1
1
mirror of https://github.com/go-gitea/gitea synced 2024-06-01 00:45:46 +00:00
gitea/models/issues/issue_search.go

490 lines
16 KiB
Go
Raw Normal View History

// Copyright 2023 The Gitea Authors. All rights reserved.
// SPDX-License-Identifier: MIT
package issues
import (
"context"
"fmt"
"strings"
"code.gitea.io/gitea/models/db"
"code.gitea.io/gitea/models/organization"
repo_model "code.gitea.io/gitea/models/repo"
"code.gitea.io/gitea/models/unit"
user_model "code.gitea.io/gitea/models/user"
"code.gitea.io/gitea/modules/util"
"xorm.io/builder"
"xorm.io/xorm"
)
// IssuesOptions represents options of an issue.
type IssuesOptions struct { //nolint
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
db.Paginator
RepoIDs []int64 // overwrites RepoCond if the length is not 0
RepoCond builder.Cond
AssigneeID int64
PosterID int64
MentionedID int64
ReviewRequestedID int64
ReviewedID int64
SubscriberID int64
MilestoneIDs []int64
ProjectID int64
ProjectBoardID int64
IsClosed util.OptionalBool
IsPull util.OptionalBool
LabelIDs []int64
IncludedLabelNames []string
ExcludedLabelNames []string
IncludeMilestones []string
SortType string
IssueIDs []int64
UpdatedAfterUnix int64
UpdatedBeforeUnix int64
// prioritize issues from this repo
PriorityRepoID int64
IsArchived util.OptionalBool
Org *organization.Organization // issues permission scope
Team *organization.Team // issues permission scope
User *user_model.User // issues permission scope
}
// applySorts sort an issues-related session based on the provided
// sortType string
func applySorts(sess *xorm.Session, sortType string, priorityRepoID int64) {
switch sortType {
case "oldest":
sess.Asc("issue.created_unix").Asc("issue.id")
case "recentupdate":
sess.Desc("issue.updated_unix").Desc("issue.created_unix").Desc("issue.id")
case "leastupdate":
sess.Asc("issue.updated_unix").Asc("issue.created_unix").Asc("issue.id")
case "mostcomment":
sess.Desc("issue.num_comments").Desc("issue.created_unix").Desc("issue.id")
case "leastcomment":
sess.Asc("issue.num_comments").Desc("issue.created_unix").Desc("issue.id")
case "priority":
sess.Desc("issue.priority").Desc("issue.created_unix").Desc("issue.id")
case "nearduedate":
// 253370764800 is 01/01/9999 @ 12:00am (UTC)
sess.Join("LEFT", "milestone", "issue.milestone_id = milestone.id").
OrderBy("CASE " +
"WHEN issue.deadline_unix = 0 AND (milestone.deadline_unix = 0 OR milestone.deadline_unix IS NULL) THEN 253370764800 " +
"WHEN milestone.deadline_unix = 0 OR milestone.deadline_unix IS NULL THEN issue.deadline_unix " +
"WHEN milestone.deadline_unix < issue.deadline_unix OR issue.deadline_unix = 0 THEN milestone.deadline_unix " +
"ELSE issue.deadline_unix END ASC").
Desc("issue.created_unix").
Desc("issue.id")
case "farduedate":
sess.Join("LEFT", "milestone", "issue.milestone_id = milestone.id").
OrderBy("CASE " +
"WHEN milestone.deadline_unix IS NULL THEN issue.deadline_unix " +
"WHEN milestone.deadline_unix < issue.deadline_unix OR issue.deadline_unix = 0 THEN milestone.deadline_unix " +
"ELSE issue.deadline_unix END DESC").
Desc("issue.created_unix").
Desc("issue.id")
case "priorityrepo":
sess.OrderBy("CASE "+
"WHEN issue.repo_id = ? THEN 1 "+
"ELSE 2 END ASC", priorityRepoID).
Desc("issue.created_unix").
Desc("issue.id")
case "project-column-sorting":
sess.Asc("project_issue.sorting").Desc("issue.created_unix").Desc("issue.id")
default:
sess.Desc("issue.created_unix").Desc("issue.id")
}
}
func applyLimit(sess *xorm.Session, opts *IssuesOptions) *xorm.Session {
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
if opts.Paginator == nil || opts.Paginator.IsListAll() {
return sess
}
// Warning: Do not use GetSkipTake() for *db.ListOptions
// Its implementation could reset the page size with setting.API.MaxResponseItems
if listOptions, ok := opts.Paginator.(*db.ListOptions); ok {
if listOptions.Page >= 0 && listOptions.PageSize > 0 {
var start int
if listOptions.Page == 0 {
start = 0
} else {
start = (listOptions.Page - 1) * listOptions.PageSize
}
sess.Limit(listOptions.PageSize, start)
}
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
return sess
}
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
start, limit := opts.Paginator.GetSkipTake()
sess.Limit(limit, start)
return sess
}
func applyLabelsCondition(sess *xorm.Session, opts *IssuesOptions) *xorm.Session {
if len(opts.LabelIDs) > 0 {
if opts.LabelIDs[0] == 0 {
sess.Where("issue.id NOT IN (SELECT issue_id FROM issue_label)")
} else {
for i, labelID := range opts.LabelIDs {
if labelID > 0 {
sess.Join("INNER", fmt.Sprintf("issue_label il%d", i),
fmt.Sprintf("issue.id = il%[1]d.issue_id AND il%[1]d.label_id = %[2]d", i, labelID))
} else if labelID < 0 { // 0 is not supported here, so just ignore it
sess.Where("issue.id not in (select issue_id from issue_label where label_id = ?)", -labelID)
}
}
}
}
if len(opts.IncludedLabelNames) > 0 {
sess.In("issue.id", BuildLabelNamesIssueIDsCondition(opts.IncludedLabelNames))
}
if len(opts.ExcludedLabelNames) > 0 {
sess.And(builder.NotIn("issue.id", BuildLabelNamesIssueIDsCondition(opts.ExcludedLabelNames)))
}
return sess
}
func applyMilestoneCondition(sess *xorm.Session, opts *IssuesOptions) *xorm.Session {
if len(opts.MilestoneIDs) == 1 && opts.MilestoneIDs[0] == db.NoConditionID {
sess.And("issue.milestone_id = 0")
} else if len(opts.MilestoneIDs) > 0 {
sess.In("issue.milestone_id", opts.MilestoneIDs)
}
if len(opts.IncludeMilestones) > 0 {
sess.In("issue.milestone_id",
builder.Select("id").
From("milestone").
Where(builder.In("name", opts.IncludeMilestones)))
}
return sess
}
func applyProjectCondition(sess *xorm.Session, opts *IssuesOptions) *xorm.Session {
if opts.ProjectID > 0 { // specific project
sess.Join("INNER", "project_issue", "issue.id = project_issue.issue_id").
And("project_issue.project_id=?", opts.ProjectID)
} else if opts.ProjectID == db.NoConditionID { // show those that are in no project
sess.And(builder.NotIn("issue.id", builder.Select("issue_id").From("project_issue").And(builder.Neq{"project_id": 0})))
}
// opts.ProjectID == 0 means all projects,
// do not need to apply any condition
return sess
}
func applyRepoConditions(sess *xorm.Session, opts *IssuesOptions) *xorm.Session {
if len(opts.RepoIDs) == 1 {
opts.RepoCond = builder.Eq{"issue.repo_id": opts.RepoIDs[0]}
} else if len(opts.RepoIDs) > 1 {
opts.RepoCond = builder.In("issue.repo_id", opts.RepoIDs)
}
if opts.RepoCond != nil {
sess.And(opts.RepoCond)
}
return sess
}
func applyConditions(sess *xorm.Session, opts *IssuesOptions) *xorm.Session {
if len(opts.IssueIDs) > 0 {
sess.In("issue.id", opts.IssueIDs)
}
applyRepoConditions(sess, opts)
if !opts.IsClosed.IsNone() {
sess.And("issue.is_closed=?", opts.IsClosed.IsTrue())
}
if opts.AssigneeID > 0 {
applyAssigneeCondition(sess, opts.AssigneeID)
} else if opts.AssigneeID == db.NoConditionID {
sess.Where("issue.id NOT IN (SELECT issue_id FROM issue_assignees)")
}
if opts.PosterID > 0 {
applyPosterCondition(sess, opts.PosterID)
}
if opts.MentionedID > 0 {
applyMentionedCondition(sess, opts.MentionedID)
}
if opts.ReviewRequestedID > 0 {
applyReviewRequestedCondition(sess, opts.ReviewRequestedID)
}
if opts.ReviewedID > 0 {
applyReviewedCondition(sess, opts.ReviewedID)
}
if opts.SubscriberID > 0 {
applySubscribedCondition(sess, opts.SubscriberID)
}
applyMilestoneCondition(sess, opts)
if opts.UpdatedAfterUnix != 0 {
sess.And(builder.Gte{"issue.updated_unix": opts.UpdatedAfterUnix})
}
if opts.UpdatedBeforeUnix != 0 {
sess.And(builder.Lte{"issue.updated_unix": opts.UpdatedBeforeUnix})
}
applyProjectCondition(sess, opts)
if opts.ProjectBoardID != 0 {
if opts.ProjectBoardID > 0 {
sess.In("issue.id", builder.Select("issue_id").From("project_issue").Where(builder.Eq{"project_board_id": opts.ProjectBoardID}))
} else {
sess.In("issue.id", builder.Select("issue_id").From("project_issue").Where(builder.Eq{"project_board_id": 0}))
}
}
switch opts.IsPull {
case util.OptionalBoolTrue:
sess.And("issue.is_pull=?", true)
case util.OptionalBoolFalse:
sess.And("issue.is_pull=?", false)
}
if opts.IsArchived != util.OptionalBoolNone {
sess.And(builder.Eq{"repository.is_archived": opts.IsArchived.IsTrue()})
}
applyLabelsCondition(sess, opts)
if opts.User != nil {
sess.And(issuePullAccessibleRepoCond("issue.repo_id", opts.User.ID, opts.Org, opts.Team, opts.IsPull.IsTrue()))
}
return sess
}
// teamUnitsRepoCond returns query condition for those repo id in the special org team with special units access
func teamUnitsRepoCond(id string, userID, orgID, teamID int64, units ...unit.Type) builder.Cond {
return builder.In(id,
builder.Select("repo_id").From("team_repo").Where(
builder.Eq{
"team_id": teamID,
}.And(
builder.Or(
// Check if the user is member of the team.
builder.In(
"team_id", builder.Select("team_id").From("team_user").Where(
builder.Eq{
"uid": userID,
},
),
),
// Check if the user is in the owner team of the organisation.
builder.Exists(builder.Select("team_id").From("team_user").
Where(builder.Eq{
"org_id": orgID,
"team_id": builder.Select("id").From("team").Where(
builder.Eq{
"org_id": orgID,
"lower_name": strings.ToLower(organization.OwnerTeamName),
}),
"uid": userID,
}),
),
)).And(
builder.In(
"team_id", builder.Select("team_id").From("team_unit").Where(
builder.Eq{
"`team_unit`.org_id": orgID,
}.And(
builder.In("`team_unit`.type", units),
),
),
),
),
))
}
// issuePullAccessibleRepoCond userID must not be zero, this condition require join repository table
func issuePullAccessibleRepoCond(repoIDstr string, userID int64, org *organization.Organization, team *organization.Team, isPull bool) builder.Cond {
cond := builder.NewCond()
unitType := unit.TypeIssues
if isPull {
unitType = unit.TypePullRequests
}
if org != nil {
if team != nil {
cond = cond.And(teamUnitsRepoCond(repoIDstr, userID, org.ID, team.ID, unitType)) // special team member repos
} else {
cond = cond.And(
builder.Or(
repo_model.UserOrgUnitRepoCond(repoIDstr, userID, org.ID, unitType), // team member repos
repo_model.UserOrgPublicUnitRepoCond(userID, org.ID), // user org public non-member repos, TODO: check repo has issues
),
)
}
} else {
cond = cond.And(
builder.Or(
repo_model.UserOwnedRepoCond(userID), // owned repos
repo_model.UserAccessRepoCond(repoIDstr, userID), // user can access repo in a unit independent way
repo_model.UserAssignedRepoCond(repoIDstr, userID), // user has been assigned accessible public repos
repo_model.UserMentionedRepoCond(repoIDstr, userID), // user has been mentioned accessible public repos
repo_model.UserCreateIssueRepoCond(repoIDstr, userID, isPull), // user has created issue/pr accessible public repos
),
)
}
return cond
}
func applyAssigneeCondition(sess *xorm.Session, assigneeID int64) *xorm.Session {
return sess.Join("INNER", "issue_assignees", "issue.id = issue_assignees.issue_id").
And("issue_assignees.assignee_id = ?", assigneeID)
}
func applyPosterCondition(sess *xorm.Session, posterID int64) *xorm.Session {
return sess.And("issue.poster_id=?", posterID)
}
func applyMentionedCondition(sess *xorm.Session, mentionedID int64) *xorm.Session {
return sess.Join("INNER", "issue_user", "issue.id = issue_user.issue_id").
And("issue_user.is_mentioned = ?", true).
And("issue_user.uid = ?", mentionedID)
}
func applyReviewRequestedCondition(sess *xorm.Session, reviewRequestedID int64) *xorm.Session {
return sess.Join("INNER", []string{"review", "r"}, "issue.id = r.issue_id").
And("issue.poster_id <> ?", reviewRequestedID).
And("r.type = ?", ReviewTypeRequest).
And("r.reviewer_id = ? and r.id in (select max(id) from review where issue_id = r.issue_id and reviewer_id = r.reviewer_id and type in (?, ?, ?))"+
" or r.reviewer_team_id in (select team_id from team_user where uid = ?)",
reviewRequestedID, ReviewTypeApprove, ReviewTypeReject, ReviewTypeRequest, reviewRequestedID)
}
func applyReviewedCondition(sess *xorm.Session, reviewedID int64) *xorm.Session {
// Query for pull requests where you are a reviewer or commenter, excluding
// any pull requests already returned by the the review requested filter.
notPoster := builder.Neq{"issue.poster_id": reviewedID}
reviewed := builder.In("issue.id", builder.
Select("issue_id").
From("review").
Where(builder.And(
builder.Neq{"type": ReviewTypeRequest},
builder.Or(
builder.Eq{"reviewer_id": reviewedID},
builder.In("reviewer_team_id", builder.
Select("team_id").
From("team_user").
Where(builder.Eq{"uid": reviewedID}),
),
),
)),
)
commented := builder.In("issue.id", builder.
Select("issue_id").
From("comment").
Where(builder.And(
builder.Eq{"poster_id": reviewedID},
builder.In("type", CommentTypeComment, CommentTypeCode, CommentTypeReview),
)),
)
return sess.And(notPoster, builder.Or(reviewed, commented))
}
func applySubscribedCondition(sess *xorm.Session, subscriberID int64) *xorm.Session {
return sess.And(
builder.
NotIn("issue.id",
builder.Select("issue_id").
From("issue_watch").
Where(builder.Eq{"is_watching": false, "user_id": subscriberID}),
),
).And(
builder.Or(
builder.In("issue.id", builder.
Select("issue_id").
From("issue_watch").
Where(builder.Eq{"is_watching": true, "user_id": subscriberID}),
),
builder.In("issue.id", builder.
Select("issue_id").
From("comment").
Where(builder.Eq{"poster_id": subscriberID}),
),
builder.Eq{"issue.poster_id": subscriberID},
builder.In("issue.repo_id", builder.
Select("id").
From("watch").
Where(builder.And(builder.Eq{"user_id": subscriberID},
builder.In("mode", repo_model.WatchModeNormal, repo_model.WatchModeAuto))),
),
),
)
}
// GetRepoIDsForIssuesOptions find all repo ids for the given options
func GetRepoIDsForIssuesOptions(opts *IssuesOptions, user *user_model.User) ([]int64, error) {
repoIDs := make([]int64, 0, 5)
e := db.GetEngine(db.DefaultContext)
sess := e.Join("INNER", "repository", "`issue`.repo_id = `repository`.id")
applyConditions(sess, opts)
accessCond := repo_model.AccessibleRepositoryCondition(user, unit.TypeInvalid)
if err := sess.Where(accessCond).
Distinct("issue.repo_id").
Table("issue").
Find(&repoIDs); err != nil {
return nil, fmt.Errorf("unable to GetRepoIDsForIssuesOptions: %w", err)
}
return repoIDs, nil
}
// Issues returns a list of issues by given conditions.
func Issues(ctx context.Context, opts *IssuesOptions) (IssueList, error) {
sess := db.GetEngine(ctx).
Join("INNER", "repository", "`issue`.repo_id = `repository`.id")
applyLimit(sess, opts)
applyConditions(sess, opts)
applySorts(sess, opts.SortType, opts.PriorityRepoID)
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
issues := IssueList{}
if err := sess.Find(&issues); err != nil {
return nil, fmt.Errorf("unable to query Issues: %w", err)
}
if err := issues.LoadAttributes(ctx); err != nil {
return nil, fmt.Errorf("unable to LoadAttributes for Issues: %w", err)
}
return issues, nil
}
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
// IssueIDs returns a list of issue ids by given conditions.
func IssueIDs(ctx context.Context, opts *IssuesOptions, otherConds ...builder.Cond) ([]int64, int64, error) {
sess := db.GetEngine(ctx).
Join("INNER", "repository", "`issue`.repo_id = `repository`.id")
applyConditions(sess, opts)
for _, cond := range otherConds {
sess.And(cond)
}
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
applyLimit(sess, opts)
applySorts(sess, opts.SortType, opts.PriorityRepoID)
var res []int64
total, err := sess.Select("`issue`.id").Table(&Issue{}).FindAndCount(&res)
if err != nil {
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
return nil, 0, err
}
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
return res, total, nil
}