MeiliSearch/meilisearch/src/search.rs

1334 lines
48 KiB
Rust

use core::fmt;
use std::cmp::min;
use std::collections::{BTreeMap, BTreeSet, HashSet};
use std::str::FromStr;
use std::sync::Arc;
use std::time::{Duration, Instant};
use deserr::Deserr;
use either::Either;
use indexmap::IndexMap;
use meilisearch_auth::IndexSearchRules;
use meilisearch_types::deserr::DeserrJsonError;
use meilisearch_types::error::deserr_codes::*;
use meilisearch_types::error::ResponseError;
use meilisearch_types::heed::RoTxn;
use meilisearch_types::index_uid::IndexUid;
use meilisearch_types::milli::score_details::{ScoreDetails, ScoringStrategy};
use meilisearch_types::milli::vector::Embedder;
use meilisearch_types::milli::{FacetValueHit, OrderBy, SearchForFacetValues, TimeBudget};
use meilisearch_types::settings::DEFAULT_PAGINATION_MAX_TOTAL_HITS;
use meilisearch_types::{milli, Document};
use milli::tokenizer::TokenizerBuilder;
use milli::{
AscDesc, FieldId, FieldsIdsMap, Filter, FormatOptions, Index, MatchBounds, MatcherBuilder,
SortError, TermsMatchingStrategy, DEFAULT_VALUES_PER_FACET,
};
use regex::Regex;
use serde::Serialize;
use serde_json::{json, Value};
use crate::error::MeilisearchHttpError;
type MatchesPosition = BTreeMap<String, Vec<MatchBounds>>;
pub const DEFAULT_SEARCH_OFFSET: fn() -> usize = || 0;
pub const DEFAULT_SEARCH_LIMIT: fn() -> usize = || 20;
pub const DEFAULT_CROP_LENGTH: fn() -> usize = || 10;
pub const DEFAULT_CROP_MARKER: fn() -> String = || "".to_string();
pub const DEFAULT_HIGHLIGHT_PRE_TAG: fn() -> String = || "<em>".to_string();
pub const DEFAULT_HIGHLIGHT_POST_TAG: fn() -> String = || "</em>".to_string();
pub const DEFAULT_SEMANTIC_RATIO: fn() -> SemanticRatio = || SemanticRatio(0.5);
#[derive(Clone, Default, PartialEq, Deserr)]
#[deserr(error = DeserrJsonError, rename_all = camelCase, deny_unknown_fields)]
pub struct SearchQuery {
#[deserr(default, error = DeserrJsonError<InvalidSearchQ>)]
pub q: Option<String>,
#[deserr(default, error = DeserrJsonError<InvalidSearchVector>)]
pub vector: Option<Vec<f32>>,
#[deserr(default, error = DeserrJsonError<InvalidHybridQuery>)]
pub hybrid: Option<HybridQuery>,
#[deserr(default = DEFAULT_SEARCH_OFFSET(), error = DeserrJsonError<InvalidSearchOffset>)]
pub offset: usize,
#[deserr(default = DEFAULT_SEARCH_LIMIT(), error = DeserrJsonError<InvalidSearchLimit>)]
pub limit: usize,
#[deserr(default, error = DeserrJsonError<InvalidSearchPage>)]
pub page: Option<usize>,
#[deserr(default, error = DeserrJsonError<InvalidSearchHitsPerPage>)]
pub hits_per_page: Option<usize>,
#[deserr(default, error = DeserrJsonError<InvalidSearchAttributesToRetrieve>)]
pub attributes_to_retrieve: Option<BTreeSet<String>>,
#[deserr(default, error = DeserrJsonError<InvalidSearchAttributesToCrop>)]
pub attributes_to_crop: Option<Vec<String>>,
#[deserr(default, error = DeserrJsonError<InvalidSearchCropLength>, default = DEFAULT_CROP_LENGTH())]
pub crop_length: usize,
#[deserr(default, error = DeserrJsonError<InvalidSearchAttributesToHighlight>)]
pub attributes_to_highlight: Option<HashSet<String>>,
#[deserr(default, error = DeserrJsonError<InvalidSearchShowMatchesPosition>, default)]
pub show_matches_position: bool,
#[deserr(default, error = DeserrJsonError<InvalidSearchShowRankingScore>, default)]
pub show_ranking_score: bool,
#[deserr(default, error = DeserrJsonError<InvalidSearchShowRankingScoreDetails>, default)]
pub show_ranking_score_details: bool,
#[deserr(default, error = DeserrJsonError<InvalidSearchFilter>)]
pub filter: Option<Value>,
#[deserr(default, error = DeserrJsonError<InvalidSearchSort>)]
pub sort: Option<Vec<String>>,
#[deserr(default, error = DeserrJsonError<InvalidSearchFacets>)]
pub facets: Option<Vec<String>>,
#[deserr(default, error = DeserrJsonError<InvalidSearchHighlightPreTag>, default = DEFAULT_HIGHLIGHT_PRE_TAG())]
pub highlight_pre_tag: String,
#[deserr(default, error = DeserrJsonError<InvalidSearchHighlightPostTag>, default = DEFAULT_HIGHLIGHT_POST_TAG())]
pub highlight_post_tag: String,
#[deserr(default, error = DeserrJsonError<InvalidSearchCropMarker>, default = DEFAULT_CROP_MARKER())]
pub crop_marker: String,
#[deserr(default, error = DeserrJsonError<InvalidSearchMatchingStrategy>, default)]
pub matching_strategy: MatchingStrategy,
#[deserr(default, error = DeserrJsonError<InvalidSearchAttributesToSearchOn>, default)]
pub attributes_to_search_on: Option<Vec<String>>,
}
// Since this structure is logged A LOT we're going to reduce the number of things it logs to the bare minimum.
// - Only what IS used, we know everything else is set to None so there is no need to print it
// - Re-order the most important field to debug first
impl fmt::Debug for SearchQuery {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
let Self {
q,
vector,
hybrid,
offset,
limit,
page,
hits_per_page,
attributes_to_retrieve,
attributes_to_crop,
crop_length,
attributes_to_highlight,
show_matches_position,
show_ranking_score,
show_ranking_score_details,
filter,
sort,
facets,
highlight_pre_tag,
highlight_post_tag,
crop_marker,
matching_strategy,
attributes_to_search_on,
} = self;
let mut debug = f.debug_struct("SearchQuery");
// First, everything related to the number of documents to retrieve
debug.field("limit", &limit).field("offset", &offset);
if let Some(page) = page {
debug.field("page", &page);
}
if let Some(hits_per_page) = hits_per_page {
debug.field("hits_per_page", &hits_per_page);
}
// Then, everything related to the queries
if let Some(q) = q {
debug.field("q", &q);
}
if let Some(v) = vector {
if v.len() < 10 {
debug.field("vector", &v);
} else {
debug.field(
"vector",
&format!("[{}, {}, {}, ... {} dimensions]", v[0], v[1], v[2], v.len()),
);
}
}
if let Some(hybrid) = hybrid {
debug.field("hybrid", &hybrid);
}
if let Some(attributes_to_search_on) = attributes_to_search_on {
debug.field("attributes_to_search_on", &attributes_to_search_on);
}
if let Some(filter) = filter {
debug.field("filter", &filter);
}
if let Some(sort) = sort {
debug.field("sort", &sort);
}
if let Some(facets) = facets {
debug.field("facets", &facets);
}
debug.field("matching_strategy", &matching_strategy);
// Then everything related to the formatting
debug.field("crop_length", &crop_length);
if *show_matches_position {
debug.field("show_matches_position", show_matches_position);
}
if *show_ranking_score {
debug.field("show_ranking_score", show_ranking_score);
}
if *show_ranking_score_details {
debug.field("self.show_ranking_score_details", show_ranking_score_details);
}
debug.field("crop_length", &crop_length);
if let Some(facets) = facets {
debug.field("facets", &facets);
}
if let Some(attributes_to_retrieve) = attributes_to_retrieve {
debug.field("attributes_to_retrieve", &attributes_to_retrieve);
}
if let Some(attributes_to_crop) = attributes_to_crop {
debug.field("attributes_to_crop", &attributes_to_crop);
}
if let Some(attributes_to_highlight) = attributes_to_highlight {
debug.field("attributes_to_highlight", &attributes_to_highlight);
}
debug.field("highlight_pre_tag", &highlight_pre_tag);
debug.field("highlight_post_tag", &highlight_post_tag);
debug.field("crop_marker", &crop_marker);
debug.finish()
}
}
#[derive(Debug, Clone, Default, PartialEq, Deserr)]
#[deserr(error = DeserrJsonError<InvalidHybridQuery>, rename_all = camelCase, deny_unknown_fields)]
pub struct HybridQuery {
#[deserr(default, error = DeserrJsonError<InvalidSearchSemanticRatio>, default)]
pub semantic_ratio: SemanticRatio,
#[deserr(default, error = DeserrJsonError<InvalidEmbedder>, default)]
pub embedder: Option<String>,
}
pub enum SearchKind {
KeywordOnly,
SemanticOnly { embedder_name: String, embedder: Arc<Embedder> },
Hybrid { embedder_name: String, embedder: Arc<Embedder>, semantic_ratio: f32 },
}
impl SearchKind {
pub(crate) fn semantic(
index_scheduler: &index_scheduler::IndexScheduler,
index: &Index,
embedder_name: Option<&str>,
vector_len: Option<usize>,
) -> Result<Self, ResponseError> {
let (embedder_name, embedder) =
Self::embedder(index_scheduler, index, embedder_name, vector_len)?;
Ok(Self::SemanticOnly { embedder_name, embedder })
}
pub(crate) fn hybrid(
index_scheduler: &index_scheduler::IndexScheduler,
index: &Index,
embedder_name: Option<&str>,
semantic_ratio: f32,
vector_len: Option<usize>,
) -> Result<Self, ResponseError> {
let (embedder_name, embedder) =
Self::embedder(index_scheduler, index, embedder_name, vector_len)?;
Ok(Self::Hybrid { embedder_name, embedder, semantic_ratio })
}
fn embedder(
index_scheduler: &index_scheduler::IndexScheduler,
index: &Index,
embedder_name: Option<&str>,
vector_len: Option<usize>,
) -> Result<(String, Arc<Embedder>), ResponseError> {
let embedder_configs = index.embedding_configs(&index.read_txn()?)?;
let embedders = index_scheduler.embedders(embedder_configs)?;
let embedder_name = embedder_name.unwrap_or_else(|| embedders.get_default_embedder_name());
let embedder = embedders.get(embedder_name);
let embedder = embedder
.ok_or(milli::UserError::InvalidEmbedder(embedder_name.to_owned()))
.map_err(milli::Error::from)?
.0;
if let Some(vector_len) = vector_len {
if vector_len != embedder.dimensions() {
return Err(meilisearch_types::milli::Error::UserError(
meilisearch_types::milli::UserError::InvalidVectorDimensions {
expected: embedder.dimensions(),
found: vector_len,
},
)
.into());
}
}
Ok((embedder_name.to_owned(), embedder))
}
}
#[derive(Debug, Clone, Copy, PartialEq, Deserr)]
#[deserr(try_from(f32) = TryFrom::try_from -> InvalidSearchSemanticRatio)]
pub struct SemanticRatio(f32);
impl Default for SemanticRatio {
fn default() -> Self {
DEFAULT_SEMANTIC_RATIO()
}
}
impl std::convert::TryFrom<f32> for SemanticRatio {
type Error = InvalidSearchSemanticRatio;
fn try_from(f: f32) -> Result<Self, Self::Error> {
// the suggested "fix" is: `!(0.0..=1.0).contains(&f)`` which is allegedly less readable
#[allow(clippy::manual_range_contains)]
if f > 1.0 || f < 0.0 {
Err(InvalidSearchSemanticRatio)
} else {
Ok(SemanticRatio(f))
}
}
}
impl std::ops::Deref for SemanticRatio {
type Target = f32;
fn deref(&self) -> &Self::Target {
&self.0
}
}
impl SearchQuery {
pub fn is_finite_pagination(&self) -> bool {
self.page.or(self.hits_per_page).is_some()
}
}
/// A `SearchQuery` + an index UID.
// This struct contains the fields of `SearchQuery` inline.
// This is because neither deserr nor serde support `flatten` when using `deny_unknown_fields.
// The `From<SearchQueryWithIndex>` implementation ensures both structs remain up to date.
#[derive(Debug, Clone, PartialEq, Deserr)]
#[deserr(error = DeserrJsonError, rename_all = camelCase, deny_unknown_fields)]
pub struct SearchQueryWithIndex {
#[deserr(error = DeserrJsonError<InvalidIndexUid>, missing_field_error = DeserrJsonError::missing_index_uid)]
pub index_uid: IndexUid,
#[deserr(default, error = DeserrJsonError<InvalidSearchQ>)]
pub q: Option<String>,
#[deserr(default, error = DeserrJsonError<InvalidSearchQ>)]
pub vector: Option<Vec<f32>>,
#[deserr(default, error = DeserrJsonError<InvalidHybridQuery>)]
pub hybrid: Option<HybridQuery>,
#[deserr(default = DEFAULT_SEARCH_OFFSET(), error = DeserrJsonError<InvalidSearchOffset>)]
pub offset: usize,
#[deserr(default = DEFAULT_SEARCH_LIMIT(), error = DeserrJsonError<InvalidSearchLimit>)]
pub limit: usize,
#[deserr(default, error = DeserrJsonError<InvalidSearchPage>)]
pub page: Option<usize>,
#[deserr(default, error = DeserrJsonError<InvalidSearchHitsPerPage>)]
pub hits_per_page: Option<usize>,
#[deserr(default, error = DeserrJsonError<InvalidSearchAttributesToRetrieve>)]
pub attributes_to_retrieve: Option<BTreeSet<String>>,
#[deserr(default, error = DeserrJsonError<InvalidSearchAttributesToCrop>)]
pub attributes_to_crop: Option<Vec<String>>,
#[deserr(default, error = DeserrJsonError<InvalidSearchCropLength>, default = DEFAULT_CROP_LENGTH())]
pub crop_length: usize,
#[deserr(default, error = DeserrJsonError<InvalidSearchAttributesToHighlight>)]
pub attributes_to_highlight: Option<HashSet<String>>,
#[deserr(default, error = DeserrJsonError<InvalidSearchShowRankingScore>, default)]
pub show_ranking_score: bool,
#[deserr(default, error = DeserrJsonError<InvalidSearchShowRankingScoreDetails>, default)]
pub show_ranking_score_details: bool,
#[deserr(default, error = DeserrJsonError<InvalidSearchShowMatchesPosition>, default)]
pub show_matches_position: bool,
#[deserr(default, error = DeserrJsonError<InvalidSearchFilter>)]
pub filter: Option<Value>,
#[deserr(default, error = DeserrJsonError<InvalidSearchSort>)]
pub sort: Option<Vec<String>>,
#[deserr(default, error = DeserrJsonError<InvalidSearchFacets>)]
pub facets: Option<Vec<String>>,
#[deserr(default, error = DeserrJsonError<InvalidSearchHighlightPreTag>, default = DEFAULT_HIGHLIGHT_PRE_TAG())]
pub highlight_pre_tag: String,
#[deserr(default, error = DeserrJsonError<InvalidSearchHighlightPostTag>, default = DEFAULT_HIGHLIGHT_POST_TAG())]
pub highlight_post_tag: String,
#[deserr(default, error = DeserrJsonError<InvalidSearchCropMarker>, default = DEFAULT_CROP_MARKER())]
pub crop_marker: String,
#[deserr(default, error = DeserrJsonError<InvalidSearchMatchingStrategy>, default)]
pub matching_strategy: MatchingStrategy,
#[deserr(default, error = DeserrJsonError<InvalidSearchAttributesToSearchOn>, default)]
pub attributes_to_search_on: Option<Vec<String>>,
}
impl SearchQueryWithIndex {
pub fn into_index_query(self) -> (IndexUid, SearchQuery) {
let SearchQueryWithIndex {
index_uid,
q,
vector,
offset,
limit,
page,
hits_per_page,
attributes_to_retrieve,
attributes_to_crop,
crop_length,
attributes_to_highlight,
show_ranking_score,
show_ranking_score_details,
show_matches_position,
filter,
sort,
facets,
highlight_pre_tag,
highlight_post_tag,
crop_marker,
matching_strategy,
attributes_to_search_on,
hybrid,
} = self;
(
index_uid,
SearchQuery {
q,
vector,
offset,
limit,
page,
hits_per_page,
attributes_to_retrieve,
attributes_to_crop,
crop_length,
attributes_to_highlight,
show_ranking_score,
show_ranking_score_details,
show_matches_position,
filter,
sort,
facets,
highlight_pre_tag,
highlight_post_tag,
crop_marker,
matching_strategy,
attributes_to_search_on,
hybrid,
// do not use ..Default::default() here,
// rather add any missing field from `SearchQuery` to `SearchQueryWithIndex`
},
)
}
}
#[derive(Debug, Copy, Clone, PartialEq, Eq, Deserr)]
#[deserr(rename_all = camelCase)]
pub enum MatchingStrategy {
/// Remove query words from last to first
Last,
/// All query words are mandatory
All,
}
impl Default for MatchingStrategy {
fn default() -> Self {
Self::Last
}
}
impl From<MatchingStrategy> for TermsMatchingStrategy {
fn from(other: MatchingStrategy) -> Self {
match other {
MatchingStrategy::Last => Self::Last,
MatchingStrategy::All => Self::All,
}
}
}
#[derive(Debug, Default, Clone, PartialEq, Eq, Deserr)]
#[deserr(rename_all = camelCase)]
pub enum FacetValuesSort {
/// Facet values are sorted in alphabetical order, ascending from A to Z.
#[default]
Alpha,
/// Facet values are sorted by decreasing count.
/// The count is the number of records containing this facet value in the results of the query.
Count,
}
impl From<FacetValuesSort> for OrderBy {
fn from(val: FacetValuesSort) -> Self {
match val {
FacetValuesSort::Alpha => OrderBy::Lexicographic,
FacetValuesSort::Count => OrderBy::Count,
}
}
}
#[derive(Debug, Clone, Serialize, PartialEq)]
pub struct SearchHit {
#[serde(flatten)]
pub document: Document,
#[serde(rename = "_formatted", skip_serializing_if = "Document::is_empty")]
pub formatted: Document,
#[serde(rename = "_matchesPosition", skip_serializing_if = "Option::is_none")]
pub matches_position: Option<MatchesPosition>,
#[serde(rename = "_rankingScore", skip_serializing_if = "Option::is_none")]
pub ranking_score: Option<f64>,
#[serde(rename = "_rankingScoreDetails", skip_serializing_if = "Option::is_none")]
pub ranking_score_details: Option<serde_json::Map<String, serde_json::Value>>,
}
#[derive(Serialize, Clone, PartialEq)]
#[serde(rename_all = "camelCase")]
pub struct SearchResult {
pub hits: Vec<SearchHit>,
pub query: String,
pub processing_time_ms: u128,
#[serde(flatten)]
pub hits_info: HitsInfo,
#[serde(skip_serializing_if = "Option::is_none")]
pub facet_distribution: Option<BTreeMap<String, IndexMap<String, u64>>>,
#[serde(skip_serializing_if = "Option::is_none")]
pub facet_stats: Option<BTreeMap<String, FacetStats>>,
#[serde(skip_serializing_if = "Option::is_none")]
pub semantic_hit_count: Option<u32>,
// These fields are only used for analytics purposes
#[serde(skip)]
pub degraded: bool,
#[serde(skip)]
pub used_negative_operator: bool,
}
impl fmt::Debug for SearchResult {
fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result {
let SearchResult {
hits,
query,
processing_time_ms,
hits_info,
facet_distribution,
facet_stats,
semantic_hit_count,
degraded,
used_negative_operator,
} = self;
let mut debug = f.debug_struct("SearchResult");
// The most important thing when looking at a search result is the time it took to process
debug.field("processing_time_ms", &processing_time_ms);
debug.field("hits", &format!("[{} hits returned]", hits.len()));
debug.field("query", &query);
debug.field("hits_info", &hits_info);
if *used_negative_operator {
debug.field("used_negative_operator", used_negative_operator);
}
if *degraded {
debug.field("degraded", degraded);
}
if let Some(facet_distribution) = facet_distribution {
debug.field("facet_distribution", &facet_distribution);
}
if let Some(facet_stats) = facet_stats {
debug.field("facet_stats", &facet_stats);
}
if let Some(semantic_hit_count) = semantic_hit_count {
debug.field("semantic_hit_count", &semantic_hit_count);
}
debug.finish()
}
}
#[derive(Serialize, Debug, Clone, PartialEq)]
#[serde(rename_all = "camelCase")]
pub struct SearchResultWithIndex {
pub index_uid: String,
#[serde(flatten)]
pub result: SearchResult,
}
#[derive(Serialize, Debug, Clone, PartialEq, Eq)]
#[serde(untagged)]
pub enum HitsInfo {
#[serde(rename_all = "camelCase")]
Pagination { hits_per_page: usize, page: usize, total_pages: usize, total_hits: usize },
#[serde(rename_all = "camelCase")]
OffsetLimit { limit: usize, offset: usize, estimated_total_hits: usize },
}
#[derive(Serialize, Debug, Clone, PartialEq)]
pub struct FacetStats {
pub min: f64,
pub max: f64,
}
#[derive(Serialize, Debug, Clone, PartialEq)]
#[serde(rename_all = "camelCase")]
pub struct FacetSearchResult {
pub facet_hits: Vec<FacetValueHit>,
pub facet_query: Option<String>,
pub processing_time_ms: u128,
}
/// Incorporate search rules in search query
pub fn add_search_rules(query: &mut SearchQuery, rules: IndexSearchRules) {
query.filter = match (query.filter.take(), rules.filter) {
(None, rules_filter) => rules_filter,
(filter, None) => filter,
(Some(filter), Some(rules_filter)) => {
let filter = match filter {
Value::Array(filter) => filter,
filter => vec![filter],
};
let rules_filter = match rules_filter {
Value::Array(rules_filter) => rules_filter,
rules_filter => vec![rules_filter],
};
Some(Value::Array([filter, rules_filter].concat()))
}
}
}
fn prepare_search<'t>(
index: &'t Index,
rtxn: &'t RoTxn,
query: &'t SearchQuery,
search_kind: &SearchKind,
time_budget: TimeBudget,
) -> Result<(milli::Search<'t>, bool, usize, usize), MeilisearchHttpError> {
let mut search = index.search(rtxn);
search.time_budget(time_budget);
match search_kind {
SearchKind::KeywordOnly => {
if let Some(q) = &query.q {
search.query(q);
}
}
SearchKind::SemanticOnly { embedder_name, embedder } => {
let vector = match query.vector.clone() {
Some(vector) => vector,
None => embedder
.embed_one(query.q.clone().unwrap())
.map_err(milli::vector::Error::from)
.map_err(milli::Error::from)?,
};
search.semantic(embedder_name.clone(), embedder.clone(), Some(vector));
}
SearchKind::Hybrid { embedder_name, embedder, semantic_ratio: _ } => {
if let Some(q) = &query.q {
search.query(q);
}
// will be embedded in hybrid search if necessary
search.semantic(embedder_name.clone(), embedder.clone(), query.vector.clone());
}
}
if let Some(ref searchable) = query.attributes_to_search_on {
search.searchable_attributes(searchable);
}
let is_finite_pagination = query.is_finite_pagination();
search.terms_matching_strategy(query.matching_strategy.into());
let max_total_hits = index
.pagination_max_total_hits(rtxn)
.map_err(milli::Error::from)?
.map(|x| x as usize)
.unwrap_or(DEFAULT_PAGINATION_MAX_TOTAL_HITS);
search.exhaustive_number_hits(is_finite_pagination);
search.scoring_strategy(if query.show_ranking_score || query.show_ranking_score_details {
ScoringStrategy::Detailed
} else {
ScoringStrategy::Skip
});
// compute the offset on the limit depending on the pagination mode.
let (offset, limit) = if is_finite_pagination {
let limit = query.hits_per_page.unwrap_or_else(DEFAULT_SEARCH_LIMIT);
let page = query.page.unwrap_or(1);
// page 0 gives a limit of 0 forcing Meilisearch to return no document.
page.checked_sub(1).map_or((0, 0), |p| (limit * p, limit))
} else {
(query.offset, query.limit)
};
// Make sure that a user can't get more documents than the hard limit,
// we align that on the offset too.
let offset = min(offset, max_total_hits);
let limit = min(limit, max_total_hits.saturating_sub(offset));
search.offset(offset);
search.limit(limit);
if let Some(ref filter) = query.filter {
if let Some(facets) = parse_filter(filter)? {
search.filter(facets);
}
}
if let Some(ref sort) = query.sort {
let sort = match sort.iter().map(|s| AscDesc::from_str(s)).collect() {
Ok(sorts) => sorts,
Err(asc_desc_error) => {
return Err(milli::Error::from(SortError::from(asc_desc_error)).into())
}
};
search.sort_criteria(sort);
}
Ok((search, is_finite_pagination, max_total_hits, offset))
}
pub fn perform_search(
index: &Index,
query: SearchQuery,
search_kind: SearchKind,
) -> Result<SearchResult, MeilisearchHttpError> {
let before_search = Instant::now();
let rtxn = index.read_txn()?;
let time_budget = match index.search_cutoff(&rtxn)? {
Some(cutoff) => TimeBudget::new(Duration::from_millis(cutoff)),
None => TimeBudget::default(),
};
let (search, is_finite_pagination, max_total_hits, offset) =
prepare_search(index, &rtxn, &query, &search_kind, time_budget)?;
let (
milli::SearchResult {
documents_ids,
matching_words,
candidates,
document_scores,
degraded,
used_negative_operator,
},
semantic_hit_count,
) = match &search_kind {
SearchKind::KeywordOnly => (search.execute()?, None),
SearchKind::SemanticOnly { .. } => {
let results = search.execute()?;
let semantic_hit_count = results.document_scores.len() as u32;
(results, Some(semantic_hit_count))
}
SearchKind::Hybrid { semantic_ratio, .. } => search.execute_hybrid(*semantic_ratio)?,
};
let fields_ids_map = index.fields_ids_map(&rtxn).unwrap();
let displayed_ids = index
.displayed_fields_ids(&rtxn)?
.map(|fields| fields.into_iter().collect::<BTreeSet<_>>())
.unwrap_or_else(|| fields_ids_map.iter().map(|(id, _)| id).collect());
let fids = |attrs: &BTreeSet<String>| {
let mut ids = BTreeSet::new();
for attr in attrs {
if attr == "*" {
ids = displayed_ids.clone();
break;
}
if let Some(id) = fields_ids_map.id(attr) {
ids.insert(id);
}
}
ids
};
// The attributes to retrieve are the ones explicitly marked as to retrieve (all by default),
// but these attributes must be also be present
// - in the fields_ids_map
// - in the displayed attributes
let to_retrieve_ids: BTreeSet<_> = query
.attributes_to_retrieve
.as_ref()
.map(fids)
.unwrap_or_else(|| displayed_ids.clone())
.intersection(&displayed_ids)
.cloned()
.collect();
let attr_to_highlight = query.attributes_to_highlight.unwrap_or_default();
let attr_to_crop = query.attributes_to_crop.unwrap_or_default();
// Attributes in `formatted_options` correspond to the attributes that will be in `_formatted`
// These attributes are:
// - the attributes asked to be highlighted or cropped (with `attributesToCrop` or `attributesToHighlight`)
// - the attributes asked to be retrieved: these attributes will not be highlighted/cropped
// But these attributes must be also present in displayed attributes
let formatted_options = compute_formatted_options(
&attr_to_highlight,
&attr_to_crop,
query.crop_length,
&to_retrieve_ids,
&fields_ids_map,
&displayed_ids,
);
let mut tokenizer_builder = TokenizerBuilder::default();
tokenizer_builder.create_char_map(true);
let script_lang_map = index.script_language(&rtxn)?;
if !script_lang_map.is_empty() {
tokenizer_builder.allow_list(&script_lang_map);
}
let separators = index.allowed_separators(&rtxn)?;
let separators: Option<Vec<_>> =
separators.as_ref().map(|x| x.iter().map(String::as_str).collect());
if let Some(ref separators) = separators {
tokenizer_builder.separators(separators);
}
let dictionary = index.dictionary(&rtxn)?;
let dictionary: Option<Vec<_>> =
dictionary.as_ref().map(|x| x.iter().map(String::as_str).collect());
if let Some(ref dictionary) = dictionary {
tokenizer_builder.words_dict(dictionary);
}
let mut formatter_builder = MatcherBuilder::new(matching_words, tokenizer_builder.build());
formatter_builder.crop_marker(query.crop_marker);
formatter_builder.highlight_prefix(query.highlight_pre_tag);
formatter_builder.highlight_suffix(query.highlight_post_tag);
let mut documents = Vec::new();
let documents_iter = index.documents(&rtxn, documents_ids)?;
for ((_id, obkv), score) in documents_iter.into_iter().zip(document_scores.into_iter()) {
// First generate a document with all the displayed fields
let displayed_document = make_document(&displayed_ids, &fields_ids_map, obkv)?;
// select the attributes to retrieve
let attributes_to_retrieve = to_retrieve_ids
.iter()
.map(|&fid| fields_ids_map.name(fid).expect("Missing field name"));
let mut document =
permissive_json_pointer::select_values(&displayed_document, attributes_to_retrieve);
let (matches_position, formatted) = format_fields(
&displayed_document,
&fields_ids_map,
&formatter_builder,
&formatted_options,
query.show_matches_position,
&displayed_ids,
)?;
if let Some(sort) = query.sort.as_ref() {
insert_geo_distance(sort, &mut document);
}
let ranking_score =
query.show_ranking_score.then(|| ScoreDetails::global_score(score.iter()));
let ranking_score_details =
query.show_ranking_score_details.then(|| ScoreDetails::to_json_map(score.iter()));
let hit = SearchHit {
document,
formatted,
matches_position,
ranking_score_details,
ranking_score,
};
documents.push(hit);
}
let number_of_hits = min(candidates.len() as usize, max_total_hits);
let hits_info = if is_finite_pagination {
let hits_per_page = query.hits_per_page.unwrap_or_else(DEFAULT_SEARCH_LIMIT);
// If hit_per_page is 0, then pages can't be computed and so we respond 0.
let total_pages = (number_of_hits + hits_per_page.saturating_sub(1))
.checked_div(hits_per_page)
.unwrap_or(0);
HitsInfo::Pagination {
hits_per_page,
page: query.page.unwrap_or(1),
total_pages,
total_hits: number_of_hits,
}
} else {
HitsInfo::OffsetLimit { limit: query.limit, offset, estimated_total_hits: number_of_hits }
};
let (facet_distribution, facet_stats) = match query.facets {
Some(ref fields) => {
let mut facet_distribution = index.facets_distribution(&rtxn);
let max_values_by_facet = index
.max_values_per_facet(&rtxn)
.map_err(milli::Error::from)?
.map(|x| x as usize)
.unwrap_or(DEFAULT_VALUES_PER_FACET);
facet_distribution.max_values_per_facet(max_values_by_facet);
let sort_facet_values_by =
index.sort_facet_values_by(&rtxn).map_err(milli::Error::from)?;
if fields.iter().all(|f| f != "*") {
let fields: Vec<_> =
fields.iter().map(|n| (n, sort_facet_values_by.get(n))).collect();
facet_distribution.facets(fields);
}
let distribution = facet_distribution
.candidates(candidates)
.default_order_by(sort_facet_values_by.get("*"))
.execute()?;
let stats = facet_distribution.compute_stats()?;
(Some(distribution), Some(stats))
}
None => (None, None),
};
let facet_stats = facet_stats.map(|stats| {
stats.into_iter().map(|(k, (min, max))| (k, FacetStats { min, max })).collect()
});
let result = SearchResult {
hits: documents,
hits_info,
query: query.q.unwrap_or_default(),
processing_time_ms: before_search.elapsed().as_millis(),
facet_distribution,
facet_stats,
degraded,
used_negative_operator,
semantic_hit_count,
};
Ok(result)
}
pub fn perform_facet_search(
index: &Index,
search_query: SearchQuery,
facet_query: Option<String>,
facet_name: String,
search_kind: SearchKind,
) -> Result<FacetSearchResult, MeilisearchHttpError> {
let before_search = Instant::now();
let rtxn = index.read_txn()?;
let time_budget = match index.search_cutoff(&rtxn)? {
Some(cutoff) => TimeBudget::new(Duration::from_millis(cutoff)),
None => TimeBudget::default(),
};
let (search, _, _, _) = prepare_search(index, &rtxn, &search_query, &search_kind, time_budget)?;
let mut facet_search = SearchForFacetValues::new(
facet_name,
search,
matches!(search_kind, SearchKind::Hybrid { .. }),
);
if let Some(facet_query) = &facet_query {
facet_search.query(facet_query);
}
if let Some(max_facets) = index.max_values_per_facet(&rtxn)? {
facet_search.max_values(max_facets as usize);
}
Ok(FacetSearchResult {
facet_hits: facet_search.execute()?,
facet_query,
processing_time_ms: before_search.elapsed().as_millis(),
})
}
fn insert_geo_distance(sorts: &[String], document: &mut Document) {
lazy_static::lazy_static! {
static ref GEO_REGEX: Regex =
Regex::new(r"_geoPoint\(\s*([[:digit:].\-]+)\s*,\s*([[:digit:].\-]+)\s*\)").unwrap();
};
if let Some(capture_group) = sorts.iter().find_map(|sort| GEO_REGEX.captures(sort)) {
// TODO: TAMO: milli encountered an internal error, what do we want to do?
let base = [capture_group[1].parse().unwrap(), capture_group[2].parse().unwrap()];
let geo_point = &document.get("_geo").unwrap_or(&json!(null));
if let Some((lat, lng)) = geo_point["lat"].as_f64().zip(geo_point["lng"].as_f64()) {
let distance = milli::distance_between_two_points(&base, &[lat, lng]);
document.insert("_geoDistance".to_string(), json!(distance.round() as usize));
}
}
}
fn compute_formatted_options(
attr_to_highlight: &HashSet<String>,
attr_to_crop: &[String],
query_crop_length: usize,
to_retrieve_ids: &BTreeSet<FieldId>,
fields_ids_map: &FieldsIdsMap,
displayed_ids: &BTreeSet<FieldId>,
) -> BTreeMap<FieldId, FormatOptions> {
let mut formatted_options = BTreeMap::new();
add_highlight_to_formatted_options(
&mut formatted_options,
attr_to_highlight,
fields_ids_map,
displayed_ids,
);
add_crop_to_formatted_options(
&mut formatted_options,
attr_to_crop,
query_crop_length,
fields_ids_map,
displayed_ids,
);
// Should not return `_formatted` if no valid attributes to highlight/crop
if !formatted_options.is_empty() {
add_non_formatted_ids_to_formatted_options(&mut formatted_options, to_retrieve_ids);
}
formatted_options
}
fn add_highlight_to_formatted_options(
formatted_options: &mut BTreeMap<FieldId, FormatOptions>,
attr_to_highlight: &HashSet<String>,
fields_ids_map: &FieldsIdsMap,
displayed_ids: &BTreeSet<FieldId>,
) {
for attr in attr_to_highlight {
let new_format = FormatOptions { highlight: true, crop: None };
if attr == "*" {
for id in displayed_ids {
formatted_options.insert(*id, new_format);
}
break;
}
if let Some(id) = fields_ids_map.id(attr) {
if displayed_ids.contains(&id) {
formatted_options.insert(id, new_format);
}
}
}
}
fn add_crop_to_formatted_options(
formatted_options: &mut BTreeMap<FieldId, FormatOptions>,
attr_to_crop: &[String],
crop_length: usize,
fields_ids_map: &FieldsIdsMap,
displayed_ids: &BTreeSet<FieldId>,
) {
for attr in attr_to_crop {
let mut split = attr.rsplitn(2, ':');
let (attr_name, attr_len) = match split.next().zip(split.next()) {
Some((len, name)) => {
let crop_len = len.parse::<usize>().unwrap_or(crop_length);
(name, crop_len)
}
None => (attr.as_str(), crop_length),
};
if attr_name == "*" {
for id in displayed_ids {
formatted_options
.entry(*id)
.and_modify(|f| f.crop = Some(attr_len))
.or_insert(FormatOptions { highlight: false, crop: Some(attr_len) });
}
}
if let Some(id) = fields_ids_map.id(attr_name) {
if displayed_ids.contains(&id) {
formatted_options
.entry(id)
.and_modify(|f| f.crop = Some(attr_len))
.or_insert(FormatOptions { highlight: false, crop: Some(attr_len) });
}
}
}
}
fn add_non_formatted_ids_to_formatted_options(
formatted_options: &mut BTreeMap<FieldId, FormatOptions>,
to_retrieve_ids: &BTreeSet<FieldId>,
) {
for id in to_retrieve_ids {
formatted_options.entry(*id).or_insert(FormatOptions { highlight: false, crop: None });
}
}
fn make_document(
displayed_attributes: &BTreeSet<FieldId>,
field_ids_map: &FieldsIdsMap,
obkv: obkv::KvReaderU16,
) -> Result<Document, MeilisearchHttpError> {
let mut document = serde_json::Map::new();
// recreate the original json
for (key, value) in obkv.iter() {
let value = serde_json::from_slice(value)?;
let key = field_ids_map.name(key).expect("Missing field name").to_string();
document.insert(key, value);
}
// select the attributes to retrieve
let displayed_attributes = displayed_attributes
.iter()
.map(|&fid| field_ids_map.name(fid).expect("Missing field name"));
let document = permissive_json_pointer::select_values(&document, displayed_attributes);
Ok(document)
}
fn format_fields<'a>(
document: &Document,
field_ids_map: &FieldsIdsMap,
builder: &'a MatcherBuilder<'a>,
formatted_options: &BTreeMap<FieldId, FormatOptions>,
compute_matches: bool,
displayable_ids: &BTreeSet<FieldId>,
) -> Result<(Option<MatchesPosition>, Document), MeilisearchHttpError> {
let mut matches_position = compute_matches.then(BTreeMap::new);
let mut document = document.clone();
// reduce the formatted option list to the attributes that should be formatted,
// instead of all the attributes to display.
let formatting_fields_options: Vec<_> = formatted_options
.iter()
.filter(|(_, option)| option.should_format())
.map(|(fid, option)| (field_ids_map.name(*fid).unwrap(), option))
.collect();
// select the attributes to retrieve
let displayable_names =
displayable_ids.iter().map(|&fid| field_ids_map.name(fid).expect("Missing field name"));
permissive_json_pointer::map_leaf_values(&mut document, displayable_names, |key, value| {
// To get the formatting option of each key we need to see all the rules that applies
// to the value and merge them together. eg. If a user said he wanted to highlight `doggo`
// and crop `doggo.name`. `doggo.name` needs to be highlighted + cropped while `doggo.age` is only
// highlighted.
// Warn: The time to compute the format list scales with the number of fields to format;
// cumulated with map_leaf_values that iterates over all the nested fields, it gives a quadratic complexity:
// d*f where d is the total number of fields to display and f is the total number of fields to format.
let format = formatting_fields_options
.iter()
.filter(|(name, _option)| {
milli::is_faceted_by(name, key) || milli::is_faceted_by(key, name)
})
.map(|(_, option)| **option)
.reduce(|acc, option| acc.merge(option));
let mut infos = Vec::new();
*value = format_value(std::mem::take(value), builder, format, &mut infos, compute_matches);
if let Some(matches) = matches_position.as_mut() {
if !infos.is_empty() {
matches.insert(key.to_owned(), infos);
}
}
});
let selectors = formatted_options
.keys()
// This unwrap must be safe since we got the ids from the fields_ids_map just
// before.
.map(|&fid| field_ids_map.name(fid).unwrap());
let document = permissive_json_pointer::select_values(&document, selectors);
Ok((matches_position, document))
}
fn format_value<'a>(
value: Value,
builder: &'a MatcherBuilder<'a>,
format_options: Option<FormatOptions>,
infos: &mut Vec<MatchBounds>,
compute_matches: bool,
) -> Value {
match value {
Value::String(old_string) => {
let mut matcher = builder.build(&old_string);
if compute_matches {
let matches = matcher.matches();
infos.extend_from_slice(&matches[..]);
}
match format_options {
Some(format_options) => {
let value = matcher.format(format_options);
Value::String(value.into_owned())
}
None => Value::String(old_string),
}
}
Value::Array(values) => Value::Array(
values
.into_iter()
.map(|v| {
format_value(
v,
builder,
format_options.map(|format_options| FormatOptions {
highlight: format_options.highlight,
crop: None,
}),
infos,
compute_matches,
)
})
.collect(),
),
Value::Object(object) => Value::Object(
object
.into_iter()
.map(|(k, v)| {
(
k,
format_value(
v,
builder,
format_options.map(|format_options| FormatOptions {
highlight: format_options.highlight,
crop: None,
}),
infos,
compute_matches,
),
)
})
.collect(),
),
Value::Number(number) => {
let s = number.to_string();
let mut matcher = builder.build(&s);
if compute_matches {
let matches = matcher.matches();
infos.extend_from_slice(&matches[..]);
}
match format_options {
Some(format_options) => {
let value = matcher.format(format_options);
Value::String(value.into_owned())
}
None => Value::String(s),
}
}
value => value,
}
}
pub(crate) fn parse_filter(facets: &Value) -> Result<Option<Filter>, MeilisearchHttpError> {
match facets {
Value::String(expr) => {
let condition = Filter::from_str(expr)?;
Ok(condition)
}
Value::Array(arr) => parse_filter_array(arr),
v => Err(MeilisearchHttpError::InvalidExpression(&["String", "Array"], v.clone())),
}
}
fn parse_filter_array(arr: &[Value]) -> Result<Option<Filter>, MeilisearchHttpError> {
let mut ands = Vec::new();
for value in arr {
match value {
Value::String(s) => ands.push(Either::Right(s.as_str())),
Value::Array(arr) => {
let mut ors = Vec::new();
for value in arr {
match value {
Value::String(s) => ors.push(s.as_str()),
v => {
return Err(MeilisearchHttpError::InvalidExpression(
&["String"],
v.clone(),
))
}
}
}
ands.push(Either::Left(ors));
}
v => {
return Err(MeilisearchHttpError::InvalidExpression(
&["String", "[String]"],
v.clone(),
))
}
}
}
Ok(Filter::from_array(ands)?)
}
#[cfg(test)]
mod test {
use super::*;
#[test]
fn test_insert_geo_distance() {
let value: Document = serde_json::from_str(
r#"{
"_geo": {
"lat": 50.629973371633746,
"lng": 3.0569447399419567
},
"city": "Lille",
"id": "1"
}"#,
)
.unwrap();
let sorters = &["_geoPoint(50.629973371633746,3.0569447399419567):desc".to_string()];
let mut document = value.clone();
insert_geo_distance(sorters, &mut document);
assert_eq!(document.get("_geoDistance"), Some(&json!(0)));
let sorters = &["_geoPoint(50.629973371633746, 3.0569447399419567):asc".to_string()];
let mut document = value.clone();
insert_geo_distance(sorters, &mut document);
assert_eq!(document.get("_geoDistance"), Some(&json!(0)));
let sorters =
&["_geoPoint( 50.629973371633746 , 3.0569447399419567 ):desc".to_string()];
let mut document = value.clone();
insert_geo_distance(sorters, &mut document);
assert_eq!(document.get("_geoDistance"), Some(&json!(0)));
let sorters = &[
"prix:asc",
"villeneuve:desc",
"_geoPoint(50.629973371633746, 3.0569447399419567):asc",
"ubu:asc",
]
.map(|s| s.to_string());
let mut document = value.clone();
insert_geo_distance(sorters, &mut document);
assert_eq!(document.get("_geoDistance"), Some(&json!(0)));
// only the first geoPoint is used to compute the distance
let sorters = &[
"chien:desc",
"_geoPoint(50.629973371633746, 3.0569447399419567):asc",
"pangolin:desc",
"_geoPoint(100.0, -80.0):asc",
"chat:asc",
]
.map(|s| s.to_string());
let mut document = value.clone();
insert_geo_distance(sorters, &mut document);
assert_eq!(document.get("_geoDistance"), Some(&json!(0)));
// there was no _geoPoint so nothing is inserted in the document
let sorters = &["chien:asc".to_string()];
let mut document = value;
insert_geo_distance(sorters, &mut document);
assert_eq!(document.get("_geoDistance"), None);
}
}