1
1
mirror of https://github.com/go-gitea/gitea synced 2024-11-18 16:14:26 +00:00
gitea/modules/indexer/issues/meilisearch/meilisearch.go
Jason Song 1e76a824bc
Refactor and enhance issue indexer to support both searching, filtering and paging (#26012)
Fix #24662.

Replace #24822 and #25708 (although it has been merged)


## Background

In the past, Gitea supported issue searching with a keyword and
conditions in a less efficient way. It worked by searching for issues
with the keyword and obtaining limited IDs (as it is heavy to get all)
on the indexer (bleve/elasticsearch/meilisearch), and then querying with
conditions on the database to find a subset of the found IDs. This is
why the results could be incomplete.

To solve this issue, we need to store all fields that could be used as
conditions in the indexer and support both keyword and additional
conditions when searching with the indexer.

## Major changes

- Redefine `IndexerData` to include all fields that could be used as
filter conditions.
- Refactor `Search(ctx context.Context, kw string, repoIDs []int64,
limit, start int, state string)` to `Search(ctx context.Context, options
*SearchOptions)`, so it supports more conditions now.
- Change the data type stored in `issueIndexerQueue`. Use
`IndexerMetadata` instead of `IndexerData` in case the data has been
updated while it is in the queue. This also reduces the storage size of
the queue.
- Enhance searching with Bleve/Elasticsearch/Meilisearch, make them
fully support `SearchOptions`. Also, update the data versions.
- Keep most logic of database indexer, but remove
`issues.SearchIssueIDsByKeyword` in `models` to avoid confusion where is
the entry point to search issues.
- Start a Meilisearch instance to test it in unit tests.
- Add unit tests with almost full coverage to test
Bleve/Elasticsearch/Meilisearch indexer.

---------

Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
2023-07-31 06:28:53 +00:00

243 lines
6.6 KiB
Go

// Copyright 2023 The Gitea Authors. All rights reserved.
// SPDX-License-Identifier: MIT
package meilisearch
import (
"context"
"strconv"
"strings"
indexer_internal "code.gitea.io/gitea/modules/indexer/internal"
inner_meilisearch "code.gitea.io/gitea/modules/indexer/internal/meilisearch"
"code.gitea.io/gitea/modules/indexer/issues/internal"
"github.com/meilisearch/meilisearch-go"
)
const (
issueIndexerLatestVersion = 2
// TODO: make this configurable if necessary
maxTotalHits = 10000
)
var _ internal.Indexer = &Indexer{}
// Indexer implements Indexer interface
type Indexer struct {
inner *inner_meilisearch.Indexer
indexer_internal.Indexer // do not composite inner_meilisearch.Indexer directly to avoid exposing too much
}
// NewIndexer creates a new meilisearch indexer
func NewIndexer(url, apiKey, indexerName string) *Indexer {
settings := &meilisearch.Settings{
// The default ranking rules of meilisearch are: ["words", "typo", "proximity", "attribute", "sort", "exactness"]
// So even if we specify the sort order, it could not be respected because the priority of "sort" is so low.
// So we need to specify the ranking rules to make sure the sort order is respected.
// See https://www.meilisearch.com/docs/learn/core_concepts/relevancy
RankingRules: []string{"sort", // make sure "sort" has the highest priority
"words", "typo", "proximity", "attribute", "exactness"},
SearchableAttributes: []string{
"title",
"content",
"comments",
},
DisplayedAttributes: []string{
"id",
},
FilterableAttributes: []string{
"repo_id",
"is_public",
"is_pull",
"is_closed",
"label_ids",
"no_label",
"milestone_id",
"project_id",
"project_board_id",
"poster_id",
"assignee_id",
"mention_ids",
"reviewed_ids",
"review_requested_ids",
"subscriber_ids",
"updated_unix",
},
SortableAttributes: []string{
"updated_unix",
"created_unix",
"deadline_unix",
"comment_count",
"id",
},
Pagination: &meilisearch.Pagination{
MaxTotalHits: maxTotalHits,
},
}
inner := inner_meilisearch.NewIndexer(url, apiKey, indexerName, issueIndexerLatestVersion, settings)
indexer := &Indexer{
inner: inner,
Indexer: inner,
}
return indexer
}
// Index will save the index data
func (b *Indexer) Index(_ context.Context, issues ...*internal.IndexerData) error {
if len(issues) == 0 {
return nil
}
for _, issue := range issues {
_, err := b.inner.Client.Index(b.inner.VersionedIndexName()).AddDocuments(issue)
if err != nil {
return err
}
}
// TODO: bulk send index data
return nil
}
// Delete deletes indexes by ids
func (b *Indexer) Delete(_ context.Context, ids ...int64) error {
if len(ids) == 0 {
return nil
}
for _, id := range ids {
_, err := b.inner.Client.Index(b.inner.VersionedIndexName()).DeleteDocument(strconv.FormatInt(id, 10))
if err != nil {
return err
}
}
// TODO: bulk send deletes
return nil
}
// Search searches for issues by given conditions.
// Returns the matching issue IDs
func (b *Indexer) Search(ctx context.Context, options *internal.SearchOptions) (*internal.SearchResult, error) {
query := inner_meilisearch.FilterAnd{}
if len(options.RepoIDs) > 0 {
q := &inner_meilisearch.FilterOr{}
q.Or(inner_meilisearch.NewFilterIn("repo_id", options.RepoIDs...))
if options.AllPublic {
q.Or(inner_meilisearch.NewFilterEq("is_public", true))
}
query.And(q)
}
if !options.IsPull.IsNone() {
query.And(inner_meilisearch.NewFilterEq("is_pull", options.IsPull.IsTrue()))
}
if !options.IsClosed.IsNone() {
query.And(inner_meilisearch.NewFilterEq("is_closed", options.IsClosed.IsTrue()))
}
if options.NoLabelOnly {
query.And(inner_meilisearch.NewFilterEq("no_label", true))
} else {
if len(options.IncludedLabelIDs) > 0 {
q := &inner_meilisearch.FilterAnd{}
for _, labelID := range options.IncludedLabelIDs {
q.And(inner_meilisearch.NewFilterEq("label_ids", labelID))
}
query.And(q)
} else if len(options.IncludedAnyLabelIDs) > 0 {
query.And(inner_meilisearch.NewFilterIn("label_ids", options.IncludedAnyLabelIDs...))
}
if len(options.ExcludedLabelIDs) > 0 {
q := &inner_meilisearch.FilterAnd{}
for _, labelID := range options.ExcludedLabelIDs {
q.And(inner_meilisearch.NewFilterNot(inner_meilisearch.NewFilterEq("label_ids", labelID)))
}
query.And(q)
}
}
if len(options.MilestoneIDs) > 0 {
query.And(inner_meilisearch.NewFilterIn("milestone_id", options.MilestoneIDs...))
}
if options.ProjectID != nil {
query.And(inner_meilisearch.NewFilterEq("project_id", *options.ProjectID))
}
if options.ProjectBoardID != nil {
query.And(inner_meilisearch.NewFilterEq("project_board_id", *options.ProjectBoardID))
}
if options.PosterID != nil {
query.And(inner_meilisearch.NewFilterEq("poster_id", *options.PosterID))
}
if options.AssigneeID != nil {
query.And(inner_meilisearch.NewFilterEq("assignee_id", *options.AssigneeID))
}
if options.MentionID != nil {
query.And(inner_meilisearch.NewFilterEq("mention_ids", *options.MentionID))
}
if options.ReviewedID != nil {
query.And(inner_meilisearch.NewFilterEq("reviewed_ids", *options.ReviewedID))
}
if options.ReviewRequestedID != nil {
query.And(inner_meilisearch.NewFilterEq("review_requested_ids", *options.ReviewRequestedID))
}
if options.SubscriberID != nil {
query.And(inner_meilisearch.NewFilterEq("subscriber_ids", *options.SubscriberID))
}
if options.UpdatedAfterUnix != nil {
query.And(inner_meilisearch.NewFilterGte("updated_unix", *options.UpdatedAfterUnix))
}
if options.UpdatedBeforeUnix != nil {
query.And(inner_meilisearch.NewFilterLte("updated_unix", *options.UpdatedBeforeUnix))
}
if options.SortBy == "" {
options.SortBy = internal.SortByCreatedAsc
}
sortBy := []string{
parseSortBy(options.SortBy),
"id:desc",
}
skip, limit := indexer_internal.ParsePaginator(options.Paginator, maxTotalHits)
searchRes, err := b.inner.Client.Index(b.inner.VersionedIndexName()).Search(options.Keyword, &meilisearch.SearchRequest{
Filter: query.Statement(),
Limit: int64(limit),
Offset: int64(skip),
Sort: sortBy,
})
if err != nil {
return nil, err
}
hits := make([]internal.Match, 0, len(searchRes.Hits))
for _, hit := range searchRes.Hits {
hits = append(hits, internal.Match{
ID: int64(hit.(map[string]any)["id"].(float64)),
})
}
return &internal.SearchResult{
Total: searchRes.EstimatedTotalHits,
Hits: hits,
}, nil
}
func parseSortBy(sortBy internal.SortBy) string {
field := strings.TrimPrefix(string(sortBy), "-")
if strings.HasPrefix(string(sortBy), "-") {
return field + ":desc"
}
return field + ":asc"
}