0
0
Fork 0
mirror of https://codeberg.org/forgejo/forgejo.git synced 2025-01-27 15:50:09 +01:00
forgejo/modules/lfs/pointer_scanner_nogogit.go
zeripath 5cb0c9aa0d
Propagate context and ensure git commands run in request context ()
This PR continues the work in  by progressively ensuring that git
commands run within the request context.

This now means that the if there is a git repo already open in the context it will be used instead of reopening it.

Signed-off-by: Andrew Thornton <art27@cantab.net>
2022-01-19 23:26:57 +00:00

113 lines
3.2 KiB
Go

// Copyright 2021 The Gitea Authors. All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
//go:build !gogit
// +build !gogit
package lfs
import (
"bufio"
"context"
"io"
"strconv"
"strings"
"sync"
"code.gitea.io/gitea/modules/git"
"code.gitea.io/gitea/modules/git/pipeline"
)
// SearchPointerBlobs scans the whole repository for LFS pointer files
func SearchPointerBlobs(ctx context.Context, repo *git.Repository, pointerChan chan<- PointerBlob, errChan chan<- error) {
basePath := repo.Path
catFileCheckReader, catFileCheckWriter := io.Pipe()
shasToBatchReader, shasToBatchWriter := io.Pipe()
catFileBatchReader, catFileBatchWriter := io.Pipe()
wg := sync.WaitGroup{}
wg.Add(4)
// Create the go-routines in reverse order.
// 4. Take the output of cat-file --batch and check if each file in turn
// to see if they're pointers to files in the LFS store
go createPointerResultsFromCatFileBatch(ctx, catFileBatchReader, &wg, pointerChan)
// 3. Take the shas of the blobs and batch read them
go pipeline.CatFileBatch(ctx, shasToBatchReader, catFileBatchWriter, &wg, basePath)
// 2. From the provided objects restrict to blobs <=1k
go pipeline.BlobsLessThan1024FromCatFileBatchCheck(catFileCheckReader, shasToBatchWriter, &wg)
// 1. Run batch-check on all objects in the repository
if git.CheckGitVersionAtLeast("2.6.0") != nil {
revListReader, revListWriter := io.Pipe()
shasToCheckReader, shasToCheckWriter := io.Pipe()
wg.Add(2)
go pipeline.CatFileBatchCheck(ctx, shasToCheckReader, catFileCheckWriter, &wg, basePath)
go pipeline.BlobsFromRevListObjects(revListReader, shasToCheckWriter, &wg)
go pipeline.RevListAllObjects(ctx, revListWriter, &wg, basePath, errChan)
} else {
go pipeline.CatFileBatchCheckAllObjects(ctx, catFileCheckWriter, &wg, basePath, errChan)
}
wg.Wait()
close(pointerChan)
close(errChan)
}
func createPointerResultsFromCatFileBatch(ctx context.Context, catFileBatchReader *io.PipeReader, wg *sync.WaitGroup, pointerChan chan<- PointerBlob) {
defer wg.Done()
defer catFileBatchReader.Close()
bufferedReader := bufio.NewReader(catFileBatchReader)
buf := make([]byte, 1025)
loop:
for {
select {
case <-ctx.Done():
break loop
default:
}
// File descriptor line: sha
sha, err := bufferedReader.ReadString(' ')
if err != nil {
_ = catFileBatchReader.CloseWithError(err)
break
}
sha = strings.TrimSpace(sha)
// Throw away the blob
if _, err := bufferedReader.ReadString(' '); err != nil {
_ = catFileBatchReader.CloseWithError(err)
break
}
sizeStr, err := bufferedReader.ReadString('\n')
if err != nil {
_ = catFileBatchReader.CloseWithError(err)
break
}
size, err := strconv.Atoi(sizeStr[:len(sizeStr)-1])
if err != nil {
_ = catFileBatchReader.CloseWithError(err)
break
}
pointerBuf := buf[:size+1]
if _, err := io.ReadFull(bufferedReader, pointerBuf); err != nil {
_ = catFileBatchReader.CloseWithError(err)
break
}
pointerBuf = pointerBuf[:size]
// Now we need to check if the pointerBuf is an LFS pointer
pointer, _ := ReadPointerFromBuffer(pointerBuf)
if !pointer.IsValid() {
continue
}
pointerChan <- PointerBlob{Hash: sha, Pointer: pointer}
}
}