Files
forge/modules/git/catfile_batch.go
zeekayandHanzo Dev aa87442fdd Record upstream authorship centrally in NOTICE instead of in every file header
Source files keep their SPDX-License-Identifier and Hanzo copyright line; the
Gitea, Gogs and LFS Test Server copyrights they used to repeat now live once in
NOTICE, alongside LICENSE. The header lint accepts any copyright line above the
SPDX line, so it no longer demands an upstream one.

Co-authored-by: Hanzo Dev <[email protected]>
2026-09-10 17:50:35 -07:00

53 lines
1.9 KiB
Go

// Copyright 2026 Hanzo AI, Inc. All rights reserved.
// SPDX-License-Identifier: MIT
package git
import (
"context"
"io"
)
type BufferedReader interface {
io.Reader
Buffered() int
Peek(n int) ([]byte, error)
Discard(n int) (int, error)
ReadString(sep byte) (string, error)
ReadSlice(sep byte) ([]byte, error)
ReadBytes(sep byte) ([]byte, error)
}
type CatFileObject struct {
ID string
Type string
Size int64
}
type CatFileBatch interface {
// QueryInfo queries the object info from the git repository by its object name using "git cat-file --batch" family commands.
// "git cat-file" accepts "<rev>" for the object name, it can be a ref name, object id, etc. https://git-scm.com/docs/gitrevisions
// In the forge, we only use the simple ref name or object id, no other complex rev syntax like "suffix" or "git describe" although they are supported by git.
QueryInfo(obj string) (*CatFileObject, error)
// QueryContent is similar to QueryInfo, it queries the object info and additionally returns a reader for its content.
// FIXME: this design still follows the old pattern: the returned BufferedReader is very fragile,
// callers should carefully maintain its lifecycle and discard all unread data.
// TODO: It needs to be refactored to a fully managed Reader stream in the future, don't let callers manually Close or Discard
QueryContent(obj string) (*CatFileObject, BufferedReader, error)
}
type CatFileBatchCloser interface {
CatFileBatch
Close()
}
// NewBatch creates a "batch object provider (CatFileBatch)" for the given repository path to retrieve object info and content efficiently.
// The CatFileBatch and the readers create by it should only be used in the same goroutine.
func NewBatch(ctx context.Context, repoPath string) (CatFileBatchCloser, error) {
if DefaultFeatures().SupportCatFileBatchCommand {
return newCatFileBatchCommand(ctx, repoPath)
}
return newCatFileBatchLegacy(ctx, repoPath)
}