forgejo-federation/modules/git/grep.go
wxiaoguang 1e7a6483b8 Support repo code search without setting up an indexer (#29998)
By using git's ability, end users (especially small instance users) do
not need to enable the indexer, they could also benefit from the code
searching feature.

Fix #29996

![image](https://github.com/go-gitea/gitea/assets/2114189/11b7e458-88a4-480d-b4d7-72ee59406dd1)

![image](https://github.com/go-gitea/gitea/assets/2114189/0fe777d5-c95c-4288-a818-0427680805b6)

---------

Co-authored-by: silverwind <me@silverwind.io>
2024-03-28 05:16:51 +00:00

112 lines
2.7 KiB
Go

// Copyright 2024 The Gitea Authors. All rights reserved.
// SPDX-License-Identifier: MIT
package git
import (
"bufio"
"bytes"
"context"
"errors"
"fmt"
"os"
"strconv"
"strings"
"code.gitea.io/gitea/modules/util"
)
type GrepResult struct {
Filename string
LineNumbers []int
LineCodes []string
}
type GrepOptions struct {
RefName string
ContextLineNumber int
IsFuzzy bool
}
func GrepSearch(ctx context.Context, repo *Repository, search string, opts GrepOptions) ([]*GrepResult, error) {
stdoutReader, stdoutWriter, err := os.Pipe()
if err != nil {
return nil, fmt.Errorf("unable to create os pipe to grep: %w", err)
}
defer func() {
_ = stdoutReader.Close()
_ = stdoutWriter.Close()
}()
/*
The output is like this ( "^@" means \x00):
HEAD:.air.toml
6^@bin = "gitea"
HEAD:.changelog.yml
2^@repo: go-gitea/gitea
*/
var results []*GrepResult
cmd := NewCommand(ctx, "grep", "--null", "--break", "--heading", "--fixed-strings", "--line-number", "--ignore-case", "--full-name")
cmd.AddOptionValues("--context", fmt.Sprint(opts.ContextLineNumber))
if opts.IsFuzzy {
words := strings.Fields(search)
for _, word := range words {
cmd.AddOptionValues("-e", strings.TrimLeft(word, "-"))
}
} else {
cmd.AddOptionValues("-e", strings.TrimLeft(search, "-"))
}
cmd.AddDynamicArguments(util.IfZero(opts.RefName, "HEAD"))
stderr := bytes.Buffer{}
err = cmd.Run(&RunOpts{
Dir: repo.Path,
Stdout: stdoutWriter,
Stderr: &stderr,
PipelineFunc: func(ctx context.Context, cancel context.CancelFunc) error {
_ = stdoutWriter.Close()
defer stdoutReader.Close()
isInBlock := false
scanner := bufio.NewScanner(stdoutReader)
var res *GrepResult
for scanner.Scan() {
line := scanner.Text()
if !isInBlock {
if _ /* ref */, filename, ok := strings.Cut(line, ":"); ok {
isInBlock = true
res = &GrepResult{Filename: filename}
results = append(results, res)
}
continue
}
if line == "" {
if len(results) >= 50 {
cancel()
break
}
isInBlock = false
continue
}
if line == "--" {
continue
}
if lineNum, lineCode, ok := strings.Cut(line, "\x00"); ok {
lineNumInt, _ := strconv.Atoi(lineNum)
res.LineNumbers = append(res.LineNumbers, lineNumInt)
res.LineCodes = append(res.LineCodes, lineCode)
}
}
return scanner.Err()
},
})
// git grep exits with 1 if no results are found
if IsErrorExitCode(err, 1) && stderr.Len() == 0 {
return nil, nil
}
if err != nil && !errors.Is(err, context.Canceled) {
return nil, fmt.Errorf("unable to run git grep: %w, stderr: %s", err, stderr.String())
}
return results, nil
}