>` → block WAND) must remain monomorphized — no boxing, no vtable dispatch in scoring loops.
- **Arbitrary implementations without recompiling tantivy.** The design must not require a fixed set of implementations known at tantivy compile time. External crates depend on tantivy, not the reverse.
- **Query code is backend-agnostic.** Adding a new `SegmentReader` implementation must not require changes to `TermWeight`, `PhraseWeight`, `AutomatonWeight`, or any other query code.
- **Non-viral API.** `Searcher`, `Index`, `Weight`, and other public types are not generic over the backend. Users don't need to thread a type parameter through their code.
## Current Design
### Trait hierarchy
- **`SegmentReader`** — trait for accessing a segment's data. Returns `Arc<dyn DynInvertedIndexReader>` from `inverted_index(field)`. `TantivySegmentReader` is the default implementation.
- **`DynInvertedIndexReader`** — object-safe trait for dynamic dispatch. Returns `Box<dyn Postings>`. Used as `Arc<dyn DynInvertedIndexReader>`.
- **`InvertedIndexReader`** — typed trait with `type Postings` and `type DocSet` associated types. `TantivyInvertedIndexReader` implements this with `Postings = SegmentPostings`. There is a blanket impl of `InvertedIndexReader` for `dyn DynInvertedIndexReader` with `Postings = Box<dyn Postings>`.
### `try_downcast_and_call!` macro
The macro attempts to downcast `&dyn DynInvertedIndexReader` to `&TantivyInvertedIndexReader`. The body is compiled twice — once with the concrete reader (typed postings, monomorphized) and once with the dyn fallback (boxed postings).
```rust
try_downcast_and_call!(inverted_index.as_ref(), |reader| {
let postings = reader.read_postings_from_terminfo(&term_info, option)?;
TermScorer::new(postings, fieldnorm_reader, similarity_weight)
})
```
This replaced the earlier `TypedInvertedIndexReaderCb` trait + struct pattern, which required creating a struct for every call site to serve as a "generic closure."
## Rejected approaches
### Specialized methods on `DynInvertedIndexReader`
Adding methods like `build_term_scorer()`, `build_phrase_scorer()`, `fill_bitset_from_terminfo()` to `DynInvertedIndexReader` was rejected. This forces every implementor to reimplement scoring logic for each query type — a combinatorial explosion that couples the reader to every query shape. The reader should only know how to produce postings, not how to build scorers. It also prevents supporting arbitrary query types without changing the trait.
### Feature-gated types for external readers
Using `#[cfg(feature = "quickwit")]` branches in the macro to add additional downcast targets. Requires recompiling tantivy for each reader and doesn't scale to arbitrary `SegmentReader` / `InvertedIndexReader` implementations.
### Reader-side dispatch with a callback trait
A method like `fn with_typed_reader(&self, cb: &mut dyn TypedCb<R>) -> R` on `DynInvertedIndexReader` would let the reader dispatch the callback with its concrete type. But the generic `R` parameter makes the trait not object-safe. Working around this with type erasure (storing results in the callback via `Any`) is complex and fragile.
## Planned: `TypedSegmentReader` trait for external fast paths
The current `try_downcast_and_call!` hardcodes `TantivyInvertedIndexReader`. To give external crates the monomorphized fast path, the downcast target should be a **trait with associated types**, not a specific concrete struct.
```rust
trait TypedSegmentReader: SegmentReader {
type InvertedIndexReader: InvertedIndexReader;
// future: type FastFieldReader: ...;
// future: type StoreReader: ...;
fn typed_inverted_index(&self, field: Field) -> &Self::InvertedIndexReader;
}
```
The dispatch downcasts `dyn SegmentReader` (via `as_any()`) to a concrete type that implements `TypedSegmentReader`, then the body works generically through the associated types. The body is compiled once per registered concrete type but is written against the trait — it never names `TantivyInvertedIndexReader` or `SegmentPostings` directly.
- External crates implement `TypedSegmentReader` with their own associated types and get the monomorphized fast path.
- One dispatch point covers all typed sub-components (inverted index, fast fields, store reader, etc.).
- Query weight code is fully generic — adding a new backend doesn't touch any query code.
- This does **not** mean query-specific methods on `SegmentReader`. The trait provides typed access to sub-components, not knowledge of query shapes.
### Open question: downcast chain registration
The concrete type must still be known for the `Any` downcast. The dispatch needs a list of concrete types to try. Since tantivy cannot depend on external crates, this list can't live in tantivy itself.
A macro invoked by the final binary could generate the downcast chain with all `TypedSegmentReader` implementors. Not yet designed.

2
examples/custom_collector.rs
View File

@@ -70,7 +70,7 @@ impl Collector for StatsCollector {
fn for_segment(
&self,
_segment_local_id: u32,
segment_reader: &dyn SegmentReader,
segment_reader: &SegmentReader,
) -> tantivy::Result<StatsSegmentCollector> {
let fast_field_reader = segment_reader.fast_fields().u64(&self.field)?;
Ok(StatsSegmentCollector {

2
examples/date_time_field.rs
View File

@@ -60,7 +60,7 @@ fn main() -> tantivy::Result<()> {
let count_docs = searcher.search(&*query, &TopDocs::with_limit(4).order_by_score())?;
assert_eq!(count_docs.len(), 1);
for (_score, doc_address) in count_docs {
let retrieved_doc = searcher.doc(doc_address)?;
let retrieved_doc = searcher.doc::<TantivyDocument>(doc_address)?;
assert!(retrieved_doc
.get_first(occurred_at)
.unwrap()

4
examples/faceted_search_with_tweaked_score.rs
View File

@@ -65,7 +65,7 @@ fn main() -> tantivy::Result<()> {
);
let top_docs_by_custom_score =
// Call TopDocs with a custom tweak score
TopDocs::with_limit(2).tweak_score(move |segment_reader: &dyn SegmentReader| {
TopDocs::with_limit(2).tweak_score(move |segment_reader: &SegmentReader| {
let ingredient_reader = segment_reader.facet_reader("ingredient").unwrap();
let facet_dict = ingredient_reader.facet_dict();
@@ -91,7 +91,7 @@ fn main() -> tantivy::Result<()> {
.iter()
.map(|(_, doc_id)| {
searcher
.doc(*doc_id)
.doc::<TantivyDocument>(*doc_id)
.unwrap()
.get_first(title)
.and_then(|v| v.as_str().map(|el| el.to_string()))

38
examples/iterating_docs_and_positions.rs
View File

@@ -91,10 +91,46 @@ fn main() -> tantivy::Result<()> {
}
}
// Some other powerful operations (especially `.seek`) may be useful to consume these
// A `Term` is a text token associated with a field.
// Let's go through all docs containing the term `title:the` and access their position
let term_the = Term::from_field_text(title, "the");
// Some other powerful operations (especially `.skip_to`) may be useful to consume these
// posting lists rapidly.
// You can check for them in the [`DocSet`](https://docs.rs/tantivy/~0/tantivy/trait.DocSet.html) trait
// and the [`Postings`](https://docs.rs/tantivy/~0/tantivy/trait.Postings.html) trait
// Also, for some VERY specific high performance use case like an OLAP analysis of logs,
// you can get better performance by accessing directly the blocks of doc ids.
for segment_reader in searcher.segment_readers() {
// A segment contains different data structure.
// Inverted index stands for the combination of
// - the term dictionary
// - the inverted lists associated with each terms and their positions
let inverted_index = segment_reader.inverted_index(title)?;
// This segment posting object is like a cursor over the documents matching the term.
// The `IndexRecordOption` arguments tells tantivy we will be interested in both term
// frequencies and positions.
//
// If you don't need all this information, you may get better performance by decompressing
// less information.
if let Some(mut block_segment_postings) =
inverted_index.read_block_postings(&term_the, IndexRecordOption::Basic)?
{
loop {
let docs = block_segment_postings.docs();
if docs.is_empty() {
break;
}
// Once again these docs MAY contains deleted documents as well.
let docs = block_segment_postings.docs();
// Prints `Docs [0, 2].`
println!("Docs {docs:?}");
block_segment_postings.advance();
}
}
}
Ok(())
}

2
examples/phrase_prefix_search.rs
View File

@@ -67,7 +67,7 @@ fn main() -> Result<()> {
let mut titles = top_docs
.into_iter()
.map(|(_score, doc_address)| {
let doc = searcher.doc(doc_address)?;
let doc = searcher.doc::<TantivyDocument>(doc_address)?;
let title = doc
.get_first(title)
.and_then(|v| v.as_str())

2
examples/snippet.rs
View File

@@ -55,7 +55,7 @@ fn main() -> tantivy::Result<()> {
let snippet_generator = SnippetGenerator::create(&searcher, &*query, body)?;
for (score, doc_address) in top_docs {
let doc = searcher.doc(doc_address)?;
let doc = searcher.doc::<TantivyDocument>(doc_address)?;
let snippet = snippet_generator.snippet_from_doc(&doc);
println!("Document score {score}:");
println!("title: {}", doc.get_first(title).unwrap().as_str().unwrap());

4
examples/warmer.rs
View File

@@ -43,7 +43,7 @@ impl DynamicPriceColumn {
}
}
pub fn price_for_segment(&self, segment_reader: &dyn SegmentReader) -> Option<Arc<Vec<Price>>> {
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()
}
@@ -157,7 +157,7 @@ fn main() -> tantivy::Result<()> {
let query = query_parser.parse_query("cooking")?;
let searcher = reader.searcher();
let score_by_price = move |segment_reader: &dyn SegmentReader| {
let score_by_price = move |segment_reader: &SegmentReader| {
let price = price_dynamic_column
.price_for_segment(segment_reader)
.unwrap();

2
query-grammar/Cargo.toml
View File

@@ -1,6 +1,6 @@
[package]
name = "tantivy-query-grammar"
version = "0.25.0"
version = "0.26.0"
authors = ["Paul Masurel <paul.masurel@gmail.com>"]
license = "MIT"
categories = ["database-implementations", "data-structures"]

93
query-grammar/src/query_grammar.rs
View File

@@ -327,7 +327,9 @@ fn exists(inp: &str) -> IResult<&str, UserInputLeaf> {
peek(alt((
value(
"",
satisfy(|c: char| c.is_whitespace() || ESCAPE_IN_WORD.contains(&c)),
satisfy(|c: char| {
c.is_whitespace() || (ESCAPE_IN_WORD.contains(&c) && c != '\\')
}),
),
eof,
))),
@@ -345,7 +347,9 @@ fn exists_precond(inp: &str) -> IResult<&str, (), ()> {
peek(alt((
value(
"",
satisfy(|c: char| c.is_whitespace() || ESCAPE_IN_WORD.contains(&c)),
satisfy(|c: char| {
c.is_whitespace() || (ESCAPE_IN_WORD.contains(&c) && c != '\\')
}),
),
eof,
))), // we need to check this isn't a wildcard query
@@ -707,6 +711,7 @@ fn regex(inp: &str) -> IResult<&str, UserInputLeaf> {
peek(alt((
value((), multispace1),
value((), char(')')),
value((), char('^')),
value((), eof),
))),
),
@@ -728,9 +733,10 @@ fn regex_infallible(inp: &str) -> JResult<&str, UserInputLeaf> {
peek(alt((
value((), multispace1),
value((), char(')')),
value((), char('^')),
value((), eof),
))),
"expected whitespace, closing parenthesis, or end of input",
"expected whitespace, closing parenthesis, boost, or end of input",
),
)(inp)
{
@@ -773,6 +779,10 @@ fn leaf(inp: &str) -> IResult<&str, UserInputAst> {
value((), multispace1),
value((), char(')')),
value((), eof),
value(
(),
satisfy(|c: char| ESCAPE_IN_WORD.contains(&c) && c != '\\'),
),
))),
),
|_| UserInputAst::from(UserInputLeaf::All),
@@ -805,6 +815,10 @@ fn leaf_infallible(inp: &str) -> JResult<&str, Option<UserInputAst>> {
value((), multispace1),
value((), char(')')),
value((), eof),
value(
(),
satisfy(|c: char| ESCAPE_IN_WORD.contains(&c) && c != '\\'),
),
))),
),
),
@@ -1045,18 +1059,43 @@ fn operand_leaf(inp: &str) -> IResult<&str, (Option<BinaryOperand>, Option<Occur
}
fn ast(inp: &str) -> IResult<&str, UserInputAst> {
let boolean_expr = map_res(
separated_pair(occur_leaf, multispace1, many1(operand_leaf)),
|(left, right)| aggregate_binary_expressions(left, right),
);
let single_leaf = map(occur_leaf, |(occur, ast)| {
if occur == Some(Occur::MustNot) {
ast.unary(Occur::MustNot)
} else {
ast
}
});
delimited(multispace0, alt((boolean_expr, single_leaf)), multispace0)(inp)
// Parse `occur_leaf` once, then conditionally extend into a boolean
// expression. The previous implementation used `alt((boolean_expr,
// single_leaf))` which, when the input was a single leaf with no
// following operand, would parse `occur_leaf` once for `boolean_expr`,
// fail at `multispace1`, backtrack, then re-parse `occur_leaf` for
// `single_leaf`. With recursively-nested groups like `(+(+(+a)))`, that
// doubling at every level produced O(2^n) parse time. Parsing once and
// peeking ahead for the operand keeps it O(n).
delimited(
multispace0,
|inp| {
let (rest, first) = occur_leaf(inp)?;
// Only fall back on `Err::Error` (recoverable), mirroring
// `alt`'s behaviour. `Err::Failure` and `Err::Incomplete`
// must propagate so cut points and streaming needs are not
// accidentally swallowed if they are ever introduced in the
// operand parsers.
match preceded(multispace1, many1(operand_leaf))(rest) {
Ok((rest, more)) => {
let combined = aggregate_binary_expressions(first, more)
.map_err(|_| nom::Err::Error(Error::new(inp, ErrorKind::MapRes)))?;
Ok((rest, combined))
}
Err(nom::Err::Error(_)) => {
let (occur, ast) = first;
let single = if occur == Some(Occur::MustNot) {
ast.unary(Occur::MustNot)
} else {
ast
};
Ok((rest, single))
}
Err(e) => Err(e),
}
},
multispace0,
)(inp)
}
fn ast_infallible(inp: &str) -> JResult<&str, UserInputAst> {
@@ -1726,6 +1765,8 @@ mod test {
test_parse_query_to_ast_helper("*", "*");
test_parse_query_to_ast_helper("(*)", "*");
test_parse_query_to_ast_helper("(* )", "*");
// All query with boost
test_parse_query_to_ast_helper("*^2", "(*)^2");
}
#[test]
@@ -1788,6 +1829,7 @@ mod test {
test_parse_query_to_ast_helper("a:b*", "\"a\":b*");
test_parse_query_to_ast_helper("a:*b", "\"a\":*b");
test_parse_query_to_ast_helper(r#"a:*def*"#, "\"a\":*def*");
test_parse_query_to_ast_helper("a:*\\:foo", "\"a\":*:foo");
}
#[test]
@@ -1852,6 +1894,8 @@ mod test {
},
_ => panic!("Expected a leaf"),
}
// Regex followed by `^boost`
test_parse_query_to_ast_helper(r#"foo:/bar/^2"#, r#"("foo":/bar/)^2"#);
}
#[test]
@@ -1891,4 +1935,23 @@ mod test {
r#"(+"field":'happy tax payer' +"other_field":1)"#,
);
}
// Regression test for https://github.com/quickwit-oss/tantivy/issues/2498:
// deeply nested parenthesized queries used to take O(2^n) time because the
// top-level `ast()` parser tried `boolean_expr` first and re-parsed the
// inner `occur_leaf` when it backtracked to `single_leaf`. Depth 60 would
// take ~10^18 operations under the regression; with the fix it parses
// instantly. We use `test_parse_query_to_ast_helper` so this test would
// never finish if the regression returned.
#[test]
fn test_parse_deeply_nested_query() {
let depth = 60;
let leading: String = "(".repeat(depth);
let trailing: String = ")".repeat(depth);
let query = format!("{leading}title:test{trailing}");
test_parse_query_to_ast_helper(&query, r#""title":test"#);
let query_with_plus = format!("+{leading}title:test{trailing}");
test_parse_query_to_ast_helper(&query_with_plus, r#""title":test"#);
}
}

6
src/aggregation/accessor_helpers.rs
View File

@@ -57,7 +57,7 @@ pub(crate) fn get_numeric_or_date_column_types() -> &'static [ColumnType] {
/// Get fast field reader or empty as default.
pub(crate) fn get_ff_reader(
reader: &dyn SegmentReader,
reader: &SegmentReader,
field_name: &str,
allowed_column_types: Option<&[ColumnType]>,
) -> crate::Result<(columnar::Column<u64>, ColumnType)> {
@@ -74,7 +74,7 @@ pub(crate) fn get_ff_reader(
}
pub(crate) fn get_dynamic_columns(
reader: &dyn SegmentReader,
reader: &SegmentReader,
field_name: &str,
) -> crate::Result<Vec<columnar::DynamicColumn>> {
let ff_fields = reader.fast_fields().dynamic_column_handles(field_name)?;
@@ -90,7 +90,7 @@ pub(crate) fn get_dynamic_columns(
///
/// Is guaranteed to return at least one column.
pub(crate) fn get_all_ff_reader_or_empty(
reader: &dyn SegmentReader,
reader: &SegmentReader,
field_name: &str,
allowed_column_types: Option<&[ColumnType]>,
fallback_type: ColumnType,

311
src/aggregation/agg_data.rs
View File

@@ -10,18 +10,18 @@ use crate::aggregation::accessor_helpers::{
};
use crate::aggregation::agg_req::{Aggregation, AggregationVariants, Aggregations};
use crate::aggregation::bucket::{
build_segment_filter_collector, build_segment_range_collector, CompositeAggReqData,
CompositeAggregation, CompositeSourceAccessors, FilterAggReqData, HistogramAggReqData,
HistogramBounds, IncludeExcludeParam, MissingTermAggReqData, RangeAggReqData,
SegmentHistogramCollector, TermMissingAgg, TermsAggReqData, TermsAggregation,
TermsAggregationInternal,
build_segment_filter_collector, build_segment_histogram_collector,
build_segment_range_collector, CompositeAggReqData, CompositeAggregation,
CompositeSourceAccessors, FilterAggReqData, HistogramAggReqData, HistogramBounds,
IncludeExcludeParam, MissingTermAggReqData, RangeAggReqData, TermMissingAgg, TermsAggReqData,
TermsAggregation, TermsAggregationInternal,
};
use crate::aggregation::metric::{
build_segment_stats_collector, AverageAggregation, CardinalityAggReqData,
CardinalityAggregationReq, CountAggregation, ExtendedStatsAggregation, MaxAggregation,
MetricAggReqData, MinAggregation, SegmentCardinalityCollector, SegmentExtendedStatsCollector,
SegmentPercentilesCollector, StatsAggregation, StatsType, SumAggregation, TopHitsAggReqData,
TopHitsSegmentCollector,
SegmentPercentilesCollector, StatsAggregation, StatsType, SumAggregation, TermOrdSet,
TopHitsAggReqData, TopHitsSegmentCollector, BITSET_MAX_TERM_ORD,
};
use crate::aggregation::segment_agg_result::{
GenericSegmentAggregationResultsCollector, SegmentAggregationCollector,
@@ -41,7 +41,7 @@ pub struct AggregationsSegmentCtx {
impl AggregationsSegmentCtx {
pub(crate) fn push_term_req_data(&mut self, data: TermsAggReqData) -> usize {
self.per_request.term_req_data.push(Some(Box::new(data)));
self.per_request.term_req_data.push(data);
self.per_request.term_req_data.len() - 1
}
pub(crate) fn push_cardinality_req_data(&mut self, data: CardinalityAggReqData) -> usize {
@@ -61,31 +61,25 @@ impl AggregationsSegmentCtx {
self.per_request.missing_term_req_data.len() - 1
}
pub(crate) fn push_histogram_req_data(&mut self, data: HistogramAggReqData) -> usize {
self.per_request
.histogram_req_data
.push(Some(Box::new(data)));
self.per_request.histogram_req_data.push(data);
self.per_request.histogram_req_data.len() - 1
}
pub(crate) fn push_range_req_data(&mut self, data: RangeAggReqData) -> usize {
self.per_request.range_req_data.push(Some(Box::new(data)));
self.per_request.range_req_data.push(data);
self.per_request.range_req_data.len() - 1
}
pub(crate) fn push_filter_req_data(&mut self, data: FilterAggReqData) -> usize {
self.per_request.filter_req_data.push(Some(Box::new(data)));
self.per_request.filter_req_data.push(data);
self.per_request.filter_req_data.len() - 1
}
pub(crate) fn push_composite_req_data(&mut self, data: CompositeAggReqData) -> usize {
self.per_request
.composite_req_data
.push(Some(Box::new(data)));
self.per_request.composite_req_data.push(data);
self.per_request.composite_req_data.len() - 1
}
#[inline]
pub(crate) fn get_term_req_data(&self, idx: usize) -> &TermsAggReqData {
self.per_request.term_req_data[idx]
.as_deref()
.expect("term_req_data slot is empty (taken)")
&self.per_request.term_req_data[idx]
}
#[inline]
pub(crate) fn get_cardinality_req_data(&self, idx: usize) -> &CardinalityAggReqData {
@@ -103,116 +97,6 @@ impl AggregationsSegmentCtx {
pub(crate) fn get_missing_term_req_data(&self, idx: usize) -> &MissingTermAggReqData {
&self.per_request.missing_term_req_data[idx]
}
#[inline]
pub(crate) fn get_histogram_req_data(&self, idx: usize) -> &HistogramAggReqData {
self.per_request.histogram_req_data[idx]
.as_deref()
.expect("histogram_req_data slot is empty (taken)")
}
#[inline]
pub(crate) fn get_range_req_data(&self, idx: usize) -> &RangeAggReqData {
self.per_request.range_req_data[idx]
.as_deref()
.expect("range_req_data slot is empty (taken)")
}
#[inline]
pub(crate) fn get_composite_req_data(&self, idx: usize) -> &CompositeAggReqData {
self.per_request.composite_req_data[idx]
.as_deref()
.expect("composite_req_data slot is empty (taken)")
}
// ---------- mutable getters ----------
#[inline]
pub(crate) fn get_metric_req_data_mut(&mut self, idx: usize) -> &mut MetricAggReqData {
&mut self.per_request.stats_metric_req_data[idx]
}
#[inline]
pub(crate) fn get_cardinality_req_data_mut(
&mut self,
idx: usize,
) -> &mut CardinalityAggReqData {
&mut self.per_request.cardinality_req_data[idx]
}
#[inline]
pub(crate) fn get_histogram_req_data_mut(&mut self, idx: usize) -> &mut HistogramAggReqData {
self.per_request.histogram_req_data[idx]
.as_deref_mut()
.expect("histogram_req_data slot is empty (taken)")
}
// ---------- take / put (terms, histogram, range) ----------
/// Move out the boxed Histogram request at `idx`, leaving `None`.
#[inline]
pub(crate) fn take_histogram_req_data(&mut self, idx: usize) -> Box<HistogramAggReqData> {
self.per_request.histogram_req_data[idx]
.take()
.expect("histogram_req_data slot is empty (taken)")
}
/// Put back a Histogram request into an empty slot at `idx`.
#[inline]
pub(crate) fn put_back_histogram_req_data(
&mut self,
idx: usize,
value: Box<HistogramAggReqData>,
) {
debug_assert!(self.per_request.histogram_req_data[idx].is_none());
self.per_request.histogram_req_data[idx] = Some(value);
}
/// Move out the boxed Range request at `idx`, leaving `None`.
#[inline]
pub(crate) fn take_range_req_data(&mut self, idx: usize) -> Box<RangeAggReqData> {
self.per_request.range_req_data[idx]
.take()
.expect("range_req_data slot is empty (taken)")
}
/// Put back a Range request into an empty slot at `idx`.
#[inline]
pub(crate) fn put_back_range_req_data(&mut self, idx: usize, value: Box<RangeAggReqData>) {
debug_assert!(self.per_request.range_req_data[idx].is_none());
self.per_request.range_req_data[idx] = Some(value);
}
/// Move out the boxed Filter request at `idx`, leaving `None`.
#[inline]
pub(crate) fn take_filter_req_data(&mut self, idx: usize) -> Box<FilterAggReqData> {
self.per_request.filter_req_data[idx]
.take()
.expect("filter_req_data slot is empty (taken)")
}
/// Put back a Filter request into an empty slot at `idx`.
#[inline]
pub(crate) fn put_back_filter_req_data(&mut self, idx: usize, value: Box<FilterAggReqData>) {
debug_assert!(self.per_request.filter_req_data[idx].is_none());
self.per_request.filter_req_data[idx] = Some(value);
}
/// Move out the Composite request at `idx`.
#[inline]
pub(crate) fn take_composite_req_data(&mut self, idx: usize) -> Box<CompositeAggReqData> {
self.per_request.composite_req_data[idx]
.take()
.expect("composite_req_data slot is empty (taken)")
}
/// Put back a Composite request into an empty slot at `idx`.
#[inline]
pub(crate) fn put_back_composite_req_data(
&mut self,
idx: usize,
value: Box<CompositeAggReqData>,
) {
debug_assert!(self.per_request.composite_req_data[idx].is_none());
self.per_request.composite_req_data[idx] = Some(value);
}
}
/// Each type of aggregation has its own request data struct. This struct holds
@@ -223,15 +107,14 @@ impl AggregationsSegmentCtx {
/// for a node with [AggKind::Terms]).
#[derive(Default)]
pub struct PerRequestAggSegCtx {
// Box for cheap take/put - Only necessary for bucket aggs that have sub-aggregations
/// TermsAggReqData contains the request data for a terms aggregation.
pub term_req_data: Vec<Option<Box<TermsAggReqData>>>,
pub term_req_data: Vec<TermsAggReqData>,
/// HistogramAggReqData contains the request data for a histogram aggregation.
pub histogram_req_data: Vec<Option<Box<HistogramAggReqData>>>,
pub histogram_req_data: Vec<HistogramAggReqData>,
/// RangeAggReqData contains the request data for a range aggregation.
pub range_req_data: Vec<Option<Box<RangeAggReqData>>>,
pub range_req_data: Vec<RangeAggReqData>,
/// FilterAggReqData contains the request data for a filter aggregation.
pub filter_req_data: Vec<Option<Box<FilterAggReqData>>>,
pub filter_req_data: Vec<FilterAggReqData>,
/// Shared by avg, min, max, sum, stats, extended_stats, count
pub stats_metric_req_data: Vec<MetricAggReqData>,
/// CardinalityAggReqData contains the request data for a cardinality aggregation.
@@ -241,7 +124,7 @@ pub struct PerRequestAggSegCtx {
/// MissingTermAggReqData contains the request data for a missing term aggregation.
pub missing_term_req_data: Vec<MissingTermAggReqData>,
/// CompositeAggReqData contains the request data for a composite aggregation.
pub composite_req_data: Vec<Option<Box<CompositeAggReqData>>>,
pub composite_req_data: Vec<CompositeAggReqData>,
/// Request tree used to build collectors.
pub agg_tree: Vec<AggRefNode>,
@@ -252,22 +135,22 @@ impl PerRequestAggSegCtx {
fn get_memory_consumption(&self) -> usize {
self.term_req_data
.iter()
.map(|b| b.as_ref().unwrap().get_memory_consumption())
.map(|t| t.get_memory_consumption())
.sum::<usize>()
+ self
.histogram_req_data
.iter()
.map(|b| b.as_ref().unwrap().get_memory_consumption())
.map(|t| t.get_memory_consumption())
.sum::<usize>()
+ self
.range_req_data
.iter()
.map(|b| b.as_ref().unwrap().get_memory_consumption())
.map(|t| t.get_memory_consumption())
.sum::<usize>()
+ self
.filter_req_data
.iter()
.map(|b| b.as_ref().unwrap().get_memory_consumption())
.map(|t| t.get_memory_consumption())
.sum::<usize>()
+ self
.stats_metric_req_data
@@ -292,7 +175,7 @@ impl PerRequestAggSegCtx {
+ self
.composite_req_data
.iter()
.map(|b| b.as_ref().map(|d| d.get_memory_consumption()).unwrap_or(0))
.map(|t| t.get_memory_consumption())
.sum::<usize>()
+ self.agg_tree.len() * std::mem::size_of::<AggRefNode>()
}
@@ -301,40 +184,16 @@ impl PerRequestAggSegCtx {
let idx = node.idx_in_req_data;
let kind = node.kind;
match kind {
AggKind::Terms => self.term_req_data[idx]
.as_deref()
.expect("term_req_data slot is empty (taken)")
.name
.as_str(),
AggKind::Terms => self.term_req_data[idx].name.as_str(),
AggKind::Cardinality => &self.cardinality_req_data[idx].name,
AggKind::StatsKind(_) => &self.stats_metric_req_data[idx].name,
AggKind::TopHits => &self.top_hits_req_data[idx].name,
AggKind::MissingTerm => &self.missing_term_req_data[idx].name,
AggKind::Histogram => self.histogram_req_data[idx]
.as_deref()
.expect("histogram_req_data slot is empty (taken)")
.name
.as_str(),
AggKind::DateHistogram => self.histogram_req_data[idx]
.as_deref()
.expect("histogram_req_data slot is empty (taken)")
.name
.as_str(),
AggKind::Range => self.range_req_data[idx]
.as_deref()
.expect("range_req_data slot is empty (taken)")
.name
.as_str(),
AggKind::Filter => self.filter_req_data[idx]
.as_deref()
.expect("filter_req_data slot is empty (taken)")
.name
.as_str(),
AggKind::Composite => self.composite_req_data[idx]
.as_deref()
.expect("composite_req_data slot is empty (taken)")
.name
.as_str(),
AggKind::Histogram => self.histogram_req_data[idx].name.as_str(),
AggKind::DateHistogram => self.histogram_req_data[idx].name.as_str(),
AggKind::Range => self.range_req_data[idx].name.as_str(),
AggKind::Filter => self.filter_req_data[idx].name.as_str(),
AggKind::Composite => self.composite_req_data[idx].name.as_str(),
}
}
@@ -412,13 +271,39 @@ pub(crate) fn build_segment_agg_collector(
Ok(Box::new(TermMissingAgg::new(req, node)?))
}
AggKind::Cardinality => {
let req_data = &mut req.get_cardinality_req_data_mut(node.idx_in_req_data);
Ok(Box::new(SegmentCardinalityCollector::from_req(
req_data.column_type,
node.idx_in_req_data,
req_data.accessor.clone(),
req_data.missing_value_for_accessor,
)))
let req_data = req.get_cardinality_req_data(node.idx_in_req_data);
// For str columns, choose the per-bucket entries representation
// based on the segment's column.max_value():
// * small (< BITSET_MAX_TERM_ORD): `BitSet`, pre-allocated, no promotion machinery.
// * large: `TermOrdSet` (sparse FxHashSet that promotes to a paged bitset).
// For non-str columns the `entries` field is unused (values go
// straight into the HLL sketch); we still pick `TermOrdSet`
// because its empty Sparse(FxHashSet) costs nothing.
let is_str = req_data.column_type == ColumnType::Str;
let max_term_ord_inclusive = if is_str {
req_data.accessor.max_value()
} else {
0
};
let collector: Box<dyn SegmentAggregationCollector> =
if is_str && max_term_ord_inclusive < BITSET_MAX_TERM_ORD {
Box::new(SegmentCardinalityCollector::<BitSet>::from_req(
req_data.column_type,
node.idx_in_req_data,
req_data.accessor.clone(),
req_data.missing_value_for_accessor,
max_term_ord_inclusive,
))
} else {
Box::new(SegmentCardinalityCollector::<TermOrdSet>::from_req(
req_data.column_type,
node.idx_in_req_data,
req_data.accessor.clone(),
req_data.missing_value_for_accessor,
max_term_ord_inclusive,
))
};
Ok(collector)
}
AggKind::StatsKind(stats_type) => {
let req_data = &mut req.per_request.stats_metric_req_data[node.idx_in_req_data];
@@ -433,7 +318,7 @@ pub(crate) fn build_segment_agg_collector(
SegmentExtendedStatsCollector::from_req(req_data, sigma),
)),
StatsType::Percentiles => {
let req_data = req.get_metric_req_data_mut(node.idx_in_req_data);
let req_data = req.get_metric_req_data(node.idx_in_req_data);
Ok(Box::new(
SegmentPercentilesCollector::from_req_and_validate(
req_data.field_type,
@@ -453,12 +338,8 @@ pub(crate) fn build_segment_agg_collector(
req_data.segment_ordinal,
)))
}
AggKind::Histogram => Ok(Box::new(SegmentHistogramCollector::from_req_and_validate(
req, node,
)?)),
AggKind::DateHistogram => Ok(Box::new(SegmentHistogramCollector::from_req_and_validate(
req, node,
)?)),
AggKind::Histogram => build_segment_histogram_collector(req, node),
AggKind::DateHistogram => build_segment_histogram_collector(req, node),
AggKind::Range => Ok(build_segment_range_collector(req, node)?),
AggKind::Filter => build_segment_filter_collector(req, node),
AggKind::Composite => Ok(Box::new(
@@ -520,7 +401,7 @@ impl AggKind {
/// Build AggregationsData by walking the request tree.
pub(crate) fn build_aggregations_data_from_req(
aggs: &Aggregations,
reader: &dyn SegmentReader,
reader: &SegmentReader,
segment_ordinal: SegmentOrdinal,
context: AggContextParams,
) -> crate::Result<AggregationsSegmentCtx> {
@@ -540,7 +421,7 @@ pub(crate) fn build_aggregations_data_from_req(
fn build_nodes(
agg_name: &str,
req: &Aggregation,
reader: &dyn SegmentReader,
reader: &SegmentReader,
segment_ordinal: SegmentOrdinal,
data: &mut AggregationsSegmentCtx,
is_top_level: bool,
@@ -773,23 +654,18 @@ fn build_nodes(
let schema = reader.schema();
let tokenizers = &data.context.tokenizers;
let query = filter_req.parse_query(schema, tokenizers)?;
let evaluator = crate::aggregation::bucket::DocumentQueryEvaluator::new(
query,
schema.clone(),
reader,
)?;
// Pre-allocate buffer for batch filtering
let max_doc = reader.max_doc();
let buffer_capacity = crate::docset::COLLECT_BLOCK_BUFFER_LEN.min(max_doc as usize);
let matching_docs_buffer = Vec::with_capacity(buffer_capacity);
let evaluator =
std::rc::Rc::new(crate::aggregation::bucket::DocumentQueryEvaluator::new(
query,
schema.clone(),
reader,
)?);
let idx_in_req_data = data.push_filter_req_data(FilterAggReqData {
name: agg_name.to_string(),
req: filter_req.clone(),
segment_reader: reader.clone_arc(),
segment_reader: reader.clone(),
evaluator,
matching_docs_buffer,
is_top_level,
});
let children = build_children(&req.sub_aggregation, reader, segment_ordinal, data)?;
@@ -804,7 +680,7 @@ fn build_nodes(
fn build_composite_node(
agg_name: &str,
reader: &dyn SegmentReader,
reader: &SegmentReader,
_segment_ordinal: SegmentOrdinal,
data: &mut AggregationsSegmentCtx,
sub_aggs: &Aggregations,
@@ -833,7 +709,7 @@ fn build_composite_node(
fn build_children(
aggs: &Aggregations,
reader: &dyn SegmentReader,
reader: &SegmentReader,
segment_ordinal: SegmentOrdinal,
data: &mut AggregationsSegmentCtx,
) -> crate::Result<Vec<AggRefNode>> {
@@ -852,7 +728,7 @@ fn build_children(
}
fn get_term_agg_accessors(
reader: &dyn SegmentReader,
reader: &SegmentReader,
field_name: &str,
missing: &Option<Key>,
) -> crate::Result<Vec<(Column<u64>, ColumnType)>> {
@@ -905,7 +781,7 @@ fn build_terms_or_cardinality_nodes(
agg_name: &str,
field_name: &str,
missing: &Option<Key>,
reader: &dyn SegmentReader,
reader: &SegmentReader,
segment_ordinal: SegmentOrdinal,
data: &mut AggregationsSegmentCtx,
sub_aggs: &Aggregations,
@@ -985,8 +861,12 @@ fn build_terms_or_cardinality_nodes(
let str_col = str_dict_column
.as_ref()
.expect("str_dict_column must exist for string column");
allowed_term_ids =
build_allowed_term_ids_for_str(str_col, &req.include, &req.exclude)?;
allowed_term_ids = build_allowed_term_ids_for_str(
str_col,
&req.include,
&req.exclude,
missing.is_some(),
)?;
};
let idx_in_req_data = data.push_term_req_data(TermsAggReqData {
accessor,
@@ -1002,10 +882,20 @@ fn build_terms_or_cardinality_nodes(
(idx_in_req_data, AggKind::Terms)
}
TermsOrCardinalityRequest::Cardinality(ref req) => {
// `str_dict_column` is computed once per field; for JSON paths
// with mixed types it's `Some` even on the numeric req_data.
// Cardinality only consults it for the str column path, so
// gate by column_type to avoid driving non-str collectors
// through the coupon-cache path.
let str_dict_column_for_req = if column_type == ColumnType::Str {
str_dict_column.clone()
} else {
None
};
let idx_in_req_data = data.push_cardinality_req_data(CardinalityAggReqData {
accessor,
column_type,
str_dict_column: str_dict_column.clone(),
str_dict_column: str_dict_column_for_req,
missing_value_for_accessor,
name: agg_name.to_string(),
req: req.clone(),
@@ -1025,16 +915,21 @@ fn build_terms_or_cardinality_nodes(
/// Builds a single BitSet of allowed term ordinals for a string dictionary column according to
/// include/exclude parameters.
///
/// When `reserve_missing_sentinel` is true, the bitset will have 1 additional slot for the missing
/// term ordinal
fn build_allowed_term_ids_for_str(
str_col: &StrColumn,
include: &Option<IncludeExcludeParam>,
exclude: &Option<IncludeExcludeParam>,
reserve_missing_sentinel: bool,
) -> crate::Result<Option<BitSet>> {
let mut allowed: Option<BitSet> = None;
let num_terms = str_col.dictionary().num_terms() as u32;
let missing_sentinel_adjustment = if reserve_missing_sentinel { 1 } else { 0 };
let allowed_capacity = str_col.dictionary().num_terms() as u32 + missing_sentinel_adjustment;
if let Some(include) = include {
// add matches
allowed = Some(BitSet::with_max_value(num_terms));
allowed = Some(BitSet::with_max_value(allowed_capacity));
let allowed = allowed.as_mut().unwrap();
for_each_matching_term_ord(str_col, include, |ord| allowed.insert(ord))?;
};
@@ -1042,7 +937,7 @@ fn build_allowed_term_ids_for_str(
if let Some(exclude) = exclude {
if allowed.is_none() {
// Start with all terms allowed
allowed = Some(BitSet::with_max_value_and_full(num_terms));
allowed = Some(BitSet::with_max_value_and_full(allowed_capacity));
}
let allowed = allowed.as_mut().unwrap();
for_each_matching_term_ord(str_col, exclude, |ord| allowed.remove(ord))?;

71
src/aggregation/agg_req.rs
View File

@@ -115,6 +115,71 @@ pub fn get_fast_field_names(aggs: &Aggregations) -> HashSet<String> {
fast_field_names
}
/// Validates that all fields referenced in the aggregation request exist in the schema
/// and are configured as fast fields.
///
/// This is a convenience function for upfront validation before executing aggregations.
/// Returns an error if any field doesn't exist or is not a fast field.
///
/// Validation is intentionally opt-in rather than baked into aggregation execution: the
/// default lenient behavior (returning empty results for missing fields) supports
/// schema evolution and federated queries where the same request runs against segments
/// or indices with different schemas.
///
/// # Example
/// ```
/// use tantivy::aggregation::agg_req::{Aggregations, validate_aggregation_fields_exist};
/// use tantivy::schema::{Schema, FAST};
/// use tantivy::Index;
///
/// # fn main() -> tantivy::Result<()> {
/// // Create a simple index
/// let mut schema_builder = Schema::builder();
/// schema_builder.add_f64_field("price", FAST);
/// let schema = schema_builder.build();
/// let index = Index::create_in_ram(schema);
///
/// // Parse aggregation request
/// let agg_req: Aggregations = serde_json::from_str(r#"{
/// "avg_price": { "avg": { "field": "price" } }
/// }"#)?;
///
/// let reader = index.reader()?;
/// let searcher = reader.searcher();
///
/// // Validate fields before executing
/// for segment_reader in searcher.segment_readers() {
/// validate_aggregation_fields_exist(&agg_req, segment_reader)?;
/// }
/// # Ok(())
/// # }
/// ```
pub fn validate_aggregation_fields_exist(
aggs: &Aggregations,
reader: &crate::SegmentReader,
) -> crate::Result<()> {
let field_names = get_fast_field_names(aggs);
let schema = reader.schema();
for field_name in field_names {
// Check if the field is either directly in the schema or could be part of a json field
// present in the schema, and verify it's a fast field.
if let Some((field, _path)) = schema.find_field(&field_name) {
let field_type = schema.get_field_entry(field).field_type();
if !field_type.is_fast() {
return Err(crate::TantivyError::SchemaError(format!(
"Field '{}' is not a fast field. Aggregations require fast fields.",
field_name
)));
}
} else {
return Err(crate::TantivyError::FieldNotFound(field_name));
}
}
Ok(())
}
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
/// All aggregation types.
pub enum AggregationVariants {
@@ -234,6 +299,12 @@ impl AggregationVariants {
_ => None,
}
}
pub(crate) fn as_sum(&self) -> Option<&SumAggregation> {
match &self {
AggregationVariants::Sum(sum) => Some(sum),
_ => None,
}
}
}
#[cfg(test)]

3
src/aggregation/agg_result.rs
View File

@@ -208,7 +208,8 @@ pub enum BucketEntries<T> {
}
impl<T> BucketEntries<T> {
fn iter<'a>(&'a self) -> Box<dyn Iterator<Item = &'a T> + 'a> {
/// Iterate over all bucket entries.
pub fn iter<'a>(&'a self) -> Box<dyn Iterator<Item = &'a T> + 'a> {
match self {
BucketEntries::Vec(vec) => Box::new(vec.iter()),
BucketEntries::HashMap(map) => Box::new(map.values()),

43
src/aggregation/agg_tests.rs
View File

@@ -1436,3 +1436,46 @@ fn test_aggregation_on_json_object_mixed_numerical_segments() {
)
);
}
#[test]
fn test_aggregation_field_validation_helper() {
// Test the standalone validation helper function for field validation
let index = get_test_index_2_segments(false).unwrap();
let reader = index.reader().unwrap();
let searcher = reader.searcher();
let segment_reader = searcher.segment_reader(0);
// Test with invalid field
let agg_req: Aggregations = serde_json::from_str(
r#"{
"avg_test": {
"avg": { "field": "nonexistent_field" }
}
}"#,
)
.unwrap();
let result =
crate::aggregation::agg_req::validate_aggregation_fields_exist(&agg_req, segment_reader);
assert!(result.is_err());
match result {
Err(crate::TantivyError::FieldNotFound(field_name)) => {
assert_eq!(field_name, "nonexistent_field");
}
_ => panic!("Expected FieldNotFound error, got: {:?}", result),
}
// Test with valid field
let agg_req: Aggregations = serde_json::from_str(
r#"{
"avg_test": {
"avg": { "field": "score" }
}
}"#,
)
.unwrap();
let result =
crate::aggregation::agg_req::validate_aggregation_fields_exist(&agg_req, segment_reader);
assert!(result.is_ok());
}

7
src/aggregation/bucket/composite/accessors.rs
View File

@@ -16,6 +16,7 @@ use crate::{SegmentReader, TantivyError};
/// Contains all information required by the SegmentCompositeCollector to perform the
/// composite aggregation on a segment.
#[derive(Debug, Clone)]
pub struct CompositeAggReqData {
/// The name of the aggregation.
pub name: String,
@@ -34,6 +35,7 @@ impl CompositeAggReqData {
}
/// Accessors for a single column in a composite source.
#[derive(Debug, Clone)]
pub struct CompositeAccessor {
/// The fast field column
pub column: Column<u64>,
@@ -48,6 +50,7 @@ pub struct CompositeAccessor {
}
/// Accessors to all the columns that belong to the field of a composite source.
#[derive(Debug, Clone)]
pub struct CompositeSourceAccessors {
/// The accessors for this source
pub accessors: Vec<CompositeAccessor>,
@@ -75,7 +78,7 @@ impl CompositeSourceAccessors {
///
/// Precomputes some values to make collection faster.
pub fn build_for_source(
reader: &dyn SegmentReader,
reader: &SegmentReader,
source: &CompositeAggregationSource,
// First option is None when no after key was set in the query, the
// second option is None when the after key was set but its value for
@@ -358,7 +361,7 @@ impl PrecomputedDateInterval {
///
/// Some column types (term, IP) might not have an exact representation of the
/// specified after key
#[derive(Debug)]
#[derive(Debug, Clone)]
pub enum PrecomputedAfterKey {
/// The after key could be exactly represented in the column space.
Exact(u64),

59
src/aggregation/bucket/composite/collector.rs
View File

@@ -21,7 +21,7 @@ use crate::aggregation::bucket::composite::map::{DynArrayHeapMap, MAX_DYN_ARRAY_
use crate::aggregation::bucket::{
CalendarInterval, CompositeAggregationSource, MissingOrder, Order,
};
use crate::aggregation::cached_sub_aggs::{CachedSubAggs, HighCardSubAggCache};
use crate::aggregation::buffered_sub_aggs::{BufferedSubAggs, HighCardSubAggBuffer};
use crate::aggregation::intermediate_agg_result::{
CompositeIntermediateKey, IntermediateAggregationResult, IntermediateAggregationResults,
IntermediateBucketResult, IntermediateCompositeBucketEntry, IntermediateCompositeBucketResult,
@@ -118,8 +118,8 @@ impl InternalValueRepr {
pub struct SegmentCompositeCollector {
/// One DynArrayHeapMap per parent bucket.
parent_buckets: Vec<DynArrayHeapMap<InternalValueRepr, CompositeBucketCollector>>,
accessor_idx: usize,
sub_agg: Option<CachedSubAggs<HighCardSubAggCache>>,
req_data: CompositeAggReqData,
sub_agg: Option<BufferedSubAggs<HighCardSubAggBuffer>>,
bucket_id_provider: BucketIdProvider,
/// Number of sources, needed when creating new DynArrayHeapMaps.
num_sources: usize,
@@ -132,10 +132,7 @@ impl SegmentAggregationCollector for SegmentCompositeCollector {
results: &mut IntermediateAggregationResults,
parent_bucket_id: BucketId,
) -> crate::Result<()> {
let name = agg_data
.get_composite_req_data(self.accessor_idx)
.name
.clone();
let name = self.req_data.name.clone();
let buckets = self.add_intermediate_bucket_result(agg_data, parent_bucket_id)?;
results.push(
@@ -152,13 +149,12 @@ impl SegmentAggregationCollector for SegmentCompositeCollector {
docs: &[crate::DocId],
agg_data: &mut AggregationsSegmentCtx,
) -> crate::Result<()> {
let mem_pre = self.get_memory_consumption();
let composite_agg_data = agg_data.take_composite_req_data(self.accessor_idx);
let mem_pre = self.get_memory_consumption(parent_bucket_id);
for doc in docs {
let mut visitor = CompositeKeyVisitor {
doc_id: *doc,
composite_agg_data: &composite_agg_data,
composite_agg_data: &self.req_data,
buckets: &mut self.parent_buckets[parent_bucket_id as usize],
sub_agg: &mut self.sub_agg,
bucket_id_provider: &mut self.bucket_id_provider,
@@ -166,13 +162,12 @@ impl SegmentAggregationCollector for SegmentCompositeCollector {
};
visitor.visit(0, true)?;
}
agg_data.put_back_composite_req_data(self.accessor_idx, composite_agg_data);
if let Some(sub_agg) = &mut self.sub_agg {
sub_agg.check_flush_local(agg_data)?;
}
let mem_delta = self.get_memory_consumption() - mem_pre;
let mem_delta = self.get_memory_consumption(parent_bucket_id) - mem_pre;
if mem_delta > 0 {
agg_data.context.limits.add_memory_consumed(mem_delta)?;
}
@@ -199,36 +194,49 @@ impl SegmentAggregationCollector for SegmentCompositeCollector {
}
Ok(())
}
fn compute_metric_value(
&self,
_bucket_id: BucketId,
_sub_agg_name: &str,
_sub_agg_property: &str,
_agg_data: &AggregationsSegmentCtx,
) -> Option<f64> {
// Composite is a multi-bucket agg with no single value to extract.
None
}
}
impl SegmentCompositeCollector {
fn get_memory_consumption(&self) -> u64 {
self.parent_buckets
.iter()
.map(|m| m.memory_consumption())
.sum()
fn get_memory_consumption(&self, parent_bucket_id: BucketId) -> u64 {
self.parent_buckets[parent_bucket_id as usize].memory_consumption()
}
pub(crate) fn from_req_and_validate(
req_data: &mut AggregationsSegmentCtx,
node: &AggRefNode,
) -> crate::Result<Self> {
validate_req(req_data, node.idx_in_req_data)?;
let composite_req_data =
req_data.per_request.composite_req_data[node.idx_in_req_data].clone();
validate_req(&composite_req_data)?;
req_data
.context
.limits
.add_memory_consumed(composite_req_data.get_memory_consumption() as u64)?;
let has_sub_aggregations = !node.children.is_empty();
let sub_agg = if has_sub_aggregations {
let sub_agg_collector = build_segment_agg_collectors(req_data, &node.children)?;
Some(CachedSubAggs::new(sub_agg_collector))
Some(BufferedSubAggs::new(sub_agg_collector))
} else {
None
};
let composite_req_data = req_data.get_composite_req_data(node.idx_in_req_data);
let num_sources = composite_req_data.req.sources.len();
Ok(SegmentCompositeCollector {
parent_buckets: vec![DynArrayHeapMap::try_new(num_sources)?],
accessor_idx: node.idx_in_req_data,
req_data: composite_req_data,
sub_agg,
bucket_id_provider: BucketIdProvider::default(),
num_sources,
@@ -250,7 +258,7 @@ impl SegmentCompositeCollector {
let mut dict: FxHashMap<Vec<CompositeIntermediateKey>, IntermediateCompositeBucketEntry> =
Default::default();
dict.reserve(heap_map.size());
let composite_data = agg_data.get_composite_req_data(self.accessor_idx);
let composite_data = &self.req_data;
for (key_internal_repr, agg) in heap_map.into_iter() {
let key = resolve_key(&key_internal_repr, composite_data)?;
let mut sub_aggregation_res = IntermediateAggregationResults::default();
@@ -290,8 +298,7 @@ impl SegmentCompositeCollector {
}
}
fn validate_req(req_data: &mut AggregationsSegmentCtx, accessor_idx: usize) -> crate::Result<()> {
let composite_data = req_data.get_composite_req_data(accessor_idx);
fn validate_req(composite_data: &CompositeAggReqData) -> crate::Result<()> {
let req = &composite_data.req;
if req.sources.is_empty() {
return Err(TantivyError::InvalidArgument(
@@ -332,7 +339,7 @@ fn collect_bucket_with_limit(
limit_num_buckets: usize,
buckets: &mut DynArrayHeapMap<InternalValueRepr, CompositeBucketCollector>,
key: &[InternalValueRepr],
sub_agg: &mut Option<CachedSubAggs<HighCardSubAggCache>>,
sub_agg: &mut Option<BufferedSubAggs<HighCardSubAggBuffer>>,
bucket_id_provider: &mut BucketIdProvider,
) {
let mut record_in_bucket = |bucket: &mut CompositeBucketCollector| {
@@ -488,7 +495,7 @@ struct CompositeKeyVisitor<'a> {
doc_id: crate::DocId,
composite_agg_data: &'a CompositeAggReqData,
buckets: &'a mut DynArrayHeapMap<InternalValueRepr, CompositeBucketCollector>,
sub_agg: &'a mut Option<CachedSubAggs<HighCardSubAggCache>>,
sub_agg: &'a mut Option<BufferedSubAggs<HighCardSubAggBuffer>>,
bucket_id_provider: &'a mut BucketIdProvider,
sub_level_values: SmallVec<[InternalValueRepr; MAX_DYN_ARRAY_SIZE]>,
}

92
src/aggregation/bucket/composite/mod.rs
View File

@@ -511,14 +511,14 @@ mod tests {
fn datetime_from_iso_str(date_str: &str) -> common::DateTime {
let dt = OffsetDateTime::parse(date_str, &Rfc3339)
.expect(&format!("Failed to parse date: {}", date_str));
.unwrap_or_else(|_| panic!("Failed to parse date: {}", date_str));
let timestamp_secs = dt.unix_timestamp_nanos();
common::DateTime::from_timestamp_nanos(timestamp_secs as i64)
}
fn ms_timestamp_from_iso_str(date_str: &str) -> i64 {
let dt = OffsetDateTime::parse(date_str, &Rfc3339)
.expect(&format!("Failed to parse date: {}", date_str));
.unwrap_or_else(|_| panic!("Failed to parse date: {}", date_str));
(dt.unix_timestamp_nanos() / 1_000_000) as i64
}
@@ -548,7 +548,7 @@ mod tests {
agg_req_json["my_composite"]["composite"]["after"] = after_key.take().unwrap();
}
let agg_req: Aggregations = serde_json::from_value(agg_req_json).unwrap();
let res = exec_request(agg_req.clone(), &index).unwrap();
let res = exec_request(agg_req.clone(), index).unwrap();
let expected_page_buckets = &expected_buckets_vec[page_idx * page_size
..std::cmp::min((page_idx + 1) * page_size, expected_buckets_vec.len())];
assert_eq!(
@@ -559,34 +559,30 @@ mod tests {
page_size,
agg_req,
);
if page_idx + 1 < page_count {
assert!(
res["my_composite"].get("after_key").is_some(),
"expected after_key on all but last page"
);
after_key = Some(res["my_composite"]["after_key"].clone());
} else if res["my_composite"].get("after_key").is_some() {
// currently we sometime have an after_key on the last page,
// check that the next "page" is empty
let agg_req_json = json!({
"my_composite": {
"composite": {
"sources": composite_agg_sources,
"size": page_size,
"after": res["my_composite"]["after_key"].clone(),
}
}
});
let agg_req: Aggregations = serde_json::from_value(agg_req_json).unwrap();
let res = exec_request(agg_req.clone(), &index).unwrap();
assert_eq!(
res["my_composite"]["buckets"],
json!([]),
"expected no buckets when using after_key from last page, query: {:?}",
agg_req
);
}
assert!(
res["my_composite"].get("after_key").is_some(),
"expected after_key on every non-empty page"
);
after_key = Some(res["my_composite"]["after_key"].clone());
}
// Using the after_key from the last page must yield an empty page.
let agg_req_json = json!({
"my_composite": {
"composite": {
"sources": composite_agg_sources,
"size": page_size,
"after": after_key,
}
}
});
let agg_req: Aggregations = serde_json::from_value(agg_req_json).unwrap();
let res = exec_request(agg_req.clone(), index).unwrap();
assert_eq!(
res["my_composite"]["buckets"],
json!([]),
"expected no buckets when using after_key from last page, query: {:?}",
agg_req
);
}
}
@@ -711,8 +707,28 @@ mod tests {
{"key": {"myterm": "terme"}, "doc_count": 1}
])
);
assert!(res["my_composite"].get("after_key").is_none());
// paginating past last page should be empty
let agg_req_json = json!({
"my_composite": {
"composite": {
"sources": [
{"myterm": {"terms": {"field": "string_id"}}}
],
"size": 3,
"after": &res["my_composite"]["after_key"]
}
}
});
let agg_req: Aggregations = serde_json::from_value(agg_req_json).unwrap();
let res = exec_request(agg_req.clone(), &index).unwrap();
assert!(res["my_composite"].get("after_key").is_none());
assert_eq!(
res["my_composite"]["buckets"],
json!([]),
"expected no buckets when using after_key from last page, query: {:?}",
agg_req
);
Ok(())
}
@@ -820,7 +836,10 @@ mod tests {
{"key": {"myterm": "apple"}, "doc_count": 1}
])
);
assert!(res["fruity_aggreg"].get("after_key").is_none());
assert_eq!(
res["fruity_aggreg"]["after_key"],
json!({"myterm": "str:apple"})
);
Ok(())
}
@@ -1792,7 +1811,14 @@ mod tests {
{"key": {"month": ms_timestamp_from_iso_str("2021-02-01T00:00:00Z"), "category": "books"}, "doc_count": 1},
]),
);
assert!(res["my_composite"].get("after_key").is_none());
let feb_2021_ns = ms_timestamp_from_iso_str("2021-02-01T00:00:00Z") * 1_000_000;
assert_eq!(
res["my_composite"]["after_key"],
json!({
"month": format!("dt:{}", feb_2021_ns),
"category": "str:books"
})
);
Ok(())
}

106
src/aggregation/bucket/filter.rs
View File

@@ -1,5 +1,5 @@
use std::fmt::Debug;
use std::sync::Arc;
use std::rc::Rc;
use common::BitSet;
use serde::{Deserialize, Deserializer, Serialize, Serializer};
@@ -7,8 +7,8 @@ use serde::{Deserialize, Deserializer, Serialize, Serializer};
use crate::aggregation::agg_data::{
build_segment_agg_collectors, AggRefNode, AggregationsSegmentCtx,
};
use crate::aggregation::cached_sub_aggs::{
CachedSubAggs, HighCardSubAggCache, LowCardSubAggCache, SubAggCache,
use crate::aggregation::buffered_sub_aggs::{
BufferedSubAggs, HighCardSubAggBuffer, LowCardSubAggBuffer, SubAggBuffer,
};
use crate::aggregation::intermediate_agg_result::{
IntermediateAggregationResult, IntermediateAggregationResults, IntermediateBucketResult,
@@ -397,29 +397,29 @@ impl PartialEq for FilterAggregation {
/// Request data for filter aggregation
/// This struct holds the per-segment data needed to execute a filter aggregation
#[derive(Clone)]
pub struct FilterAggReqData {
/// The name of the filter aggregation
pub name: String,
/// The filter aggregation
pub req: FilterAggregation,
/// The segment reader
pub segment_reader: Arc<dyn SegmentReader>,
/// Document evaluator for the filter query (precomputed BitSet)
/// This is built once when the request data is created
pub evaluator: DocumentQueryEvaluator,
/// Reusable buffer for matching documents to minimize allocations during collection
pub matching_docs_buffer: Vec<DocId>,
pub segment_reader: SegmentReader,
/// Document evaluator for the filter query (precomputed BitSet).
/// Wrapped in `Rc` so cloning the request data does not duplicate the (potentially large)
/// underlying BitSet.
pub evaluator: Rc<DocumentQueryEvaluator>,
/// True if this filter aggregation is at the top level of the aggregation tree (not nested).
pub is_top_level: bool,
}
impl FilterAggReqData {
pub(crate) fn get_memory_consumption(&self) -> usize {
// Estimate: name + segment reader reference + bitset + buffer capacity
// Estimate: name + segment reader reference + bitset
self.name.len()
+ self.evaluator.bitset.get_memory_consumption()
+ self.matching_docs_buffer.capacity() * std::mem::size_of::<DocId>()
+ std::mem::size_of::<bool>()
+ std::mem::size_of::<SegmentReader>()
+ self.evaluator.bitset.len() / 8 // BitSet memory (bits to bytes)
+ std::mem::size_of::<bool>()
}
}
@@ -438,7 +438,7 @@ impl DocumentQueryEvaluator {
pub(crate) fn new(
query: Box<dyn Query>,
schema: Schema,
segment_reader: &dyn SegmentReader,
segment_reader: &SegmentReader,
) -> crate::Result<Self> {
let max_doc = segment_reader.max_doc();
@@ -503,21 +503,24 @@ struct DocCount {
}
/// Segment collector for filter aggregation
pub struct SegmentFilterCollector<C: SubAggCache> {
pub struct SegmentFilterCollector<B: SubAggBuffer> {
/// Document counts per parent bucket
parent_buckets: Vec<DocCount>,
/// Sub-aggregation collectors
sub_aggregations: Option<CachedSubAggs<C>>,
sub_aggregations: Option<BufferedSubAggs<B>>,
bucket_id_provider: BucketIdProvider,
/// Accessor index for this filter aggregation (to access FilterAggReqData)
accessor_idx: usize,
/// Per-segment filter request data, owned by this collector.
req_data: FilterAggReqData,
/// Reusable buffer for matching documents to minimize allocations during collection.
matching_docs_buffer: Vec<DocId>,
}
impl<C: SubAggCache> SegmentFilterCollector<C> {
impl<B: SubAggBuffer> SegmentFilterCollector<B> {
/// Create a new filter segment collector following the new agg_data pattern
pub(crate) fn from_req_and_validate(
req: &mut AggregationsSegmentCtx,
node: &AggRefNode,
req_data: FilterAggReqData,
) -> crate::Result<Self> {
// Build sub-aggregation collectors if any
let sub_agg_collector = if !node.children.is_empty() {
@@ -525,13 +528,17 @@ impl<C: SubAggCache> SegmentFilterCollector<C> {
} else {
None
};
let sub_agg_collector = sub_agg_collector.map(CachedSubAggs::new);
let sub_agg_collector = sub_agg_collector.map(BufferedSubAggs::new);
let max_doc = req_data.segment_reader.max_doc();
let buffer_capacity = crate::docset::COLLECT_BLOCK_BUFFER_LEN.min(max_doc as usize);
Ok(SegmentFilterCollector {
parent_buckets: Vec::new(),
sub_aggregations: sub_agg_collector,
accessor_idx: node.idx_in_req_data,
req_data,
bucket_id_provider: BucketIdProvider::default(),
matching_docs_buffer: Vec::with_capacity(buffer_capacity),
})
}
}
@@ -540,33 +547,38 @@ pub(crate) fn build_segment_filter_collector(
req: &mut AggregationsSegmentCtx,
node: &AggRefNode,
) -> crate::Result<Box<dyn SegmentAggregationCollector>> {
let is_top_level = req.per_request.filter_req_data[node.idx_in_req_data]
.as_ref()
.expect("filter_req_data slot is empty")
.is_top_level;
let req_data = req.per_request.filter_req_data[node.idx_in_req_data].clone();
req.context
.limits
.add_memory_consumed(req_data.get_memory_consumption() as u64)?;
let is_top_level = req_data.is_top_level;
if is_top_level {
Ok(Box::new(
SegmentFilterCollector::<LowCardSubAggCache>::from_req_and_validate(req, node)?,
SegmentFilterCollector::<LowCardSubAggBuffer>::from_req_and_validate(
req, node, req_data,
)?,
))
} else {
Ok(Box::new(
SegmentFilterCollector::<HighCardSubAggCache>::from_req_and_validate(req, node)?,
SegmentFilterCollector::<HighCardSubAggBuffer>::from_req_and_validate(
req, node, req_data,
)?,
))
}
}
impl<C: SubAggCache> Debug for SegmentFilterCollector<C> {
impl<B: SubAggBuffer> Debug for SegmentFilterCollector<B> {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
f.debug_struct("SegmentFilterCollector")
.field("buckets", &self.parent_buckets)
.field("has_sub_aggs", &self.sub_aggregations.is_some())
.field("accessor_idx", &self.accessor_idx)
.field("name", &self.req_data.name)
.finish()
}
}
impl<C: SubAggCache> SegmentAggregationCollector for SegmentFilterCollector<C> {
impl<B: SubAggBuffer> SegmentAggregationCollector for SegmentFilterCollector<B> {
fn add_intermediate_aggregation_result(
&mut self,
agg_data: &AggregationsSegmentCtx,
@@ -598,11 +610,7 @@ impl<C: SubAggCache> SegmentAggregationCollector for SegmentFilterCollector<C> {
};
// Get the name of this filter aggregation
let name = agg_data.per_request.filter_req_data[self.accessor_idx]
.as_ref()
.expect("filter_req_data slot is empty")
.name
.clone();
let name = self.req_data.name.clone();
results.push(
name,
@@ -623,27 +631,24 @@ impl<C: SubAggCache> SegmentAggregationCollector for SegmentFilterCollector<C> {
}
let mut bucket = self.parent_buckets[parent_bucket_id as usize];
// Take the request data to avoid borrow checker issues with sub-aggregations
let mut req = agg_data.take_filter_req_data(self.accessor_idx);
// Use batch filtering with O(1) BitSet lookups
req.matching_docs_buffer.clear();
req.evaluator
.filter_batch(docs, &mut req.matching_docs_buffer);
self.matching_docs_buffer.clear();
self.req_data
.evaluator
.filter_batch(docs, &mut self.matching_docs_buffer);
bucket.doc_count += req.matching_docs_buffer.len() as u64;
bucket.doc_count += self.matching_docs_buffer.len() as u64;
// Batch process sub-aggregations if we have matches
if !req.matching_docs_buffer.is_empty() {
if !self.matching_docs_buffer.is_empty() {
if let Some(sub_aggs) = &mut self.sub_aggregations {
for &doc_id in &req.matching_docs_buffer {
for &doc_id in &self.matching_docs_buffer {
sub_aggs.push(bucket.bucket_id, doc_id);
}
}
}
// Put the request data back
agg_data.put_back_filter_req_data(self.accessor_idx, req);
if let Some(sub_aggs) = &mut self.sub_aggregations {
sub_aggs.check_flush_local(agg_data)?;
}
@@ -674,6 +679,17 @@ impl<C: SubAggCache> SegmentAggregationCollector for SegmentFilterCollector<C> {
}
Ok(())
}
fn compute_metric_value(
&self,
_bucket_id: BucketId,
_sub_agg_name: &str,
_sub_agg_property: &str,
_agg_data: &AggregationsSegmentCtx,
) -> Option<f64> {
// TODO: forward into the inner `sub_agg` for nested order paths (`filter.metric`).
None
}
}
/// Intermediate result for filter aggregation

568
src/aggregation/bucket/histogram/histogram.rs
View File

@@ -10,7 +10,7 @@ use crate::aggregation::agg_data::{
};
use crate::aggregation::agg_req::Aggregations;
use crate::aggregation::agg_result::BucketEntry;
use crate::aggregation::cached_sub_aggs::{CachedSubAggs, HighCardCachedSubAggs};
use crate::aggregation::buffered_sub_aggs::{BufferedSubAggs, HighCardBufferedSubAggs};
use crate::aggregation::intermediate_agg_result::{
IntermediateAggregationResult, IntermediateAggregationResults, IntermediateBucketResult,
IntermediateHistogramBucketEntry,
@@ -21,6 +21,7 @@ use crate::TantivyError;
/// Contains all information required by the SegmentHistogramCollector to perform the
/// histogram or date_histogram aggregation on a segment.
#[derive(Debug, Clone)]
pub struct HistogramAggReqData {
/// The column accessor to access the fast field values.
pub accessor: Column<u64>,
@@ -243,22 +244,55 @@ impl Display for HistogramBounds {
}
impl HistogramBounds {
fn contains(&self, val: f64) -> bool {
pub(crate) fn contains(&self, val: f64) -> bool {
val >= self.min && val <= self.max
}
}
#[derive(Default, Clone, Debug, PartialEq)]
pub(crate) struct SegmentHistogramBucketEntry {
pub key: f64,
pub doc_count: u64,
pub bucket_id: BucketId,
/// The per-bucket identifier stored in a [`SegmentHistogramBucketEntry`].
///
/// It is [`BucketId`] when the histogram has sub aggregations (which key their state by it), and
/// the zero-sized `()` when it does not. Without sub aggregations the id is never read, so storing
/// `()` drops 8 bytes per bucket (24 -> 16) and turns id assignment into a no-op.
pub trait BucketIdSlot: Copy + Default + std::fmt::Debug + PartialEq {
/// Assigns the next id from the provider, called once when a bucket is first filled.
fn assign(provider: &mut BucketIdProvider) -> Self;
/// Resolves to the `BucketId` for sub-aggregation bookkeeping.
///
/// Only ever called for the [`BucketId`] slot: the `()` slot is used exactly when there are no
/// sub aggregations, so every call site is guarded by `sub_agg.is_some()` and is dead for `()`.
fn to_bucket_id(self) -> BucketId;
}
impl BucketIdSlot for BucketId {
#[inline(always)]
fn assign(provider: &mut BucketIdProvider) -> Self {
provider.next_bucket_id()
}
#[inline(always)]
fn to_bucket_id(self) -> BucketId {
self
}
}
impl BucketIdSlot for () {
#[inline(always)]
fn assign(_provider: &mut BucketIdProvider) -> Self {}
#[inline(always)]
fn to_bucket_id(self) -> BucketId {
unreachable!("bucket ids are only resolved when sub aggregations are present")
}
}
impl SegmentHistogramBucketEntry {
#[derive(Default, Clone, Debug, PartialEq)]
pub(crate) struct SegmentHistogramBucketEntry<B> {
pub key: f64,
pub doc_count: u64,
pub bucket_id: B,
}
impl<B: BucketIdSlot> SegmentHistogramBucketEntry<B> {
pub(crate) fn into_intermediate_bucket_entry(
self,
sub_aggregation: &mut Option<HighCardCachedSubAggs>,
sub_aggregation: &mut Option<HighCardBufferedSubAggs>,
agg_data: &AggregationsSegmentCtx,
) -> crate::Result<IntermediateHistogramBucketEntry> {
let mut sub_aggregation_res = IntermediateAggregationResults::default();
@@ -268,7 +302,7 @@ impl SegmentHistogramBucketEntry {
.add_intermediate_aggregation_result(
agg_data,
&mut sub_aggregation_res,
self.bucket_id,
self.bucket_id.to_bucket_id(),
)?;
}
Ok(IntermediateHistogramBucketEntry {
@@ -279,34 +313,147 @@ impl SegmentHistogramBucketEntry {
}
}
#[derive(Clone, Debug, Default)]
struct HistogramBuckets {
pub buckets: FxHashMap<i64, SegmentHistogramBucketEntry>,
/// The contiguous bucket range a histogram can span, derived from the column min/max (clamped to
/// the histogram bounds). Buckets in `[base_pos, base_pos + len)` can be stored in a flat `Vec`
/// indexed by `bucket_pos - base_pos`, avoiding the hash map on the hot path.
#[derive(Clone, Copy, Debug)]
pub(crate) struct DenseRange {
/// `bucket_pos` mapped to index 0 of the dense `Vec`.
pub(crate) base_pos: i64,
/// Number of bucket positions in the range.
pub(crate) len: usize,
}
/// Storage for the histogram buckets of a single parent bucket.
///
/// Starts out sparse (a hash map keyed by `bucket_pos`). Once enough distinct buckets have been
/// filled that we are clearly going to cover most of the column's theoretical range, it switches
/// to a dense `Vec` indexed by `bucket_pos - base_pos`, which removes hashing from the hot loop.
#[derive(Clone, Debug)]
enum HistogramBuckets<B> {
Sparse(FxHashMap<i64, SegmentHistogramBucketEntry<B>>),
Dense {
base_pos: i64,
/// One slot per bucket position; a slot with `doc_count == 0` has not been hit yet.
buckets: Vec<SegmentHistogramBucketEntry<B>>,
},
}
impl<B> Default for HistogramBuckets<B> {
fn default() -> Self {
HistogramBuckets::Sparse(FxHashMap::default())
}
}
impl<B: BucketIdSlot> HistogramBuckets<B> {
fn memory_consumption(&self) -> u64 {
let num_slots = match self {
HistogramBuckets::Sparse(map) => map.capacity(),
HistogramBuckets::Dense { buckets, .. } => buckets.capacity(),
};
num_slots as u64 * std::mem::size_of::<SegmentHistogramBucketEntry<B>>() as u64
}
/// Switches from sparse to dense storage once the dense `Vec` would use no more memory than the
/// hash map does now, so the switch never increases memory. Called at block boundaries.
///
/// The `Vec` holds one `Entry` per bucket position in the range. The map additionally stores
/// the key and a control byte per slot, at a load factor of 7/16..7/8, so for a dense histogram
/// its footprint grows past the `Vec` well before full coverage. And since the `Vec` never
/// grows afterwards while the map would keep growing, dense only gets relatively cheaper — so
/// no upper bound on the range is needed: a large but sparse range simply never crosses over.
#[inline]
fn maybe_densify(&mut self, dense_range: Option<DenseRange>) {
let Some(range) = dense_range else { return };
let HistogramBuckets::Sparse(map) = self else {
return;
};
let dense_bytes = range
.len
.saturating_mul(std::mem::size_of::<SegmentHistogramBucketEntry<B>>());
let sparse_bytes = map
.capacity()
.saturating_mul(std::mem::size_of::<(i64, SegmentHistogramBucketEntry<B>)>() + 1);
if dense_bytes > sparse_bytes {
return;
}
let map = std::mem::take(map);
let mut buckets = vec![SegmentHistogramBucketEntry::<B>::default(); range.len];
for (bucket_pos, entry) in map {
buckets[(bucket_pos - range.base_pos) as usize] = entry;
}
*self = HistogramBuckets::Dense {
base_pos: range.base_pos,
buckets,
};
}
/// Returns the bucket entry for `bucket_pos`, setting its key (and `bucket_id`, when `B` is
/// [`BucketId`]) on first use.
///
/// For the dense variant `bucket_pos` is guaranteed to be inside the range, since it is
/// derived from the column min/max that bounds every value (see [`compute_dense_range`]).
#[inline]
fn get_or_create(
&mut self,
bucket_pos: i64,
bucket_id_provider: &mut BucketIdProvider,
key_from_pos: impl FnOnce(i64) -> f64,
) -> &mut SegmentHistogramBucketEntry<B> {
match self {
HistogramBuckets::Sparse(map) => {
map.entry(bucket_pos)
.or_insert_with(|| SegmentHistogramBucketEntry {
key: key_from_pos(bucket_pos),
doc_count: 0,
bucket_id: B::assign(bucket_id_provider),
})
}
HistogramBuckets::Dense { base_pos, buckets } => {
let idx = (bucket_pos - *base_pos) as usize;
debug_assert!(idx < buckets.len(), "bucket_pos outside the dense range");
let entry = &mut buckets[idx];
if entry.doc_count == 0 {
entry.key = key_from_pos(bucket_pos);
entry.bucket_id = B::assign(bucket_id_provider);
}
entry
}
}
}
/// Consumes the storage, yielding all non-empty bucket entries.
fn into_filled_entries(self) -> Vec<SegmentHistogramBucketEntry<B>> {
match self {
HistogramBuckets::Sparse(map) => map.into_values().collect(),
HistogramBuckets::Dense { buckets, .. } => {
buckets.into_iter().filter(|b| b.doc_count > 0).collect()
}
}
}
}
/// The collector puts values from the fast field into the correct buckets and does a conversion to
/// the correct datatype.
#[derive(Debug)]
pub struct SegmentHistogramCollector {
pub struct SegmentHistogramCollector<B> {
/// The buckets containing the aggregation data.
/// One Histogram bucket per parent bucket id.
parent_buckets: Vec<HistogramBuckets>,
sub_agg: Option<HighCardCachedSubAggs>,
accessor_idx: usize,
parent_buckets: Vec<HistogramBuckets<B>>,
sub_agg: Option<HighCardBufferedSubAggs>,
req_data: HistogramAggReqData,
bucket_id_provider: BucketIdProvider,
/// Theoretical bucket range derived from the column min/max, if dense `Vec` storage is
/// viable. `None` keeps every parent bucket in the sparse hash map.
dense_range: Option<DenseRange>,
}
impl SegmentAggregationCollector for SegmentHistogramCollector {
impl<B: BucketIdSlot> SegmentAggregationCollector for SegmentHistogramCollector<B> {
fn add_intermediate_aggregation_result(
&mut self,
agg_data: &AggregationsSegmentCtx,
results: &mut IntermediateAggregationResults,
parent_bucket_id: BucketId,
) -> crate::Result<()> {
let name = agg_data
.get_histogram_req_data(self.accessor_idx)
.name
.clone();
let name = self.req_data.name.clone();
// TODO: avoid prepare_max_bucket here and handle empty buckets.
self.prepare_max_bucket(parent_bucket_id, agg_data)?;
let histogram = std::mem::take(&mut self.parent_buckets[parent_bucket_id as usize]);
@@ -323,10 +470,13 @@ impl SegmentAggregationCollector for SegmentHistogramCollector {
docs: &[crate::DocId],
agg_data: &mut AggregationsSegmentCtx,
) -> crate::Result<()> {
let req = agg_data.take_histogram_req_data(self.accessor_idx);
let mem_pre = self.get_memory_consumption();
let buckets = &mut self.parent_buckets[parent_bucket_id as usize].buckets;
let mem_pre = self.get_memory_consumption(parent_bucket_id);
let dense_range = self.dense_range;
let store = &mut self.parent_buckets[parent_bucket_id as usize];
// Upgrade to dense storage before processing the block if the buckets are dense enough.
store.maybe_densify(dense_range);
let req = &self.req_data;
let bounds = req.bounds;
let interval = req.req.interval;
let offset = req.offset;
@@ -335,35 +485,43 @@ impl SegmentAggregationCollector for SegmentHistogramCollector {
agg_data
.column_block_accessor
.fetch_block(docs, &req.accessor);
for (doc, val) in agg_data
.column_block_accessor
.iter_docid_vals(docs, &req.accessor)
{
let val = f64_from_fastfield_u64(val, req.field_type);
let bucket_pos = get_bucket_pos(val);
if bounds.contains(val) {
let bucket = buckets.entry(bucket_pos).or_insert_with(|| {
let key = get_bucket_key_from_pos(bucket_pos as f64, interval, offset);
SegmentHistogramBucketEntry {
key,
doc_count: 0,
bucket_id: self.bucket_id_provider.next_bucket_id(),
}
});
bucket.doc_count += 1;
if let Some(sub_agg) = &mut self.sub_agg {
sub_agg.push(bucket.bucket_id, doc);
// special path for nested buckets
if let Some(sub_agg) = &mut self.sub_agg {
for (doc, val) in agg_data
.column_block_accessor
.iter_docid_vals(docs, &req.accessor)
{
let val = f64_from_fastfield_u64(val, req.field_type);
if bounds.contains(val) {
let bucket = store.get_or_create(
get_bucket_pos(val),
&mut self.bucket_id_provider,
|pos| get_bucket_key_from_pos(pos as f64, interval, offset),
);
bucket.doc_count += 1;
sub_agg.push(bucket.bucket_id.to_bucket_id(), doc);
}
}
} else {
for val in agg_data.column_block_accessor.iter_vals() {
let val = f64_from_fastfield_u64(val, req.field_type);
if bounds.contains(val) {
let bucket = store.get_or_create(
get_bucket_pos(val),
&mut self.bucket_id_provider,
|pos| get_bucket_key_from_pos(pos as f64, interval, offset),
);
bucket.doc_count += 1;
}
}
}
agg_data.put_back_histogram_req_data(self.accessor_idx, req);
let mem_delta = self.get_memory_consumption() - mem_pre;
if mem_delta > 0 {
agg_data
.context
.limits
.add_memory_consumed(mem_delta as u64)?;
// `checked_sub` is `None` when densifying shrank the accounted memory; only account growth.
if let Some(mem_delta) = self
.get_memory_consumption(parent_bucket_id)
.checked_sub(mem_pre)
{
agg_data.context.limits.add_memory_consumed(mem_delta)?;
}
if let Some(sub_agg) = &mut self.sub_agg {
@@ -386,39 +544,45 @@ impl SegmentAggregationCollector for SegmentHistogramCollector {
_agg_data: &AggregationsSegmentCtx,
) -> crate::Result<()> {
while self.parent_buckets.len() <= max_bucket as usize {
self.parent_buckets.push(HistogramBuckets {
buckets: FxHashMap::default(),
});
self.parent_buckets.push(HistogramBuckets::default());
}
Ok(())
}
fn compute_metric_value(
&self,
_bucket_id: BucketId,
_sub_agg_name: &str,
_sub_agg_property: &str,
_agg_data: &AggregationsSegmentCtx,
) -> Option<f64> {
// Histogram is a multi-bucket agg with no single value to extract.
None
}
}
impl SegmentHistogramCollector {
fn get_memory_consumption(&self) -> usize {
let self_mem = std::mem::size_of::<Self>();
let buckets_mem = self.parent_buckets.len() * std::mem::size_of::<HistogramBuckets>();
self_mem + buckets_mem
impl<B: BucketIdSlot> SegmentHistogramCollector<B> {
fn get_memory_consumption(&self, parent_bucket_id: BucketId) -> u64 {
self.parent_buckets[parent_bucket_id as usize].memory_consumption()
}
/// Converts the collector result into a intermediate bucket result.
fn add_intermediate_bucket_result(
&mut self,
agg_data: &AggregationsSegmentCtx,
histogram: HistogramBuckets,
histogram: HistogramBuckets<B>,
) -> crate::Result<IntermediateBucketResult> {
let mut buckets = Vec::with_capacity(histogram.buckets.len());
let filled = histogram.into_filled_entries();
let mut buckets = Vec::with_capacity(filled.len());
for bucket in histogram.buckets.into_values() {
for bucket in filled {
let bucket_res = bucket.into_intermediate_bucket_entry(&mut self.sub_agg, agg_data);
buckets.push(bucket_res?);
}
buckets.sort_unstable_by(|b1, b2| b1.key.total_cmp(&b2.key));
let is_date_agg = agg_data
.get_histogram_req_data(self.accessor_idx)
.field_type
== ColumnType::DateTime;
let is_date_agg = self.req_data.field_type == ColumnType::DateTime;
Ok(IntermediateBucketResult::Histogram {
buckets,
is_date_agg,
@@ -434,32 +598,175 @@ impl SegmentHistogramCollector {
} else {
None
};
let req_data = agg_data.get_histogram_req_data_mut(node.idx_in_req_data);
req_data.req.validate()?;
if req_data.field_type == ColumnType::DateTime && !req_data.is_date_histogram {
req_data.req.normalize_date_time();
}
req_data.bounds = req_data.req.hard_bounds.unwrap_or(HistogramBounds {
min: f64::MIN,
max: f64::MAX,
});
req_data.offset = req_data.req.offset.unwrap_or(0.0);
let sub_agg = sub_agg.map(CachedSubAggs::new);
let mut req_data = agg_data.per_request.histogram_req_data[node.idx_in_req_data].clone();
normalize_histogram_req(&mut req_data)?;
agg_data
.context
.limits
.add_memory_consumed(req_data.get_memory_consumption() as u64)?;
let dense_range = compute_dense_range(
&req_data.accessor,
req_data.field_type,
req_data.req.interval,
req_data.offset,
req_data.bounds,
);
let sub_agg = sub_agg.map(BufferedSubAggs::new);
Ok(Self {
parent_buckets: Default::default(),
sub_agg,
accessor_idx: node.idx_in_req_data,
req_data,
bucket_id_provider: BucketIdProvider::default(),
dense_range,
})
}
}
impl SegmentHistogramCollector<()> {
/// Builds a histogram collector whose parent `t` is a dense histogram filled from
/// `counts[t * num_time_buckets .. (t + 1) * num_time_buckets]` (row-major). Used by the fused
/// terms×histogram collector to turn its flat 2D counters into the regular intermediate result,
/// so cross-segment merging is shared with the general path.
pub(crate) fn from_dense_rows(
req_data: HistogramAggReqData,
base_pos: i64,
num_time_buckets: usize,
counts: &[u32],
) -> Self {
let interval = req_data.req.interval;
let offset = req_data.offset;
let num_parents = counts.len().checked_div(num_time_buckets).unwrap_or(0);
let parent_buckets = (0..num_parents)
.map(|t| {
let row = &counts[t * num_time_buckets..(t + 1) * num_time_buckets];
let buckets = row
.iter()
.enumerate()
.map(|(b, &doc_count)| SegmentHistogramBucketEntry {
key: get_bucket_key_from_pos(
(base_pos + b as i64) as f64,
interval,
offset,
),
doc_count: doc_count as u64,
bucket_id: (),
})
.collect();
HistogramBuckets::Dense { base_pos, buckets }
})
.collect();
Self {
parent_buckets,
sub_agg: None,
req_data,
bucket_id_provider: BucketIdProvider::default(),
dense_range: None,
}
}
}
/// Validates and normalizes a histogram request in place: applies date ns-normalization (for a
/// `histogram` on a date column) and resolves `bounds`/`offset` from the request.
fn normalize_histogram_req(req_data: &mut HistogramAggReqData) -> crate::Result<()> {
req_data.req.validate()?;
if req_data.field_type == ColumnType::DateTime && !req_data.is_date_histogram {
req_data.req.normalize_date_time();
}
req_data.bounds = req_data.req.hard_bounds.unwrap_or(HistogramBounds {
min: f64::MIN,
max: f64::MAX,
});
req_data.offset = req_data.req.offset.unwrap_or(0.0);
// Drop `hard_bounds` that can't exclude any value (the column's range already sits inside
// them): the per-doc `bounds.contains` check is then a no-op, so collapsing to the unbounded
// sentinel lets the histogram hot loop skip it and the fused term×histogram path derive
// per-term counts from the grid. Only this collect-time filter is touched — empty-bucket
// emission reads `req.hard_bounds` directly (see `get_req_min_max`), and `hard_bounds` only
// ever clips that range, so a wider-than-data bound leaves the result unchanged.
if req_data.req.hard_bounds.is_some() {
let col_min = f64_from_fastfield_u64(req_data.accessor.min_value(), req_data.field_type);
let col_max = f64_from_fastfield_u64(req_data.accessor.max_value(), req_data.field_type);
if col_min >= req_data.bounds.min && col_max <= req_data.bounds.max {
req_data.bounds = HistogramBounds {
min: f64::MIN,
max: f64::MAX,
};
}
}
Ok(())
}
/// Clones and normalizes (resolving interval/offset/bounds) the histogram request at `node`, and
/// returns it together with its dense bucket range — or `None` if the column has no usable range.
/// Used by the fused terms×histogram collector, which then owns the normalized request.
pub(crate) fn prepare_histogram_dense_range(
agg_data: &AggregationsSegmentCtx,
node: &AggRefNode,
) -> crate::Result<Option<(HistogramAggReqData, DenseRange)>> {
let mut req_data = agg_data.per_request.histogram_req_data[node.idx_in_req_data].clone();
normalize_histogram_req(&mut req_data)?;
let dense_range = compute_dense_range(
&req_data.accessor,
req_data.field_type,
req_data.req.interval,
req_data.offset,
req_data.bounds,
);
Ok(dense_range.map(|range| (req_data, range)))
}
/// Builds a boxed histogram (or date histogram) segment collector, picking the bucket-id storage
/// based on whether there are sub aggregations: `()` (no id stored) when there are none, otherwise
/// [`BucketId`].
pub(crate) fn build_segment_histogram_collector(
agg_data: &mut AggregationsSegmentCtx,
node: &AggRefNode,
) -> crate::Result<Box<dyn SegmentAggregationCollector>> {
if node.children.is_empty() {
Ok(Box::new(
SegmentHistogramCollector::<()>::from_req_and_validate(agg_data, node)?,
))
} else {
Ok(Box::new(
SegmentHistogramCollector::<BucketId>::from_req_and_validate(agg_data, node)?,
))
}
}
#[inline]
fn get_bucket_pos_f64(val: f64, interval: f64, offset: f64) -> f64 {
pub(crate) fn get_bucket_pos_f64(val: f64, interval: f64, offset: f64) -> f64 {
((val - offset) / interval).floor()
}
/// Computes the dense bucket range for a column from its min/max value (clamped to the histogram
/// bounds), or `None` if there are no values within bounds (or the range overflows `usize`).
///
/// There is no upper bound on the range: whether dense storage is actually used is decided later,
/// per parent bucket, by [`HistogramBuckets::maybe_densify`] based on the memory it would save.
///
/// The column min/max bound every value the collector can see, so a `Vec` sized to this range can
/// be indexed by `bucket_pos - base_pos` without any out-of-bounds check on the hot path.
fn compute_dense_range(
accessor: &Column<u64>,
field_type: ColumnType,
interval: f64,
offset: f64,
bounds: HistogramBounds,
) -> Option<DenseRange> {
let col_min = f64_from_fastfield_u64(accessor.min_value(), field_type);
let col_max = f64_from_fastfield_u64(accessor.max_value(), field_type);
let lo = col_min.max(bounds.min);
let hi = col_max.min(bounds.max);
if lo > hi {
return None;
}
let base_pos = get_bucket_pos_f64(lo, interval, offset) as i64;
let top_pos = get_bucket_pos_f64(hi, interval, offset) as i64;
let len = usize::try_from(top_pos.checked_sub(base_pos)?.checked_add(1)?).ok()?;
(len > 0).then_some(DenseRange { base_pos, len })
}
#[inline]
fn get_bucket_key_from_pos(bucket_pos: f64, interval: f64, offset: f64) -> f64 {
bucket_pos * interval + offset
@@ -764,6 +1071,62 @@ mod tests {
Ok(())
}
#[test]
fn histogram_dense_storage_test() -> crate::Result<()> {
histogram_dense_storage_test_with_opt(false)?;
histogram_dense_storage_test_with_opt(true)?;
Ok(())
}
/// Exercises the switch from sparse hash map to dense `Vec` storage. The switch happens at a
/// block boundary (a block is `COLLECT_BLOCK_BUFFER_LEN` = 64 docs), so we need many docs in a
/// single segment, densely covering the bucket range. `with_sub_agg` toggles the `iter_vals`
/// fast path vs. the `iter_docid_vals` path used when there is a sub aggregation.
fn histogram_dense_storage_test_with_opt(with_sub_agg: bool) -> crate::Result<()> {
let num_buckets = 50usize;
let docs_per_bucket = 10usize;
// Value `k` repeated `docs_per_bucket` times for each bucket `k`, so every value in bucket
// `k` equals `k` and the per-bucket average is exactly `k`.
let values: Vec<f64> = (0..num_buckets * docs_per_bucket)
.map(|i| (i % num_buckets) as f64)
.collect();
// `merge_segments = true` collapses the per-value segments into a single segment with all
// the docs, which is collected in 64-doc blocks and therefore switches to dense storage.
let index = get_test_index_from_values(true, &values)?;
let agg_req: Aggregations = serde_json::from_value(if with_sub_agg {
json!({
"histogram": {
"histogram": { "field": "score_f64", "interval": 1.0 },
"aggs": { "avg": { "avg": { "field": "score_f64" } } }
}
})
} else {
json!({
"histogram": {
"histogram": { "field": "score_f64", "interval": 1.0 }
}
})
})
.unwrap();
let res = exec_request(agg_req, &index)?;
for k in 0..num_buckets {
assert_eq!(res["histogram"]["buckets"][k]["key"], k as f64);
assert_eq!(
res["histogram"]["buckets"][k]["doc_count"],
docs_per_bucket as u64
);
if with_sub_agg {
assert_eq!(res["histogram"]["buckets"][k]["avg"]["value"], k as f64);
}
}
assert_eq!(res["histogram"]["buckets"][num_buckets], Value::Null);
Ok(())
}
#[test]
fn histogram_memory_limit() -> crate::Result<()> {
let index = get_test_index_with_num_docs(true, 100)?;
@@ -1058,6 +1421,55 @@ mod tests {
Ok(())
}
#[test]
fn histogram_non_binding_hard_bounds_test_multi_segment() -> crate::Result<()> {
histogram_non_binding_hard_bounds_test_with_opt(false)
}
#[test]
fn histogram_non_binding_hard_bounds_test_single_segment() -> crate::Result<()> {
histogram_non_binding_hard_bounds_test_with_opt(true)
}
/// `hard_bounds` wider than the data (here with mid-interval edges, to cover the "bound cuts a
/// bucket" case) can't exclude any value, so the result must be identical to the same request
/// without bounds. Guards the normalization that collapses such bounds to the unbounded
/// sentinel so the hot loop / fused path can skip the per-doc bounds check.
fn histogram_non_binding_hard_bounds_test_with_opt(merge_segments: bool) -> crate::Result<()> {
let values = vec![10.0, 12.0, 14.0, 16.0, 10.0, 13.0, 10.0, 12.0];
let index = get_test_index_from_values(merge_segments, &values)?;
// Mid-interval edges, but wider than the data range [10, 16] -> they exclude nothing.
let with_bounds: Aggregations = serde_json::from_value(json!({
"histogram": {
"histogram": {
"field": "score_f64",
"interval": 1.0,
"hard_bounds": { "min": 9.5, "max": 16.5 }
}
}
}))
.unwrap();
let no_bounds: Aggregations = serde_json::from_value(json!({
"histogram": {
"histogram": { "field": "score_f64", "interval": 1.0 }
}
}))
.unwrap();
let res_bounds = exec_request(with_bounds, &index)?;
let res_plain = exec_request(no_bounds, &index)?;
// Dropping a non-binding bound must not change anything.
assert_eq!(res_bounds, res_plain);
// Sanity: buckets span the data range with gaps filled (min_doc_count defaults to 0).
assert_eq!(res_bounds["histogram"]["buckets"][0]["key"], 10.0);
assert_eq!(res_bounds["histogram"]["buckets"][0]["doc_count"], 3);
assert_eq!(res_bounds["histogram"]["buckets"][6]["key"], 16.0);
assert_eq!(res_bounds["histogram"]["buckets"][6]["doc_count"], 1);
assert_eq!(res_bounds["histogram"]["buckets"][7], Value::Null);
Ok(())
}
#[test]
fn histogram_empty_result_behaviour_test_single_segment() -> crate::Result<()> {
histogram_empty_result_behaviour_test_with_opt(true)

137
src/aggregation/bucket/range.rs
View File

@@ -9,8 +9,9 @@ use crate::aggregation::agg_data::{
build_segment_agg_collectors, AggRefNode, AggregationsSegmentCtx,
};
use crate::aggregation::agg_limits::AggregationLimitsGuard;
use crate::aggregation::cached_sub_aggs::{
CachedSubAggs, HighCardSubAggCache, LowCardCachedSubAggs, LowCardSubAggCache, SubAggCache,
use crate::aggregation::buffered_sub_aggs::{
BufferedSubAggs, HighCardSubAggBuffer, LowCardBufferedSubAggs, LowCardSubAggBuffer,
SubAggBuffer,
};
use crate::aggregation::intermediate_agg_result::{
IntermediateAggregationResult, IntermediateAggregationResults, IntermediateBucketResult,
@@ -22,6 +23,7 @@ use crate::TantivyError;
/// Contains all information required by the SegmentRangeCollector to perform the
/// range aggregation on a segment.
#[derive(Debug, Clone)]
pub struct RangeAggReqData {
/// The column accessor to access the fast field values.
pub accessor: Column<u64>,
@@ -155,13 +157,13 @@ pub(crate) struct SegmentRangeAndBucketEntry {
/// The collector puts values from the fast field into the correct buckets and does a conversion to
/// the correct datatype.
pub struct SegmentRangeCollector<C: SubAggCache> {
pub struct SegmentRangeCollector<B: SubAggBuffer> {
/// The buckets containing the aggregation data.
/// One for each ParentBucketId
parent_buckets: Vec<Vec<SegmentRangeAndBucketEntry>>,
column_type: ColumnType,
pub(crate) accessor_idx: usize,
sub_agg: Option<CachedSubAggs<C>>,
pub(crate) req_data: RangeAggReqData,
sub_agg: Option<BufferedSubAggs<B>>,
/// Here things get a bit weird. We need to assign unique bucket ids across all
/// parent buckets. So we keep track of the next available bucket id here.
/// This allows a kind of flattening of the bucket ids across all parent buckets.
@@ -178,12 +180,12 @@ pub struct SegmentRangeCollector<C: SubAggCache> {
limits: AggregationLimitsGuard,
}
impl<C: SubAggCache> Debug for SegmentRangeCollector<C> {
impl<B: SubAggBuffer> Debug for SegmentRangeCollector<B> {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
f.debug_struct("SegmentRangeCollector")
.field("parent_buckets_len", &self.parent_buckets.len())
.field("column_type", &self.column_type)
.field("accessor_idx", &self.accessor_idx)
.field("name", &self.req_data.name)
.field("has_sub_agg", &self.sub_agg.is_some())
.finish()
}
@@ -229,7 +231,7 @@ impl SegmentRangeBucketEntry {
}
}
impl<C: SubAggCache> SegmentAggregationCollector for SegmentRangeCollector<C> {
impl<B: SubAggBuffer> SegmentAggregationCollector for SegmentRangeCollector<B> {
fn add_intermediate_aggregation_result(
&mut self,
agg_data: &AggregationsSegmentCtx,
@@ -238,10 +240,7 @@ impl<C: SubAggCache> SegmentAggregationCollector for SegmentRangeCollector<C> {
) -> crate::Result<()> {
self.prepare_max_bucket(parent_bucket_id, agg_data)?;
let field_type = self.column_type;
let name = agg_data
.get_range_req_data(self.accessor_idx)
.name
.to_string();
let name = self.req_data.name.to_string();
let buckets = std::mem::take(&mut self.parent_buckets[parent_bucket_id as usize]);
@@ -280,17 +279,15 @@ impl<C: SubAggCache> SegmentAggregationCollector for SegmentRangeCollector<C> {
docs: &[crate::DocId],
agg_data: &mut AggregationsSegmentCtx,
) -> crate::Result<()> {
let req = agg_data.take_range_req_data(self.accessor_idx);
agg_data
.column_block_accessor
.fetch_block(docs, &req.accessor);
.fetch_block(docs, &self.req_data.accessor);
let buckets = &mut self.parent_buckets[parent_bucket_id as usize];
for (doc, val) in agg_data
.column_block_accessor
.iter_docid_vals(docs, &req.accessor)
.iter_docid_vals(docs, &self.req_data.accessor)
{
let bucket_pos = get_bucket_pos(val, buckets);
let bucket = &mut buckets[bucket_pos];
@@ -300,7 +297,6 @@ impl<C: SubAggCache> SegmentAggregationCollector for SegmentRangeCollector<C> {
}
}
agg_data.put_back_range_req_data(self.accessor_idx, req);
if let Some(sub_agg) = self.sub_agg.as_mut() {
sub_agg.check_flush_local(agg_data)?;
}
@@ -318,15 +314,26 @@ impl<C: SubAggCache> SegmentAggregationCollector for SegmentRangeCollector<C> {
fn prepare_max_bucket(
&mut self,
max_bucket: BucketId,
agg_data: &AggregationsSegmentCtx,
_agg_data: &AggregationsSegmentCtx,
) -> crate::Result<()> {
while self.parent_buckets.len() <= max_bucket as usize {
let new_buckets = self.create_new_buckets(agg_data)?;
let new_buckets = self.create_new_buckets()?;
self.parent_buckets.push(new_buckets);
}
Ok(())
}
fn compute_metric_value(
&self,
_bucket_id: BucketId,
_sub_agg_name: &str,
_sub_agg_property: &str,
_agg_data: &AggregationsSegmentCtx,
) -> Option<f64> {
// Range is a multi-bucket agg with no single value to extract.
None
}
}
/// Build a concrete `SegmentRangeCollector` with either a Vec- or HashMap-backed
/// bucket storage, depending on the column type and aggregation level.
@@ -334,8 +341,11 @@ pub(crate) fn build_segment_range_collector(
agg_data: &mut AggregationsSegmentCtx,
node: &AggRefNode,
) -> crate::Result<Box<dyn SegmentAggregationCollector>> {
let accessor_idx = node.idx_in_req_data;
let req_data = agg_data.get_range_req_data(node.idx_in_req_data);
let req_data = agg_data.per_request.range_req_data[node.idx_in_req_data].clone();
agg_data
.context
.limits
.add_memory_consumed(req_data.get_memory_consumption() as u64)?;
let field_type = req_data.field_type;
// TODO: A better metric instead of is_top_level would be the number of buckets expected.
@@ -350,19 +360,19 @@ pub(crate) fn build_segment_range_collector(
};
if is_low_card {
Ok(Box::new(SegmentRangeCollector::<LowCardSubAggCache> {
sub_agg: sub_agg.map(LowCardCachedSubAggs::new),
Ok(Box::new(SegmentRangeCollector::<LowCardSubAggBuffer> {
sub_agg: sub_agg.map(LowCardBufferedSubAggs::new),
column_type: field_type,
accessor_idx,
req_data,
parent_buckets: Vec::new(),
bucket_id_provider: BucketIdProvider::default(),
limits: agg_data.context.limits.clone(),
}))
} else {
Ok(Box::new(SegmentRangeCollector::<HighCardSubAggCache> {
sub_agg: sub_agg.map(CachedSubAggs::new),
Ok(Box::new(SegmentRangeCollector::<HighCardSubAggBuffer> {
sub_agg: sub_agg.map(BufferedSubAggs::new),
column_type: field_type,
accessor_idx,
req_data,
parent_buckets: Vec::new(),
bucket_id_provider: BucketIdProvider::default(),
limits: agg_data.context.limits.clone(),
@@ -370,13 +380,10 @@ pub(crate) fn build_segment_range_collector(
}
}
impl<C: SubAggCache> SegmentRangeCollector<C> {
pub(crate) fn create_new_buckets(
&mut self,
agg_data: &AggregationsSegmentCtx,
) -> crate::Result<Vec<SegmentRangeAndBucketEntry>> {
impl<B: SubAggBuffer> SegmentRangeCollector<B> {
pub(crate) fn create_new_buckets(&mut self) -> crate::Result<Vec<SegmentRangeAndBucketEntry>> {
let field_type = self.column_type;
let req_data = agg_data.get_range_req_data(self.accessor_idx);
let req_data = &self.req_data;
// 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.
@@ -551,17 +558,16 @@ mod tests {
get_test_index_with_num_docs,
};
pub fn get_collector_from_ranges(
ranges: Vec<RangeAggregationRange>,
pub fn build_test_buckets(
ranges: &[RangeAggregationRange],
field_type: ColumnType,
) -> SegmentRangeCollector<HighCardSubAggCache> {
) -> Vec<SegmentRangeAndBucketEntry> {
let req = RangeAggregation {
field: "dummy".to_string(),
ranges,
ranges: ranges.to_vec(),
..Default::default()
};
// Build buckets directly as in from_req_and_validate without AggregationsData
let buckets: Vec<_> = extend_validate_ranges(&req.ranges, &field_type)
extend_validate_ranges(&req.ranges, &field_type)
.expect("unexpected error in extend_validate_ranges")
.iter()
.map(|range| {
@@ -592,16 +598,7 @@ mod tests {
},
}
})
.collect();
SegmentRangeCollector {
parent_buckets: vec![buckets],
column_type: field_type,
accessor_idx: 0,
sub_agg: None,
bucket_id_provider: Default::default(),
limits: AggregationLimitsGuard::default(),
}
.collect()
}
#[test]
@@ -844,10 +841,10 @@ mod tests {
#[test]
fn bucket_test_extend_range_hole() {
let buckets = vec![(10f64..20f64).into(), (30f64..40f64).into()];
let collector = get_collector_from_ranges(buckets, ColumnType::F64);
let buckets = [(10f64..20f64).into(), (30f64..40f64).into()];
let parent_buckets = [build_test_buckets(&buckets, ColumnType::F64)];
let buckets = collector.parent_buckets[0].clone();
let buckets = parent_buckets[0].clone();
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());
@@ -863,14 +860,14 @@ mod tests {
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![
let buckets = [
(f64::MIN..10f64).into(),
(10f64..20f64).into(),
(20f64..f64::MAX).into(),
];
let collector = get_collector_from_ranges(buckets, ColumnType::F64);
let parent_buckets = [build_test_buckets(&buckets, ColumnType::F64)];
let buckets = collector.parent_buckets[0].clone();
let buckets = parent_buckets[0].clone();
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());
@@ -882,28 +879,28 @@ mod tests {
#[test]
fn bucket_range_test_negative_vals() {
let buckets = vec![(-10f64..-1f64).into()];
let collector = get_collector_from_ranges(buckets, ColumnType::F64);
let buckets = [(-10f64..-1f64).into()];
let parent_buckets = [build_test_buckets(&buckets, ColumnType::F64)];
let buckets = collector.parent_buckets[0].clone();
let buckets = parent_buckets[0].clone();
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, ColumnType::F64);
let buckets = [(0f64..10f64).into()];
let parent_buckets = [build_test_buckets(&buckets, ColumnType::F64)];
let buckets = collector.parent_buckets[0].clone();
let buckets = parent_buckets[0].clone();
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, ColumnType::U64);
let search = |val: u64| get_bucket_pos(val, &collector.parent_buckets[0]);
let check_ranges = |ranges: &[RangeAggregationRange]| {
let parent_buckets = [build_test_buckets(ranges, ColumnType::U64)];
let search = |val: u64| get_bucket_pos(val, &parent_buckets[0]);
assert_eq!(search(u64::MIN), 0);
assert_eq!(search(9), 0);
@@ -916,7 +913,7 @@ mod tests {
};
let ranges = vec![(10.0..100.0).into()];
check_ranges(ranges);
check_ranges(&ranges);
let ranges = vec![
RangeAggregationRange {
@@ -926,7 +923,7 @@ mod tests {
},
(10.0..100.0).into(),
];
check_ranges(ranges);
check_ranges(&ranges);
let ranges = vec![
RangeAggregationRange {
@@ -941,15 +938,15 @@ mod tests {
from: Some(100.0),
},
];
check_ranges(ranges);
check_ranges(&ranges);
}
#[test]
fn range_binary_search_test_f64() {
let ranges = vec![(10.0..100.0).into()];
let ranges = [(10.0..100.0).into()];
let collector = get_collector_from_ranges(ranges, ColumnType::F64);
let search = |val: u64| get_bucket_pos(val, &collector.parent_buckets[0]);
let parent_buckets = [build_test_buckets(&ranges, ColumnType::F64)];
let search = |val: u64| get_bucket_pos(val, &parent_buckets[0]);
assert_eq!(search(u64::MIN), 0);
assert_eq!(search(9f64.to_u64()), 0);

636
src/aggregation/bucket/term_agg.rs → src/aggregation/bucket/term_agg/mod.rs
View File

@@ -1,5 +1,4 @@
use std::fmt::Debug;
use std::io;
use std::net::Ipv6Addr;
use columnar::column_values::CompactSpaceU64Accessor;
@@ -17,8 +16,9 @@ use crate::aggregation::agg_data::{
};
use crate::aggregation::agg_limits::MemoryConsumption;
use crate::aggregation::agg_req::Aggregations;
use crate::aggregation::cached_sub_aggs::{
CachedSubAggs, HighCardSubAggCache, LowCardCachedSubAggs, LowCardSubAggCache, SubAggCache,
use crate::aggregation::buffered_sub_aggs::{
BufferedSubAggs, HighCardSubAggBuffer, LowCardBufferedSubAggs, LowCardSubAggBuffer,
SubAggBuffer,
};
use crate::aggregation::intermediate_agg_result::{
IntermediateAggregationResult, IntermediateAggregationResults, IntermediateBucketResult,
@@ -29,6 +29,8 @@ use crate::aggregation::{format_date, BucketId, Key};
use crate::error::DataCorruption;
use crate::TantivyError;
mod term_histogram;
/// Contains all information required by the SegmentTermCollector to perform the
/// terms aggregation on a segment.
#[derive(Debug, Clone)]
@@ -61,7 +63,7 @@ impl TermsAggReqData {
+ self
.allowed_term_ids
.as_ref()
.map(|bs| bs.get_memory_consumption())
.map(|bs| bs.len() / 8)
.unwrap_or(0)
}
}
@@ -352,19 +354,15 @@ pub(crate) fn build_segment_term_collector(
)));
}
// Validate sub aggregation exists when ordering by sub-aggregation.
{
if let OrderTarget::SubAggregation(sub_agg_name) = &terms_req_data.req.order.target {
let (agg_name, _agg_property) = get_agg_name_and_property(sub_agg_name);
node.get_sub_agg(agg_name, &req_data.per_request)
.ok_or_else(|| {
TantivyError::InvalidArgument(format!(
"could not find aggregation with name {agg_name} in metric \
sub_aggregations"
))
})?;
}
// Validate that the referenced sub-aggregation exists when ordering by one.
if let OrderTarget::SubAggregation(sub_agg_name) = &terms_req_data.req.order.target {
let (agg_name, _agg_property) = get_agg_name_and_property(sub_agg_name);
node.get_sub_agg(agg_name, &req_data.per_request)
.ok_or_else(|| {
TantivyError::InvalidArgument(format!(
"could not find aggregation with name {agg_name} in metric sub_aggregations"
))
})?;
}
// Build sub-aggregation blueprint if there are children.
@@ -378,9 +376,21 @@ pub(crate) fn build_segment_term_collector(
// Let's see if we can use a vec to aggregate our data
// instead of a hashmap.
let col_max_value = terms_req_data.accessor.max_value();
let max_term_id: u64 =
let max_column_val: u64 =
col_max_value.max(terms_req_data.missing_value_for_accessor.unwrap_or(0u64));
// Fused fast path: low-cardinality terms × a single `histogram`/`date_histogram` leaf over full
// columns with a small enough bucket grid. Anything else falls through to the general path.
if let Some(collector) = term_histogram::maybe_build_collector(
req_data,
node,
&terms_req_data,
max_column_val,
is_top_level,
)? {
return Ok(collector);
}
let sub_agg_collector = if has_sub_aggregations {
Some(build_segment_agg_collectors(req_data, &node.children)?)
} else {
@@ -389,51 +399,51 @@ pub(crate) fn build_segment_term_collector(
let mut bucket_id_provider = BucketIdProvider::default();
// Decide which bucket storage is best suited for this aggregation.
if is_top_level && max_term_id < MAX_NUM_TERMS_FOR_VEC && !has_sub_aggregations {
let term_buckets = VecTermBucketsNoAgg::new(max_term_id + 1, &mut bucket_id_provider);
let collector: SegmentTermCollector<_, HighCardSubAggCache> = SegmentTermCollector {
if is_top_level && max_column_val < MAX_NUM_TERMS_FOR_VEC && !has_sub_aggregations {
let term_buckets = VecTermBucketsNoAgg::new(max_column_val + 1, &mut bucket_id_provider);
let collector: SegmentTermCollector<_, HighCardSubAggBuffer> = SegmentTermCollector {
parent_buckets: vec![term_buckets],
sub_agg: None,
bucket_id_provider,
max_term_id,
max_term_id: max_column_val,
terms_req_data,
};
Ok(Box::new(collector))
} else if is_top_level && max_term_id < MAX_NUM_TERMS_FOR_VEC {
let term_buckets = VecTermBuckets::new(max_term_id + 1, &mut bucket_id_provider);
let sub_agg = sub_agg_collector.map(LowCardCachedSubAggs::new);
let collector: SegmentTermCollector<_, LowCardSubAggCache> = SegmentTermCollector {
} else if is_top_level && max_column_val < MAX_NUM_TERMS_FOR_VEC {
let term_buckets = VecTermBuckets::new(max_column_val + 1, &mut bucket_id_provider);
let sub_agg = sub_agg_collector.map(LowCardBufferedSubAggs::new);
let collector: SegmentTermCollector<_, LowCardSubAggBuffer> = SegmentTermCollector {
parent_buckets: vec![term_buckets],
sub_agg,
bucket_id_provider,
max_term_id,
max_term_id: max_column_val,
terms_req_data,
};
Ok(Box::new(collector))
} else if max_term_id < 8_000_000 && is_top_level {
} else if max_column_val < 8_000_000 && is_top_level {
let term_buckets: PagedTermMap =
PagedTermMap::new(max_term_id + 1, &mut bucket_id_provider);
PagedTermMap::new(max_column_val + 1, &mut bucket_id_provider);
// Build sub-aggregation blueprint (flat pairs)
let sub_agg = sub_agg_collector.map(CachedSubAggs::new);
let collector: SegmentTermCollector<PagedTermMap, HighCardSubAggCache> =
let sub_agg = sub_agg_collector.map(BufferedSubAggs::new);
let collector: SegmentTermCollector<PagedTermMap, HighCardSubAggBuffer> =
SegmentTermCollector {
parent_buckets: vec![term_buckets],
sub_agg,
bucket_id_provider,
max_term_id,
max_term_id: max_column_val,
terms_req_data,
};
Ok(Box::new(collector))
} else {
let term_buckets: HashMapTermBuckets = HashMapTermBuckets::default();
// Build sub-aggregation blueprint (flat pairs)
let sub_agg = sub_agg_collector.map(CachedSubAggs::new);
let collector: SegmentTermCollector<HashMapTermBuckets, HighCardSubAggCache> =
let sub_agg = sub_agg_collector.map(BufferedSubAggs::new);
let collector: SegmentTermCollector<HashMapTermBuckets, HighCardSubAggBuffer> =
SegmentTermCollector {
parent_buckets: vec![term_buckets],
sub_agg,
bucket_id_provider,
max_term_id,
max_term_id: max_column_val,
terms_req_data,
};
Ok(Box::new(collector))
@@ -758,10 +768,10 @@ impl TermAggregationMap for VecTermBuckets {
/// The collector puts values from the fast field into the correct buckets and does a conversion to
/// the correct datatype.
#[derive(Debug)]
struct SegmentTermCollector<TermMap: TermAggregationMap, C: SubAggCache> {
struct SegmentTermCollector<TermMap: TermAggregationMap, B: SubAggBuffer> {
/// The buckets containing the aggregation data.
parent_buckets: Vec<TermMap>,
sub_agg: Option<CachedSubAggs<C>>,
sub_agg: Option<BufferedSubAggs<B>>,
bucket_id_provider: BucketIdProvider,
max_term_id: u64,
terms_req_data: TermsAggReqData,
@@ -772,8 +782,8 @@ pub(crate) fn get_agg_name_and_property(name: &str) -> (&str, &str) {
(agg_name, agg_property)
}
impl<TermMap: TermAggregationMap, C: SubAggCache> SegmentAggregationCollector
for SegmentTermCollector<TermMap, C>
impl<TermMap: TermAggregationMap, B: SubAggBuffer> SegmentAggregationCollector
for SegmentTermCollector<TermMap, B>
{
fn add_intermediate_aggregation_result(
&mut self,
@@ -790,8 +800,14 @@ impl<TermMap: TermAggregationMap, C: SubAggCache> SegmentAggregationCollector
let term_req = &self.terms_req_data;
let name = term_req.name.clone();
let bucket =
Self::into_intermediate_bucket_result(term_req, &mut self.sub_agg, bucket, agg_data)?;
let bucket = Self::into_intermediate_bucket_result(
term_req,
self.sub_agg
.as_mut()
.map(BufferedSubAggs::get_sub_agg_collector),
bucket,
agg_data,
)?;
results.push(name, IntermediateAggregationResult::Bucket(bucket))?;
Ok(())
}
@@ -803,7 +819,7 @@ impl<TermMap: TermAggregationMap, C: SubAggCache> SegmentAggregationCollector
docs: &[crate::DocId],
agg_data: &mut AggregationsSegmentCtx,
) -> crate::Result<()> {
let mem_pre = self.get_memory_consumption();
let mem_pre = self.get_memory_consumption(parent_bucket_id);
let req_data = &mut self.terms_req_data;
@@ -847,7 +863,7 @@ impl<TermMap: TermAggregationMap, C: SubAggCache> SegmentAggregationCollector
}
}
let mem_delta = self.get_memory_consumption() - mem_pre;
let mem_delta = self.get_memory_consumption(parent_bucket_id) - mem_pre;
if mem_delta > 0 {
agg_data
.context
@@ -881,6 +897,17 @@ impl<TermMap: TermAggregationMap, C: SubAggCache> SegmentAggregationCollector
}
Ok(())
}
fn compute_metric_value(
&self,
_bucket_id: BucketId,
_sub_agg_name: &str,
_sub_agg_property: &str,
_agg_data: &AggregationsSegmentCtx,
) -> Option<f64> {
// Terms is a multi-bucket agg with no single value to extract.
None
}
}
/// Missing value are represented as a sentinel value in the column.
@@ -907,30 +934,53 @@ fn extract_missing_value<T>(
Some((key, bucket))
}
impl<TermMap, C> SegmentTermCollector<TermMap, C>
fn reborrow_opt_collector<'a>(
opt: &'a mut Option<&mut dyn SegmentAggregationCollector>,
) -> Option<&'a mut dyn SegmentAggregationCollector> {
match opt {
Some(inner) => Some(*inner),
None => None,
}
}
fn into_intermediate_bucket_entry(
bucket: Bucket,
sub_agg_collector: Option<&mut dyn SegmentAggregationCollector>,
agg_data: &AggregationsSegmentCtx,
) -> crate::Result<IntermediateTermBucketEntry> {
let mut sub_aggregation_res = IntermediateAggregationResults::default();
if let Some(sub_agg_collector) = sub_agg_collector {
sub_agg_collector.add_intermediate_aggregation_result(
agg_data,
&mut sub_aggregation_res,
bucket.bucket_id,
)?;
}
Ok(IntermediateTermBucketEntry {
doc_count: bucket.count,
sub_aggregation: sub_aggregation_res,
})
}
impl<TermMap, B> SegmentTermCollector<TermMap, B>
where
TermMap: TermAggregationMap,
C: SubAggCache,
B: SubAggBuffer,
{
fn get_memory_consumption(&self) -> usize {
self.parent_buckets
.iter()
.map(|b| b.get_memory_consumption())
.sum()
#[inline]
fn get_memory_consumption(&self, parent_bucket_id: BucketId) -> usize {
self.parent_buckets[parent_bucket_id as usize].get_memory_consumption()
}
#[inline]
pub(crate) fn into_intermediate_bucket_result(
term_req: &TermsAggReqData,
sub_agg: &mut Option<CachedSubAggs<C>>,
mut sub_agg_collector: Option<&mut dyn SegmentAggregationCollector>,
term_buckets: TermMap,
agg_data: &AggregationsSegmentCtx,
) -> crate::Result<IntermediateBucketResult> {
let mut entries: Vec<(u64, Bucket)> = term_buckets.into_vec();
let order_by_sub_aggregation =
matches!(term_req.req.order.target, OrderTarget::SubAggregation(_));
match &term_req.req.order.target {
OrderTarget::Key => {
// We rely on the fact, that term ordinals match the order of the strings
@@ -942,10 +992,37 @@ where
entries.sort_unstable_by_key(|bucket| bucket.0);
}
}
OrderTarget::SubAggregation(_name) => {
// don't sort and cut off since it's hard to make assumptions on the quality of the
// results when cutting off du to unknown nature of the sub_aggregation (possible
// to check).
OrderTarget::SubAggregation(sub_agg_path) => {
// Peek segment-level metric values, sort, then fall through to
// `cut_off_buckets`. Like Elasticsearch, we always cut off when ordering
// by a sub-agg: top-K results are approximate and may differ from the
// global ordering, especially for non-monotonic metrics like avg/min.
let coll = sub_agg_collector.as_deref().ok_or_else(|| {
TantivyError::InvalidArgument(format!(
"Could not find sub-aggregation collector for path {sub_agg_path}"
))
})?;
let (agg_name, agg_prop) = get_agg_name_and_property(sub_agg_path);
// Fetch values up-front; otherwise sort would re-compute per comparison
let mut keyed: Vec<(f64, (u64, Bucket))> = entries
.into_iter()
.map(|bucket| {
let metric_value = coll
.compute_metric_value(bucket.1.bucket_id, agg_name, agg_prop, agg_data)
.unwrap_or(0.0);
(metric_value, bucket)
})
.collect();
if term_req.req.order.order == Order::Desc {
keyed.sort_unstable_by(|a, b| {
b.0.partial_cmp(&a.0).unwrap_or(std::cmp::Ordering::Equal)
});
} else {
keyed.sort_unstable_by(|a, b| {
a.0.partial_cmp(&b.0).unwrap_or(std::cmp::Ordering::Equal)
});
}
entries = keyed.into_iter().map(|(_, e)| e).collect();
}
OrderTarget::Count => {
if term_req.req.order.order == Order::Desc {
@@ -956,40 +1033,12 @@ where
}
}
let (term_doc_count_before_cutoff, sum_other_doc_count) = if order_by_sub_aggregation {
(0, 0)
} else {
cut_off_buckets(&mut entries, term_req.req.segment_size as usize)
};
let (term_doc_count_before_cutoff, sum_other_doc_count) =
cut_off_buckets(&mut entries, term_req.req.segment_size as usize);
let mut dict: FxHashMap<IntermediateKey, IntermediateTermBucketEntry> = Default::default();
dict.reserve(entries.len());
let into_intermediate_bucket_entry =
|bucket: Bucket,
sub_agg: &mut Option<CachedSubAggs<C>>|
-> crate::Result<IntermediateTermBucketEntry> {
if let Some(sub_agg) = sub_agg {
let mut sub_aggregation_res = IntermediateAggregationResults::default();
sub_agg
.get_sub_agg_collector()
.add_intermediate_aggregation_result(
agg_data,
&mut sub_aggregation_res,
bucket.bucket_id,
)?;
Ok(IntermediateTermBucketEntry {
doc_count: bucket.count,
sub_aggregation: sub_aggregation_res,
})
} else {
Ok(IntermediateTermBucketEntry {
doc_count: bucket.count,
sub_aggregation: Default::default(),
})
}
};
if term_req.column_type == ColumnType::Str {
let fallback_dict = Dictionary::empty();
let term_dict = term_req
@@ -1000,7 +1049,11 @@ where
if let Some((intermediate_key, bucket)) = extract_missing_value(&mut entries, term_req)
{
let intermediate_entry = into_intermediate_bucket_entry(bucket, sub_agg)?;
let intermediate_entry = into_intermediate_bucket_entry(
bucket,
reborrow_opt_collector(&mut sub_agg_collector),
agg_data,
)?;
dict.insert(intermediate_key, intermediate_entry);
}
@@ -1008,19 +1061,28 @@ where
entries.sort_unstable_by_key(|bucket| bucket.0);
let (term_ids, buckets): (Vec<u64>, Vec<Bucket>) = entries.into_iter().unzip();
let mut buckets_it = buckets.into_iter();
term_dict.sorted_ords_to_term_cb(term_ids.into_iter(), |term| {
let bucket = buckets_it.next().unwrap();
let intermediate_entry =
into_intermediate_bucket_entry(bucket, sub_agg).map_err(io::Error::other)?;
let intermediate_entries: Vec<IntermediateTermBucketEntry> = buckets
.into_iter()
.map(|bucket| {
into_intermediate_bucket_entry(
bucket,
reborrow_opt_collector(&mut sub_agg_collector),
agg_data,
)
})
.collect::<crate::Result<_>>()?;
let mut intermediate_entry_it = intermediate_entries.into_iter();
term_dict.sorted_ords_to_term_cb(&term_ids[..], |term| {
let intermediate_entry = intermediate_entry_it.next().unwrap();
dict.insert(
IntermediateKey::Str(
String::from_utf8(term.to_vec()).expect("could not convert to String"),
),
intermediate_entry,
);
Ok(())
})?;
if term_req.req.min_doc_count == 0 {
@@ -1055,14 +1117,22 @@ where
}
} else if term_req.column_type == ColumnType::DateTime {
for (val, doc_count) in entries {
let intermediate_entry = into_intermediate_bucket_entry(doc_count, sub_agg)?;
let intermediate_entry = into_intermediate_bucket_entry(
doc_count,
reborrow_opt_collector(&mut sub_agg_collector),
agg_data,
)?;
let val = i64::from_u64(val);
let date = format_date(val)?;
dict.insert(IntermediateKey::Str(date), intermediate_entry);
}
} else if term_req.column_type == ColumnType::Bool {
for (val, doc_count) in entries {
let intermediate_entry = into_intermediate_bucket_entry(doc_count, sub_agg)?;
let intermediate_entry = into_intermediate_bucket_entry(
doc_count,
reborrow_opt_collector(&mut sub_agg_collector),
agg_data,
)?;
let val = bool::from_u64(val);
dict.insert(IntermediateKey::Bool(val), intermediate_entry);
}
@@ -1082,14 +1152,22 @@ where
})?;
for (val, doc_count) in entries {
let intermediate_entry = into_intermediate_bucket_entry(doc_count, sub_agg)?;
let intermediate_entry = into_intermediate_bucket_entry(
doc_count,
reborrow_opt_collector(&mut sub_agg_collector),
agg_data,
)?;
let val: u128 = compact_space_accessor.compact_to_u128(val as u32);
let val = Ipv6Addr::from_u128(val);
dict.insert(IntermediateKey::IpAddr(val), intermediate_entry);
}
} else {
for (val, doc_count) in entries {
let intermediate_entry = into_intermediate_bucket_entry(doc_count, sub_agg)?;
let intermediate_entry = into_intermediate_bucket_entry(
doc_count,
reborrow_opt_collector(&mut sub_agg_collector),
agg_data,
)?;
if term_req.column_type == ColumnType::U64 {
dict.insert(IntermediateKey::U64(val), intermediate_entry);
} else if term_req.column_type == ColumnType::I64 {
@@ -1123,13 +1201,13 @@ where
}
}
impl<TermMap: TermAggregationMap, C: SubAggCache> SegmentTermCollector<TermMap, C> {
impl<TermMap: TermAggregationMap, B: SubAggBuffer> SegmentTermCollector<TermMap, B> {
#[inline]
fn collect_terms_with_docs(
iter: impl Iterator<Item = (crate::DocId, u64)>,
term_buckets: &mut TermMap,
bucket_id_provider: &mut BucketIdProvider,
sub_agg: &mut CachedSubAggs<C>,
sub_agg: &mut BufferedSubAggs<B>,
) {
for (doc, term_id) in iter {
let bucket_id = term_buckets.term_entry(term_id, bucket_id_provider);
@@ -1202,7 +1280,7 @@ mod tests {
use crate::aggregation::{AggregationLimitsGuard, DistributedAggregationCollector};
use crate::indexer::NoMergePolicy;
use crate::query::AllQuery;
use crate::schema::{IntoIpv6Addr, Schema, FAST, STRING};
use crate::schema::{IntoIpv6Addr, Schema, FAST, INDEXED, STRING, TEXT};
use crate::{Index, IndexWriter};
#[test]
@@ -1731,6 +1809,263 @@ mod tests {
Ok(())
}
#[test]
fn terms_aggregation_order_by_cardinality_desc_single_segment() -> crate::Result<()> {
terms_aggregation_order_by_cardinality_desc(true)
}
#[test]
fn terms_aggregation_order_by_cardinality_desc_multi_segment() -> crate::Result<()> {
terms_aggregation_order_by_cardinality_desc(false)
}
fn terms_aggregation_order_by_cardinality_desc(merge_segments: bool) -> crate::Result<()> {
// Distinct score values per bucket key: A→5, B→1, C→3.
// Order by cardinality desc must yield A, C, B.
let segment_and_terms = vec![vec![
(1.0, "A".to_string()),
(2.0, "A".to_string()),
(3.0, "A".to_string()),
(4.0, "A".to_string()),
(5.0, "A".to_string()),
(1.0, "B".to_string()),
(1.0, "B".to_string()),
(1.0, "B".to_string()),
(1.0, "C".to_string()),
(2.0, "C".to_string()),
(3.0, "C".to_string()),
]];
let index = get_test_index_from_values_and_terms(merge_segments, &segment_and_terms)?;
let agg_req: Aggregations = serde_json::from_value(json!({
"my_texts": {
"terms": {
"field": "string_id",
"order": { "card": "desc" }
},
"aggs": {
"card": { "cardinality": { "field": "score" } }
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
assert_eq!(res["my_texts"]["buckets"][0]["key"], "A");
assert_eq!(res["my_texts"]["buckets"][0]["card"]["value"], 5.0);
assert_eq!(res["my_texts"]["buckets"][1]["key"], "C");
assert_eq!(res["my_texts"]["buckets"][1]["card"]["value"], 3.0);
assert_eq!(res["my_texts"]["buckets"][2]["key"], "B");
assert_eq!(res["my_texts"]["buckets"][2]["card"]["value"], 1.0);
// Asc engages the segment-cutoff path too (monotonic-safe: discarded buckets had
// local card >= cutoff, so merged card >= cutoff and they cannot be globally smallest).
let agg_req: Aggregations = serde_json::from_value(json!({
"my_texts": {
"terms": {
"field": "string_id",
"order": { "card": "asc" }
},
"aggs": {
"card": { "cardinality": { "field": "score" } }
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
assert_eq!(res["my_texts"]["buckets"][0]["key"], "B");
assert_eq!(res["my_texts"]["buckets"][1]["key"], "C");
assert_eq!(res["my_texts"]["buckets"][2]["key"], "A");
// size=2 with desc engages the segment cutoff: must keep top-2 by cardinality (A, C),
// and `sum_other_doc_count` reflects the dropped B (3 docs).
let agg_req: Aggregations = serde_json::from_value(json!({
"my_texts": {
"terms": {
"field": "string_id",
"size": 2,
"order": { "card": "desc" }
},
"aggs": {
"card": { "cardinality": { "field": "score" } }
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
assert_eq!(res["my_texts"]["buckets"][0]["key"], "A");
assert_eq!(res["my_texts"]["buckets"][1]["key"], "C");
assert_eq!(res["my_texts"]["buckets"].as_array().unwrap().len(), 2);
// size=2 with asc engages the segment cutoff: must keep bottom-2 by cardinality (B, C).
let agg_req: Aggregations = serde_json::from_value(json!({
"my_texts": {
"terms": {
"field": "string_id",
"size": 2,
"order": { "card": "asc" }
},
"aggs": {
"card": { "cardinality": { "field": "score" } }
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
assert_eq!(res["my_texts"]["buckets"][0]["key"], "B");
assert_eq!(res["my_texts"]["buckets"][1]["key"], "C");
assert_eq!(res["my_texts"]["buckets"].as_array().unwrap().len(), 2);
Ok(())
}
#[test]
fn terms_aggregation_order_by_sum_single_segment() -> crate::Result<()> {
terms_aggregation_order_by_sum(true)
}
#[test]
fn terms_aggregation_order_by_sum_multi_segment() -> crate::Result<()> {
terms_aggregation_order_by_sum(false)
}
fn terms_aggregation_order_by_sum(merge_segments: bool) -> crate::Result<()> {
// Per-bucket sums on the U64 `score` column (non-negative => sum is monotonic):
// A → 1+2+3+4+5 = 15, B → 1+1+1 = 3, C → 1+2+3 = 6.
let segment_and_terms = vec![
vec![
(1.0, "A".to_string()),
(2.0, "A".to_string()),
(3.0, "A".to_string()),
(1.0, "B".to_string()),
(1.0, "C".to_string()),
],
vec![
(4.0, "A".to_string()),
(5.0, "A".to_string()),
(1.0, "B".to_string()),
(1.0, "B".to_string()),
(2.0, "C".to_string()),
(3.0, "C".to_string()),
],
];
let index = get_test_index_from_values_and_terms(merge_segments, &segment_and_terms)?;
// Desc on a Sum metric engages the fast path (column is U64).
let agg_req: Aggregations = serde_json::from_value(json!({
"my_texts": {
"terms": {
"field": "string_id",
"order": { "total": "desc" }
},
"aggs": {
"total": { "sum": { "field": "score" } }
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
assert_eq!(res["my_texts"]["buckets"][0]["key"], "A");
assert_eq!(res["my_texts"]["buckets"][0]["total"]["value"], 15.0);
assert_eq!(res["my_texts"]["buckets"][1]["key"], "C");
assert_eq!(res["my_texts"]["buckets"][1]["total"]["value"], 6.0);
assert_eq!(res["my_texts"]["buckets"][2]["key"], "B");
assert_eq!(res["my_texts"]["buckets"][2]["total"]["value"], 3.0);
// Asc engages the fast path too — discarded buckets had local sum >= cutoff,
// and merged sum >= local (non-negative addends), so they cannot be globally smallest.
let agg_req: Aggregations = serde_json::from_value(json!({
"my_texts": {
"terms": {
"field": "string_id",
"order": { "total": "asc" }
},
"aggs": {
"total": { "sum": { "field": "score" } }
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
assert_eq!(res["my_texts"]["buckets"][0]["key"], "B");
assert_eq!(res["my_texts"]["buckets"][1]["key"], "C");
assert_eq!(res["my_texts"]["buckets"][2]["key"], "A");
// size=2 desc with cutoff: top-2 by sum (A, C).
let agg_req: Aggregations = serde_json::from_value(json!({
"my_texts": {
"terms": {
"field": "string_id",
"size": 2,
"order": { "total": "desc" }
},
"aggs": {
"total": { "sum": { "field": "score" } }
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
assert_eq!(res["my_texts"]["buckets"][0]["key"], "A");
assert_eq!(res["my_texts"]["buckets"][1]["key"], "C");
assert_eq!(res["my_texts"]["buckets"].as_array().unwrap().len(), 2);
// Stats sub-property: ordering by `mystats.sum` on a U64 column also engages.
let agg_req: Aggregations = serde_json::from_value(json!({
"my_texts": {
"terms": {
"field": "string_id",
"order": { "mystats.sum": "desc" }
},
"aggs": {
"mystats": { "stats": { "field": "score" } }
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
assert_eq!(res["my_texts"]["buckets"][0]["key"], "A");
assert_eq!(res["my_texts"]["buckets"][1]["key"], "C");
assert_eq!(res["my_texts"]["buckets"][2]["key"], "B");
// Sum on a signed column (I64) takes the same cutoff path. Results may be
// approximate near the boundary on adversarial data, but for this dataset the
// top-K is unambiguous.
let agg_req: Aggregations = serde_json::from_value(json!({
"my_texts": {
"terms": {
"field": "string_id",
"order": { "total": "desc" }
},
"aggs": {
"total": { "sum": { "field": "score_i64" } }
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
assert_eq!(res["my_texts"]["buckets"][0]["key"], "A");
assert_eq!(res["my_texts"]["buckets"][1]["key"], "C");
assert_eq!(res["my_texts"]["buckets"][2]["key"], "B");
// Order by extended_stats sub-property exercises compute_metric_value on the
// ExtendedStats collector. A→max=5, B→max=1, C→max=3, so desc by max → A, C, B.
let agg_req: Aggregations = serde_json::from_value(json!({
"my_texts": {
"terms": {
"field": "string_id",
"order": { "ext.max": "desc" }
},
"aggs": {
"ext": { "extended_stats": { "field": "score" } }
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
assert_eq!(res["my_texts"]["buckets"][0]["key"], "A");
assert_eq!(res["my_texts"]["buckets"][1]["key"], "C");
assert_eq!(res["my_texts"]["buckets"][2]["key"], "B");
Ok(())
}
#[test]
fn terms_aggregation_test_order_key_single_segment() -> crate::Result<()> {
terms_aggregation_test_order_key_merge_segment(true)
@@ -2896,4 +3231,101 @@ mod tests {
Ok(())
}
fn prep_index_with_n_unique_terms_plus_one_null(n: u64) -> crate::Result<Index> {
let mut schema_builder = Schema::builder();
let id_field = schema_builder.add_u64_field("id", INDEXED);
let title_field = schema_builder.add_text_field("title", TEXT | FAST);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema.clone());
// set to one thread to guarantee all docs end up in the same segment
let mut writer = index.writer_with_num_threads(1, 50_000_000)?;
writer.add_document(doc!(
id_field => 0u64,
))?;
for i in 1u64..=n {
let title = format!("foo{i}");
writer.add_document(doc!(
id_field => i,
title_field => title,
))?;
}
writer.commit()?;
Ok(index)
}
#[test]
fn null_bitset_bounds_check_regression() -> crate::Result<()> {
// include cases
for i in 0..=4 {
let index = prep_index_with_n_unique_terms_plus_one_null(i * 64)?;
let normal_req: Aggregations = serde_json::from_value(json!({
"my_bool": {
"terms": {
"field": "title",
"missing": "__NULL__",
"size": 1000,
}
}
}))?;
let include_req: Aggregations = serde_json::from_value(json!({
"my_bool": {
"terms": {
"field": "title",
"include": "foo(.*)",
"missing": "__NULL__",
"size": 1000,
}
}
}))?;
let exclude_req: Aggregations = serde_json::from_value(json!({
"my_bool": {
"terms": {
"field": "title",
"exclude": "foo(.*)",
"missing": "__NULL__",
"size": 1000,
}
}
}))?;
let normal_res = exec_request(normal_req, &index)?;
let normal_buckets = normal_res["my_bool"]["buckets"].as_array().unwrap();
assert_eq!(
normal_buckets.len(),
(i * 64) as usize + 1,
"The normal request should return all 'foo' buckets, plus the missing term bucket",
);
let include_res = exec_request(include_req, &index)?;
eprintln!("include_res: {include_res:?}");
let include_buckets = include_res["my_bool"]["buckets"].as_array().unwrap();
assert_eq!(
include_buckets.len(),
(i * 64) as usize,
"The include request should return all 'foo' buckets, and not the missing term \
bucket",
);
assert!(include_buckets
.iter()
.all(|b| b["key"].as_str().unwrap().starts_with("foo")));
let exclude_res = exec_request(exclude_req, &index)?;
let exclude_buckets = exclude_res["my_bool"]["buckets"].as_array().unwrap();
if i != 0 {
// TODO: Remove this if after fixing exclude + missing bug
assert_eq!(
exclude_buckets.len(),
1,
"The exclude request should exclude all 'foo' buckets, and only the missing \
term bucket",
);
assert_eq!(exclude_buckets[0]["key"], "__NULL__");
}
}
Ok(())
}
}

585
src/aggregation/bucket/term_agg/term_histogram.rs Normal file
View File

@@ -0,0 +1,585 @@
//! Fused collector for the very common shape `terms` (low cardinality) × a single
//! `histogram`/`date_histogram` sub-aggregation with nothing nested below it.
//!
//! See [`SegmentTermHistogramCollector`] for the approach and [`maybe_build_collector`] for the
//! conditions under which it is used.
use columnar::ColumnBlockAccessor;
use super::{Bucket, SegmentTermCollector, TermsAggReqData, VecTermBuckets};
use crate::aggregation::agg_data::{AggKind, AggRefNode, AggregationsSegmentCtx};
use crate::aggregation::bucket::{
get_bucket_pos_f64, prepare_histogram_dense_range, HistogramAggReqData,
SegmentHistogramCollector,
};
use crate::aggregation::buffered_sub_aggs::LowCardSubAggBuffer;
use crate::aggregation::intermediate_agg_result::{
IntermediateAggregationResult, IntermediateAggregationResults,
};
use crate::aggregation::segment_agg_result::{BucketIdProvider, SegmentAggregationCollector};
use crate::aggregation::{f64_from_fastfield_u64, BucketId};
/// Maximum number of cells (`num_terms × num_time_buckets`) in the fused flat 2D grid. Above this
/// the grid would be too large/cache-unfriendly, so we fall back to the general buffered path.
/// `1 << 14` cells = 128 KB of `u64` counters, comfortably L2-resident.
///
/// Since we are only at the top-level, this won't be multiplied by any parent buckets.
const MAX_FUSED_GRID_BUCKETS: usize = 16384;
/// Fused collector for `terms` (low cardinality) × a single `histogram`/`date_histogram` leaf with
/// nothing nested below it, when the resulting `num_terms × num_time_buckets` grid is small (see
/// [`MAX_FUSED_GRID_BUCKETS`]).
///
/// It keeps a flat, fully dense 2D counter grid (`counts[term * num_time_buckets + bucket]`) and a
/// per-term total. A single pass reads both the term and histogram columns in document order and
/// bumps the counters directly — no doc-id buffering, no per-term scattered re-fetch, no dynamic
/// dispatch on flush, no per-bucket key/id storage during collection (keys are derived from the
/// index at the end).
///
/// At result time the flat grid is expanded back into the regular term map + histogram storage and
/// handed to the shared intermediate-result builders, so cross-segment merging is identical to the
/// general path.
#[derive(Debug)]
pub(crate) struct SegmentTermHistogramCollector {
/// Per-term count of docs *outside* `hard_bounds` (still in `doc_count`, but in no bucket).
/// Per-term total = this + the term's `counts` row-sum; left empty when there are no hard
/// bounds (every doc is in-bounds, so there's no remainder to track).
term_counts: Vec<u32>,
/// Flattened `[num_terms * num_time_buckets]` histogram counters (`u32`, see
/// `term_counts`).
///
/// Each term id get its own contiguous slice of `num_time_buckets` histogram counter.
/// When we count all docs (#nofilter), we can derive the per-term total as the sum over that
/// term's slice.
counts: Vec<u32>,
/// Histogram buckets per term (the dense time-range length).
num_time_buckets: usize,
/// `bucket_pos` mapped to time-bucket index 0.
base_pos: i64,
terms_req_data: TermsAggReqData,
/// The (cloned, normalized) histogram request: its column + interval/offset/bounds.
hist_req_data: HistogramAggReqData,
/// Private block accessors for both columns. We read them together, so each needs its own
/// (the shared `agg_data` scratch accessor only holds one block at a time). Owning them keeps
/// `collect` independent of `agg_data`.
term_block: ColumnBlockAccessor<u64>,
hist_block: ColumnBlockAccessor<u64>,
/// No hard bounds, so every doc is in-bounds.
all_docs_in_bounds: bool,
/// Both columns are full (fused-path precondition); cached so `collect` skips the per-block
/// cardinality lookup in `fetch_block`.
is_full: bool,
}
impl SegmentAggregationCollector for SegmentTermHistogramCollector {
fn add_intermediate_aggregation_result(
&mut self,
agg_data: &AggregationsSegmentCtx,
results: &mut IntermediateAggregationResults,
parent_bucket_id: BucketId,
) -> crate::Result<()> {
debug_assert_eq!(
parent_bucket_id, 0,
"fused term-histogram collector is top-level only"
);
// Expand the flat grid back into the regular structures and reuse the shared builders, so
// ordering/cut-off/dict handling and cross-segment merging match the general path exactly.
let mut bucket_id_provider = BucketIdProvider::default();
// Per-term total = histogram row-sum (in-bounds) + `term_counts` (out-of-bounds remainder,
// empty when there are no hard bounds).
let term_buckets = VecTermBuckets {
buckets: self
.counts
.chunks_exact(self.num_time_buckets)
.enumerate()
.map(|(term_id, row)| {
let in_bounds: u32 = row.iter().sum();
let out_of_bounds = self.term_counts.get(term_id).copied().unwrap_or(0);
Bucket {
count: in_bounds + out_of_bounds,
bucket_id: bucket_id_provider.next_bucket_id(),
}
})
.collect(),
};
let mut histogram = SegmentHistogramCollector::<()>::from_dense_rows(
self.hist_req_data.clone(),
self.base_pos,
self.num_time_buckets,
&self.counts,
);
let name = self.terms_req_data.name.clone();
let bucket = SegmentTermCollector::<VecTermBuckets, LowCardSubAggBuffer>::into_intermediate_bucket_result(
&self.terms_req_data,
Some(&mut histogram as &mut dyn SegmentAggregationCollector),
term_buckets,
agg_data,
)?;
results.push(name, IntermediateAggregationResult::Bucket(bucket))?;
Ok(())
}
#[inline]
fn collect(
&mut self,
parent_bucket_id: BucketId,
docs: &[crate::DocId],
_agg_data: &mut AggregationsSegmentCtx,
) -> crate::Result<()> {
debug_assert_eq!(
parent_bucket_id, 0,
"fused term-histogram collector is top-level only"
);
// Fetch both columns into our own accessors (we read them together, so they can't share the
// single `agg_data` scratch accessor). The collector owns all its inputs, so `collect`
// doesn't touch `agg_data`.
self.term_block
.fetch_block_with_is_full(docs, &self.terms_req_data.accessor, self.is_full);
self.hist_block
.fetch_block_with_is_full(docs, &self.hist_req_data.accessor, self.is_full);
// Hoist the loop-invariant fields into locals: the optimizer can't prove the
// `self.counts`/`self.term_counts` writes don't alias these `self` fields, so it can't keep
// them in registers and re-reads them from memory every iteration — ~15% slower on
// `terms_status_with_date_histogram` when read straight from `self`.
// Note: check which are actually relevant.
let field_type = self.hist_req_data.field_type;
let bounds = self.hist_req_data.bounds;
let interval = self.hist_req_data.req.interval;
let offset = self.hist_req_data.offset;
let base_pos = self.base_pos;
let num_time_buckets = self.num_time_buckets;
let all_docs_in_bounds = self.all_docs_in_bounds;
let term_counts = &mut self.term_counts;
let counts = &mut self.counts;
// Both columns are full (checked at construction), so values align with `docs` positionally
// and are read together in one pass.
// In-bounds docs bump the `counts` grid, out-of-bounds bump `term_counts`; deriving the
// total at flush avoids a per-doc `term_counts` RMW that serializes on
// store-to-load forwarding.
for (term_id, hist_raw) in self.term_block.iter_vals().zip(self.hist_block.iter_vals()) {
let term_id = term_id as usize;
let val = f64_from_fastfield_u64(hist_raw, field_type);
if all_docs_in_bounds || bounds.contains(val) {
let bucket = (get_bucket_pos_f64(val, interval, offset) as i64 - base_pos) as usize;
debug_assert!(
bucket < num_time_buckets,
"histogram bucket outside dense range"
);
counts[term_id * num_time_buckets + bucket] += 1;
} else {
term_counts[term_id] += 1;
}
}
Ok(())
}
fn flush(&mut self, _agg_data: &mut AggregationsSegmentCtx) -> crate::Result<()> {
// Nothing is buffered: `collect` writes the flat grid directly.
Ok(())
}
fn prepare_max_bucket(
&mut self,
_max_bucket: BucketId,
_agg_data: &AggregationsSegmentCtx,
) -> crate::Result<()> {
// Top-level: the flat grid is allocated up front.
Ok(())
}
fn compute_metric_value(
&self,
_bucket_id: BucketId,
_sub_agg_name: &str,
_sub_agg_property: &str,
_agg_data: &AggregationsSegmentCtx,
) -> Option<f64> {
None
}
}
/// Builds the fused terms×histogram collector for a single top-level parent, when the shape is
/// eligible. Returns `Ok(None)` to fall back to the general buffered terms path.
///
/// Eligibility: top-level, low-cardinality terms over a full column with no missing/include-exclude
/// handling; a single `histogram`/`date_histogram` leaf (no nesting below it) over a full column;
/// and a `num_terms × num_time_buckets` grid no larger than [`MAX_FUSED_GRID_BUCKETS`].
pub(super) fn maybe_build_collector(
agg_data: &mut AggregationsSegmentCtx,
node: &AggRefNode,
terms_req_data: &TermsAggReqData,
col_max_val: u64,
is_top_level: bool,
) -> crate::Result<Option<Box<dyn SegmentAggregationCollector>>> {
// Both columns must be full (one value per doc) so their values align positionally with `docs`
// and we can zip them. Requiring full columns also makes the terms agg's `missing` config a
// no-op (`fetch_block_with_missing` early-returns on full columns), so we needn't check for it.
//
// We don't cap the term cardinality here: the flat grid is bounded by the total cell count
// (`num_terms * num_time_buckets <= MAX_FUSED_GRID_BUCKETS`) checked below, which subsumes it.
//
// We only allow this at the top-level, since we don't know how many buckets are created. We
// are less likely to get enough docs for the preallocation to be worth and there's a risk of
// using too much memory. We could check the maximum theoretical buckets up-front and pass
// them down.
let fuseable = is_top_level
// TODO: We can easily support this
&& terms_req_data.allowed_term_ids.is_none()
&& terms_req_data.accessor.get_cardinality().is_full()
// The flat counters are `u32`, bumped once per value, so no count can exceed the column's
// value count. (Essentially always true here: the column is full, so its value count
// equals the doc count, and `DocId` is `u32`.)
&& terms_req_data.accessor.values.num_vals() < u32::MAX
&& node.children.len() == 1
&& matches!(
node.children[0].kind,
AggKind::Histogram | AggKind::DateHistogram
)
&& node.children[0].children.is_empty()
&& agg_data.per_request.histogram_req_data[node.children[0].idx_in_req_data]
.accessor
.get_cardinality()
.is_full();
if !fuseable {
return Ok(None);
}
// Clone + normalize the histogram request and get its dense bucket range; only take the fused
// path when the flat `num_terms × num_time_buckets` grid is small enough.
let Some((hist_req_data, range)) = prepare_histogram_dense_range(agg_data, &node.children[0])?
else {
return Ok(None);
};
let num_terms = col_max_val.saturating_add(1) as usize;
if num_terms.saturating_mul(range.len) > MAX_FUSED_GRID_BUCKETS {
return Ok(None);
}
// No hard bounds means every doc is in-bounds, letting `collect` short-circuit the bounds
// check — and leaving `term_counts` (the out-of-bounds remainder) unused, so we skip allocating
// it.
let all_docs_in_bounds =
hist_req_data.bounds.min == f64::MIN && hist_req_data.bounds.max == f64::MAX;
let counts = vec![0u32; num_terms * range.len];
let term_counts = if all_docs_in_bounds {
Vec::new()
} else {
vec![0u32; num_terms]
};
// Charge both grids to the aggregation memory limit.
agg_data.context.limits.add_memory_consumed(
((counts.len() + term_counts.len()) * std::mem::size_of::<u32>()) as u64,
)?;
Ok(Some(Box::new(SegmentTermHistogramCollector {
term_counts,
counts,
num_time_buckets: range.len,
base_pos: range.base_pos,
terms_req_data: terms_req_data.clone(),
hist_req_data,
term_block: ColumnBlockAccessor::default(),
hist_block: ColumnBlockAccessor::default(),
all_docs_in_bounds,
is_full: terms_req_data.accessor.get_cardinality().is_full(),
})))
}
#[cfg(test)]
mod tests {
use crate::aggregation::agg_req::Aggregations;
use crate::aggregation::tests::{
exec_request, exec_request_with_query_and_memory_limit,
get_test_index_from_values_and_terms,
};
use crate::aggregation::AggregationLimitsGuard;
/// Hand-computed correctness check for the fused terms×histogram fast path
/// ([`super::SegmentTermHistogramCollector`]): low-cardinality terms × a histogram leaf over
/// full columns, exercised single- and multi-segment.
#[test]
fn fused_term_histogram_test() -> crate::Result<()> {
fused_term_histogram_with_opt(false)?;
fused_term_histogram_with_opt(true)?;
Ok(())
}
fn fused_term_histogram_with_opt(merge_segments: bool) -> crate::Result<()> {
// 300 docs: term = {a, b, c} by i % 3, histogram value = i % 20 (interval 1 => buckets
// 0..19). gcd(3, 20) = 1, so every (term, bucket) pair occurs exactly 300 / 60 = 5 times.
let docs: Vec<(f64, String)> = (0..300u64)
.map(|i| {
(
(i % 20) as f64,
["a", "b", "c"][(i % 3) as usize].to_string(),
)
})
.collect();
// Two segments, to also exercise cross-segment merging of the fused per-term histograms.
let segments = vec![docs[..150].to_vec(), docs[150..].to_vec()];
let index = get_test_index_from_values_and_terms(merge_segments, &segments)?;
let agg_req: Aggregations = serde_json::from_value(serde_json::json!({
"by_term": {
"terms": { "field": "string_id", "order": { "_key": "asc" } },
"aggs": {
"histo": { "histogram": { "field": "score_f64", "interval": 1.0 } }
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
for (term_idx, term) in ["a", "b", "c"].iter().enumerate() {
assert_eq!(res["by_term"]["buckets"][term_idx]["key"], *term);
assert_eq!(res["by_term"]["buckets"][term_idx]["doc_count"], 100);
let histo = &res["by_term"]["buckets"][term_idx]["histo"]["buckets"];
for b in 0..20usize {
assert_eq!(histo[b]["key"], b as f64, "term {term} bucket {b}");
assert_eq!(histo[b]["doc_count"], 5, "term {term} bucket {b}");
}
assert_eq!(histo[20], serde_json::Value::Null);
}
assert_eq!(res["by_term"]["buckets"][3], serde_json::Value::Null);
Ok(())
}
/// A `missing` config on a *full* term column still takes the fused path (the string sentinel
/// is just `col_max + 1`, so the column stays low-cardinality). Since no doc is missing, the
/// real term buckets must be exactly as without `missing`.
#[test]
fn fused_term_histogram_with_missing_on_full_column() -> crate::Result<()> {
let docs: Vec<(f64, String)> = (0..300u64)
.map(|i| {
(
(i % 20) as f64,
["a", "b", "c"][(i % 3) as usize].to_string(),
)
})
.collect();
let index = get_test_index_from_values_and_terms(true, &[docs])?;
let agg_req: Aggregations = serde_json::from_value(serde_json::json!({
"by_term": {
"terms": { "field": "string_id", "missing": "MISSING", "order": { "_key": "asc" } },
"aggs": {
"histo": { "histogram": { "field": "score_f64", "interval": 1.0 } }
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
// Column is full, so "MISSING" never applies: a, b, c are unchanged (100 docs, 5 per
// bucket).
for (term_idx, term) in ["a", "b", "c"].iter().enumerate() {
assert_eq!(res["by_term"]["buckets"][term_idx]["key"], *term);
assert_eq!(res["by_term"]["buckets"][term_idx]["doc_count"], 100);
let histo = &res["by_term"]["buckets"][term_idx]["histo"]["buckets"];
for b in 0..20usize {
assert_eq!(histo[b]["doc_count"], 5, "term {term} bucket {b}");
}
}
Ok(())
}
/// Term cardinality above the general path's `MAX_NUM_TERMS_FOR_VEC` (100) still fuses: the
/// flat grid is bounded by the total cell count (`num_terms * num_time_buckets`), not the
/// term count.
#[test]
fn fused_term_histogram_many_terms() -> crate::Result<()> {
let num_terms = 150usize;
let docs_per_term = 2usize;
// All docs share histogram value 0 (a single bucket), so the grid is 150 x 1 = 150 cells.
let docs: Vec<(f64, String)> = (0..num_terms * docs_per_term)
.map(|i| (0.0, format!("t{:03}", i % num_terms)))
.collect();
let index = get_test_index_from_values_and_terms(true, &[docs])?;
let agg_req: Aggregations = serde_json::from_value(serde_json::json!({
"by_term": {
"terms": { "field": "string_id", "size": 1000, "order": { "_key": "asc" } },
"aggs": {
"histo": { "histogram": { "field": "score_f64", "interval": 1.0 } }
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
let buckets = res["by_term"]["buckets"].as_array().unwrap();
assert_eq!(buckets.len(), num_terms);
for (i, bucket) in buckets.iter().enumerate() {
assert_eq!(bucket["key"], format!("t{i:03}"));
assert_eq!(bucket["doc_count"], docs_per_term as u64);
assert_eq!(bucket["histo"]["buckets"][0]["key"], 0.0);
assert_eq!(
bucket["histo"]["buckets"][0]["doc_count"],
docs_per_term as u64
);
}
Ok(())
}
/// `hard_bounds` exercises the non-derived `term_counts` branch: a term's `doc_count` must
/// count *every* doc with that term, including docs whose histogram value is outside the
/// bounds (those are excluded from the histogram buckets but still counted for the term). This
/// is the case where the per-doc `term_counts` increment cannot be replaced by the grid
/// row-sum.
#[test]
fn fused_term_histogram_with_hard_bounds() -> crate::Result<()> {
// 300 docs: term = {a, b, c} by i % 3, value = i % 20. Per term: 100 docs, each value in
// 0..=19 occurring 5 times.
let docs: Vec<(f64, String)> = (0..300u64)
.map(|i| {
(
(i % 20) as f64,
["a", "b", "c"][(i % 3) as usize].to_string(),
)
})
.collect();
let index = get_test_index_from_values_and_terms(true, &[docs])?;
// hard_bounds [5, 14] (inclusive) keeps only values 5..=14 in the histogram (10 buckets);
// values 0..=4 and 15..=19 are out of bounds.
let agg_req: Aggregations = serde_json::from_value(serde_json::json!({
"by_term": {
"terms": { "field": "string_id", "order": { "_key": "asc" } },
"aggs": {
"histo": {
"histogram": {
"field": "score_f64",
"interval": 1.0,
"hard_bounds": { "min": 5.0, "max": 14.0 }
}
}
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
for (term_idx, term) in ["a", "b", "c"].iter().enumerate() {
assert_eq!(res["by_term"]["buckets"][term_idx]["key"], *term);
// doc_count includes the 50 per-term docs whose value is outside [5, 14].
assert_eq!(res["by_term"]["buckets"][term_idx]["doc_count"], 100);
let histo = &res["by_term"]["buckets"][term_idx]["histo"]["buckets"];
for b in 0..10usize {
let key = 5 + b;
assert_eq!(histo[b]["key"], key as f64, "term {term} bucket key {key}");
assert_eq!(histo[b]["doc_count"], 5, "term {term} bucket {key}");
}
// Only the 10 in-bounds buckets exist.
assert_eq!(histo[10], serde_json::Value::Null);
}
Ok(())
}
/// Non-binding `hard_bounds` (wider than the data, with mid-interval edges) must still produce
/// exact results via the derive-from-grid path: since no doc is out of bounds, normalization
/// drops the bound, every doc lands in the dense range, and each term's total equals its
/// histogram row-sum. This is the case that previously fell back to the per-doc counter only
/// because `bounds != [MIN, MAX]`.
#[test]
fn fused_term_histogram_with_non_binding_hard_bounds() -> crate::Result<()> {
// 300 docs: term = {a, b, c} by i % 3, value = i % 20. Data values span [0, 19].
let docs: Vec<(f64, String)> = (0..300u64)
.map(|i| {
(
(i % 20) as f64,
["a", "b", "c"][(i % 3) as usize].to_string(),
)
})
.collect();
let index = get_test_index_from_values_and_terms(true, &[docs])?;
// Bounds wider than [0, 19], with mid-interval edges -> they exclude nothing.
let agg_req: Aggregations = serde_json::from_value(serde_json::json!({
"by_term": {
"terms": { "field": "string_id", "order": { "_key": "asc" } },
"aggs": {
"histo": {
"histogram": {
"field": "score_f64",
"interval": 1.0,
"hard_bounds": { "min": -0.5, "max": 19.5 }
}
}
}
}
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
for (term_idx, term) in ["a", "b", "c"].iter().enumerate() {
assert_eq!(res["by_term"]["buckets"][term_idx]["key"], *term);
// Every doc is in-bounds, so the per-term total is the full 100 (as without bounds).
assert_eq!(res["by_term"]["buckets"][term_idx]["doc_count"], 100);
let histo = &res["by_term"]["buckets"][term_idx]["histo"]["buckets"];
for b in 0..20usize {
assert_eq!(histo[b]["key"], b as f64, "term {term} bucket {b}");
assert_eq!(histo[b]["doc_count"], 5, "term {term} bucket {b}");
}
assert_eq!(histo[20], serde_json::Value::Null);
}
Ok(())
}
/// Regression: with hard bounds the fused path allocates `term_counts` (one `u32`/term) on top
/// of the grid, and that allocation must be charged to the memory limit. With many terms and a
/// single time bucket the two are equal in size, so a limit admitting the grid alone but not
/// grid + `term_counts` must fail.
#[test]
fn fused_term_histogram_hard_bounds_charges_term_counts() -> crate::Result<()> {
// 16k distinct terms, one doc each; values alternate in/out of the single-bucket bounds
// [5, 5] so the bounds bind and `term_counts` is allocated. num_terms=16000,
// num_time_buckets=1 => `counts` and `term_counts` are ~64 KB each.
let docs: Vec<(f64, String)> = (0..16_000u64)
.map(|i| (if i % 2 == 0 { 5.0 } else { 10.0 }, format!("t{i:05}")))
.collect();
let index = get_test_index_from_values_and_terms(true, &[docs])?;
let agg_req: Aggregations = serde_json::from_value(serde_json::json!({
"by_term": {
"terms": { "field": "string_id" },
"aggs": {
"histo": {
"histogram": {
"field": "score_f64",
"interval": 1.0,
"hard_bounds": { "min": 5.0, "max": 5.0 }
}
}
}
}
}))
.unwrap();
// ~96 KB admits the grid (~64 KB) but not grid + `term_counts` (~128 KB).
let err = exec_request_with_query_and_memory_limit(
agg_req,
&index,
None,
AggregationLimitsGuard::new(Some(96_000), None),
)
.unwrap_err();
assert!(
err.to_string().contains("memory limit was exceeded"),
"expected a memory-limit error, got: {err}"
);
Ok(())
}
}

17
src/aggregation/bucket/term_missing_agg.rs
View File

@@ -5,7 +5,7 @@ use crate::aggregation::agg_data::{
build_segment_agg_collectors, AggRefNode, AggregationsSegmentCtx,
};
use crate::aggregation::bucket::term_agg::TermsAggregation;
use crate::aggregation::cached_sub_aggs::{CachedSubAggs, HighCardCachedSubAggs};
use crate::aggregation::buffered_sub_aggs::{BufferedSubAggs, HighCardBufferedSubAggs};
use crate::aggregation::intermediate_agg_result::{
IntermediateAggregationResult, IntermediateAggregationResults, IntermediateBucketResult,
IntermediateKey, IntermediateTermBucketEntry, IntermediateTermBucketResult,
@@ -47,7 +47,7 @@ struct MissingCount {
#[derive(Default, Debug)]
pub struct TermMissingAgg {
accessor_idx: usize,
sub_agg: Option<HighCardCachedSubAggs>,
sub_agg: Option<HighCardBufferedSubAggs>,
/// Idx = parent bucket id, Value = missing count for that bucket
missing_count_per_bucket: Vec<MissingCount>,
bucket_id_provider: BucketIdProvider,
@@ -66,7 +66,7 @@ impl TermMissingAgg {
None
};
let sub_agg = sub_agg.map(CachedSubAggs::new);
let sub_agg = sub_agg.map(BufferedSubAggs::new);
let bucket_id_provider = BucketIdProvider::default();
Ok(Self {
@@ -177,6 +177,17 @@ impl SegmentAggregationCollector for TermMissingAgg {
}
Ok(())
}
fn compute_metric_value(
&self,
_bucket_id: BucketId,
_sub_agg_name: &str,
_sub_agg_property: &str,
_agg_data: &AggregationsSegmentCtx,
) -> Option<f64> {
// TODO: forward to `sub_agg` for nested order paths (`missing_agg>metric`).
None
}
}
#[cfg(test)]

46
src/aggregation/cached_sub_aggs.rs → src/aggregation/buffered_sub_aggs.rs
View File

@@ -6,7 +6,7 @@ use crate::aggregation::bucket::MAX_NUM_TERMS_FOR_VEC;
use crate::aggregation::BucketId;
use crate::DocId;
/// A cache for sub-aggregations, storing doc ids per bucket id.
/// A buffer for sub-aggregations, storing doc ids per bucket id.
/// Depending on the cardinality of the parent aggregation, we use different
/// storage strategies.
///
@@ -24,21 +24,21 @@ use crate::DocId;
/// aggregations.
/// What this datastructure does in general is to group docs by bucket id.
#[derive(Debug)]
pub(crate) struct CachedSubAggs<C: SubAggCache> {
cache: C,
pub(crate) struct BufferedSubAggs<B: SubAggBuffer> {
buffer: B,
sub_agg_collector: Box<dyn SegmentAggregationCollector>,
num_docs: usize,
}
pub type LowCardCachedSubAggs = CachedSubAggs<LowCardSubAggCache>;
pub type HighCardCachedSubAggs = CachedSubAggs<HighCardSubAggCache>;
pub type LowCardBufferedSubAggs = BufferedSubAggs<LowCardSubAggBuffer>;
pub type HighCardBufferedSubAggs = BufferedSubAggs<HighCardSubAggBuffer>;
const FLUSH_THRESHOLD: usize = 2048;
/// A trait for caching sub-aggregation doc ids per bucket id.
/// A trait for buffering sub-aggregation doc ids per bucket id.
/// Different implementations can be used depending on the cardinality
/// of the parent aggregation.
pub trait SubAggCache: Debug {
pub trait SubAggBuffer: Debug {
fn new() -> Self;
fn push(&mut self, bucket_id: BucketId, doc_id: DocId);
fn flush_local(
@@ -49,22 +49,22 @@ pub trait SubAggCache: Debug {
) -> crate::Result<()>;
}
impl<Backend: SubAggCache + Debug> CachedSubAggs<Backend> {
impl<Backend: SubAggBuffer + Debug> BufferedSubAggs<Backend> {
pub fn new(sub_agg: Box<dyn SegmentAggregationCollector>) -> Self {
Self {
cache: Backend::new(),
buffer: Backend::new(),
sub_agg_collector: sub_agg,
num_docs: 0,
}
}
pub fn get_sub_agg_collector(&mut self) -> &mut Box<dyn SegmentAggregationCollector> {
&mut self.sub_agg_collector
pub fn get_sub_agg_collector(&mut self) -> &mut dyn SegmentAggregationCollector {
&mut *self.sub_agg_collector
}
#[inline]
pub fn push(&mut self, bucket_id: BucketId, doc_id: DocId) {
self.cache.push(bucket_id, doc_id);
self.buffer.push(bucket_id, doc_id);
self.num_docs += 1;
}
@@ -75,7 +75,7 @@ impl<Backend: SubAggCache + Debug> CachedSubAggs<Backend> {
agg_data: &mut AggregationsSegmentCtx,
) -> crate::Result<()> {
if self.num_docs >= FLUSH_THRESHOLD {
self.cache
self.buffer
.flush_local(&mut self.sub_agg_collector, agg_data, false)?;
self.num_docs = 0;
}
@@ -85,7 +85,7 @@ impl<Backend: SubAggCache + Debug> CachedSubAggs<Backend> {
/// Note: this _does_ flush the sub aggregations.
pub fn flush(&mut self, agg_data: &mut AggregationsSegmentCtx) -> crate::Result<()> {
if self.num_docs != 0 {
self.cache
self.buffer
.flush_local(&mut self.sub_agg_collector, agg_data, true)?;
self.num_docs = 0;
}
@@ -94,11 +94,11 @@ impl<Backend: SubAggCache + Debug> CachedSubAggs<Backend> {
}
}
/// Number of partitions for high cardinality sub-aggregation cache.
/// Number of partitions for high cardinality sub-aggregation buffer.
const NUM_PARTITIONS: usize = 16;
#[derive(Debug)]
pub(crate) struct HighCardSubAggCache {
pub(crate) struct HighCardSubAggBuffer {
/// This weird partitioning is used to do some cheap grouping on the bucket ids.
/// bucket ids are dense, e.g. when we don't detect the cardinality as low cardinality,
/// but there are just 16 bucket ids, each bucket id will go to its own partition.
@@ -108,7 +108,7 @@ pub(crate) struct HighCardSubAggCache {
partitions: Box<[PartitionEntry; NUM_PARTITIONS]>,
}
impl HighCardSubAggCache {
impl HighCardSubAggBuffer {
#[inline]
fn clear(&mut self) {
for partition in self.partitions.iter_mut() {
@@ -131,13 +131,14 @@ impl PartitionEntry {
}
}
impl SubAggCache for HighCardSubAggCache {
impl SubAggBuffer for HighCardSubAggBuffer {
fn new() -> Self {
Self {
partitions: Box::new(core::array::from_fn(|_| PartitionEntry::default())),
}
}
#[inline]
fn push(&mut self, bucket_id: BucketId, doc_id: DocId) {
let idx = bucket_id % NUM_PARTITIONS as u32;
let slot = &mut self.partitions[idx as usize];
@@ -173,14 +174,14 @@ impl SubAggCache for HighCardSubAggCache {
}
#[derive(Debug)]
pub(crate) struct LowCardSubAggCache {
/// Cache doc ids per bucket for sub-aggregations.
pub(crate) struct LowCardSubAggBuffer {
/// Buffer doc ids per bucket for sub-aggregations.
///
/// The outer Vec is indexed by BucketId.
per_bucket_docs: Vec<Vec<DocId>>,
}
impl LowCardSubAggCache {
impl LowCardSubAggBuffer {
#[inline]
fn clear(&mut self) {
for v in &mut self.per_bucket_docs {
@@ -189,13 +190,14 @@ impl LowCardSubAggCache {
}
}
impl SubAggCache for LowCardSubAggCache {
impl SubAggBuffer for LowCardSubAggBuffer {
fn new() -> Self {
Self {
per_bucket_docs: Vec::new(),
}
}
#[inline]
fn push(&mut self, bucket_id: BucketId, doc_id: DocId) {
let idx = bucket_id as usize;
if self.per_bucket_docs.len() <= idx {

12
src/aggregation/collector.rs
View File

@@ -1,6 +1,6 @@
use super::agg_req::Aggregations;
use super::agg_result::AggregationResults;
use super::cached_sub_aggs::LowCardCachedSubAggs;
use super::buffered_sub_aggs::LowCardBufferedSubAggs;
use super::intermediate_agg_result::IntermediateAggregationResults;
use super::AggContextParams;
// group buffering strategy is chosen explicitly by callers; no need to hash-group on the fly.
@@ -66,7 +66,7 @@ impl Collector for DistributedAggregationCollector {
fn for_segment(
&self,
segment_local_id: crate::SegmentOrdinal,
reader: &dyn SegmentReader,
reader: &crate::SegmentReader,
) -> crate::Result<Self::Child> {
AggregationSegmentCollector::from_agg_req_and_reader(
&self.agg,
@@ -96,7 +96,7 @@ impl Collector for AggregationCollector {
fn for_segment(
&self,
segment_local_id: crate::SegmentOrdinal,
reader: &dyn SegmentReader,
reader: &crate::SegmentReader,
) -> crate::Result<Self::Child> {
AggregationSegmentCollector::from_agg_req_and_reader(
&self.agg,
@@ -136,7 +136,7 @@ fn merge_fruits(
/// `AggregationSegmentCollector` does the aggregation collection on a segment.
pub struct AggregationSegmentCollector {
aggs_with_accessor: AggregationsSegmentCtx,
agg_collector: LowCardCachedSubAggs,
agg_collector: LowCardBufferedSubAggs,
error: Option<TantivyError>,
}
@@ -145,14 +145,14 @@ impl AggregationSegmentCollector {
/// reader. Also includes validation, e.g. checking field types and existence.
pub fn from_agg_req_and_reader(
agg: &Aggregations,
reader: &dyn SegmentReader,
reader: &SegmentReader,
segment_ordinal: SegmentOrdinal,
context: &AggContextParams,
) -> crate::Result<Self> {
let mut agg_data =
build_aggregations_data_from_req(agg, reader, segment_ordinal, context.clone())?;
let mut result =
LowCardCachedSubAggs::new(build_segment_agg_collectors_root(&mut agg_data)?);
LowCardBufferedSubAggs::new(build_segment_agg_collectors_root(&mut agg_data)?);
result
.get_sub_agg_collector()
.prepare_max_bucket(0, &agg_data)?; // prepare for bucket zero

49
src/aggregation/intermediate_agg_result.rs
View File

@@ -377,7 +377,22 @@ impl IntermediateMetricResult {
MetricResult::ExtendedStats(intermediate_stats.finalize())
}
IntermediateMetricResult::Sum(intermediate_sum) => {
MetricResult::Sum(intermediate_sum.finalize().into())
// By default match Elasticsearch: empty / all-missing sum
// buckets serialize as `"value": 0`, not `"value": null`.
// The non-ES `none_if_no_match` flag on `SumAggregation`
// opts into SQL-style `null` for downstream consumers.
let none_if_no_match = req
.agg
.as_sum()
.and_then(|sum| sum.none_if_no_match)
.unwrap_or(false);
let value = intermediate_sum.finalize();
if none_if_no_match {
MetricResult::Sum(value.into())
} else {
let value = Some(value.unwrap_or(0.0));
MetricResult::Sum(value.into())
}
}
IntermediateMetricResult::Percentiles(percentiles) => MetricResult::Percentiles(
percentiles
@@ -1004,24 +1019,20 @@ impl IntermediateCompositeBucketResult {
) -> crate::Result<BucketResult> {
let trimmed_entry_vec =
trim_composite_buckets(self.entries, &self.orders, self.target_size)?;
let after_key = if trimmed_entry_vec.len() == req.size as usize {
trimmed_entry_vec
.last()
.map(|bucket| {
let (intermediate_key, _entry) = bucket;
intermediate_key
.iter()
.enumerate()
.map(|(idx, intermediate_key)| {
let source = &req.sources[idx];
(source.name().to_string(), intermediate_key.clone().into())
})
.collect()
})
.unwrap()
} else {
FxHashMap::default()
};
let after_key = trimmed_entry_vec
.last()
.map(|bucket| {
let (intermediate_key, _entry) = bucket;
intermediate_key
.iter()
.enumerate()
.map(|(idx, intermediate_key)| {
let source = &req.sources[idx];
(source.name().to_string(), intermediate_key.clone().into())
})
.collect()
})
.unwrap_or_default();
let buckets = trimmed_entry_vec
.into_iter()

970
src/aggregation/metric/cardinality.rs
View File

File diff suppressed because it is too large Load Diff

20
src/aggregation/metric/extended_stats.rs
View File

@@ -399,6 +399,26 @@ impl SegmentAggregationCollector for SegmentExtendedStatsCollector {
}
Ok(())
}
fn compute_metric_value(
&self,
bucket_id: BucketId,
sub_agg_name: &str,
sub_agg_property: &str,
_agg_data: &AggregationsSegmentCtx,
) -> Option<f64> {
if self.name != sub_agg_name {
return None;
}
let extended = self.buckets.get(bucket_id as usize)?;
// Finalize is a pure read of accumulators — calling it here for the cutoff sort
// doesn't disturb the eventual intermediate result.
extended
.finalize()
.get_value(sub_agg_property)
.ok()
.flatten()
}
}
#[cfg(test)]

5
src/aggregation/metric/mod.rs
View File

@@ -107,10 +107,9 @@ pub enum PercentileValues {
#[derive(Clone, Debug, PartialEq, Serialize, Deserialize)]
/// The entry when requesting percentiles with keyed: false
pub struct PercentileValuesVecEntry {
/// Percentile
/// The percentile key (e.g. 1.0, 5.0, 25.0).
pub key: f64,
/// Value at the percentile
/// The percentile value. `NaN` when there are no values.
pub value: f64,
}

20
src/aggregation/metric/percentiles.rs
View File

@@ -312,6 +312,26 @@ impl SegmentAggregationCollector for SegmentPercentilesCollector {
}
Ok(())
}
fn compute_metric_value(
&self,
bucket_id: BucketId,
sub_agg_name: &str,
sub_agg_property: &str,
agg_data: &AggregationsSegmentCtx,
) -> Option<f64> {
if agg_data.get_metric_req_data(self.accessor_idx).name != sub_agg_name {
return None;
}
let percentile: f64 = sub_agg_property.parse().ok()?;
if !(0.0..=100.0).contains(&percentile) {
return None;
}
let bucket = self.buckets.get(bucket_id as usize)?;
// DDSketch.quantile is a pure read; calling it here for the cutoff sort does
// not affect the intermediate state used for the final result.
bucket.sketch.quantile(percentile / 100.0).ok().flatten()
}
}
#[cfg(test)]

34
src/aggregation/metric/stats.rs
View File

@@ -321,6 +321,40 @@ impl<const COLUMN_TYPE_ID: u8> SegmentAggregationCollector
}
Ok(())
}
fn compute_metric_value(
&self,
bucket_id: BucketId,
sub_agg_name: &str,
sub_agg_property: &str,
_agg_data: &AggregationsSegmentCtx,
) -> Option<f64> {
if self.name != sub_agg_name {
return None;
}
let stats = self.buckets.get(bucket_id as usize)?;
// The property depends on what we're collecting:
// - StatsType::Stats exposes count/sum/min/max/avg via dotted property.
// - Single-value kinds (Sum/Count/Min/Max/Average) expect an empty property and return
// the value they were configured to collect.
let prop = match self.collecting_for {
StatsType::Stats if !sub_agg_property.is_empty() => sub_agg_property,
StatsType::Sum if sub_agg_property.is_empty() => "sum",
StatsType::Count if sub_agg_property.is_empty() => "count",
StatsType::Max if sub_agg_property.is_empty() => "max",
StatsType::Min if sub_agg_property.is_empty() => "min",
StatsType::Average if sub_agg_property.is_empty() => "avg",
_ => return None,
};
match prop {
"count" => Some(stats.count as f64),
"sum" => Some(stats.sum),
"min" if stats.count > 0 => Some(stats.min),
"max" if stats.count > 0 => Some(stats.max),
"avg" if stats.count > 0 => Some(stats.sum / stats.count as f64),
_ => None,
}
}
}
#[inline]

111
src/aggregation/metric/sum.rs
View File

@@ -27,6 +27,16 @@ pub struct SumAggregation {
/// { "field": "my_numbers", "missing": "10.0" }
#[serde(default, deserialize_with = "deserialize_option_f64")]
pub missing: Option<f64>,
/// Non-Elasticsearch extension. When `Some(true)`, the serialized result
/// returns `"value": null` if no values were collected (all documents had
/// missing/NULL values for the field), matching the behavior of `min`,
/// `max`, and `avg`. When `None` or `Some(false)` (the default) the
/// result returns `"value": 0`, matching Elasticsearch.
///
/// Intended for SQL-style consumers where `SUM` of zero rows is `NULL`
/// and must be distinguishable from a bucket that genuinely sums to `0`.
#[serde(default, skip_serializing_if = "Option::is_none")]
pub none_if_no_match: Option<bool>,
}
impl SumAggregation {
@@ -35,6 +45,7 @@ impl SumAggregation {
Self {
field: field_name,
missing: None,
none_if_no_match: None,
}
}
/// Returns the field name the aggregation is computed on.
@@ -59,8 +70,104 @@ impl IntermediateSum {
pub fn merge_fruits(&mut self, other: IntermediateSum) {
self.stats.merge_fruits(other.stats);
}
/// Computes the final minimum value.
/// Computes the final sum value.
///
/// Returns `None` when no values were collected, matching the Rust-side
/// behavior of `IntermediateMin`, `IntermediateMax`, and
/// `IntermediateAvg`. The Elasticsearch-vs-SQL choice for the
/// user-visible result is made at the boundary in
/// [`IntermediateMetricResult::into_final_metric_result`]: by default
/// `None` is coerced to `Some(0.0)` to match Elasticsearch
/// (`"value": 0`), and the [`SumAggregation::none_if_no_match`] flag
/// opts out of that coercion for SQL-style consumers.
pub fn finalize(&self) -> Option<f64> {
Some(self.stats.finalize().sum)
let stats = self.stats.finalize();
if stats.count == 0 {
None
} else {
Some(stats.sum)
}
}
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn test_sum_finalize_returns_none_when_no_values() {
// Default IntermediateSum has count=0 — finalize should return None,
// matching MIN/MAX/AVG behavior for all-NULL groups.
let sum = IntermediateSum::default();
assert_eq!(sum.finalize(), None);
}
#[test]
fn test_sum_finalize_returns_value_when_has_values() {
let mut sum = IntermediateSum::default();
// Merge in a result that has actual values
let stats = IntermediateStats {
count: 3,
sum: 42.0,
min: 10.0,
max: 20.0,
..Default::default()
};
let other = IntermediateSum::from_stats(stats);
sum.merge_fruits(other);
assert_eq!(sum.finalize(), Some(42.0));
}
#[test]
fn test_sum_merge_two_empty_still_none() {
let mut a = IntermediateSum::default();
let b = IntermediateSum::default();
a.merge_fruits(b);
assert_eq!(a.finalize(), None);
}
#[test]
fn test_sum_aggregation_empty_index_default_matches_es() -> crate::Result<()> {
use serde_json::json;
use crate::aggregation::agg_req::Aggregations;
use crate::aggregation::tests::{exec_request, get_test_index_from_terms};
// Empty index — sum has no values to collect.
let values: Vec<Vec<&str>> = vec![];
let index = get_test_index_from_terms(false, &values)?;
let agg_req: Aggregations = serde_json::from_value(json!({
"score_sum": { "sum": { "field": "score" } }
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
// Default: match Elasticsearch — empty sum serializes as 0, not null.
assert_eq!(res["score_sum"]["value"], 0.0);
Ok(())
}
#[test]
fn test_sum_aggregation_empty_index_none_if_no_match_opt_in() -> crate::Result<()> {
use serde_json::json;
use crate::aggregation::agg_req::Aggregations;
use crate::aggregation::tests::{exec_request, get_test_index_from_terms};
let values: Vec<Vec<&str>> = vec![];
let index = get_test_index_from_terms(false, &values)?;
let agg_req: Aggregations = serde_json::from_value(json!({
"score_sum": { "sum": { "field": "score", "none_if_no_match": true } }
}))
.unwrap();
let res = exec_request(agg_req, &index)?;
// Opt-in non-ES extension — empty sum serializes as null.
assert!(
res["score_sum"]["value"].is_null(),
"expected null, got {:?}",
res["score_sum"]["value"]
);
Ok(())
}
}

11
src/aggregation/metric/top_hits.rs
View File

@@ -644,6 +644,17 @@ impl SegmentAggregationCollector for TopHitsSegmentCollector {
);
Ok(())
}
fn compute_metric_value(
&self,
_bucket_id: BucketId,
_sub_agg_name: &str,
_sub_agg_property: &str,
_agg_data: &AggregationsSegmentCtx,
) -> Option<f64> {
// top_hits is not a numeric metric and cannot be used as an order target.
None
}
}
#[cfg(test)]

2
src/aggregation/mod.rs
View File

@@ -133,7 +133,7 @@ mod agg_limits;
pub mod agg_req;
pub mod agg_result;
pub mod bucket;
pub(crate) mod cached_sub_aggs;
pub(crate) mod buffered_sub_aggs;
mod collector;
mod date;
mod error;

42
src/aggregation/segment_agg_result.rs
View File

@@ -76,6 +76,31 @@ pub trait SegmentAggregationCollector: Debug {
fn flush(&mut self, _agg_data: &mut AggregationsSegmentCtx) -> crate::Result<()> {
Ok(())
}
/// Compute the segment-level metric value of the named direct-child metric for `bucket_id`.
///
/// Used by parent term aggs that order by a sub-aggregation: the parent sorts on
/// this value and cuts off at segment time, matching the approximation tradeoff
/// Elasticsearch makes for any sub-agg ordering.
///
/// `sub_agg_property` is the dotted suffix (e.g. `"sum"` in `mystats.sum`); empty when
/// the metric is a single-value kind such as cardinality.
///
/// Returns `None` only on name mismatch, unknown property, or empty bucket. Implementations
/// may finalize their per-bucket state (e.g. compute a percentile from a sketch); calls
/// must be idempotent so the final intermediate result is unaffected.
///
/// No default impl on purpose: every collector must decide explicitly whether it
/// produces a metric value, forwards into children (single-bucket aggs), or rejects
/// the lookup. A silent `None` default would let a parent term agg's cutoff sort all
/// buckets to the same key and drop arbitrary winners.
fn compute_metric_value(
&self,
bucket_id: BucketId,
sub_agg_name: &str,
sub_agg_property: &str,
agg_data: &AggregationsSegmentCtx,
) -> Option<f64>;
}
#[derive(Default)]
@@ -137,4 +162,21 @@ impl SegmentAggregationCollector for GenericSegmentAggregationResultsCollector {
}
Ok(())
}
fn compute_metric_value(
&self,
bucket_id: BucketId,
sub_agg_name: &str,
sub_agg_property: &str,
agg_data: &AggregationsSegmentCtx,
) -> Option<f64> {
for agg in &self.aggs {
if let Some(value) =
agg.compute_metric_value(bucket_id, sub_agg_name, sub_agg_property, agg_data)
{
return Some(value);
}
}
None
}
}

12
src/collector/count_collector.rs
View File

@@ -1,5 +1,6 @@
use super::Collector;
use crate::collector::SegmentCollector;
use crate::query::Weight;
use crate::{DocId, Score, SegmentOrdinal, SegmentReader};
/// `CountCollector` collector only counts how many
@@ -43,7 +44,7 @@ impl Collector for Count {
fn for_segment(
&self,
_: SegmentOrdinal,
_: &dyn SegmentReader,
_: &SegmentReader,
) -> crate::Result<SegmentCountCollector> {
Ok(SegmentCountCollector::default())
}
@@ -55,6 +56,15 @@ impl Collector for Count {
fn merge_fruits(&self, segment_counts: Vec<usize>) -> crate::Result<usize> {
Ok(segment_counts.into_iter().sum())
}
fn collect_segment(
&self,
weight: &dyn Weight,
_segment_ord: u32,
reader: &SegmentReader,
) -> crate::Result<usize> {
Ok(weight.count(reader)? as usize)
}
}
#[derive(Default)]

4
src/collector/docset_collector.rs
View File

@@ -1,7 +1,7 @@
use std::collections::HashSet;
use super::{Collector, SegmentCollector};
use crate::{DocAddress, DocId, Score, SegmentReader};
use crate::{DocAddress, DocId, Score};
/// Collectors that returns the set of DocAddress that matches the query.
///
@@ -15,7 +15,7 @@ impl Collector for DocSetCollector {
fn for_segment(
&self,
segment_local_id: crate::SegmentOrdinal,
_segment: &dyn SegmentReader,
_segment: &crate::SegmentReader,
) -> crate::Result<Self::Child> {
Ok(DocSetChildCollector {
segment_local_id,

66
src/collector/facet_collector.rs
View File

@@ -265,7 +265,7 @@ impl Collector for FacetCollector {
fn for_segment(
&self,
_: SegmentOrdinal,
reader: &dyn SegmentReader,
reader: &SegmentReader,
) -> crate::Result<FacetSegmentCollector> {
let facet_reader = reader.facet_reader(&self.field_name)?;
let facet_dict = facet_reader.facet_dict();
@@ -389,6 +389,13 @@ impl SegmentCollector for FacetSegmentCollector {
}
let mut facet = vec![];
let (facet_ord, facet_depth) = self.unique_facet_ords[collapsed_facet_ord];
// u64::MAX is used as a sentinel for unmapped ordinals (e.g. when a
// document has the exact registered facet, not a child of it).
// Passing it to ord_to_term would resolve to the last dictionary
// entry and produce a spurious facet from an unrelated branch.
if facet_ord == u64::MAX {
continue;
}
// TODO handle errors.
if facet_dict.ord_to_term(facet_ord, &mut facet).is_ok() {
if let Some((end_collapsed_facet, _)) = facet
@@ -814,6 +821,63 @@ mod tests {
assert!(!super::is_child_facet(&b"foo\0bar"[..], &b"foo"[..]));
assert!(!super::is_child_facet(&b"foo"[..], &b"foobar\0baz"[..]));
}
// Regression test for https://github.com/quickwit-oss/tantivy/issues/2494
// When a document has the exact registered facet path (not just a child),
// harvest() must not turn the unmapped sentinel into a spurious root entry.
#[test]
fn test_facet_collector_wrong_root() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
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: IndexWriter = index.writer_for_tests()?;
let facets: Vec<&str> = vec![
"/science-fiction/asimov",
"/science-fiction/clarke",
"/science-fiction/dick",
"/science-fiction/herbert",
"/science-fiction/orwell",
// This exact match on the registered facet is the bug trigger:
// its ordinal maps to the sentinel (u64::MAX, 0) in the collapse
// mapping, which without the fix resolves to an unrelated term.
"/fantasy/epic-fantasy",
"/fantasy/epic-fantasy/tolkien",
"/fantasy/epic-fantasy/martin",
];
for facet_str in &facets {
index_writer.add_document(doc!(
facet_field => Facet::from(*facet_str)
))?;
}
index_writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let term = Term::from_facet(facet_field, &Facet::from("/fantasy/epic-fantasy"));
let query = TermQuery::new(term, IndexRecordOption::Basic);
let mut facet_collector = FacetCollector::for_field("facet");
facet_collector.add_facet("/fantasy/epic-fantasy");
let counts: FacetCounts = searcher.search(&query, &facet_collector)?;
let result: Vec<(String, u64)> = counts
.get("/")
.map(|(facet, count)| (facet.to_string(), count))
.collect();
// Only children of /fantasy/epic-fantasy should appear, not /science-fiction
assert_eq!(
result,
vec![
("/fantasy/epic-fantasy/martin".to_string(), 1),
("/fantasy/epic-fantasy/tolkien".to_string(), 1),
]
);
Ok(())
}
}
#[cfg(all(test, feature = "unstable"))]

4
src/collector/filter_collector_wrapper.rs
View File

@@ -113,7 +113,7 @@ where
fn for_segment(
&self,
segment_local_id: u32,
segment_reader: &dyn SegmentReader,
segment_reader: &SegmentReader,
) -> crate::Result<Self::Child> {
let column_opt = segment_reader.fast_fields().column_opt(&self.field)?;
@@ -287,7 +287,7 @@ where
fn for_segment(
&self,
segment_local_id: u32,
segment_reader: &dyn SegmentReader,
segment_reader: &SegmentReader,
) -> crate::Result<Self::Child> {
let column_opt = segment_reader.fast_fields().bytes(&self.field)?;

4
src/collector/histogram_collector.rs
View File

@@ -6,7 +6,7 @@ use fastdivide::DividerU64;
use crate::collector::{Collector, SegmentCollector};
use crate::fastfield::{FastFieldNotAvailableError, FastValue};
use crate::schema::Type;
use crate::{DocId, Score, SegmentReader};
use crate::{DocId, Score};
/// Histogram builds an histogram of the values of a fastfield for the
/// collected DocSet.
@@ -110,7 +110,7 @@ impl Collector for HistogramCollector {
fn for_segment(
&self,
_segment_local_id: crate::SegmentOrdinal,
segment: &dyn SegmentReader,
segment: &crate::SegmentReader,
) -> crate::Result<Self::Child> {
let column_opt = segment.fast_fields().u64_lenient(&self.field)?;
let (column, _column_type) = column_opt.ok_or_else(|| FastFieldNotAvailableError {

14
src/collector/mod.rs
View File

@@ -156,7 +156,7 @@ pub trait Collector: Sync + Send {
fn for_segment(
&self,
segment_local_id: SegmentOrdinal,
segment: &dyn SegmentReader,
segment: &SegmentReader,
) -> crate::Result<Self::Child>;
/// Returns true iff the collector requires to compute scores for documents.
@@ -174,7 +174,7 @@ pub trait Collector: Sync + Send {
&self,
weight: &dyn Weight,
segment_ord: u32,
reader: &dyn SegmentReader,
reader: &SegmentReader,
) -> crate::Result<<Self::Child as SegmentCollector>::Fruit> {
let with_scoring = self.requires_scoring();
let mut segment_collector = self.for_segment(segment_ord, reader)?;
@@ -186,7 +186,7 @@ pub trait Collector: Sync + Send {
pub(crate) fn default_collect_segment_impl<TSegmentCollector: SegmentCollector>(
segment_collector: &mut TSegmentCollector,
weight: &dyn Weight,
reader: &dyn SegmentReader,
reader: &SegmentReader,
with_scoring: bool,
) -> crate::Result<()> {
match (reader.alive_bitset(), with_scoring) {
@@ -255,7 +255,7 @@ impl<TCollector: Collector> Collector for Option<TCollector> {
fn for_segment(
&self,
segment_local_id: SegmentOrdinal,
segment: &dyn SegmentReader,
segment: &SegmentReader,
) -> crate::Result<Self::Child> {
Ok(if let Some(inner) = self {
let inner_segment_collector = inner.for_segment(segment_local_id, segment)?;
@@ -336,7 +336,7 @@ where
fn for_segment(
&self,
segment_local_id: u32,
segment: &dyn SegmentReader,
segment: &SegmentReader,
) -> crate::Result<Self::Child> {
let left = self.0.for_segment(segment_local_id, segment)?;
let right = self.1.for_segment(segment_local_id, segment)?;
@@ -407,7 +407,7 @@ where
fn for_segment(
&self,
segment_local_id: u32,
segment: &dyn SegmentReader,
segment: &SegmentReader,
) -> crate::Result<Self::Child> {
let one = self.0.for_segment(segment_local_id, segment)?;
let two = self.1.for_segment(segment_local_id, segment)?;
@@ -487,7 +487,7 @@ where
fn for_segment(
&self,
segment_local_id: u32,
segment: &dyn SegmentReader,
segment: &SegmentReader,
) -> crate::Result<Self::Child> {
let one = self.0.for_segment(segment_local_id, segment)?;
let two = self.1.for_segment(segment_local_id, segment)?;

4
src/collector/multi_collector.rs
View File

@@ -24,7 +24,7 @@ impl<TCollector: Collector> Collector for CollectorWrapper<TCollector> {
fn for_segment(
&self,
segment_local_id: u32,
reader: &dyn SegmentReader,
reader: &SegmentReader,
) -> crate::Result<Box<dyn BoxableSegmentCollector>> {
let child = self.0.for_segment(segment_local_id, reader)?;
Ok(Box::new(SegmentCollectorWrapper(child)))
@@ -209,7 +209,7 @@ impl Collector for MultiCollector<'_> {
fn for_segment(
&self,
segment_local_id: SegmentOrdinal,
segment: &dyn SegmentReader,
segment: &SegmentReader,
) -> crate::Result<MultiCollectorChild> {
let children = self
.collector_wrappers

6
src/collector/sort_key/order.rs
View File

@@ -5,7 +5,7 @@ use serde::{Deserialize, Serialize};
use crate::collector::{SegmentSortKeyComputer, SortKeyComputer};
use crate::schema::{OwnedValue, Schema};
use crate::{DocId, Order, Score, SegmentReader};
use crate::{DocId, Order, Score};
fn compare_owned_value<const NULLS_FIRST: bool>(lhs: &OwnedValue, rhs: &OwnedValue) -> Ordering {
match (lhs, rhs) {
@@ -430,7 +430,7 @@ where
fn segment_sort_key_computer(
&self,
segment_reader: &dyn SegmentReader,
segment_reader: &crate::SegmentReader,
) -> crate::Result<Self::Child> {
let child = self.0.segment_sort_key_computer(segment_reader)?;
Ok(SegmentSortKeyComputerWithComparator {
@@ -468,7 +468,7 @@ where
fn segment_sort_key_computer(
&self,
segment_reader: &dyn SegmentReader,
segment_reader: &crate::SegmentReader,
) -> crate::Result<Self::Child> {
let child = self.0.segment_sort_key_computer(segment_reader)?;
Ok(SegmentSortKeyComputerWithComparator {

2
src/collector/sort_key/sort_by_bytes.rs
View File

@@ -32,7 +32,7 @@ impl SortKeyComputer for SortByBytes {
fn segment_sort_key_computer(
&self,
segment_reader: &dyn crate::SegmentReader,
segment_reader: &crate::SegmentReader,
) -> crate::Result<Self::Child> {
let bytes_column_opt = segment_reader.fast_fields().bytes(&self.column_name)?;
Ok(ByBytesColumnSegmentSortKeyComputer { bytes_column_opt })

4
src/collector/sort_key/sort_by_erased_type.rs
View File

@@ -6,7 +6,7 @@ use crate::collector::sort_key::{
use crate::collector::{SegmentSortKeyComputer, SortKeyComputer};
use crate::fastfield::FastFieldNotAvailableError;
use crate::schema::OwnedValue;
use crate::{DateTime, DocId, Score, SegmentReader};
use crate::{DateTime, DocId, Score};
/// Sort by the boxed / OwnedValue representation of either a fast field, or of the score.
///
@@ -86,7 +86,7 @@ impl SortKeyComputer for SortByErasedType {
fn segment_sort_key_computer(
&self,
segment_reader: &dyn SegmentReader,
segment_reader: &crate::SegmentReader,
) -> crate::Result<Self::Child> {
let inner: Box<dyn ErasedSegmentSortKeyComputer> = match self {
Self::Field(column_name) => {

222
src/collector/sort_key/sort_by_score.rs
View File

@@ -1,6 +1,9 @@
use std::cmp::{Ordering, Reverse};
use std::collections::BinaryHeap;
use crate::collector::sort_key::NaturalComparator;
use crate::collector::{SegmentSortKeyComputer, SortKeyComputer, TopNComputer};
use crate::{DocAddress, DocId, Score, SegmentReader};
use crate::collector::{SegmentSortKeyComputer, SortKeyComputer};
use crate::{DocAddress, DocId, Score};
/// Sort by similarity score.
#[derive(Clone, Debug, Copy)]
@@ -19,25 +22,27 @@ impl SortKeyComputer for SortBySimilarityScore {
fn segment_sort_key_computer(
&self,
_segment_reader: &dyn SegmentReader,
_segment_reader: &crate::SegmentReader,
) -> crate::Result<Self::Child> {
Ok(SortBySimilarityScore)
}
// Sorting by score is special in that it allows for the Block-Wand optimization.
//
// We use a BinaryHeap (TopNHeap) instead of TopNComputer here so that the
// threshold is always the exact K-th best score. TopNComputer only updates its
// threshold every K docs (at truncation), giving Block-WAND a stale bound.
fn collect_segment_top_k(
&self,
k: usize,
weight: &dyn crate::query::Weight,
reader: &dyn SegmentReader,
reader: &crate::SegmentReader,
segment_ord: u32,
) -> crate::Result<Vec<(Self::SortKey, DocAddress)>> {
let mut top_n: TopNComputer<Score, DocId, Self::Comparator> =
TopNComputer::new_with_comparator(k, self.comparator());
let mut top_n = TopNHeap::new(k);
if let Some(alive_bitset) = reader.alive_bitset() {
let mut threshold = Score::MIN;
top_n.threshold = Some(threshold);
weight.for_each_pruning(Score::MIN, reader, &mut |doc, score| {
if alive_bitset.is_deleted(doc) {
return threshold;
@@ -56,7 +61,7 @@ impl SortKeyComputer for SortBySimilarityScore {
Ok(top_n
.into_vec()
.into_iter()
.map(|cid| (cid.sort_key, DocAddress::new(segment_ord, cid.doc)))
.map(|(score, doc)| (score, DocAddress::new(segment_ord, doc)))
.collect())
}
}
@@ -75,3 +80,204 @@ impl SegmentSortKeyComputer for SortBySimilarityScore {
score
}
}
/// Min-heap entry: higher score = greater, lower doc wins ties.
struct ScoreHeapEntry {
score: Score,
doc: DocId,
}
impl Eq for ScoreHeapEntry {}
impl PartialEq for ScoreHeapEntry {
fn eq(&self, other: &Self) -> bool {
self.cmp(other) == Ordering::Equal
}
}
impl PartialOrd for ScoreHeapEntry {
fn partial_cmp(&self, other: &Self) -> Option<Ordering> {
Some(self.cmp(other))
}
}
impl Ord for ScoreHeapEntry {
fn cmp(&self, other: &Self) -> Ordering {
self.score
.partial_cmp(&other.score)
.unwrap_or(Ordering::Equal)
.then_with(|| other.doc.cmp(&self.doc))
}
}
/// Heap-based top-K for score collection. O(log K) per insert, but the threshold
/// is always tight, so Block-WAND prunes better than with [`TopNComputer`]'s
/// buffer/median approach.
///
/// Like [`TopNComputer`], items must arrive in ascending doc order, and equal
/// scores are rejected (strict `>`) so that lower doc IDs win ties.
///
/// [`TopNComputer`]: crate::collector::TopNComputer
struct TopNHeap {
heap: BinaryHeap<Reverse<ScoreHeapEntry>>,
top_n: usize,
threshold: Option<Score>,
}
impl TopNHeap {
fn new(top_n: usize) -> Self {
TopNHeap {
heap: BinaryHeap::with_capacity(top_n),
top_n,
threshold: None,
}
}
#[inline]
fn push(&mut self, score: Score, doc: DocId) {
if self.heap.len() < self.top_n {
self.heap.push(Reverse(ScoreHeapEntry { score, doc }));
if self.heap.len() == self.top_n {
self.threshold = self.heap.peek().map(|Reverse(entry)| entry.score);
}
} else if let Some(threshold) = self.threshold {
if score > threshold {
// peek_mut + assign is a single sift-down, vs pop + push = two sifts.
if let Some(mut min) = self.heap.peek_mut() {
*min = Reverse(ScoreHeapEntry { score, doc });
}
self.threshold = self.heap.peek().map(|Reverse(entry)| entry.score);
}
}
}
fn into_vec(self) -> Vec<(Score, DocId)> {
self.heap
.into_vec()
.into_iter()
.map(|Reverse(entry)| (entry.score, entry.doc))
.collect()
}
}
#[cfg(test)]
mod tests {
use proptest::prelude::*;
use super::*;
use crate::collector::sort_key::NaturalComparator;
use crate::collector::TopNComputer;
#[test]
fn test_top_n_heap_zero_capacity() {
let mut heap = TopNHeap::new(0);
heap.push(1.0, 0);
heap.push(2.0, 1);
assert!(heap.into_vec().is_empty());
}
#[test]
fn test_top_n_heap_basic() {
let mut heap = TopNHeap::new(2);
heap.push(1.0, 0);
heap.push(3.0, 1);
heap.push(2.0, 2);
let mut results = heap.into_vec();
results.sort_by(|a, b| b.0.partial_cmp(&a.0).unwrap().then_with(|| a.1.cmp(&b.1)));
assert_eq!(results, vec![(3.0, 1), (2.0, 2)]);
}
#[test]
fn test_top_n_heap_threshold_always_accurate() {
let mut heap = TopNHeap::new(2);
assert_eq!(heap.threshold, None);
heap.push(1.0, 0);
assert_eq!(heap.threshold, None);
heap.push(3.0, 1);
assert_eq!(heap.threshold, Some(1.0));
heap.push(2.0, 2); // evicts 1.0
assert_eq!(heap.threshold, Some(2.0));
heap.push(4.0, 3); // evicts 2.0
assert_eq!(heap.threshold, Some(3.0));
}
#[test]
fn test_top_n_heap_tiebreaking_lower_doc_wins() {
let mut heap = TopNHeap::new(2);
heap.push(5.0, 0);
heap.push(5.0, 1);
heap.push(5.0, 2); // rejected: not strictly > threshold
let mut results = heap.into_vec();
results.sort_by_key(|&(_, doc)| doc);
assert_eq!(results, vec![(5.0, 0), (5.0, 1)]);
}
#[test]
fn test_top_n_heap_single_element() {
let mut heap = TopNHeap::new(1);
heap.push(1.0, 0);
assert_eq!(heap.threshold, Some(1.0));
heap.push(0.5, 1); // rejected
heap.push(2.0, 2); // accepted
assert_eq!(heap.threshold, Some(2.0));
let results = heap.into_vec();
assert_eq!(results, vec![(2.0, 2)]);
}
#[test]
fn test_top_n_heap_under_capacity() {
let mut heap = TopNHeap::new(5);
heap.push(3.0, 0);
heap.push(1.0, 1);
heap.push(2.0, 2);
// Only 3 elements, capacity is 5 — all should be kept
assert_eq!(heap.threshold, None);
let mut results = heap.into_vec();
results.sort_by(|a, b| b.0.partial_cmp(&a.0).unwrap().then_with(|| a.1.cmp(&b.1)));
assert_eq!(results, vec![(3.0, 0), (2.0, 2), (1.0, 1)]);
}
proptest! {
#[test]
fn test_top_n_heap_matches_top_n_computer(
limit in 0..20_usize,
mut docs in proptest::collection::vec((0..1000_u32, 0..1000_u32), 0..200_usize),
) {
// Both require ascending doc order.
docs.sort_by_key(|(_, doc_id)| *doc_id);
docs.dedup_by_key(|(_, doc_id)| *doc_id);
let mut heap = TopNHeap::new(limit);
let mut computer: TopNComputer<Score, DocId, NaturalComparator> =
TopNComputer::new_with_comparator(limit, NaturalComparator);
for &(score_u32, doc) in &docs {
let score = score_u32 as Score;
heap.push(score, doc);
computer.push(score, doc);
}
let mut heap_results = heap.into_vec();
heap_results.sort_by(|a, b| {
b.0.partial_cmp(&a.0).unwrap().then_with(|| a.1.cmp(&b.1))
});
let computer_results: Vec<(Score, DocId)> = computer
.into_sorted_vec()
.into_iter()
.map(|cd| (cd.sort_key, cd.doc))
.collect();
prop_assert_eq!(heap_results, computer_results);
}
}
}

4
src/collector/sort_key/sort_by_static_fast_value.rs
View File

@@ -52,7 +52,7 @@ impl<T: FastValue> SortKeyComputer for SortByStaticFastValue<T> {
if schema_type != T::to_type() {
return Err(crate::TantivyError::SchemaError(format!(
"Field `{}` is of type {schema_type:?}, not of the type {:?}.",
&self.field,
self.field,
T::to_type()
)));
}
@@ -61,7 +61,7 @@ impl<T: FastValue> SortKeyComputer for SortByStaticFastValue<T> {
fn segment_sort_key_computer(
&self,
segment_reader: &dyn SegmentReader,
segment_reader: &SegmentReader,
) -> crate::Result<Self::Child> {
let sort_column_opt = segment_reader.fast_fields().u64_lenient(&self.field)?;
let (sort_column, _sort_column_type) =

4
src/collector/sort_key/sort_by_string.rs
View File

@@ -3,7 +3,7 @@ use columnar::StrColumn;
use crate::collector::sort_key::NaturalComparator;
use crate::collector::{SegmentSortKeyComputer, SortKeyComputer};
use crate::termdict::TermOrdinal;
use crate::{DocId, Score, SegmentReader};
use crate::{DocId, Score};
/// Sort by the first value of a string column.
///
@@ -35,7 +35,7 @@ impl SortKeyComputer for SortByString {
fn segment_sort_key_computer(
&self,
segment_reader: &dyn SegmentReader,
segment_reader: &crate::SegmentReader,
) -> crate::Result<Self::Child> {
let str_column_opt = segment_reader.fast_fields().str(&self.column_name)?;
Ok(ByStringColumnSegmentSortKeyComputer { str_column_opt })

22
src/collector/sort_key/sort_key_computer.rs
View File

@@ -119,7 +119,7 @@ pub trait SortKeyComputer: Sync {
&self,
k: usize,
weight: &dyn crate::query::Weight,
reader: &dyn SegmentReader,
reader: &crate::SegmentReader,
segment_ord: u32,
) -> crate::Result<Vec<(Self::SortKey, DocAddress)>> {
let with_scoring = self.requires_scoring();
@@ -135,7 +135,7 @@ pub trait SortKeyComputer: Sync {
}
/// Builds a child sort key computer for a specific segment.
fn segment_sort_key_computer(&self, segment_reader: &dyn SegmentReader) -> Result<Self::Child>;
fn segment_sort_key_computer(&self, segment_reader: &SegmentReader) -> Result<Self::Child>;
}
impl<HeadSortKeyComputer, TailSortKeyComputer> SortKeyComputer
@@ -156,7 +156,7 @@ where
(self.0.comparator(), self.1.comparator())
}
fn segment_sort_key_computer(&self, segment_reader: &dyn SegmentReader) -> Result<Self::Child> {
fn segment_sort_key_computer(&self, segment_reader: &SegmentReader) -> Result<Self::Child> {
Ok((
self.0.segment_sort_key_computer(segment_reader)?,
self.1.segment_sort_key_computer(segment_reader)?,
@@ -357,7 +357,7 @@ where
)
}
fn segment_sort_key_computer(&self, segment_reader: &dyn SegmentReader) -> Result<Self::Child> {
fn segment_sort_key_computer(&self, segment_reader: &SegmentReader) -> Result<Self::Child> {
let sort_key_computer1 = self.0.segment_sort_key_computer(segment_reader)?;
let sort_key_computer2 = self.1.segment_sort_key_computer(segment_reader)?;
let sort_key_computer3 = self.2.segment_sort_key_computer(segment_reader)?;
@@ -420,7 +420,7 @@ where
SortKeyComputer4::Comparator,
);
fn segment_sort_key_computer(&self, segment_reader: &dyn SegmentReader) -> Result<Self::Child> {
fn segment_sort_key_computer(&self, segment_reader: &SegmentReader) -> Result<Self::Child> {
let sort_key_computer1 = self.0.segment_sort_key_computer(segment_reader)?;
let sort_key_computer2 = self.1.segment_sort_key_computer(segment_reader)?;
let sort_key_computer3 = self.2.segment_sort_key_computer(segment_reader)?;
@@ -454,7 +454,7 @@ where
impl<F, SegmentF, TSortKey> SortKeyComputer for F
where
F: 'static + Send + Sync + Fn(&dyn SegmentReader) -> SegmentF,
F: 'static + Send + Sync + Fn(&SegmentReader) -> SegmentF,
SegmentF: 'static + FnMut(DocId) -> TSortKey,
TSortKey: 'static + PartialOrd + Clone + Send + Sync + std::fmt::Debug,
{
@@ -462,7 +462,7 @@ where
type Child = SegmentF;
type Comparator = NaturalComparator;
fn segment_sort_key_computer(&self, segment_reader: &dyn SegmentReader) -> Result<Self::Child> {
fn segment_sort_key_computer(&self, segment_reader: &SegmentReader) -> Result<Self::Child> {
Ok((self)(segment_reader))
}
}
@@ -509,10 +509,10 @@ mod tests {
#[test]
fn test_lazy_score_computer() {
let score_computer_primary = |_segment_reader: &dyn SegmentReader| |_doc: DocId| 200u32;
let score_computer_primary = |_segment_reader: &SegmentReader| |_doc: DocId| 200u32;
let call_count = Arc::new(AtomicUsize::new(0));
let call_count_clone = call_count.clone();
let score_computer_secondary = move |_segment_reader: &dyn SegmentReader| {
let score_computer_secondary = move |_segment_reader: &SegmentReader| {
let call_count_new_clone = call_count_clone.clone();
move |_doc: DocId| {
call_count_new_clone.fetch_add(1, AtomicOrdering::SeqCst);
@@ -572,10 +572,10 @@ mod tests {
#[test]
fn test_lazy_score_computer_dynamic_ordering() {
let score_computer_primary = |_segment_reader: &dyn SegmentReader| |_doc: DocId| 200u32;
let score_computer_primary = |_segment_reader: &SegmentReader| |_doc: DocId| 200u32;
let call_count = Arc::new(AtomicUsize::new(0));
let call_count_clone = call_count.clone();
let score_computer_secondary = move |_segment_reader: &dyn SegmentReader| {
let score_computer_secondary = move |_segment_reader: &SegmentReader| {
let call_count_new_clone = call_count_clone.clone();
move |_doc: DocId| {
call_count_new_clone.fetch_add(1, AtomicOrdering::SeqCst);

8
src/collector/sort_key_top_collector.rs
View File

@@ -32,11 +32,7 @@ where TSortKeyComputer: SortKeyComputer + Send + Sync + 'static
self.sort_key_computer.check_schema(schema)
}
fn for_segment(
&self,
segment_ord: u32,
segment_reader: &dyn SegmentReader,
) -> Result<Self::Child> {
fn for_segment(&self, segment_ord: u32, segment_reader: &SegmentReader) -> Result<Self::Child> {
let segment_sort_key_computer = self
.sort_key_computer
.segment_sort_key_computer(segment_reader)?;
@@ -67,7 +63,7 @@ where TSortKeyComputer: SortKeyComputer + Send + Sync + 'static
&self,
weight: &dyn Weight,
segment_ord: u32,
reader: &dyn SegmentReader,
reader: &SegmentReader,
) -> crate::Result<Vec<(TSortKeyComputer::SortKey, DocAddress)>> {
let k = self.doc_range.end;
let docs = self

8
src/collector/tests.rs
View File

@@ -5,7 +5,7 @@ use crate::query::{AllQuery, QueryParser};
use crate::schema::{Schema, FAST, TEXT};
use crate::time::format_description::well_known::Rfc3339;
use crate::time::OffsetDateTime;
use crate::{DateTime, DocAddress, Index, Searcher, SegmentReader, TantivyDocument};
use crate::{DateTime, DocAddress, Index, Searcher, TantivyDocument};
pub const TEST_COLLECTOR_WITH_SCORE: TestCollector = TestCollector {
compute_score: true,
@@ -109,7 +109,7 @@ impl Collector for TestCollector {
fn for_segment(
&self,
segment_id: SegmentOrdinal,
_reader: &dyn SegmentReader,
_reader: &SegmentReader,
) -> crate::Result<TestSegmentCollector> {
Ok(TestSegmentCollector {
segment_id,
@@ -180,7 +180,7 @@ impl Collector for FastFieldTestCollector {
fn for_segment(
&self,
_: SegmentOrdinal,
segment_reader: &dyn SegmentReader,
segment_reader: &SegmentReader,
) -> crate::Result<FastFieldSegmentCollector> {
let reader = segment_reader
.fast_fields()
@@ -243,7 +243,7 @@ impl Collector for BytesFastFieldTestCollector {
fn for_segment(
&self,
_segment_local_id: u32,
segment_reader: &dyn SegmentReader,
segment_reader: &SegmentReader,
) -> crate::Result<BytesFastFieldSegmentCollector> {
let column_opt = segment_reader.fast_fields().bytes(&self.field)?;
Ok(BytesFastFieldSegmentCollector {

14
src/collector/top_score_collector.rs
View File

@@ -393,7 +393,7 @@ impl TopDocs {
/// // This is where we build our collector with our custom score.
/// let top_docs_by_custom_score = TopDocs
/// ::with_limit(10)
/// .tweak_score(move |segment_reader: &dyn SegmentReader| {
/// .tweak_score(move |segment_reader: &SegmentReader| {
/// // The argument is a function that returns our scoring
/// // function.
/// //
@@ -442,7 +442,7 @@ pub struct TweakScoreFn<F>(F);
impl<F, TTweakScoreSortKeyFn, TSortKey> SortKeyComputer for TweakScoreFn<F>
where
F: 'static + Send + Sync + Fn(&dyn SegmentReader) -> TTweakScoreSortKeyFn,
F: 'static + Send + Sync + Fn(&SegmentReader) -> TTweakScoreSortKeyFn,
TTweakScoreSortKeyFn: 'static + Fn(DocId, Score) -> TSortKey,
TweakScoreSegmentSortKeyComputer<TTweakScoreSortKeyFn>:
SegmentSortKeyComputer<SortKey = TSortKey, SegmentSortKey = TSortKey>,
@@ -458,7 +458,7 @@ where
fn segment_sort_key_computer(
&self,
segment_reader: &dyn SegmentReader,
segment_reader: &SegmentReader,
) -> crate::Result<Self::Child> {
Ok({
TweakScoreSegmentSortKeyComputer {
@@ -513,7 +513,9 @@ pub struct TopNComputer<Score, D, C> {
/// The buffer reverses sort order to get top-semantics instead of bottom-semantics
buffer: Vec<ComparableDoc<Score, D>>,
top_n: usize,
pub(crate) threshold: Option<Score>,
/// The current threshold for pruning. Documents with scores at or below
/// this value are skipped by `push()`. Updated when the buffer is truncated.
pub threshold: Option<Score>,
comparator: C,
}
@@ -1525,7 +1527,7 @@ mod tests {
let text_query = query_parser.parse_query("droopy tax")?;
let collector = TopDocs::with_limit(2)
.and_offset(1)
.order_by(move |_segment_reader: &dyn SegmentReader| move |doc: DocId| doc);
.order_by(move |_segment_reader: &SegmentReader| move |doc: DocId| doc);
let score_docs: Vec<(u32, DocAddress)> =
index.reader()?.searcher().search(&text_query, &collector)?;
assert_eq!(
@@ -1543,7 +1545,7 @@ mod tests {
let text_query = query_parser.parse_query("droopy tax").unwrap();
let collector = TopDocs::with_limit(2)
.and_offset(1)
.order_by(move |_segment_reader: &dyn SegmentReader| move |doc: DocId| doc);
.order_by(move |_segment_reader: &SegmentReader| move |doc: DocId| doc);
let score_docs: Vec<(u32, DocAddress)> = index
.reader()
.unwrap()

6
src/core/json_utils.rs
View File

@@ -4,7 +4,7 @@ use common::{replace_in_place, JsonPathWriter};
use rustc_hash::FxHashMap;
use crate::indexer::indexing_term::IndexingTerm;
use crate::postings::{IndexingContext, IndexingPosition, PostingsWriter as _, PostingsWriterEnum};
use crate::postings::{IndexingContext, IndexingPosition, PostingsWriter};
use crate::schema::document::{ReferenceValue, ReferenceValueLeaf, Value};
use crate::schema::{Type, DATE_TIME_PRECISION_INDEXED};
use crate::time::format_description::well_known::Rfc3339;
@@ -80,7 +80,7 @@ fn index_json_object<'a, V: Value<'a>>(
text_analyzer: &mut TextAnalyzer,
term_buffer: &mut IndexingTerm,
json_path_writer: &mut JsonPathWriter,
postings_writer: &mut PostingsWriterEnum,
postings_writer: &mut dyn PostingsWriter,
ctx: &mut IndexingContext,
positions_per_path: &mut IndexingPositionsPerPath,
) {
@@ -110,7 +110,7 @@ pub(crate) fn index_json_value<'a, V: Value<'a>>(
text_analyzer: &mut TextAnalyzer,
term_buffer: &mut IndexingTerm,
json_path_writer: &mut JsonPathWriter,
postings_writer: &mut PostingsWriterEnum,
postings_writer: &mut dyn PostingsWriter,
ctx: &mut IndexingContext,
positions_per_path: &mut IndexingPositionsPerPath,
) {

2
src/core/mod.rs
View File

@@ -8,7 +8,7 @@ use std::path::Path;
use once_cell::sync::Lazy;
pub use self::executor::Executor;
pub use self::searcher::{Searcher, SearcherContext, SearcherGeneration};
pub use self::searcher::{Searcher, SearcherGeneration};
/// The meta file contains all the information about the list of segments and the schema
/// of the index.

204
src/core/searcher.rs
View File

@@ -4,13 +4,13 @@ use std::{fmt, io};
use crate::collector::Collector;
use crate::core::Executor;
use crate::index::{Index, SegmentId, SegmentReader};
use crate::index::{SegmentId, SegmentReader};
use crate::query::{Bm25StatisticsProvider, EnableScoring, Query};
use crate::schema::{Field, FieldType, Schema, TantivyDocument, Term};
use crate::schema::document::DocumentDeserialize;
use crate::schema::{Schema, Term};
use crate::space_usage::SearcherSpaceUsage;
use crate::store::{CacheStats, StoreReader, DOCSTORE_CACHE_CAPACITY};
use crate::tokenizer::{TextAnalyzer, TokenizerManager};
use crate::{DocAddress, Inventory, Opstamp, TantivyError, TrackedObject};
use crate::store::{CacheStats, StoreReader};
use crate::{DocAddress, Index, Opstamp, TrackedObject};
/// Identifies the searcher generation accessed by a [`Searcher`].
///
@@ -36,7 +36,7 @@ pub struct SearcherGeneration {
impl SearcherGeneration {
pub(crate) fn from_segment_readers(
segment_readers: &[Arc<dyn SegmentReader>],
segment_readers: &[SegmentReader],
generation_id: u64,
) -> Self {
let mut segment_id_to_del_opstamp = BTreeMap::new();
@@ -61,103 +61,6 @@ impl SearcherGeneration {
}
}
/// Search-time context required by a [`Searcher`].
#[derive(Clone)]
pub struct SearcherContext {
schema: Schema,
executor: Executor,
tokenizers: TokenizerManager,
fast_field_tokenizers: TokenizerManager,
}
impl SearcherContext {
/// Creates a context from explicit search-time components.
pub fn new(
schema: Schema,
executor: Executor,
tokenizers: TokenizerManager,
fast_field_tokenizers: TokenizerManager,
) -> SearcherContext {
SearcherContext {
schema,
executor,
tokenizers,
fast_field_tokenizers,
}
}
/// Creates a context from an index.
pub fn from_index(index: &Index) -> SearcherContext {
SearcherContext::new(
index.schema(),
index.search_executor().clone(),
index.tokenizers().clone(),
index.fast_field_tokenizer().clone(),
)
}
/// Access the schema associated with this context.
pub fn schema(&self) -> &Schema {
&self.schema
}
/// Access the executor associated with this context.
pub fn search_executor(&self) -> &Executor {
&self.executor
}
/// Access the tokenizer manager associated with this context.
pub fn tokenizers(&self) -> &TokenizerManager {
&self.tokenizers
}
/// Access the fast field tokenizer manager associated with this context.
pub fn fast_field_tokenizer(&self) -> &TokenizerManager {
&self.fast_field_tokenizers
}
/// Get the tokenizer associated with a specific field.
pub fn tokenizer_for_field(&self, field: Field) -> crate::Result<TextAnalyzer> {
let field_entry = self.schema.get_field_entry(field);
let field_type = field_entry.field_type();
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()
)))
}
};
let indexing_options = indexing_options_opt.ok_or_else(|| {
TantivyError::InvalidArgument(format!(
"No indexing options set for field {field_entry:?}"
))
})?;
self.tokenizers
.get(indexing_options.tokenizer())
.ok_or_else(|| {
TantivyError::InvalidArgument(format!(
"No Tokenizer found for field {field_entry:?}"
))
})
}
}
impl From<&Index> for SearcherContext {
fn from(index: &Index) -> Self {
SearcherContext::from_index(index)
}
}
impl From<Index> for SearcherContext {
fn from(index: Index) -> Self {
SearcherContext::from(&index)
}
}
/// Holds a list of `SegmentReader`s ready for search.
///
/// It guarantees that the `Segment` will not be removed before
@@ -168,51 +71,9 @@ pub struct Searcher {
}
impl Searcher {
/// Creates a `Searcher` from an arbitrary list of segment readers.
///
/// This is useful when segment readers are not opened from
/// `IndexReader` / `meta.json` (e.g. external segment sources).
/// The generated [`SearcherGeneration`] uses `generation_id = 0`.
pub fn from_segment_readers<Ctx: Into<SearcherContext>>(
context: Ctx,
segment_readers: Vec<Arc<dyn SegmentReader>>,
) -> crate::Result<Searcher> {
Self::from_segment_readers_with_generation_id(context, segment_readers, 0)
}
/// Same as [`Searcher::from_segment_readers`] but allows setting
/// a custom generation id.
pub fn from_segment_readers_with_generation_id<Ctx: Into<SearcherContext>>(
context: Ctx,
segment_readers: Vec<Arc<dyn SegmentReader>>,
generation_id: u64,
) -> crate::Result<Searcher> {
let context = context.into();
let generation = SearcherGeneration::from_segment_readers(&segment_readers, generation_id);
let tracked_generation = Inventory::default().track(generation);
let inner = SearcherInner::new(
context,
segment_readers,
tracked_generation,
DOCSTORE_CACHE_CAPACITY,
)?;
Ok(Arc::new(inner).into())
}
/// Returns the search context associated with the `Searcher`.
pub fn context(&self) -> &SearcherContext {
&self.inner.context
}
/// Deprecated alias for [`Searcher::context`].
#[deprecated(note = "use Searcher::context()")]
pub fn index(&self) -> &SearcherContext {
self.context()
}
/// Access the schema associated with the index of this searcher.
pub fn schema(&self) -> &Schema {
self.context().schema()
/// Returns the `Index` associated with the `Searcher`
pub fn index(&self) -> &Index {
&self.inner.index
}
/// [`SearcherGeneration`] which identifies the version of the snapshot held by this `Searcher`.
@@ -224,7 +85,7 @@ impl Searcher {
///
/// The searcher uses the segment ordinal to route the
/// request to the right `Segment`.
pub fn doc(&self, doc_address: DocAddress) -> crate::Result<TantivyDocument> {
pub fn doc<D: DocumentDeserialize>(&self, doc_address: DocAddress) -> crate::Result<D> {
let store_reader = &self.inner.store_readers[doc_address.segment_ord as usize];
store_reader.get(doc_address.doc_id)
}
@@ -244,12 +105,20 @@ impl Searcher {
/// Fetches a document in an asynchronous manner.
#[cfg(feature = "quickwit")]
pub async fn doc_async(&self, doc_address: DocAddress) -> crate::Result<TantivyDocument> {
let executor = self.context().search_executor();
pub async fn doc_async<D: DocumentDeserialize>(
&self,
doc_address: DocAddress,
) -> crate::Result<D> {
let executor = self.inner.index.search_executor();
let store_reader = &self.inner.store_readers[doc_address.segment_ord as usize];
store_reader.get_async(doc_address.doc_id, executor).await
}
/// Access the schema associated with the index of this searcher.
pub fn schema(&self) -> &Schema {
&self.inner.schema
}
/// Returns the overall number of documents in the index.
pub fn num_docs(&self) -> u64 {
self.inner
@@ -285,13 +154,13 @@ impl Searcher {
}
/// Return the list of segment readers
pub fn segment_readers(&self) -> &[Arc<dyn SegmentReader>] {
pub fn segment_readers(&self) -> &[SegmentReader] {
&self.inner.segment_readers
}
/// Returns the segment_reader associated with the given segment_ord
pub fn segment_reader(&self, segment_ord: u32) -> &dyn SegmentReader {
self.inner.segment_readers[segment_ord as usize].as_ref()
pub fn segment_reader(&self, segment_ord: u32) -> &SegmentReader {
&self.inner.segment_readers[segment_ord as usize]
}
/// Runs a query on the segment readers wrapped by the searcher.
@@ -332,7 +201,7 @@ impl Searcher {
} else {
EnableScoring::disabled_from_searcher(self)
};
let executor = self.context().search_executor();
let executor = self.inner.index.search_executor();
self.search_with_executor(query, collector, executor, enabled_scoring)
}
@@ -360,11 +229,7 @@ impl Searcher {
let segment_readers = self.segment_readers();
let fruits = executor.map(
|(segment_ord, segment_reader)| {
collector.collect_segment(
weight.as_ref(),
segment_ord as u32,
segment_reader.as_ref(),
)
collector.collect_segment(weight.as_ref(), segment_ord as u32, segment_reader)
},
segment_readers.iter().enumerate(),
)?;
@@ -392,17 +257,19 @@ impl From<Arc<SearcherInner>> for Searcher {
/// It guarantees that the `Segment` will not be removed before
/// the destruction of the `Searcher`.
pub(crate) struct SearcherInner {
context: SearcherContext,
segment_readers: Vec<Arc<dyn SegmentReader>>,
store_readers: Vec<Box<dyn StoreReader>>,
schema: Schema,
index: Index,
segment_readers: Vec<SegmentReader>,
store_readers: Vec<StoreReader>,
generation: TrackedObject<SearcherGeneration>,
}
impl SearcherInner {
/// Creates a new `Searcher`
pub(crate) fn new(
context: SearcherContext,
segment_readers: Vec<Arc<dyn SegmentReader>>,
schema: Schema,
index: Index,
segment_readers: Vec<SegmentReader>,
generation: TrackedObject<SearcherGeneration>,
doc_store_cache_num_blocks: usize,
) -> io::Result<SearcherInner> {
@@ -414,13 +281,14 @@ impl SearcherInner {
generation.segments(),
"Set of segments referenced by this Searcher and its SearcherGeneration must match"
);
let store_readers: Vec<Box<dyn StoreReader>> = segment_readers
let store_readers: Vec<StoreReader> = segment_readers
.iter()
.map(|segment_reader| segment_reader.get_store_reader(doc_store_cache_num_blocks))
.collect::<io::Result<Vec<_>>>()?;
Ok(SearcherInner {
context,
schema,
index,
segment_readers,
store_readers,
generation,
@@ -433,7 +301,7 @@ impl fmt::Debug for Searcher {
let segment_ids = self
.segment_readers()
.iter()
.map(|segment_reader| segment_reader.segment_id())
.map(SegmentReader::segment_id)
.collect::<Vec<_>>();
write!(f, "Searcher({segment_ids:?})")
}

52
src/core/tests.rs
View File

@@ -7,10 +7,24 @@ use crate::query::TermQuery;
use crate::schema::{Field, IndexRecordOption, Schema, INDEXED, STRING, TEXT};
use crate::tokenizer::TokenizerManager;
use crate::{
Directory, DocSet, Executor, Index, IndexBuilder, IndexReader, IndexSettings, IndexWriter,
ReloadPolicy, Searcher, SearcherContext, TantivyDocument, Term,
Directory, DocSet, Index, IndexBuilder, IndexReader, IndexSettings, IndexWriter, ReloadPolicy,
TantivyDocument, Term,
};
#[test]
fn test_indexer_for_field() {
let mut schema_builder = Schema::builder();
let num_likes_field = schema_builder.add_u64_field("num_likes", INDEXED);
let body_field = schema_builder.add_text_field("body", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
assert!(index.tokenizer_for_field(body_field).is_ok());
assert_eq!(
format!("{:?}", index.tokenizer_for_field(num_likes_field).err()),
"Some(SchemaError(\"\\\"num_likes\\\" is not a text field.\"))"
);
}
#[test]
fn test_set_tokenizer_manager() {
let mut schema_builder = Schema::builder();
@@ -286,40 +300,6 @@ fn test_single_segment_index_writer() -> crate::Result<()> {
Ok(())
}
#[test]
fn test_searcher_from_external_segment_readers() -> crate::Result<()> {
let mut schema_builder = Schema::builder();
let text_field = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema.clone());
let mut writer: IndexWriter = index.writer_for_tests()?;
writer.add_document(doc!(text_field => "hello"))?;
writer.add_document(doc!(text_field => "hello"))?;
writer.commit()?;
let reader = index.reader()?;
let searcher = reader.searcher();
let segment_readers = searcher.segment_readers().to_vec();
let context = SearcherContext::new(
schema,
Executor::single_thread(),
TokenizerManager::default(),
TokenizerManager::default(),
);
let custom_searcher =
Searcher::from_segment_readers_with_generation_id(context, segment_readers, 42)?;
let term_query = TermQuery::new(
Term::from_field_text(text_field, "hello"),
IndexRecordOption::Basic,
);
let count = custom_searcher.search(&term_query, &Count)?;
assert_eq!(count, 2);
assert_eq!(custom_searcher.generation().generation_id(), 42);
assert_eq!(custom_searcher.segment_readers().len(), 1);
Ok(())
}
#[test]
fn test_merging_segment_update_docfreq() {
let mut schema_builder = Schema::builder();

4
src/directory/composite_file.rs
View File

@@ -167,9 +167,7 @@ impl CompositeFile {
.map(|byte_range| self.data.slice(byte_range.clone()))
}
/// Returns per-field byte usage for all slices stored in this composite file.
///
/// The provided `schema` is used to resolve field ids into field names.
/// Returns the space usage per field in this composite file.
pub fn space_usage(&self, schema: &Schema) -> PerFieldSpaceUsage {
let mut fields = Vec::new();
for (&field_addr, byte_range) in &self.offsets_index {

110
src/docset.rs
View File

@@ -1,7 +1,6 @@
use std::borrow::BorrowMut;
use std::ops::{Deref as _, DerefMut as _};
use std::borrow::{Borrow, BorrowMut};
use common::BitSet;
use common::TinySet;
use crate::fastfield::AliveBitSet;
use crate::DocId;
@@ -17,6 +16,12 @@ pub const TERMINATED: DocId = i32::MAX as u32;
/// exactly this size as long as we can fill the buffer.
pub const COLLECT_BLOCK_BUFFER_LEN: usize = 64;
/// Number of `TinySet` (64-bit) buckets in a block used by [`DocSet::fill_bitset_block`].
pub const BLOCK_NUM_TINYBITSETS: usize = 16;
/// Number of doc IDs covered by one block: `BLOCK_NUM_TINYBITSETS * 64 = 1024`.
pub const BLOCK_WINDOW: u32 = BLOCK_NUM_TINYBITSETS as u32 * 64;
/// Represents an iterable set of sorted doc ids.
pub trait DocSet: Send {
/// Goes to the next element.
@@ -133,19 +138,6 @@ pub trait DocSet: Send {
buffer.len()
}
/// Fills the given bitset with the documents in the docset.
///
/// If the docset max_doc is smaller than the largest doc, this function might not consume the
/// docset entirely.
fn fill_bitset(&mut self, bitset: &mut BitSet) {
let bitset_max_value: u32 = bitset.max_value();
let mut doc = self.doc();
while doc < bitset_max_value {
bitset.insert(doc);
doc = self.advance();
}
}
/// Returns the current document
/// Right after creating a new `DocSet`, the docset points to the first document.
///
@@ -176,6 +168,31 @@ pub trait DocSet: Send {
self.size_hint() as u64
}
/// Fills a bitmask representing which documents in `[min_doc, min_doc + BLOCK_WINDOW)` are
/// present in this docset.
///
/// The window is divided into `BLOCK_NUM_TINYBITSETS` buckets of 64 docs each.
/// Returns the next doc `>= min_doc + BLOCK_WINDOW`, or `TERMINATED` if exhausted.
fn fill_bitset_block(
&mut self,
min_doc: DocId,
mask: &mut [TinySet; BLOCK_NUM_TINYBITSETS],
) -> DocId {
self.seek(min_doc);
let horizon = min_doc + BLOCK_WINDOW;
loop {
let doc = self.doc();
if doc >= horizon {
return doc;
}
let delta = doc - min_doc;
mask[(delta / 64) as usize].insert_mut(delta % 64);
if self.advance() == TERMINATED {
return TERMINATED;
}
}
}
/// Returns the number documents matching.
/// Calling this method consumes the `DocSet`.
fn count(&mut self, alive_bitset: &AliveBitSet) -> u32 {
@@ -230,6 +247,18 @@ impl DocSet for &mut dyn DocSet {
(**self).seek_danger(target)
}
fn fill_buffer(&mut self, buffer: &mut [DocId; COLLECT_BLOCK_BUFFER_LEN]) -> usize {
(**self).fill_buffer(buffer)
}
fn fill_bitset_block(
&mut self,
min_doc: DocId,
mask: &mut [TinySet; BLOCK_NUM_TINYBITSETS],
) -> DocId {
(**self).fill_bitset_block(min_doc, mask)
}
fn doc(&self) -> u32 {
(**self).doc()
}
@@ -249,59 +278,60 @@ impl DocSet for &mut dyn DocSet {
fn count_including_deleted(&mut self) -> u32 {
(**self).count_including_deleted()
}
fn fill_bitset(&mut self, bitset: &mut BitSet) {
(**self).fill_bitset(bitset);
}
}
impl<TDocSet: DocSet + ?Sized> DocSet for Box<TDocSet> {
#[inline]
fn advance(&mut self) -> DocId {
self.deref_mut().advance()
let unboxed: &mut TDocSet = self.borrow_mut();
unboxed.advance()
}
#[inline]
fn seek(&mut self, target: DocId) -> DocId {
self.deref_mut().seek(target)
let unboxed: &mut TDocSet = self.borrow_mut();
unboxed.seek(target)
}
#[inline]
fn seek_danger(&mut self, target: DocId) -> SeekDangerResult {
let unboxed: &mut TDocSet = self.borrow_mut();
unboxed.seek_danger(target)
}
#[inline]
fn fill_buffer(&mut self, buffer: &mut [DocId; COLLECT_BLOCK_BUFFER_LEN]) -> usize {
self.deref_mut().fill_buffer(buffer)
let unboxed: &mut TDocSet = self.borrow_mut();
unboxed.fill_buffer(buffer)
}
fn fill_bitset_block(
&mut self,
min_doc: DocId,
mask: &mut [TinySet; BLOCK_NUM_TINYBITSETS],
) -> DocId {
let unboxed: &mut TDocSet = self.borrow_mut();
unboxed.fill_bitset_block(min_doc, mask)
}
#[inline]
fn doc(&self) -> DocId {
self.deref().doc()
let unboxed: &TDocSet = self.borrow();
unboxed.doc()
}
#[inline]
fn size_hint(&self) -> u32 {
self.deref().size_hint()
let unboxed: &TDocSet = self.borrow();
unboxed.size_hint()
}
#[inline]
fn cost(&self) -> u64 {
self.deref().cost()
let unboxed: &TDocSet = self.borrow();
unboxed.cost()
}
#[inline]
fn count(&mut self, alive_bitset: &AliveBitSet) -> u32 {
self.deref_mut().count(alive_bitset)
let unboxed: &mut TDocSet = self.borrow_mut();
unboxed.count(alive_bitset)
}
fn count_including_deleted(&mut self) -> u32 {
self.deref_mut().count_including_deleted()
}
fn fill_bitset(&mut self, bitset: &mut BitSet) {
self.deref_mut().fill_bitset(bitset);
let unboxed: &mut TDocSet = self.borrow_mut();
unboxed.count_including_deleted()
}
}

6
src/fastfield/facet_reader.rs
View File

@@ -84,7 +84,9 @@ mod tests {
let mut facet = Facet::default();
facet_reader.facet_from_ord(0, &mut facet).unwrap();
assert_eq!(facet.to_path_string(), "/a/b");
let doc = searcher.doc(DocAddress::new(0u32, 0u32)).unwrap();
let doc = searcher
.doc::<TantivyDocument>(DocAddress::new(0u32, 0u32))
.unwrap();
let value = doc
.get_first(facet_field)
.and_then(|v| v.as_value().as_facet());
@@ -143,7 +145,7 @@ mod tests {
let mut facet_ords = Vec::new();
facet_ords.extend(facet_reader.facet_ords(0u32));
assert_eq!(&facet_ords, &[0u64]);
let doc = searcher.doc(DocAddress::new(0u32, 0u32))?;
let doc = searcher.doc::<TantivyDocument>(DocAddress::new(0u32, 0u32))?;
let value: Option<Facet> = doc
.get_first(facet_field)
.and_then(|v| v.as_facet())

4
src/fastfield/mod.rs
View File

@@ -96,7 +96,7 @@ mod tests {
};
use crate::time::OffsetDateTime;
use crate::tokenizer::{LowerCaser, RawTokenizer, TextAnalyzer, TokenizerManager};
use crate::{Index, IndexWriter};
use crate::{Index, IndexWriter, SegmentReader};
pub static SCHEMA: Lazy<Schema> = Lazy::new(|| {
let mut schema_builder = Schema::builder();
@@ -430,7 +430,7 @@ mod tests {
.searcher()
.segment_readers()
.iter()
.map(|segment_reader| segment_reader.segment_id())
.map(SegmentReader::segment_id)
.collect();
assert_eq!(segment_ids.len(), 2);
index_writer.merge(&segment_ids[..]).wait().unwrap();

9
src/fastfield/readers.rs
View File

@@ -25,8 +25,7 @@ pub struct FastFieldReaders {
}
impl FastFieldReaders {
/// Opens the segment fast-field container and binds it to a schema.
pub fn open(fast_field_file: FileSlice, schema: Schema) -> io::Result<FastFieldReaders> {
pub(crate) fn open(fast_field_file: FileSlice, schema: Schema) -> io::Result<FastFieldReaders> {
let columnar = Arc::new(ColumnarReader::open(fast_field_file)?);
Ok(FastFieldReaders { columnar, schema })
}
@@ -40,8 +39,7 @@ impl FastFieldReaders {
self.resolve_column_name_given_default_field(column_name, default_field_opt)
}
/// Returns per-field space usage for all loaded fast-field columns.
pub fn space_usage(&self) -> io::Result<PerFieldSpaceUsage> {
pub(crate) fn space_usage(&self) -> io::Result<PerFieldSpaceUsage> {
let mut per_field_usages: Vec<FieldUsage> = Default::default();
for (mut field_name, column_handle) in self.columnar.iter_columns()? {
json_path_sep_to_dot(&mut field_name);
@@ -53,8 +51,7 @@ impl FastFieldReaders {
Ok(PerFieldSpaceUsage::new(per_field_usages))
}
/// Returns the underlying `ColumnarReader`.
pub fn columnar(&self) -> &ColumnarReader {
pub(crate) fn columnar(&self) -> &ColumnarReader {
self.columnar.as_ref()
}

60
src/index/index.rs
View File

@@ -7,7 +7,7 @@ use std::thread::available_parallelism;
use super::segment::Segment;
use super::segment_reader::merge_field_meta_data;
use super::{FieldMetadata, IndexSettings, TantivySegmentReader};
use super::{FieldMetadata, IndexSettings};
use crate::core::{Executor, META_FILEPATH};
use crate::directory::error::OpenReadError;
#[cfg(feature = "mmap")]
@@ -22,8 +22,9 @@ use crate::indexer::segment_updater::save_metas;
use crate::indexer::{IndexWriter, SingleSegmentIndexWriter};
use crate::reader::{IndexReader, IndexReaderBuilder};
use crate::schema::document::Document;
use crate::schema::Schema;
use crate::tokenizer::TokenizerManager;
use crate::schema::{Field, FieldType, Schema};
use crate::tokenizer::{TextAnalyzer, TokenizerManager};
use crate::SegmentReader;
fn load_metas(
directory: &dyn Directory,
@@ -243,12 +244,9 @@ impl IndexBuilder {
/// Creates a new index given an implementation of the trait `Directory`.
///
/// If a directory previously existed, it will be erased.
pub fn create<T: Into<Box<dyn Directory>>>(self, dir: T) -> crate::Result<Index> {
self.create_avoid_monomorphization(dir.into())
}
fn create_avoid_monomorphization(self, dir: Box<dyn Directory>) -> crate::Result<Index> {
fn create<T: Into<Box<dyn Directory>>>(self, dir: T) -> crate::Result<Index> {
self.validate()?;
let dir = dir.into();
let directory = ManagedDirectory::wrap(dir)?;
save_new_metas(
self.get_expect_schema()?,
@@ -257,7 +255,7 @@ impl IndexBuilder {
)?;
let mut metas = IndexMeta::with_schema(self.get_expect_schema()?);
metas.index_settings = self.index_settings;
let mut 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);
index.set_fast_field_tokenizers(self.fast_field_tokenizer_manager);
Ok(index)
@@ -383,9 +381,9 @@ impl Index {
directory: ManagedDirectory,
metas: &IndexMeta,
inventory: SegmentMetaInventory,
) -> crate::Result<Index> {
) -> Index {
let schema = metas.schema.clone();
Ok(Index {
Index {
settings: metas.index_settings.clone(),
directory,
schema,
@@ -393,7 +391,7 @@ impl Index {
fast_field_tokenizers: TokenizerManager::default(),
executor: Executor::single_thread(),
inventory,
})
}
}
/// Setter for the tokenizer manager.
@@ -416,6 +414,36 @@ impl Index {
&self.fast_field_tokenizers
}
/// Get the tokenizer associated with a specific field.
pub fn tokenizer_for_field(&self, field: Field) -> crate::Result<TextAnalyzer> {
let field_entry = self.schema.get_field_entry(field);
let field_type = field_entry.field_type();
let tokenizer_manager: &TokenizerManager = self.tokenizers();
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()
)))
}
};
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.
///
/// See [`Index.reader_builder()`].
@@ -464,10 +492,7 @@ impl Index {
let segments = self.searchable_segments()?;
let fields_metadata: Vec<Vec<FieldMetadata>> = segments
.into_iter()
.map(|segment| {
let reader = TantivySegmentReader::open(&segment)?;
reader.fields_metadata()
})
.map(|segment| SegmentReader::open(&segment)?.fields_metadata())
.collect::<Result<_, _>>()?;
Ok(merge_field_meta_data(fields_metadata))
}
@@ -487,7 +512,8 @@ impl Index {
let directory = ManagedDirectory::wrap(directory)?;
let inventory = SegmentMetaInventory::default();
let metas = load_metas(&directory, &inventory)?;
Index::open_from_metas(directory, &metas, inventory)
let index = Index::open_from_metas(directory, &metas, inventory);
Ok(index)
}
/// Reads the index meta file from the directory.

65
src/index/index_meta.rs
View File

@@ -379,36 +379,13 @@ mod tests {
opstamp: 0u64,
payload: None,
};
let json_value: serde_json::Value =
serde_json::to_value(&index_metas).expect("serialization failed");
let json = serde_json::ser::to_string(&index_metas).expect("serialization failed");
assert_eq!(
&json_value,
&serde_json::json!(
{
"index_settings": {
"docstore_compression": "none",
"docstore_blocksize": 16384
},
"segments": [],
"schema": [
{
"name": "text",
"type": "text",
"options": {
"indexing": {
"record": "position",
"fieldnorms": true,
"tokenizer": "default"
},
"stored": false,
"fast": false
}
}
],
"opstamp": 0
})
json,
r#"{"index_settings":{"docstore_compression":"none","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_value(json_value).unwrap();
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);
@@ -435,37 +412,13 @@ mod tests {
opstamp: 0u64,
payload: None,
};
let json_value = serde_json::to_value(&index_metas).expect("serialization failed");
let json = serde_json::ser::to_string(&index_metas).expect("serialization failed");
assert_eq!(
&json_value,
&serde_json::json!(
{
"index_settings": {
"docstore_compression": "zstd(compression_level=4)",
"docstore_blocksize": 1000000
},
"segments": [],
"schema": [
{
"name": "text",
"type": "text",
"options": {
"indexing": {
"record": "position",
"fieldnorms": true,
"tokenizer": "default"
},
"stored": false,
"fast": false
}
}
],
"opstamp": 0
}
)
json,
r#"{"index_settings":{"docstore_compression":"zstd(compression_level=4)","docstore_blocksize":1000000},"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_value(json_value).unwrap();
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);

886
src/index/inverted_index_reader.rs
View File

@@ -1,12 +1,7 @@
use std::any::Any;
#[cfg(feature = "quickwit")]
use std::future::Future;
use std::io;
#[cfg(feature = "quickwit")]
use std::pin::Pin;
use common::json_path_writer::JSON_END_OF_PATH;
use common::{BinarySerializable, BitSet, ByteCount, OwnedBytes};
use common::{BinarySerializable, ByteCount};
#[cfg(feature = "quickwit")]
use futures_util::{FutureExt, StreamExt, TryStreamExt};
#[cfg(feature = "quickwit")]
@@ -15,252 +10,37 @@ use itertools::Itertools;
use tantivy_fst::automaton::{AlwaysMatch, Automaton};
use crate::directory::FileSlice;
use crate::docset::DocSet;
use crate::postings::{
load_postings_from_raw_data, Postings, RawPostingsData, SegmentPostings, TermInfo,
};
use crate::positions::PositionReader;
use crate::postings::{BlockSegmentPostings, SegmentPostings, TermInfo};
use crate::schema::{IndexRecordOption, Term, Type};
#[cfg(feature = "quickwit")]
pub use crate::termdict::BoxedAutomaton;
use crate::termdict::TermDictionary;
#[cfg(feature = "quickwit")]
pub type TermRangeBounds = (std::ops::Bound<Term>, std::ops::Bound<Term>);
/// Trait defining the contract for a dynamically dispatched inverted index reader.
pub trait DynInvertedIndexReader: Send + Sync {
/// Downcasts to the concrete reader type when possible.
fn as_any(&self) -> &dyn Any;
/// Returns the term info associated with the term.
fn get_term_info(&self, term: &Term) -> io::Result<Option<TermInfo>> {
self.terms().get(term.serialized_value_bytes())
}
/// Return the term dictionary datastructure.
fn terms(&self) -> &TermDictionary;
/// Return the fields and types encoded in the dictionary in lexicographic order.
/// Only valid on JSON fields.
///
/// Notice: This requires a full scan and therefore **very expensive**.
fn list_encoded_json_fields(&self) -> io::Result<Vec<InvertedIndexFieldSpace>>;
/// Returns the total number of tokens recorded for all documents
/// (including deleted documents).
fn total_num_tokens(&self) -> u64;
/// Returns the segment postings associated with the term, and with the given option,
/// or `None` if the term has never been encountered and indexed.
fn read_postings(
&self,
term: &Term,
option: IndexRecordOption,
) -> io::Result<Option<Box<dyn Postings>>> {
self.get_term_info(term)?
.map(move |term_info| self.read_postings_from_terminfo(&term_info, option))
.transpose()
}
/// Returns the postings for a given `TermInfo`.
fn read_postings_from_terminfo(
&self,
term_info: &TermInfo,
option: IndexRecordOption,
) -> io::Result<Box<dyn Postings>>;
/// Returns the number of documents containing the term.
fn doc_freq(&self, term: &Term) -> io::Result<u32>;
/// Returns the number of documents containing the term asynchronously.
#[cfg(feature = "quickwit")]
fn doc_freq_async<'a>(
&'a self,
term: &'a Term,
) -> Pin<Box<dyn Future<Output = io::Result<u32>> + Send + 'a>>;
/// Warmup fieldnorm readers for this inverted index field.
#[cfg(feature = "quickwit")]
fn warm_fieldnorms_readers<'a>(
&'a self,
) -> Pin<Box<dyn Future<Output = io::Result<()>> + Send + 'a>>;
/// Warmup the block postings for all terms.
///
/// Default implementation is a no-op.
#[cfg(feature = "quickwit")]
fn warm_postings_full<'a>(
&'a self,
_with_positions: bool,
) -> Pin<Box<dyn Future<Output = io::Result<()>> + Send + 'a>> {
Box::pin(async { Ok(()) })
}
/// Warmup a block postings given a `Term`.
///
/// Returns whether the term was found in the dictionary.
#[cfg(feature = "quickwit")]
fn warm_postings<'a>(
&'a self,
term: &'a Term,
with_positions: bool,
) -> Pin<Box<dyn Future<Output = io::Result<bool>> + Send + 'a>>;
/// Warmup block postings for terms in a range.
///
/// Returns whether at least one matching term was found.
#[cfg(feature = "quickwit")]
fn warm_postings_range<'a>(
&'a self,
terms: TermRangeBounds,
limit: Option<u64>,
with_positions: bool,
) -> Pin<Box<dyn Future<Output = io::Result<bool>> + Send + 'a>>;
/// Warmup block postings for terms matching an automaton.
///
/// Returns whether at least one matching term was found.
#[cfg(feature = "quickwit")]
fn warm_postings_automaton<'a>(
&'a self,
automaton: BoxedAutomaton,
) -> Pin<Box<dyn Future<Output = io::Result<bool>> + Send + 'a>>;
}
/// Trait defining the contract for a typed inverted index reader.
pub trait InvertedIndexReader: DynInvertedIndexReader {
/// The concrete postings type returned by this reader.
type Postings: Postings;
/// A lighter doc-id-only iterator returned when frequencies and positions are not needed.
type DocSet: DocSet;
/// Returns a posting object given a `term_info`.
fn read_postings_from_terminfo(
&self,
term_info: &TermInfo,
option: IndexRecordOption,
) -> io::Result<Self::Postings>;
/// Returns a doc-id-only iterator for the given term.
///
/// Always reads with `IndexRecordOption::Basic` — no frequency decoding,
/// no position reader.
fn read_docset_from_terminfo(&self, term_info: &TermInfo) -> io::Result<Self::DocSet>;
/// Fills a bitset with the doc ids for the given term.
fn fill_bitset_from_terminfo(
&self,
term_info: &TermInfo,
doc_bitset: &mut BitSet,
) -> io::Result<()> {
let mut docset = self.read_docset_from_terminfo(term_info)?;
docset.fill_bitset(doc_bitset);
Ok(())
}
}
impl InvertedIndexReader for dyn DynInvertedIndexReader + '_ {
type Postings = Box<dyn Postings>;
type DocSet = Box<dyn Postings>;
fn read_postings_from_terminfo(
&self,
term_info: &TermInfo,
option: IndexRecordOption,
) -> io::Result<Self::Postings> {
DynInvertedIndexReader::read_postings_from_terminfo(self, term_info, option)
}
fn read_docset_from_terminfo(&self, term_info: &TermInfo) -> io::Result<Self::DocSet> {
DynInvertedIndexReader::read_postings_from_terminfo(
self,
term_info,
IndexRecordOption::Basic,
)
}
}
/// Attempts to downcast a `DynInvertedIndexReader` to tantivy's concrete
/// `TantivyInvertedIndexReader` before falling back to the dynamic path.
///
/// The body is compiled twice: once with the concrete reader (yielding typed
/// postings such as `SegmentPostings`) and once with the dynamic reader
/// (yielding `Box<dyn Postings>`). The body must therefore be generic
/// enough to work with both postings types.
///
/// # Example
///
/// ```ignore
/// let postings = try_downcast_and_call!(inverted_index.as_ref(), |reader| {
/// let postings = reader.read_postings_from_terminfo(&term_info, option)?;
/// io::Result::Ok(Box::new(postings) as Box<dyn Postings>)
/// })?;
/// ```
#[macro_export]
macro_rules! try_downcast_and_call {
($reader:expr, |$reader_var:ident| $body:expr) => {{
#[allow(unused_imports)]
use $crate::index::InvertedIndexReader as _;
let __dyn_reader: &dyn $crate::index::DynInvertedIndexReader = $reader;
if let Some($reader_var) = __dyn_reader
.as_any()
.downcast_ref::<$crate::index::TantivyInvertedIndexReader>()
{
$body
} else {
let $reader_var = __dyn_reader;
$body
}
}};
}
pub(crate) fn load_postings_from_terminfo(
reader: &dyn DynInvertedIndexReader,
term_info: &TermInfo,
option: IndexRecordOption,
) -> io::Result<Box<dyn Postings>> {
try_downcast_and_call!(reader, |reader| {
let postings = InvertedIndexReader::read_postings_from_terminfo(reader, term_info, option)?;
Ok(Box::new(postings) as Box<dyn Postings>)
})
}
/// Tantivy's default inverted index reader implementation.
///
/// The inverted index reader is in charge of accessing
/// the inverted index associated with a specific field.
///
/// # Note
///
/// It is safe to delete the segment associated with
/// an `InvertedIndexReader` implementation. As long as it is open,
/// an `InvertedIndexReader`. As long as it is open,
/// the [`FileSlice`] it is relying on should
/// stay available.
///
/// `TantivyInvertedIndexReader` instances are created by calling
/// `InvertedIndexReader` are created by calling
/// [`SegmentReader::inverted_index()`](crate::SegmentReader::inverted_index).
pub struct TantivyInvertedIndexReader {
pub struct InvertedIndexReader {
termdict: TermDictionary,
postings_file_slice: FileSlice,
positions_file_slice: FileSlice,
#[cfg_attr(not(feature = "quickwit"), allow(dead_code))]
fieldnorms_file_slice: FileSlice,
record_option: IndexRecordOption,
total_num_tokens: u64,
}
/// Object that records the amount of space used by a field in an inverted index.
pub struct InvertedIndexFieldSpace {
/// Field name as encoded in the term dictionary.
pub(crate) struct InvertedIndexFieldSpace {
pub field_name: String,
/// Value type for the encoded field.
pub field_type: Type,
/// Total bytes used by postings for this field.
pub postings_size: ByteCount,
/// Total bytes used by positions for this field.
pub positions_size: ByteCount,
/// Number of terms in the field.
pub num_terms: u64,
}
@@ -282,82 +62,52 @@ impl InvertedIndexFieldSpace {
}
}
impl TantivyInvertedIndexReader {
/// Returns the raw postings bytes and metadata for a term.
pub fn read_raw_postings_data(
&self,
term_info: &TermInfo,
option: IndexRecordOption,
) -> io::Result<RawPostingsData> {
let effective_option = option.downgrade(self.record_option);
let postings_data = self
.postings_file_slice
.slice(term_info.postings_range.clone())
.read_bytes()?;
let positions_data: Option<OwnedBytes> = if effective_option.has_positions() {
let positions_data = self
.positions_file_slice
.slice(term_info.positions_range.clone())
.read_bytes()?;
Some(positions_data)
} else {
None
};
Ok(RawPostingsData {
postings_data,
positions_data,
record_option: self.record_option,
effective_option,
})
}
/// Opens an inverted index reader from already-loaded term/postings/positions slices.
///
/// The first 8 bytes of `postings_file_slice` are expected to contain
/// the serialized total token count.
pub fn new(
impl InvertedIndexReader {
pub(crate) fn new(
termdict: TermDictionary,
postings_file_slice: FileSlice,
positions_file_slice: FileSlice,
fieldnorms_file_slice: FileSlice,
record_option: IndexRecordOption,
) -> io::Result<TantivyInvertedIndexReader> {
) -> io::Result<InvertedIndexReader> {
let (total_num_tokens_slice, postings_body) = postings_file_slice.split(8);
let total_num_tokens = u64::deserialize(&mut total_num_tokens_slice.read_bytes()?)?;
Ok(TantivyInvertedIndexReader {
Ok(InvertedIndexReader {
termdict,
postings_file_slice: postings_body,
positions_file_slice,
fieldnorms_file_slice,
record_option,
total_num_tokens,
})
}
/// Creates an empty `TantivyInvertedIndexReader` object, which
/// Creates an empty `InvertedIndexReader` object, which
/// contains no terms at all.
pub fn empty(record_option: IndexRecordOption) -> TantivyInvertedIndexReader {
TantivyInvertedIndexReader {
pub fn empty(record_option: IndexRecordOption) -> InvertedIndexReader {
InvertedIndexReader {
termdict: TermDictionary::empty(),
postings_file_slice: FileSlice::empty(),
positions_file_slice: FileSlice::empty(),
fieldnorms_file_slice: FileSlice::empty(),
record_option,
total_num_tokens: 0u64,
}
}
}
impl DynInvertedIndexReader for TantivyInvertedIndexReader {
fn as_any(&self) -> &dyn Any {
self
/// Returns the term info associated with the term.
pub fn get_term_info(&self, term: &Term) -> io::Result<Option<TermInfo>> {
self.termdict.get(term.serialized_value_bytes())
}
fn terms(&self) -> &TermDictionary {
/// Return the term dictionary datastructure.
pub fn terms(&self) -> &TermDictionary {
&self.termdict
}
fn list_encoded_json_fields(&self) -> io::Result<Vec<InvertedIndexFieldSpace>> {
/// Return the fields and types encoded in the dictionary in lexicographic order.
/// Only valid on JSON fields.
///
/// Notice: This requires a full scan and therefore **very expensive**.
/// TODO: Move to sstable to use the index.
pub(crate) fn list_encoded_json_fields(&self) -> io::Result<Vec<InvertedIndexFieldSpace>> {
let mut stream = self.termdict.stream()?;
let mut fields: Vec<InvertedIndexFieldSpace> = Vec::new();
@@ -410,325 +160,136 @@ impl DynInvertedIndexReader for TantivyInvertedIndexReader {
Ok(fields)
}
fn read_postings_from_terminfo(
/// Resets the block segment to another position of the postings
/// file.
///
/// This is useful for enumerating through a list of terms,
/// and consuming the associated posting lists while avoiding
/// reallocating a [`BlockSegmentPostings`].
///
/// # Warning
///
/// This does not reset the positions list.
pub fn reset_block_postings_from_terminfo(
&self,
term_info: &TermInfo,
block_postings: &mut BlockSegmentPostings,
) -> io::Result<()> {
let postings_slice = self
.postings_file_slice
.slice(term_info.postings_range.clone());
let postings_bytes = postings_slice.read_bytes()?;
block_postings.reset(term_info.doc_freq, postings_bytes)?;
Ok(())
}
/// Returns a block postings given a `Term`.
/// This method is for an advanced usage only.
///
/// Most users should prefer using [`Self::read_postings()`] instead.
pub fn read_block_postings(
&self,
term: &Term,
option: IndexRecordOption,
) -> io::Result<Option<BlockSegmentPostings>> {
self.get_term_info(term)?
.map(move |term_info| self.read_block_postings_from_terminfo(&term_info, option))
.transpose()
}
/// Returns a block postings given a `term_info`.
/// This method is for an advanced usage only.
///
/// Most users should prefer using [`Self::read_postings()`] instead.
pub fn read_block_postings_from_terminfo(
&self,
term_info: &TermInfo,
requested_option: IndexRecordOption,
) -> io::Result<BlockSegmentPostings> {
let postings_data = self
.postings_file_slice
.slice(term_info.postings_range.clone());
BlockSegmentPostings::open(
term_info.doc_freq,
postings_data,
self.record_option,
requested_option,
)
}
/// Returns a posting object given a `term_info`.
/// This method is for an advanced usage only.
///
/// Most users should prefer using [`Self::read_postings()`] instead.
pub fn read_postings_from_terminfo(
&self,
term_info: &TermInfo,
option: IndexRecordOption,
) -> io::Result<Box<dyn Postings>> {
let postings_data = self.read_raw_postings_data(term_info, option)?;
let postings = load_postings_from_raw_data(term_info.doc_freq, postings_data)?;
Ok(Box::new(postings))
) -> io::Result<SegmentPostings> {
let option = option.downgrade(self.record_option);
let block_postings = self.read_block_postings_from_terminfo(term_info, option)?;
let position_reader = {
if option.has_positions() {
let positions_data = self
.positions_file_slice
.read_bytes_slice(term_info.positions_range.clone())?;
let position_reader = PositionReader::open(positions_data)?;
Some(position_reader)
} else {
None
}
};
Ok(SegmentPostings::from_block_postings(
block_postings,
position_reader,
))
}
fn total_num_tokens(&self) -> u64 {
/// Returns the total number of tokens recorded for all documents
/// (including deleted documents).
pub fn total_num_tokens(&self) -> u64 {
self.total_num_tokens
}
fn doc_freq(&self, term: &Term) -> io::Result<u32> {
/// Returns the segment postings associated with the term, and with the given option,
/// or `None` if the term has never been encountered and indexed.
///
/// If the field was not indexed with the indexing options that cover
/// the requested options, the returned [`SegmentPostings`] the method does not fail
/// and returns a `SegmentPostings` with as much information as possible.
///
/// For instance, requesting [`IndexRecordOption::WithFreqs`] for a
/// [`TextOptions`](crate::schema::TextOptions) that does not index position
/// will return a [`SegmentPostings`] with `DocId`s and frequencies.
pub fn read_postings(
&self,
term: &Term,
option: IndexRecordOption,
) -> io::Result<Option<SegmentPostings>> {
self.get_term_info(term)?
.map(move |term_info| self.read_postings_from_terminfo(&term_info, option))
.transpose()
}
/// Returns the number of documents containing the term.
pub fn doc_freq(&self, term: &Term) -> io::Result<u32> {
Ok(self
.get_term_info(term)?
.map(|term_info| term_info.doc_freq)
.unwrap_or(0u32))
}
#[cfg(feature = "quickwit")]
fn doc_freq_async<'a>(
&'a self,
term: &'a Term,
) -> Pin<Box<dyn Future<Output = io::Result<u32>> + Send + 'a>> {
Box::pin(async move {
Ok(self
.get_term_info_async(term)
.await?
.map(|term_info| term_info.doc_freq)
.unwrap_or(0u32))
})
}
#[cfg(feature = "quickwit")]
fn warm_fieldnorms_readers<'a>(
&'a self,
) -> Pin<Box<dyn Future<Output = io::Result<()>> + Send + 'a>> {
Box::pin(async move {
self.fieldnorms_file_slice.read_bytes_async().await?;
Ok(())
})
}
#[cfg(feature = "quickwit")]
fn warm_postings_full<'a>(
&'a self,
with_positions: bool,
) -> Pin<Box<dyn Future<Output = io::Result<()>> + Send + 'a>> {
Box::pin(async move {
self.postings_file_slice.read_bytes_async().await?;
if with_positions {
self.positions_file_slice.read_bytes_async().await?;
}
Ok(())
})
}
#[cfg(feature = "quickwit")]
fn warm_postings<'a>(
&'a self,
term: &'a Term,
with_positions: bool,
) -> Pin<Box<dyn Future<Output = io::Result<bool>> + Send + 'a>> {
Box::pin(async move {
let term_info_opt: Option<TermInfo> = self.get_term_info_async(term).await?;
if let Some(term_info) = term_info_opt {
let postings = self
.postings_file_slice
.read_bytes_slice_async(term_info.postings_range.clone());
if with_positions {
let positions = self
.positions_file_slice
.read_bytes_slice_async(term_info.positions_range.clone());
futures_util::future::try_join(postings, positions).await?;
} else {
postings.await?;
}
Ok(true)
} else {
Ok(false)
}
})
}
#[cfg(feature = "quickwit")]
fn warm_postings_range<'a>(
&'a self,
terms: TermRangeBounds,
limit: Option<u64>,
with_positions: bool,
) -> Pin<Box<dyn Future<Output = io::Result<bool>> + Send + 'a>> {
Box::pin(async move {
let mut term_info = self
.get_term_range_async(terms, AlwaysMatch, limit, 0)
.await?;
let Some(first_terminfo) = term_info.next() else {
// no key matches, nothing more to load
return Ok(false);
};
let last_terminfo = term_info.last().unwrap_or_else(|| first_terminfo.clone());
let postings_range =
first_terminfo.postings_range.start..last_terminfo.postings_range.end;
let positions_range =
first_terminfo.positions_range.start..last_terminfo.positions_range.end;
let postings = self
.postings_file_slice
.read_bytes_slice_async(postings_range);
if with_positions {
let positions = self
.positions_file_slice
.read_bytes_slice_async(positions_range);
futures_util::future::try_join(postings, positions).await?;
} else {
postings.await?;
}
Ok(true)
})
}
#[cfg(feature = "quickwit")]
fn warm_postings_automaton<'a>(
&'a self,
automaton: BoxedAutomaton,
) -> Pin<Box<dyn Future<Output = io::Result<bool>> + Send + 'a>> {
Box::pin(async move {
// merge holes under 4MiB, that's how many bytes we can hope to receive during a TTFB
// from S3 (~80MiB/s, and 50ms latency)
const MERGE_HOLES_UNDER_BYTES: usize = (80 * 1024 * 1024 * 50) / 1000;
// Trigger async prefetch of relevant termdict blocks.
let _term_info_iter = self
.get_term_range_async(
(std::ops::Bound::Unbounded, std::ops::Bound::Unbounded),
automaton.clone(),
None,
MERGE_HOLES_UNDER_BYTES,
)
.await?;
drop(_term_info_iter);
// Build a 2nd stream without merged holes so we only scan matching blocks.
// This assumes the storage layer caches data fetched by the first pass.
let mut stream = self.termdict.search(automaton).into_stream()?;
let posting_ranges_iter =
std::iter::from_fn(move || stream.next().map(|(_k, v)| v.postings_range.clone()));
let merged_posting_ranges: Vec<std::ops::Range<usize>> = posting_ranges_iter
.coalesce(|range1, range2| {
if range1.end + MERGE_HOLES_UNDER_BYTES >= range2.start {
Ok(range1.start..range2.end)
} else {
Err((range1, range2))
}
})
.collect();
if merged_posting_ranges.is_empty() {
return Ok(false);
}
let slices_downloaded: Vec<()> =
futures_util::stream::iter(merged_posting_ranges.into_iter())
.map(|posting_slice| {
self.postings_file_slice
.read_bytes_slice_async(posting_slice)
.map(|result| result.map(|_slice| ()))
})
.buffer_unordered(5)
.try_collect()
.await?;
Ok(!slices_downloaded.is_empty())
})
}
}
impl InvertedIndexReader for TantivyInvertedIndexReader {
type Postings = SegmentPostings;
type DocSet = SegmentPostings;
#[inline]
fn read_postings_from_terminfo(
&self,
term_info: &TermInfo,
option: IndexRecordOption,
) -> io::Result<Self::Postings> {
let postings_data = self.read_raw_postings_data(term_info, option)?;
load_postings_from_raw_data(term_info.doc_freq, postings_data)
}
#[inline]
fn read_docset_from_terminfo(&self, term_info: &TermInfo) -> io::Result<Self::DocSet> {
let postings_data = self.read_raw_postings_data(term_info, IndexRecordOption::Basic)?;
load_postings_from_raw_data(term_info.doc_freq, postings_data)
}
}
#[cfg(test)]
mod tests {
use super::*;
struct OnlyDynReader {
termdict: TermDictionary,
}
impl Default for OnlyDynReader {
fn default() -> Self {
Self {
termdict: TermDictionary::empty(),
}
}
}
impl DynInvertedIndexReader for OnlyDynReader {
fn as_any(&self) -> &dyn Any {
self
}
fn terms(&self) -> &TermDictionary {
&self.termdict
}
fn list_encoded_json_fields(&self) -> io::Result<Vec<InvertedIndexFieldSpace>> {
Ok(Vec::new())
}
fn read_postings_from_terminfo(
&self,
_term_info: &TermInfo,
_option: IndexRecordOption,
) -> io::Result<Box<dyn Postings>> {
unreachable!("not used in downcast helper tests")
}
fn total_num_tokens(&self) -> u64 {
0
}
fn doc_freq(&self, _term: &Term) -> io::Result<u32> {
Ok(0)
}
#[cfg(feature = "quickwit")]
fn doc_freq_async<'a>(
&'a self,
_term: &'a Term,
) -> Pin<Box<dyn Future<Output = io::Result<u32>> + Send + 'a>> {
Box::pin(async { Ok(0) })
}
#[cfg(feature = "quickwit")]
fn warm_fieldnorms_readers<'a>(
&'a self,
) -> Pin<Box<dyn Future<Output = io::Result<()>> + Send + 'a>> {
Box::pin(async { Ok(()) })
}
#[cfg(feature = "quickwit")]
fn warm_postings<'a>(
&'a self,
_term: &'a Term,
_with_positions: bool,
) -> Pin<Box<dyn Future<Output = io::Result<bool>> + Send + 'a>> {
Box::pin(async { Ok(false) })
}
#[cfg(feature = "quickwit")]
fn warm_postings_range<'a>(
&'a self,
_terms: TermRangeBounds,
_limit: Option<u64>,
_with_positions: bool,
) -> Pin<Box<dyn Future<Output = io::Result<bool>> + Send + 'a>> {
Box::pin(async { Ok(false) })
}
#[cfg(feature = "quickwit")]
fn warm_postings_automaton<'a>(
&'a self,
_automaton: BoxedAutomaton,
) -> Pin<Box<dyn Future<Output = io::Result<bool>> + Send + 'a>> {
Box::pin(async { Ok(false) })
}
}
#[test]
fn try_downcast_and_call_uses_tantivy_reader() {
let reader = TantivyInvertedIndexReader::empty(IndexRecordOption::Basic);
let dyn_reader: &dyn DynInvertedIndexReader = &reader;
let used_concrete = try_downcast_and_call!(dyn_reader, |r| {
r.as_any().is::<TantivyInvertedIndexReader>()
});
assert!(used_concrete);
}
#[test]
fn try_downcast_and_call_uses_dynamic_fallback_for_other_readers() {
let reader = OnlyDynReader::default();
let dyn_reader: &dyn DynInvertedIndexReader = &reader;
let used_concrete = try_downcast_and_call!(dyn_reader, |r| {
r.as_any().is::<TantivyInvertedIndexReader>()
});
assert!(!used_concrete);
}
}
#[cfg(feature = "quickwit")]
impl TantivyInvertedIndexReader {
impl InvertedIndexReader {
pub(crate) async fn get_term_info_async(&self, term: &Term) -> io::Result<Option<TermInfo>> {
self.termdict.get_async(term.serialized_value_bytes()).await
}
async fn get_term_range_async<'a, A: Automaton + 'a>(
&'a self,
terms: TermRangeBounds,
terms: impl std::ops::RangeBounds<Term>,
automaton: A,
limit: Option<u64>,
merge_holes_under_bytes: usize,
@@ -736,17 +297,17 @@ impl TantivyInvertedIndexReader {
where
A::State: Clone,
{
use std::ops::Bound;
let range_builder = self.termdict.search(automaton);
let (start_bound, end_bound) = terms;
let range_builder = match start_bound {
std::ops::Bound::Included(bound) => range_builder.ge(bound.serialized_value_bytes()),
std::ops::Bound::Excluded(bound) => range_builder.gt(bound.serialized_value_bytes()),
std::ops::Bound::Unbounded => range_builder,
let range_builder = match terms.start_bound() {
Bound::Included(bound) => range_builder.ge(bound.serialized_value_bytes()),
Bound::Excluded(bound) => range_builder.gt(bound.serialized_value_bytes()),
Bound::Unbounded => range_builder,
};
let range_builder = match end_bound {
std::ops::Bound::Included(bound) => range_builder.le(bound.serialized_value_bytes()),
std::ops::Bound::Excluded(bound) => range_builder.lt(bound.serialized_value_bytes()),
std::ops::Bound::Unbounded => range_builder,
let range_builder = match terms.end_bound() {
Bound::Included(bound) => range_builder.le(bound.serialized_value_bytes()),
Bound::Excluded(bound) => range_builder.lt(bound.serialized_value_bytes()),
Bound::Unbounded => range_builder,
};
let range_builder = if let Some(limit) = limit {
range_builder.limit(limit)
@@ -767,4 +328,167 @@ impl TantivyInvertedIndexReader {
Ok(iter)
}
/// Warmup a block postings given a `Term`.
/// This method is for an advanced usage only.
///
/// returns a boolean, whether the term was found in the dictionary
pub async fn warm_postings(&self, term: &Term, with_positions: bool) -> io::Result<bool> {
let term_info_opt: Option<TermInfo> = self.get_term_info_async(term).await?;
if let Some(term_info) = term_info_opt {
let postings = self
.postings_file_slice
.read_bytes_slice_async(term_info.postings_range.clone());
if with_positions {
let positions = self
.positions_file_slice
.read_bytes_slice_async(term_info.positions_range.clone());
futures_util::future::try_join(postings, positions).await?;
} else {
postings.await?;
}
Ok(true)
} else {
Ok(false)
}
}
/// Warmup a block postings given a range of `Term`s.
/// This method is for an advanced usage only.
///
/// returns a boolean, whether a term matching the range was found in the dictionary
pub async fn warm_postings_range(
&self,
terms: impl std::ops::RangeBounds<Term>,
limit: Option<u64>,
with_positions: bool,
) -> io::Result<bool> {
let mut term_info = self
.get_term_range_async(terms, AlwaysMatch, limit, 0)
.await?;
let Some(first_terminfo) = term_info.next() else {
// no key matches, nothing more to load
return Ok(false);
};
let last_terminfo = term_info.last().unwrap_or_else(|| first_terminfo.clone());
let postings_range = first_terminfo.postings_range.start..last_terminfo.postings_range.end;
let positions_range =
first_terminfo.positions_range.start..last_terminfo.positions_range.end;
let postings = self
.postings_file_slice
.read_bytes_slice_async(postings_range);
if with_positions {
let positions = self
.positions_file_slice
.read_bytes_slice_async(positions_range);
futures_util::future::try_join(postings, positions).await?;
} else {
postings.await?;
}
Ok(true)
}
/// Warmup a block postings given a range of `Term`s.
/// This method is for an advanced usage only.
///
/// returns a boolean, whether a term matching the range was found in the dictionary
pub async fn warm_postings_automaton<
A: Automaton + Clone + Send + 'static,
E: FnOnce(Box<dyn FnOnce() -> io::Result<()> + Send>) -> F,
F: std::future::Future<Output = io::Result<()>>,
>(
&self,
automaton: A,
// with_positions: bool, at the moment we have no use for it, and supporting it would add
// complexity to the coalesce
executor: E,
) -> io::Result<bool>
where
A::State: Clone,
{
// merge holes under 4MiB, that's how many bytes we can hope to receive during a TTFB from
// S3 (~80MiB/s, and 50ms latency)
const MERGE_HOLES_UNDER_BYTES: usize = (80 * 1024 * 1024 * 50) / 1000;
// we build a first iterator to download everything. Simply calling the function already
// download everything we need from the sstable, but doesn't start iterating over it.
let _term_info_iter = self
.get_term_range_async(.., automaton.clone(), None, MERGE_HOLES_UNDER_BYTES)
.await?;
let (sender, posting_ranges_to_load_stream) = futures_channel::mpsc::unbounded();
let termdict = self.termdict.clone();
let cpu_bound_task = move || {
// then we build a 2nd iterator, this one with no holes, so we don't go through blocks
// we can't match.
// This makes the assumption there is a caching layer below us, which gives sync read
// for free after the initial async access. This might not always be true, but is in
// Quickwit.
// We build things from this closure otherwise we get into lifetime issues that can only
// be solved with self referential strucs. Returning an io::Result from here is a bit
// more leaky abstraction-wise, but a lot better than the alternative
let mut stream = termdict.search(automaton).into_stream()?;
// we could do without an iterator, but this allows us access to coalesce which simplify
// things
let posting_ranges_iter =
std::iter::from_fn(move || stream.next().map(|(_k, v)| v.postings_range.clone()));
let merged_posting_ranges_iter = posting_ranges_iter.coalesce(|range1, range2| {
if range1.end + MERGE_HOLES_UNDER_BYTES >= range2.start {
Ok(range1.start..range2.end)
} else {
Err((range1, range2))
}
});
for posting_range in merged_posting_ranges_iter {
if let Err(_) = sender.unbounded_send(posting_range) {
// this should happen only when search is cancelled
return Err(io::Error::other("failed to send posting range back"));
}
}
Ok(())
};
let task_handle = executor(Box::new(cpu_bound_task));
let posting_downloader = posting_ranges_to_load_stream
.map(|posting_slice| {
self.postings_file_slice
.read_bytes_slice_async(posting_slice)
.map(|result| result.map(|_slice| ()))
})
.buffer_unordered(5)
.try_collect::<Vec<()>>();
let (_, slices_downloaded) =
futures_util::future::try_join(task_handle, posting_downloader).await?;
Ok(!slices_downloaded.is_empty())
}
/// Warmup the block postings for all terms.
/// This method is for an advanced usage only.
///
/// If you know which terms to pre-load, prefer using [`Self::warm_postings`] or
/// [`Self::warm_postings`] instead.
pub async fn warm_postings_full(&self, with_positions: bool) -> io::Result<()> {
self.postings_file_slice.read_bytes_async().await?;
if with_positions {
self.positions_file_slice.read_bytes_async().await?;
}
Ok(())
}
/// Returns the number of documents containing the term asynchronously.
pub async fn doc_freq_async(&self, term: &Term) -> io::Result<u32> {
Ok(self
.get_term_info_async(term)
.await?
.map(|term_info| term_info.doc_freq)
.unwrap_or(0u32))
}
}

8
src/index/mod.rs
View File

@@ -13,12 +13,8 @@ mod segment_reader;
pub use self::index::{Index, IndexBuilder};
pub(crate) use self::index_meta::SegmentMetaInventory;
pub use self::index_meta::{IndexMeta, IndexSettings, Order, SegmentMeta};
pub(crate) use self::inverted_index_reader::load_postings_from_terminfo;
pub use self::inverted_index_reader::{
DynInvertedIndexReader, InvertedIndexFieldSpace, InvertedIndexReader,
TantivyInvertedIndexReader,
};
pub use self::inverted_index_reader::InvertedIndexReader;
pub use self::segment::Segment;
pub use self::segment_component::SegmentComponent;
pub use self::segment_id::SegmentId;
pub use self::segment_reader::{FieldMetadata, SegmentReader, TantivySegmentReader};
pub use self::segment_reader::{FieldMetadata, SegmentReader};

2
src/index/segment.rs
View File

@@ -16,7 +16,7 @@ pub struct Segment {
}
impl fmt::Debug for Segment {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
write!(f, "Segment({:?})", self.id().uuid_string())
}
}

2
src/index/segment_id.rs
View File

@@ -44,7 +44,7 @@ fn create_uuid() -> Uuid {
}
impl SegmentId {
/// Generates a new random `SegmentId`.
#[doc(hidden)]
pub fn generate_random() -> SegmentId {
SegmentId(create_uuid())
}

413
src/index/segment_reader.rs
View File

@@ -6,90 +6,19 @@ use common::{ByteCount, HasLen};
use fnv::FnvHashMap;
use itertools::Itertools;
use crate::directory::{CompositeFile, Directory, FileSlice};
use crate::directory::error::OpenReadError;
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::index::{
DynInvertedIndexReader, Segment, SegmentComponent, SegmentId, SegmentMeta,
TantivyInvertedIndexReader,
};
use crate::index::{InvertedIndexReader, Segment, SegmentComponent, SegmentId};
use crate::json_utils::json_path_sep_to_dot;
use crate::schema::{Field, IndexRecordOption, Schema, Type};
use crate::space_usage::SegmentSpaceUsage;
use crate::store::{StoreReader, TantivyStoreReader};
use crate::store::StoreReader;
use crate::termdict::TermDictionary;
use crate::{DocId, Opstamp};
/// Trait defining the contract for a segment reader.
pub trait SegmentReader: Send + Sync {
/// Returns the highest document id ever attributed in this segment + 1.
fn max_doc(&self) -> DocId;
/// Returns the number of alive documents. Deleted documents are not counted.
fn num_docs(&self) -> DocId;
/// Returns the schema of the index this segment belongs to.
fn schema(&self) -> &Schema;
/// Return the number of documents that have been deleted in the segment.
fn num_deleted_docs(&self) -> DocId;
/// Returns true if some of the documents of the segment have been deleted.
fn has_deletes(&self) -> bool;
/// Accessor to a segment's fast field reader given a field.
fn fast_fields(&self) -> &FastFieldReaders;
/// Accessor to the `FacetReader` associated with a given `Field`.
fn facet_reader(&self, field_name: &str) -> crate::Result<FacetReader> {
let field = self.schema().get_field(field_name)?;
let field_entry = self.schema().get_field_entry(field);
if field_entry.field_type().value_type() != Type::Facet {
return Err(crate::TantivyError::SchemaError(format!(
"`{field_name}` is not a facet field.`"
)));
}
let Some(facet_column) = self.fast_fields().str(field_name)? else {
panic!("Facet Field `{field_name}` is missing. This should not happen");
};
Ok(FacetReader::new(facet_column))
}
/// Accessor to the segment's `Field norms`'s reader.
fn get_fieldnorms_reader(&self, field: Field) -> crate::Result<FieldNormReader>;
/// Accessor to the segment's [`StoreReader`](crate::store::StoreReader).
fn get_store_reader(&self, cache_num_blocks: usize) -> io::Result<Box<dyn StoreReader>>;
/// Returns a field reader associated with the field given in argument.
fn inverted_index(&self, field: Field) -> crate::Result<Arc<dyn DynInvertedIndexReader>>;
/// Returns the list of fields that have been indexed in the segment.
fn fields_metadata(&self) -> crate::Result<Vec<FieldMetadata>>;
/// Returns the segment id.
fn segment_id(&self) -> SegmentId;
/// Returns the delete opstamp.
fn delete_opstamp(&self) -> Option<Opstamp>;
/// Returns the bitset representing the alive `DocId`s.
fn alive_bitset(&self) -> Option<&AliveBitSet>;
/// Returns true if the `doc` is marked as deleted.
fn is_deleted(&self, doc: DocId) -> bool;
/// Returns an iterator that will iterate over the alive document ids.
fn doc_ids_alive(&self) -> Box<dyn Iterator<Item = DocId> + Send + '_>;
/// Summarize total space usage of this segment.
fn space_usage(&self) -> io::Result<SegmentSpaceUsage>;
/// Clones this reader into a shared trait object.
fn clone_arc(&self) -> Arc<dyn SegmentReader>;
}
/// Entry point to access all of the datastructures of the `Segment`
///
/// - term dictionary
@@ -101,8 +30,8 @@ pub trait SegmentReader: Send + Sync {
/// The segment reader has a very low memory footprint,
/// as close to all of the memory data is mmapped.
#[derive(Clone)]
pub struct TantivySegmentReader {
inv_idx_reader_cache: Arc<RwLock<HashMap<Field, Arc<dyn DynInvertedIndexReader>>>>,
pub struct SegmentReader {
inv_idx_reader_cache: Arc<RwLock<HashMap<Field, Arc<InvertedIndexReader>>>>,
segment_id: SegmentId,
delete_opstamp: Option<Opstamp>,
@@ -121,123 +50,73 @@ pub struct TantivySegmentReader {
schema: Schema,
}
impl TantivySegmentReader {
/// Open a new segment for reading.
pub fn open(segment: &Segment) -> crate::Result<Arc<dyn 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<Arc<dyn SegmentReader>> {
let reader = Self::open_with_custom_alive_set_from_directory(
segment.index().directory(),
segment.meta(),
segment.schema(),
custom_bitset,
)?;
Ok(Arc::new(reader))
}
pub(crate) fn open_with_custom_alive_set_from_directory(
directory: &dyn Directory,
segment_meta: &SegmentMeta,
schema: Schema,
custom_bitset: Option<AliveBitSet>,
) -> crate::Result<TantivySegmentReader> {
let termdict_file =
directory.open_read(&segment_meta.relative_path(SegmentComponent::Terms))?;
let termdict_composite = CompositeFile::open(&termdict_file)?;
let store_file =
directory.open_read(&segment_meta.relative_path(SegmentComponent::Store))?;
crate::fail_point!("SegmentReader::open#middle");
let postings_file =
directory.open_read(&segment_meta.relative_path(SegmentComponent::Postings))?;
let postings_composite = CompositeFile::open(&postings_file)?;
let positions_composite = {
if let Ok(positions_file) =
directory.open_read(&segment_meta.relative_path(SegmentComponent::Positions))
{
CompositeFile::open(&positions_file)?
} else {
CompositeFile::empty()
}
};
let fast_fields_data =
directory.open_read(&segment_meta.relative_path(SegmentComponent::FastFields))?;
let fast_fields_readers = FastFieldReaders::open(fast_fields_data, schema.clone())?;
let fieldnorm_data =
directory.open_read(&segment_meta.relative_path(SegmentComponent::FieldNorms))?;
let fieldnorm_readers = FieldNormReaders::open(fieldnorm_data)?;
let original_bitset = if segment_meta.has_deletes() {
let alive_doc_file_slice =
directory.open_read(&segment_meta.relative_path(SegmentComponent::Delete))?;
let alive_doc_data = alive_doc_file_slice.read_bytes()?;
Some(AliveBitSet::open(alive_doc_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(TantivySegmentReader {
inv_idx_reader_cache: Default::default(),
num_docs,
max_doc,
termdict_composite,
postings_composite,
fast_fields_readers,
fieldnorm_readers,
segment_id: segment_meta.id(),
delete_opstamp: segment_meta.delete_opstamp(),
store_file,
alive_bitset_opt,
positions_composite,
schema,
})
}
}
impl SegmentReader for TantivySegmentReader {
fn max_doc(&self) -> DocId {
impl SegmentReader {
/// Returns the highest document id ever attributed in
/// this segment + 1.
pub fn max_doc(&self) -> DocId {
self.max_doc
}
fn num_docs(&self) -> DocId {
/// Returns the number of alive documents.
/// Deleted documents are not counted.
pub fn num_docs(&self) -> DocId {
self.num_docs
}
fn schema(&self) -> &Schema {
/// Returns the schema of the index this segment belongs to.
pub fn schema(&self) -> &Schema {
&self.schema
}
fn num_deleted_docs(&self) -> DocId {
/// Return the number of documents that have been
/// deleted in the segment.
pub fn num_deleted_docs(&self) -> DocId {
self.max_doc - self.num_docs
}
fn has_deletes(&self) -> bool {
self.num_docs != self.max_doc
/// Returns true if some of the documents of the segment have been deleted.
pub fn has_deletes(&self) -> bool {
self.num_deleted_docs() > 0
}
fn fast_fields(&self) -> &FastFieldReaders {
/// Accessor to a segment's fast field reader given a field.
///
/// Returns the u64 fast value reader if the field
/// is a u64 field indexed as "fast".
///
/// Return a FastFieldNotAvailableError if the field is not
/// declared as a fast field in the schema.
///
/// # Panics
/// May panic if the index is corrupted.
pub fn fast_fields(&self) -> &FastFieldReaders {
&self.fast_fields_readers
}
fn get_fieldnorms_reader(&self, field: Field) -> crate::Result<FieldNormReader> {
/// Accessor to the `FacetReader` associated with a given `Field`.
pub fn facet_reader(&self, field_name: &str) -> crate::Result<FacetReader> {
let schema = self.schema();
let field = schema.get_field(field_name)?;
let field_entry = schema.get_field_entry(field);
if field_entry.field_type().value_type() != Type::Facet {
return Err(crate::TantivyError::SchemaError(format!(
"`{field_name}` is not a facet field.`"
)));
}
let Some(facet_column) = self.fast_fields().str(field_name)? else {
panic!("Facet Field `{field_name}` is missing. This should not happen");
};
Ok(FacetReader::new(facet_column))
}
/// Accessor to the segment's `Field norms`'s reader.
///
/// Field norms are the length (in tokens) of the fields.
/// It is used in the computation of the [TfIdf](https://fulmicoton.gitbooks.io/tantivy-doc/content/tfidf.html).
///
/// They are simply stored as a fast field, serialized in
/// the `.fieldnorm` file of the segment.
pub fn get_fieldnorms_reader(&self, field: Field) -> crate::Result<FieldNormReader> {
self.fieldnorm_readers.get_field(field)?.ok_or_else(|| {
let field_name = self.schema.get_field_name(field);
let err_msg = format!(
@@ -248,14 +127,98 @@ impl SegmentReader for TantivySegmentReader {
})
}
fn get_store_reader(&self, cache_num_blocks: usize) -> io::Result<Box<dyn StoreReader>> {
Ok(Box::new(TantivyStoreReader::open(
self.store_file.clone(),
cache_num_blocks,
)?))
#[doc(hidden)]
pub fn fieldnorms_readers(&self) -> &FieldNormReaders {
&self.fieldnorm_readers
}
fn inverted_index(&self, field: Field) -> crate::Result<Arc<dyn DynInvertedIndexReader>> {
/// Accessor to the segment's [`StoreReader`](crate::store::StoreReader).
///
/// `cache_num_blocks` sets the number of decompressed blocks to be cached in an LRU.
/// The size of blocks is configurable, this should be reflexted in the
pub fn get_store_reader(&self, cache_num_blocks: usize) -> io::Result<StoreReader> {
StoreReader::open(self.store_file.clone(), cache_num_blocks)
}
/// 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)?;
let store_file = segment.open_read(SegmentComponent::Store)?;
crate::fail_point!("SegmentReader::open#middle");
let postings_file = segment.open_read(SegmentComponent::Postings)?;
let postings_composite = CompositeFile::open(&postings_file)?;
let positions_composite = match segment.open_read(SegmentComponent::Positions) {
Ok(positions_file) => CompositeFile::open(&positions_file)?,
Err(OpenReadError::FileDoesNotExist(_)) => CompositeFile::empty(),
Err(open_read_error) => return Err(open_read_error.into()),
};
let schema = segment.schema();
let fast_fields_data = segment.open_read(SegmentComponent::FastFields)?;
let fast_fields_readers = FastFieldReaders::open(fast_fields_data, schema.clone())?;
let fieldnorm_data = segment.open_read(SegmentComponent::FieldNorms)?;
let fieldnorm_readers = FieldNormReaders::open(fieldnorm_data)?;
let original_bitset = if segment.meta().has_deletes() {
let alive_doc_file_slice = segment.open_read(SegmentComponent::Delete)?;
let alive_doc_data = alive_doc_file_slice.read_bytes()?;
Some(AliveBitSet::open(alive_doc_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(),
num_docs,
max_doc,
termdict_composite,
postings_composite,
fast_fields_readers,
fieldnorm_readers,
segment_id: segment.id(),
delete_opstamp: segment.meta().delete_opstamp(),
store_file,
alive_bitset_opt,
positions_composite,
schema,
})
}
/// Returns a field reader associated with the field given in argument.
/// If the field was not present in the index during indexing time,
/// the InvertedIndexReader is empty.
///
/// The field reader is in charge of iterating through the
/// term dictionary associated with a specific field,
/// and opening the posting list associated with any term.
///
/// If the field is not marked as index, a warning is logged and an empty `InvertedIndexReader`
/// is returned.
/// Similarly, if the field is marked as indexed but no term has been indexed for the given
/// index, an empty `InvertedIndexReader` is returned (but no warning is logged).
pub fn inverted_index(&self, field: Field) -> crate::Result<Arc<InvertedIndexReader>> {
if let Some(inv_idx_reader) = self
.inv_idx_reader_cache
.read()
@@ -280,9 +243,7 @@ impl SegmentReader for TantivySegmentReader {
//
// Returns an empty inverted index.
let record_option = record_option_opt.unwrap_or(IndexRecordOption::Basic);
let inv_idx_reader: Arc<dyn DynInvertedIndexReader> =
Arc::new(TantivyInvertedIndexReader::empty(record_option));
return Ok(inv_idx_reader);
return Ok(Arc::new(InvertedIndexReader::empty(record_option)));
}
let record_option = record_option_opt.unwrap();
@@ -305,20 +266,13 @@ impl SegmentReader for TantivySegmentReader {
);
DataCorruption::comment_only(error_msg)
})?;
let fieldnorms_file = self
.fieldnorm_readers
.get_inner_file()
.open_read(field)
.unwrap_or_else(FileSlice::empty);
let inv_idx_reader: Arc<dyn DynInvertedIndexReader> =
Arc::new(TantivyInvertedIndexReader::new(
TermDictionary::open(termdict_file)?,
postings_file,
positions_file,
fieldnorms_file,
record_option,
)?);
let inv_idx_reader = Arc::new(InvertedIndexReader::new(
TermDictionary::open(termdict_file)?,
postings_file,
positions_file,
record_option,
)?);
// by releasing the lock in between, we may end up opening the inverting index
// twice, but this is fine.
@@ -330,10 +284,23 @@ impl SegmentReader for TantivySegmentReader {
Ok(inv_idx_reader)
}
fn fields_metadata(&self) -> crate::Result<Vec<FieldMetadata>> {
/// Returns the list of fields that have been indexed in the segment.
/// The field list includes the field defined in the schema as well as the fields
/// that have been indexed as a part of a JSON field.
/// The returned field name is the full field name, including the name of the JSON field.
///
/// The returned field names can be used in queries.
///
/// Notice: If your data contains JSON fields this is **very expensive**, as it requires
/// browsing through the inverted index term dictionary and the columnar field dictionary.
///
/// Disclaimer: Some fields may not be listed here. For instance, if the schema contains a json
/// field that is not indexed nor a fast field but is stored, it is possible for the field
/// to not be listed.
pub fn fields_metadata(&self) -> crate::Result<Vec<FieldMetadata>> {
let mut indexed_fields: Vec<FieldMetadata> = Vec::new();
let mut map_to_canonical = FnvHashMap::default();
for (field, field_entry) in self.schema.fields() {
for (field, field_entry) in self.schema().fields() {
let field_name = field_entry.name().to_string();
let is_indexed = field_entry.is_indexed();
if is_indexed {
@@ -355,7 +322,7 @@ impl SegmentReader for TantivySegmentReader {
// Without expand dots enabled dots need to be escaped.
let escaped_json_path = json_path.replace('.', "\\.");
let full_path = format!("{field_name}.{escaped_json_path}");
let full_path_unescaped = format!("{}.{}", field_name, &json_path);
let full_path_unescaped = format!("{}.{}", field_name, json_path);
map_to_canonical.insert(full_path_unescaped, full_path.to_string());
full_path
} else {
@@ -423,7 +390,7 @@ impl SegmentReader for TantivySegmentReader {
}
}
let fast_fields: Vec<FieldMetadata> = self
.fast_fields_readers
.fast_fields()
.columnar()
.iter_columns()?
.map(|(mut field_name, handle)| {
@@ -451,26 +418,31 @@ impl SegmentReader for TantivySegmentReader {
Ok(merged_field_metadatas)
}
fn segment_id(&self) -> SegmentId {
/// Returns the segment id
pub fn segment_id(&self) -> SegmentId {
self.segment_id
}
fn delete_opstamp(&self) -> Option<Opstamp> {
/// Returns the delete opstamp
pub fn delete_opstamp(&self) -> Option<Opstamp> {
self.delete_opstamp
}
fn alive_bitset(&self) -> Option<&AliveBitSet> {
/// Returns the bitset representing the alive `DocId`s.
pub fn alive_bitset(&self) -> Option<&AliveBitSet> {
self.alive_bitset_opt.as_ref()
}
fn is_deleted(&self, doc: DocId) -> bool {
self.alive_bitset_opt
.as_ref()
/// Returns true if the `doc` is marked
/// as deleted.
pub fn is_deleted(&self, doc: DocId) -> bool {
self.alive_bitset()
.map(|alive_bitset| alive_bitset.is_deleted(doc))
.unwrap_or(false)
}
fn doc_ids_alive(&self) -> Box<dyn Iterator<Item = DocId> + Send + '_> {
/// Returns an iterator that will iterate over the alive document ids
pub fn doc_ids_alive(&self) -> Box<dyn Iterator<Item = DocId> + Send + '_> {
if let Some(alive_bitset) = &self.alive_bitset_opt {
Box::new(alive_bitset.iter_alive())
} else {
@@ -478,25 +450,22 @@ impl SegmentReader for TantivySegmentReader {
}
}
fn space_usage(&self) -> io::Result<SegmentSpaceUsage> {
/// Summarize total space usage of this segment.
pub fn space_usage(&self) -> io::Result<SegmentSpaceUsage> {
Ok(SegmentSpaceUsage::new(
self.num_docs,
self.termdict_composite.space_usage(&self.schema),
self.postings_composite.space_usage(&self.schema),
self.positions_composite.space_usage(&self.schema),
self.num_docs(),
self.termdict_composite.space_usage(self.schema()),
self.postings_composite.space_usage(self.schema()),
self.positions_composite.space_usage(self.schema()),
self.fast_fields_readers.space_usage()?,
self.fieldnorm_readers.space_usage(&self.schema),
TantivyStoreReader::open(self.store_file.clone(), 0)?.space_usage(),
self.fieldnorm_readers.space_usage(self.schema()),
self.get_store_reader(0)?.space_usage(),
self.alive_bitset_opt
.as_ref()
.map(AliveBitSet::space_usage)
.unwrap_or_default(),
))
}
fn clone_arc(&self) -> Arc<dyn SegmentReader> {
Arc::new(self.clone())
}
}
#[derive(Clone, Debug, PartialEq, Eq, PartialOrd, Ord)]
@@ -606,7 +575,7 @@ fn intersect_alive_bitset(
}
}
impl fmt::Debug for TantivySegmentReader {
impl fmt::Debug for SegmentReader {
fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
write!(f, "SegmentReader({:?})", self.segment_id)
}

8
src/indexer/delete_queue.rs
View File

@@ -250,15 +250,11 @@ mod tests {
struct DummyWeight;
impl Weight for DummyWeight {
fn scorer(
&self,
_reader: &dyn SegmentReader,
_boost: Score,
) -> crate::Result<Box<dyn Scorer>> {
fn scorer(&self, _reader: &SegmentReader, _boost: Score) -> crate::Result<Box<dyn Scorer>> {
Err(crate::TantivyError::InternalError("dummy impl".to_owned()))
}
fn explain(&self, _reader: &dyn SegmentReader, _doc: DocId) -> crate::Result<Explanation> {
fn explain(&self, _reader: &SegmentReader, _doc: DocId) -> crate::Result<Explanation> {
Err(crate::TantivyError::InternalError("dummy impl".to_owned()))
}
}

38
src/indexer/index_writer.rs
View File

@@ -12,9 +12,7 @@ use super::{AddBatch, AddBatchReceiver, AddBatchSender, PreparedCommit};
use crate::directory::{DirectoryLock, GarbageCollectionResult, TerminatingWrite};
use crate::error::TantivyError;
use crate::fastfield::write_alive_bitset;
use crate::index::{
Index, Segment, SegmentComponent, SegmentId, SegmentMeta, SegmentReader, TantivySegmentReader,
};
use crate::index::{Index, Segment, SegmentComponent, SegmentId, SegmentMeta, SegmentReader};
use crate::indexer::delete_queue::{DeleteCursor, DeleteQueue};
use crate::indexer::doc_opstamp_mapping::DocToOpstampMapping;
use crate::indexer::index_writer_status::IndexWriterStatus;
@@ -96,7 +94,7 @@ pub struct IndexWriter<D: Document = TantivyDocument> {
fn compute_deleted_bitset(
alive_bitset: &mut BitSet,
segment_reader: &dyn SegmentReader,
segment_reader: &SegmentReader,
delete_cursor: &mut DeleteCursor,
doc_opstamps: &DocToOpstampMapping,
target_opstamp: Opstamp,
@@ -145,13 +143,7 @@ pub fn advance_deletes(
return Ok(());
}
let segment_reader = TantivySegmentReader::open_with_custom_alive_set_from_directory(
segment.index().directory(),
segment.meta(),
segment.schema(),
None,
)?;
let segment_reader: Arc<dyn SegmentReader> = Arc::new(segment_reader);
let segment_reader = SegmentReader::open(&segment)?;
let max_doc = segment_reader.max_doc();
let mut alive_bitset: BitSet = match segment_entry.alive_bitset() {
@@ -163,7 +155,7 @@ pub fn advance_deletes(
compute_deleted_bitset(
&mut alive_bitset,
segment_reader.as_ref(),
&segment_reader,
segment_entry.delete_cursor(),
&DocToOpstampMapping::None,
target_opstamp,
@@ -251,20 +243,14 @@ fn apply_deletes(
.max()
.expect("Empty DocOpstamp is forbidden");
let segment_reader = TantivySegmentReader::open_with_custom_alive_set_from_directory(
segment.index().directory(),
segment.meta(),
segment.schema(),
None,
)?;
let segment_reader: Arc<dyn SegmentReader> = Arc::new(segment_reader);
let segment_reader = SegmentReader::open(segment)?;
let doc_to_opstamps = DocToOpstampMapping::WithMap(doc_opstamps);
let max_doc = segment.meta().max_doc();
let mut deleted_bitset = BitSet::with_max_value_and_full(max_doc);
let may_have_deletes = compute_deleted_bitset(
&mut deleted_bitset,
segment_reader.as_ref(),
&segment_reader,
delete_cursor,
&doc_to_opstamps,
max_doc_opstamp,
@@ -1979,9 +1965,9 @@ mod tests {
.get_store_reader(DOCSTORE_CACHE_CAPACITY)
.unwrap();
// test store iterator
for doc_id in segment_reader.doc_ids_alive() {
let doc = store_reader.get(doc_id).unwrap();
for doc in store_reader.iter::<TantivyDocument>(segment_reader.alive_bitset()) {
let id = doc
.unwrap()
.get_first(id_field)
.unwrap()
.as_value()
@@ -1992,7 +1978,7 @@ mod tests {
// test store random access
for doc_id in segment_reader.doc_ids_alive() {
let id = store_reader
.get(doc_id)
.get::<TantivyDocument>(doc_id)
.unwrap()
.get_first(id_field)
.unwrap()
@@ -2001,7 +1987,7 @@ mod tests {
assert!(expected_ids_and_num_occurrences.contains_key(&id));
if id_is_full_doc(id) {
let id2 = store_reader
.get(doc_id)
.get::<TantivyDocument>(doc_id)
.unwrap()
.get_first(multi_numbers)
.unwrap()
@@ -2009,13 +1995,13 @@ mod tests {
.unwrap();
assert_eq!(id, id2);
let bool = store_reader
.get(doc_id)
.get::<TantivyDocument>(doc_id)
.unwrap()
.get_first(bool_field)
.unwrap()
.as_bool()
.unwrap();
let doc = store_reader.get(doc_id).unwrap();
let doc = store_reader.get::<TantivyDocument>(doc_id).unwrap();
let mut bool2 = doc.get_all(multi_bools);
assert_eq!(bool, bool2.next().unwrap().as_bool().unwrap());
assert_ne!(bool, bool2.next().unwrap().as_bool().unwrap());

27
src/indexer/merge_index_test.rs
View File

@@ -3,7 +3,7 @@ mod tests {
use crate::collector::TopDocs;
use crate::fastfield::AliveBitSet;
use crate::index::Index;
use crate::postings::{DocFreq, Postings};
use crate::postings::Postings;
use crate::query::QueryParser;
use crate::schema::{
self, BytesOptions, Facet, FacetOptions, IndexRecordOption, NumericOptions,
@@ -121,32 +121,21 @@ mod tests {
let my_text_field = index.schema().get_field("text_field").unwrap();
let term_a = Term::from_field_text(my_text_field, "text");
let inverted_index = segment_reader.inverted_index(my_text_field).unwrap();
let term_info = inverted_index.get_term_info(&term_a).unwrap().unwrap();
let postings_for_test = crate::index::load_postings_from_terminfo(
inverted_index.as_ref(),
&term_info,
IndexRecordOption::WithFreqsAndPositions,
)
.unwrap();
let mut postings = inverted_index
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)
.unwrap()
.unwrap();
assert_eq!(postings.doc_freq(), 2);
let fallback_bitset = AliveBitSet::for_test_from_deleted_docs(&[0], 100);
assert_eq!(
crate::indexer::merger::doc_freq_given_deletes(
postings_for_test,
postings.doc_freq_given_deletes(
segment_reader.alive_bitset().unwrap_or(&fallback_bitset)
),
2
);
let postings = inverted_index
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
assert_eq!(postings.unwrap().doc_freq(), DocFreq::Exact(2));
let postings = inverted_index
.read_postings(&term_a, IndexRecordOption::WithFreqsAndPositions)
.unwrap();
let mut postings = postings.unwrap();
assert_eq!(postings.term_freq(), 1);
let mut output = Vec::new();
let mut output = vec![];
postings.positions(&mut output);
assert_eq!(output, vec![1]);
postings.advance();

201
src/indexer/merger.rs
View File

@@ -1,4 +1,3 @@
use std::io;
use std::sync::Arc;
use columnar::{
@@ -16,11 +15,11 @@ use crate::fieldnorm::{FieldNormReader, FieldNormReaders, FieldNormsSerializer,
use crate::index::{Segment, SegmentComponent, SegmentReader};
use crate::indexer::doc_id_mapping::{MappingType, SegmentDocIdMapping};
use crate::indexer::SegmentSerializer;
use crate::postings::{InvertedIndexSerializer, Postings, TermInfo};
use crate::schema::{value_type_to_column_type, Field, FieldType, IndexRecordOption, Schema};
use crate::postings::{InvertedIndexSerializer, Postings, SegmentPostings};
use crate::schema::{value_type_to_column_type, Field, FieldType, Schema};
use crate::store::StoreWriter;
use crate::termdict::{TermMerger, TermOrdinal};
use crate::{DocAddress, DocId, DynInvertedIndexReader};
use crate::{DocAddress, DocId, InvertedIndexReader};
/// Segment's max doc must be `< MAX_DOC_LIMIT`.
///
@@ -28,7 +27,7 @@ use crate::{DocAddress, DocId, DynInvertedIndexReader};
pub const MAX_DOC_LIMIT: u32 = 1 << 31;
fn estimate_total_num_tokens_in_single_segment(
reader: &dyn SegmentReader,
reader: &SegmentReader,
field: Field,
) -> crate::Result<u64> {
// There are no deletes. We can simply use the exact value saved into the posting list.
@@ -40,7 +39,7 @@ fn estimate_total_num_tokens_in_single_segment(
// When there are deletes, we use an approximation either
// by using the fieldnorm.
if let Ok(fieldnorm_reader) = reader.get_fieldnorms_reader(field) {
if let Some(fieldnorm_reader) = reader.fieldnorms_readers().get_field(field)? {
let mut count: [usize; 256] = [0; 256];
for doc in reader.doc_ids_alive() {
let fieldnorm_id = fieldnorm_reader.fieldnorm_id(doc);
@@ -69,20 +68,17 @@ fn estimate_total_num_tokens_in_single_segment(
Ok((segment_num_tokens as f64 * ratio) as u64)
}
fn estimate_total_num_tokens(
readers: &[Arc<dyn SegmentReader>],
field: Field,
) -> crate::Result<u64> {
fn estimate_total_num_tokens(readers: &[SegmentReader], field: Field) -> crate::Result<u64> {
let mut total_num_tokens: u64 = 0;
for reader in readers {
total_num_tokens += estimate_total_num_tokens_in_single_segment(reader.as_ref(), field)?;
total_num_tokens += estimate_total_num_tokens_in_single_segment(reader, field)?;
}
Ok(total_num_tokens)
}
pub struct IndexMerger {
schema: Schema,
pub(crate) readers: Vec<Arc<dyn SegmentReader>>,
pub(crate) readers: Vec<SegmentReader>,
max_doc: u32,
}
@@ -166,25 +162,16 @@ impl IndexMerger {
// This can be used to merge but also apply an additional filter.
// One use case is demux, which is basically taking a list of
// segments and partitions them e.g. by a value in a field.
//
// # Panics if segments is empty.
pub fn open_with_custom_alive_set(
schema: Schema,
segments: &[Segment],
alive_bitset_opt: Vec<Option<AliveBitSet>>,
) -> crate::Result<IndexMerger> {
assert!(!segments.is_empty());
let mut readers = vec![];
for (segment, new_alive_bitset_opt) in segments.iter().zip(alive_bitset_opt) {
if segment.meta().num_docs() > 0 {
let reader =
crate::TantivySegmentReader::open_with_custom_alive_set_from_directory(
segment.index().directory(),
segment.meta(),
segment.schema(),
new_alive_bitset_opt,
)?;
let reader: Arc<dyn SegmentReader> = Arc::new(reader);
SegmentReader::open_with_custom_alive_set(segment, new_alive_bitset_opt)?;
readers.push(reader);
}
}
@@ -275,7 +262,7 @@ impl IndexMerger {
}),
);
let has_deletes: bool = self.readers.iter().any(|reader| reader.has_deletes());
let has_deletes: bool = self.readers.iter().any(SegmentReader::has_deletes);
let mapping_type = if has_deletes {
MappingType::StackedWithDeletes
} else {
@@ -310,7 +297,7 @@ impl IndexMerger {
let mut max_term_ords: Vec<TermOrdinal> = Vec::new();
let field_readers: Vec<Arc<dyn DynInvertedIndexReader>> = self
let field_readers: Vec<Arc<InvertedIndexReader>> = self
.readers
.iter()
.map(|reader| reader.inverted_index(indexed_field))
@@ -368,8 +355,7 @@ impl IndexMerger {
indexed. Have you modified the schema?",
);
let mut segment_postings_containing_the_term: Vec<(usize, Box<dyn Postings>)> =
Vec::with_capacity(self.readers.len());
let mut segment_postings_containing_the_term: Vec<(usize, SegmentPostings)> = vec![];
while merged_terms.advance() {
segment_postings_containing_the_term.clear();
@@ -380,15 +366,18 @@ impl IndexMerger {
// Let's compute the list of non-empty posting lists
for (segment_ord, term_info) in merged_terms.current_segment_ords_and_term_infos() {
let segment_reader = &self.readers[segment_ord];
let inverted_index = &field_readers[segment_ord];
if let Some((doc_freq, postings)) = postings_for_merge(
inverted_index.as_ref(),
&term_info,
segment_postings_option,
segment_reader.alive_bitset(),
)? {
let inverted_index: &InvertedIndexReader = &field_readers[segment_ord];
let segment_postings = inverted_index
.read_postings_from_terminfo(&term_info, segment_postings_option)?;
let alive_bitset_opt = segment_reader.alive_bitset();
let doc_freq = if let Some(alive_bitset) = alive_bitset_opt {
segment_postings.doc_freq_given_deletes(alive_bitset)
} else {
segment_postings.doc_freq()
};
if doc_freq > 0u32 {
total_doc_freq += doc_freq;
segment_postings_containing_the_term.push((segment_ord, postings));
segment_postings_containing_the_term.push((segment_ord, segment_postings));
}
}
@@ -406,7 +395,11 @@ impl IndexMerger {
assert!(!segment_postings_containing_the_term.is_empty());
let has_term_freq = {
let has_term_freq = segment_postings_containing_the_term[0].1.has_freq();
let has_term_freq = !segment_postings_containing_the_term[0]
.1
.block_cursor
.freqs()
.is_empty();
for (_, postings) in &segment_postings_containing_the_term[1..] {
// This may look at a strange way to test whether we have term freq or not.
// With JSON object, the schema is not sufficient to know whether a term
@@ -422,7 +415,7 @@ impl IndexMerger {
//
// Overall the reliable way to know if we have actual frequencies loaded or not
// is to check whether the actual decoded array is empty or not.
if postings.has_freq() != has_term_freq {
if has_term_freq == postings.block_cursor.freqs().is_empty() {
return Err(DataCorruption::comment_only(
"Term freqs are inconsistent across segments",
)
@@ -497,7 +490,33 @@ impl IndexMerger {
debug_time!("write-storable-fields");
debug!("write-storable-field");
store_writer.merge_segment_readers(&self.readers)?;
for reader in &self.readers {
let store_reader = reader.get_store_reader(1)?;
if reader.has_deletes()
// If there is not enough data in the store, we avoid stacking in order to
// avoid creating many small blocks in the doc store. Once we have 5 full blocks,
// we start stacking. In the worst case 2/7 of the blocks would be very small.
// [segment 1 - {1 doc}][segment 2 - {fullblock * 5}{1doc}]
// => 5 * full blocks, 2 * 1 document blocks
//
// In a more realistic scenario the segments are of the same size, so 1/6 of
// the doc stores would be on average half full, given total randomness (which
// is not the case here, but not sure how it behaves exactly).
//
// https://github.com/quickwit-oss/tantivy/issues/1053
//
// take 7 in order to not walk over all checkpoints.
|| store_reader.block_checkpoints().take(7).count() < 6
|| store_reader.decompressor() != store_writer.compressor().into()
{
for doc_bytes_res in store_reader.iter_raw(reader.alive_bitset()) {
let doc_bytes = doc_bytes_res?;
store_writer.store_bytes(&doc_bytes)?;
}
} else {
store_writer.stack(store_reader)?;
}
}
Ok(())
}
@@ -534,75 +553,6 @@ impl IndexMerger {
}
}
/// Compute the number of non-deleted documents.
///
/// This method will scan through the posting lists, consuming them.
/// (this is a rather expensive operation).
pub(crate) fn doc_freq_given_deletes(
mut postings: Box<dyn Postings>,
alive_bitset: &AliveBitSet,
) -> u32 {
let mut doc_freq = 0;
loop {
let doc = postings.doc();
if doc == TERMINATED {
return doc_freq;
}
if alive_bitset.is_alive(doc) {
doc_freq += 1u32;
}
postings.advance();
}
}
fn read_postings_for_merge(
inverted_index: &dyn DynInvertedIndexReader,
term_info: &TermInfo,
option: IndexRecordOption,
) -> io::Result<Box<dyn Postings>> {
crate::index::load_postings_from_terminfo(inverted_index, term_info, option)
}
fn postings_for_merge(
inverted_index: &dyn DynInvertedIndexReader,
term_info: &TermInfo,
option: IndexRecordOption,
alive_bitset_opt: Option<&AliveBitSet>,
) -> io::Result<Option<(u32, Box<dyn Postings>)>> {
// TODO: avoid loading postings twice — once for counting, once for writing
let count_postings = read_postings_for_merge(inverted_index, term_info, option)?;
let doc_freq = if let Some(alive_bitset) = alive_bitset_opt {
doc_freq_given_deletes(count_postings, alive_bitset)
} else {
// We do not need an exact document frequency here.
match count_postings.doc_freq() {
crate::postings::DocFreq::Exact(doc_freq) => doc_freq,
crate::postings::DocFreq::Approximate(_) => exact_doc_freq(count_postings),
}
};
if doc_freq == 0u32 {
return Ok(None);
}
let postings = read_postings_for_merge(inverted_index, term_info, option)?;
Ok(Some((doc_freq, postings)))
}
/// If the postings is not able to inform us of the document frequency,
/// we just scan through it.
pub(crate) fn exact_doc_freq(mut postings: Box<dyn Postings>) -> u32 {
let mut doc_freq = 0;
loop {
let doc = postings.doc();
if doc == TERMINATED {
return doc_freq;
}
doc_freq += 1u32;
postings.advance();
}
}
#[cfg(test)]
mod tests {
@@ -615,10 +565,8 @@ mod tests {
BytesFastFieldTestCollector, FastFieldTestCollector, TEST_COLLECTOR_WITH_SCORE,
};
use crate::collector::{Count, FacetCollector};
use crate::fastfield::AliveBitSet;
use crate::index::{Index, SegmentId};
use crate::indexer::NoMergePolicy;
use crate::postings::{DocFreq, Postings as _, SegmentPostings};
use crate::query::{AllQuery, BooleanQuery, EnableScoring, Scorer, TermQuery};
use crate::schema::{
Facet, FacetOptions, IndexRecordOption, NumericOptions, TantivyDocument, Term,
@@ -733,32 +681,32 @@ mod tests {
);
}
{
let doc = searcher.doc(DocAddress::new(0, 0))?;
let doc = searcher.doc::<TantivyDocument>(DocAddress::new(0, 0))?;
assert_eq!(
doc.get_first(text_field).unwrap().as_value().as_str(),
Some("af b")
);
}
{
let doc = searcher.doc(DocAddress::new(0, 1))?;
let doc = searcher.doc::<TantivyDocument>(DocAddress::new(0, 1))?;
assert_eq!(
doc.get_first(text_field).unwrap().as_value().as_str(),
Some("a b c")
);
}
{
let doc = searcher.doc(DocAddress::new(0, 2))?;
let doc = searcher.doc::<TantivyDocument>(DocAddress::new(0, 2))?;
assert_eq!(
doc.get_first(text_field).unwrap().as_value().as_str(),
Some("a b c d")
);
}
{
let doc = searcher.doc(DocAddress::new(0, 3))?;
let doc = searcher.doc::<TantivyDocument>(DocAddress::new(0, 3))?;
assert_eq!(doc.get_first(text_field).unwrap().as_str(), Some("af b"));
}
{
let doc = searcher.doc(DocAddress::new(0, 4))?;
let doc = searcher.doc::<TantivyDocument>(DocAddress::new(0, 4))?;
assert_eq!(doc.get_first(text_field).unwrap().as_str(), Some("a b c g"));
}
@@ -1570,10 +1518,10 @@ mod tests {
let searcher = reader.searcher();
let mut term_scorer = term_query
.specialized_weight(EnableScoring::enabled_from_searcher(&searcher))?
.term_scorer_for_test(searcher.segment_reader(0u32), 1.0)
.term_scorer_for_test(searcher.segment_reader(0u32), 1.0)?
.unwrap();
assert_eq!(term_scorer.doc(), 0);
assert_nearly_equals!(term_scorer.seek_block_max(0), 0.0079681855);
assert_nearly_equals!(term_scorer.block_max_score(), 0.0079681855);
assert_nearly_equals!(term_scorer.score(), 0.0079681855);
for _ in 0..81 {
writer.add_document(doc!(text=>"hello happy tax payer"))?;
@@ -1586,13 +1534,13 @@ mod tests {
for segment_reader in searcher.segment_readers() {
let mut term_scorer = term_query
.specialized_weight(EnableScoring::enabled_from_searcher(&searcher))?
.term_scorer_for_test(segment_reader.as_ref(), 1.0)
.term_scorer_for_test(segment_reader, 1.0)?
.unwrap();
// the difference compared to before is intrinsic to the bm25 formula. no worries
// there.
for doc in segment_reader.doc_ids_alive() {
assert_eq!(term_scorer.doc(), doc);
assert_nearly_equals!(term_scorer.seek_block_max(doc), 0.003478312);
assert_nearly_equals!(term_scorer.block_max_score(), 0.003478312);
assert_nearly_equals!(term_scorer.score(), 0.003478312);
term_scorer.advance();
}
@@ -1612,12 +1560,12 @@ mod tests {
let segment_reader = searcher.segment_reader(0u32);
let mut term_scorer = term_query
.specialized_weight(EnableScoring::enabled_from_searcher(&searcher))?
.term_scorer_for_test(segment_reader, 1.0)
.term_scorer_for_test(segment_reader, 1.0)?
.unwrap();
// the difference compared to before is intrinsic to the bm25 formula. no worries there.
for doc in segment_reader.doc_ids_alive() {
assert_eq!(term_scorer.doc(), doc);
assert_nearly_equals!(term_scorer.seek_block_max(doc), 0.003478312);
assert_nearly_equals!(term_scorer.block_max_score(), 0.003478312);
assert_nearly_equals!(term_scorer.score(), 0.003478312);
term_scorer.advance();
}
@@ -1631,19 +1579,4 @@ mod tests {
assert!(((super::MAX_DOC_LIMIT - 1) as i32) >= 0);
assert!((super::MAX_DOC_LIMIT as i32) < 0);
}
#[test]
fn test_doc_freq_given_delete() {
let docs = SegmentPostings::create_from_docs(&[0, 2, 10]);
assert_eq!(docs.doc_freq(), DocFreq::Exact(3));
let alive_bitset = AliveBitSet::for_test_from_deleted_docs(&[2], 12);
let docs_boxed: Box<dyn crate::postings::Postings> =
Box::new(SegmentPostings::create_from_docs(&[0, 2, 10]));
assert_eq!(super::doc_freq_given_deletes(docs_boxed, &alive_bitset), 2);
let all_deleted =
AliveBitSet::for_test_from_deleted_docs(&[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], 12);
let docs_boxed: Box<dyn crate::postings::Postings> =
Box::new(SegmentPostings::create_from_docs(&[0, 2, 10]));
assert_eq!(super::doc_freq_given_deletes(docs_boxed, &all_deleted), 0);
}
}

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