rust/tantivy
1
0
Fork 0
mirror of https://github.com/quickwit-oss/tantivy.git synced 2026-05-23 19:50:42 +00:00
Code Issues Packages Projects Releases Wiki Activity

Issue/36 (#559)

Browse Source 
* Added explanation

* Explain

* Splitting weight and idf

* Added comments

Closes #36
This commit is contained in:
Paul Masurel
2019-06-06 10:03:54 +09:00
committed by GitHub
parent d590f4c6b0
commit 4822940b19
17 changed files with 493 additions and 86 deletions
Download Patch File Download Diff File Expand all files Collapse all files

43
src/query/boolean_query/boolean_weight.rs
View File

@@ -1,5 +1,5 @@
use core::SegmentReader;
use query::intersect_scorers;
use query::explanation::does_not_match;
use query::score_combiner::{DoNothingCombiner, ScoreCombiner, SumWithCoordsCombiner};
use query::term_query::TermScorer;
use query::EmptyScorer;
@@ -9,8 +9,10 @@ use query::RequiredOptionalScorer;
use query::Scorer;
use query::Union;
use query::Weight;
use query::{intersect_scorers, Explanation};
use std::collections::HashMap;
use Result;
use {DocId, SkipResult};
fn scorer_union<TScoreCombiner>(scorers: Vec<Box<Scorer>>) -> Box<Scorer>
where
@@ -50,10 +52,10 @@ impl BooleanWeight {
}
}
fn complex_scorer<TScoreCombiner: ScoreCombiner>(
fn per_occur_scorers(
&self,
reader: &SegmentReader,
) -> Result<Box<Scorer>> {
) -> Result<HashMap<Occur, Vec<Box<Scorer>>>> {
let mut per_occur_scorers: HashMap<Occur, Vec<Box<Scorer>>> = HashMap::new();
for &(ref occur, ref subweight) in &self.weights {
let sub_scorer: Box<Scorer> = subweight.scorer(reader)?;
@@ -62,6 +64,14 @@ impl BooleanWeight {
.or_insert_with(Vec::new)
.push(sub_scorer);
}
Ok(per_occur_scorers)
}
fn complex_scorer<TScoreCombiner: ScoreCombiner>(
&self,
reader: &SegmentReader,
) -> Result<Box<Scorer>> {
let mut per_occur_scorers = self.per_occur_scorers(reader)?;
let should_scorer_opt: Option<Box<Scorer>> = per_occur_scorers
.remove(&Occur::Should)
@@ -118,4 +128,31 @@ impl Weight for BooleanWeight {
self.complex_scorer::<DoNothingCombiner>(reader)
}
}
fn explain(&self, reader: &SegmentReader, doc: DocId) -> Result<Explanation> {
let mut scorer = self.scorer(reader)?;
if scorer.skip_next(doc) != SkipResult::Reached {
return Err(does_not_match(doc));
}
if !self.scoring_enabled {
return Ok(Explanation::new("BooleanQuery with no scoring", 1f32));
}
let mut explanation = Explanation::new("BooleanClause. Sum of ...", scorer.score());
for &(ref occur, ref subweight) in &self.weights {
if is_positive_occur(*occur) {
if let Ok(child_explanation) = subweight.explain(reader, doc) {
explanation.add_detail(child_explanation);
}
}
}
Ok(explanation)
}
}
fn is_positive_occur(occur: Occur) -> bool {
match occur {
Occur::Must | Occur::Should => true,
Occur::MustNot => false,
}
}

165
src/query/boolean_query/mod.rs
View File

@@ -18,8 +18,8 @@ mod tests {
use query::Scorer;
use query::TermQuery;
use schema::*;
use DocId;
use Index;
use {DocAddress, DocId};
fn aux_test_helper() -> (Index, Field) {
let mut schema_builder = Schema::builder();
@@ -205,4 +205,167 @@ mod tests {
assert_eq!(score_docs(&boolean_query), vec![0.977973, 0.84699446]);
}
}
// motivated by #554
#[test]
fn test_bm25_several_fields() {
let mut schema_builder = Schema::builder();
let title = schema_builder.add_text_field("title", TEXT);
let text = schema_builder.add_text_field("text", TEXT);
let schema = schema_builder.build();
let index = Index::create_in_ram(schema);
let mut index_writer = index.writer_with_num_threads(1, 3_000_000).unwrap();
index_writer.add_document(doc!(
// tf = 1 0
title => "Законы притяжения Оксана Кулакова",
// tf = 1 0
text => "Законы притяжения Оксана Кулакова] \n\nТема: Сексуальное искусство, Женственность\nТип товара: Запись вебинара (аудио)\nПродолжительность: 1,5 часа\n\nСсылка на вебинар:\n ",
));
index_writer.add_document(doc!(
// tf = 1 0
title => "Любимые русские пироги (Оксана Путан)",
// tf = 2 0
text => "http://i95.fastpic.ru/big/2017/0628/9a/615b9c8504d94a3893d7f496ac53539a.jpg \n\nОт издателя\nОксана Путан профессиональный повар, автор кулинарных книг и известный кулинарный блогер. Ее рецепты отличаются практичностью, доступностью и пользуются огромной популярностью в русскоязычном интернете. Это третья книга автора о самом вкусном и ароматном настоящих русских пирогах и выпечке!\nДаже новички на кухне легко готовят по ее рецептам. Оксана описывает процесс приготовления настолько подробно и понятно, что вам остается только наслаждаться готовкой и не тратить время на лишние усилия. Готовьте легко и просто!\n\nhttps://www.ozon.ru/context/detail/id/139872462/"
));
index_writer.add_document(doc!(
// tf = 1 1
title => "PDF Мастер Класс \"Морячок\" (Оксана Лифенко)",
// tf = 0 0
text => "https://i.ibb.co/pzvHrDN/I3d U T6 Gg TM.jpg\nhttps://i.ibb.co/NFrb6v6/N0ls Z9nwjb U.jpg\nВ описание входит штаны, кофта, берет, матросский воротник. Описание продается в формате PDF, состоит из 12 страниц формата А4 и может быть напечатано на любом принтере.\nОписание предназначено для кукол BJD RealPuki от FairyLand, но может подойти и другим подобным куклам. Также вы можете вязать этот наряд из обычной пряжи, и он подойдет для куколок побольше.\nhttps://vk.com/market 95724412?w=product 95724412_2212"
));
for _ in 0..1_000 {
index_writer.add_document(doc!(
title => "a b d e f g",
text => "maitre corbeau sur un arbre perche tenait dans son bec un fromage Maitre rnard par lodeur alleche lui tint a peu pres ce langage."
));
}
index_writer.commit().unwrap();
let reader = index.reader().unwrap();
let searcher = reader.searcher();
let query_parser = QueryParser::for_index(&index, vec![title, text]);
let query = query_parser
.parse_query("Оксана Лифенко")
.unwrap();
let weight = query.weight(&searcher, true).unwrap();
let mut scorer = weight.scorer(searcher.segment_reader(0u32)).unwrap();
scorer.advance();
let explanation = query.explain(&searcher, DocAddress(0u32, 0u32)).unwrap();
assert_eq!(
explanation.to_pretty_json(),
r#"{
"value": 12.997711,
"description": "BooleanClause. Sum of ...",
"details": [
{
"value": 12.997711,
"description": "BooleanClause. Sum of ...",
"details": [
{
"value": 6.551476,
"description": "TermQuery, product of...",
"details": [
{
"value": 2.2,
"description": "(K1+1)"
},
{
"value": 5.658984,
"description": "idf, computed as log(1 + (N - n + 0.5) / (n + 0.5))",
"details": [
{
"value": 3.0,
"description": "n, number of docs containing this term"
},
{
"value": 1003.0,
"description": "N, total number of docs"
}
]
},
{
"value": 0.5262329,
"description": "freq / (freq + k1 * (1 - b + b * dl / avgdl))",
"details": [
{
"value": 1.0,
"description": "freq, occurrences of term within document"
},
{
"value": 1.2,
"description": "k1, term saturation parameter"
},
{
"value": 0.75,
"description": "b, length normalization parameter"
},
{
"value": 4.0,
"description": "dl, length of field"
},
{
"value": 5.997009,
"description": "avgdl, average length of field"
}
]
}
]
},
{
"value": 6.446235,
"description": "TermQuery, product of...",
"details": [
{
"value": 2.2,
"description": "(K1+1)"
},
{
"value": 5.9954567,
"description": "idf, computed as log(1 + (N - n + 0.5) / (n + 0.5))",
"details": [
{
"value": 2.0,
"description": "n, number of docs containing this term"
},
{
"value": 1003.0,
"description": "N, total number of docs"
}
]
},
{
"value": 0.4887212,
"description": "freq / (freq + k1 * (1 - b + b * dl / avgdl))",
"details": [
{
"value": 1.0,
"description": "freq, occurrences of term within document"
},
{
"value": 1.2,
"description": "k1, term saturation parameter"
},
{
"value": 0.75,
"description": "b, length normalization parameter"
},
{
"value": 20.0,
"description": "dl, length of field"
},
{
"value": 24.123629,
"description": "avgdl, average length of field"
}
]
}
]
}
]
}
]
}"#
);
}
}