Merge pull request #707 from pjbgf/experimental-sha256

*: Add support for initializing SHA256 repositories

.github/workflows/git.yml

@@ -42,6 +42,9 @@ jobs:
     - name: Test
       run: make test-coverage
 
+    - name: Test SHA256
+      run: make test-sha256
+
     - name: Build go-git with CGO disabled
       run: go build ./...
       env:

Makefile

@@ -29,6 +29,12 @@ test:
 	@echo "running against `git version`"; \
 	$(GOTEST) -race ./...
 
+TEMP_REPO := $(shell mktemp)
+test-sha256:
+	$(GOCMD) run -tags sha256 _examples/sha256/main.go $(TEMP_REPO)
+	cd $(TEMP_REPO) && git fsck
+	rm -rf $(TEMP_REPO)
+
 test-coverage:
 	@echo "running against `git version`"; \
 	echo "" > $(COVERAGE_REPORT); \

_examples/sha256/main.go

@@ -0,0 +1,66 @@
+package main
+
+import (
+	"fmt"
+	"io/ioutil"
+	"os"
+	"path/filepath"
+	"time"
+
+	"github.com/go-git/go-git/v5"
+	. "github.com/go-git/go-git/v5/_examples"
+	"github.com/go-git/go-git/v5/plumbing/format/config"
+	"github.com/go-git/go-git/v5/plumbing/object"
+)
+
+// This example requires building with the sha256 tag for it to work:
+// go run -tags sha256 main.go /tmp/repository
+
+// Basic example of how to initialise a repository using sha256 as the hashing algorithmn.
+func main() {
+	CheckArgs("<directory>")
+	directory := os.Args[1]
+
+	os.RemoveAll(directory)
+
+	// Init a new repository using the ObjectFormat SHA256.
+	r, err := git.PlainInitWithOptions(directory, &git.PlainInitOptions{ObjectFormat: config.SHA256})
+	CheckIfError(err)
+
+	w, err := r.Worktree()
+	CheckIfError(err)
+
+	// ... we need a file to commit so let's create a new file inside of the
+	// worktree of the project using the go standard library.
+	Info("echo \"hello world!\" > example-git-file")
+	filename := filepath.Join(directory, "example-git-file")
+	err = ioutil.WriteFile(filename, []byte("hello world!"), 0644)
+	CheckIfError(err)
+
+	// Adds the new file to the staging area.
+	Info("git add example-git-file")
+	_, err = w.Add("example-git-file")
+	CheckIfError(err)
+
+	// Commits the current staging area to the repository, with the new file
+	// just created. We should provide the object.Signature of Author of the
+	// commit Since version 5.0.1, we can omit the Author signature, being read
+	// from the git config files.
+	Info("git commit -m \"example go-git commit\"")
+	commit, err := w.Commit("example go-git commit", &git.CommitOptions{
+		Author: &object.Signature{
+			Name:  "John Doe",
+			Email: "john@doe.org",
+			When:  time.Now(),
+		},
+	})
+
+	CheckIfError(err)
+
+	// Prints the current HEAD to verify that all worked well.
+	Info("git show -s")
+	obj, err := r.CommitObject(commit)
+	CheckIfError(err)
+
+	fmt.Println(obj)
+}

config/config.go

@@ -59,6 +59,8 @@ type Config struct {
 		// CommentChar is the character indicating the start of a
 		// comment for commands like commit and tag
 		CommentChar string
+		// RepositoryFormatVersion identifies the repository format and layout version.
+		RepositoryFormatVersion format.RepositoryFormatVersion
 	}
 
 	User struct {
@@ -96,6 +98,17 @@ type Config struct {
 		DefaultBranch string
 	}
 
+	Extensions struct {
+		// ObjectFormat specifies the hash algorithm to use. The
+		// acceptable values are sha1 and sha256. If not specified,
+		// sha1 is assumed. It is an error to specify this key unless
+		// core.repositoryFormatVersion is 1.
+		//
+		// This setting must not be changed after repository initialization
+		// (e.g. clone or init).
+		ObjectFormat format.ObjectFormat
+	}
+
 	// Remotes list of repository remotes, the key of the map is the name
 	// of the remote, should equal to RemoteConfig.Name.
 	Remotes map[string]*RemoteConfig
@@ -226,28 +239,31 @@ func (c *Config) Validate() error {
 }
 
 const (
-	remoteSection    = "remote"
-	submoduleSection = "submodule"
-	branchSection    = "branch"
-	coreSection      = "core"
-	packSection      = "pack"
-	userSection      = "user"
-	authorSection    = "author"
-	committerSection = "committer"
-	initSection      = "init"
-	urlSection       = "url"
-	fetchKey         = "fetch"
-	urlKey           = "url"
-	bareKey          = "bare"
-	worktreeKey      = "worktree"
-	commentCharKey   = "commentChar"
-	windowKey        = "window"
-	mergeKey         = "merge"
-	rebaseKey        = "rebase"
-	nameKey          = "name"
-	emailKey         = "email"
-	descriptionKey   = "description"
-	defaultBranchKey = "defaultBranch"
+	remoteSection              = "remote"
+	submoduleSection           = "submodule"
+	branchSection              = "branch"
+	coreSection                = "core"
+	packSection                = "pack"
+	userSection                = "user"
+	authorSection              = "author"
+	committerSection           = "committer"
+	initSection                = "init"
+	urlSection                 = "url"
+	extensionsSection          = "extensions"
+	fetchKey                   = "fetch"
+	urlKey                     = "url"
+	bareKey                    = "bare"
+	worktreeKey                = "worktree"
+	commentCharKey             = "commentChar"
+	windowKey                  = "window"
+	mergeKey                   = "merge"
+	rebaseKey                  = "rebase"
+	nameKey                    = "name"
+	emailKey                   = "email"
+	descriptionKey             = "description"
+	defaultBranchKey           = "defaultBranch"
+	repositoryFormatVersionKey = "repositoryformatversion"
+	objectFormat               = "objectformat"
 
 	// DefaultPackWindow holds the number of previous objects used to
 	// generate deltas. The value 10 is the same used by git command.
@@ -391,6 +407,7 @@ func (c *Config) unmarshalInit() {
 // Marshal returns Config encoded as a git-config file.
 func (c *Config) Marshal() ([]byte, error) {
 	c.marshalCore()
+	c.marshalExtensions()
 	c.marshalUser()
 	c.marshalPack()
 	c.marshalRemotes()
@@ -410,12 +427,24 @@ func (c *Config) Marshal() ([]byte, error) {
 func (c *Config) marshalCore() {
 	s := c.Raw.Section(coreSection)
 	s.SetOption(bareKey, fmt.Sprintf("%t", c.Core.IsBare))
+	if string(c.Core.RepositoryFormatVersion) != "" {
+		s.SetOption(repositoryFormatVersionKey, string(c.Core.RepositoryFormatVersion))
+	}
 
 	if c.Core.Worktree != "" {
 		s.SetOption(worktreeKey, c.Core.Worktree)
 	}
 }
 
+func (c *Config) marshalExtensions() {
+	// Extensions are only supported on Version 1, therefore
+	// ignore them otherwise.
+	if c.Core.RepositoryFormatVersion == format.Version_1 {
+		s := c.Raw.Section(extensionsSection)
+		s.SetOption(objectFormat, string(c.Extensions.ObjectFormat))
+	}
+}
+
 func (c *Config) marshalUser() {
 	s := c.Raw.Section(userSection)
 	if c.User.Name != "" {

options.go

@@ -10,6 +10,7 @@ import (
 	"github.com/ProtonMail/go-crypto/openpgp"
 	"github.com/go-git/go-git/v5/config"
 	"github.com/go-git/go-git/v5/plumbing"
+	formatcfg "github.com/go-git/go-git/v5/plumbing/format/config"
 	"github.com/go-git/go-git/v5/plumbing/object"
 	"github.com/go-git/go-git/v5/plumbing/protocol/packp/sideband"
 	"github.com/go-git/go-git/v5/plumbing/transport"
@@ -672,3 +673,10 @@ type PlainOpenOptions struct {
 
 // Validate validates the fields and sets the default values.
 func (o *PlainOpenOptions) Validate() error { return nil }
+
+type PlainInitOptions struct {
+	ObjectFormat formatcfg.ObjectFormat
+}
+
+// Validate validates the fields and sets the default values.
+func (o *PlainInitOptions) Validate() error { return nil }

plumbing/format/commitgraph/encoder.go

@@ -1,7 +1,6 @@
 package commitgraph
 
 import (
-	"crypto"
 	"io"
 
 	"github.com/go-git/go-git/v5/plumbing"
@@ -17,7 +16,7 @@ type Encoder struct {
 
 // NewEncoder returns a new stream encoder that writes to w.
 func NewEncoder(w io.Writer) *Encoder {
-	h := hash.New(crypto.SHA1)
+	h := hash.New(hash.CryptoType)
 	mw := io.MultiWriter(w, h)
 	return &Encoder{mw, h}
 }
@@ -31,7 +30,7 @@ func (e *Encoder) Encode(idx Index) error {
 	hashToIndex, fanout, extraEdgesCount := e.prepare(idx, hashes)
 
 	chunkSignatures := [][]byte{oidFanoutSignature, oidLookupSignature, commitDataSignature}
-	chunkSizes := []uint64{4 * 256, uint64(len(hashes)) * 20, uint64(len(hashes)) * 36}
+	chunkSizes := []uint64{4 * 256, uint64(len(hashes)) * hash.Size, uint64(len(hashes)) * 36}
 	if extraEdgesCount > 0 {
 		chunkSignatures = append(chunkSignatures, extraEdgeListSignature)
 		chunkSizes = append(chunkSizes, uint64(extraEdgesCount)*4)
@@ -183,6 +182,6 @@ func (e *Encoder) encodeExtraEdges(extraEdges []uint32) (err error) {
 }
 
 func (e *Encoder) encodeChecksum() error {
-	_, err := e.Write(e.hash.Sum(nil)[:20])
+	_, err := e.Write(e.hash.Sum(nil)[:hash.Size])
 	return err
 }

plumbing/format/config/format.go

@@ -0,0 +1,53 @@
+package config
+
+// RepositoryFormatVersion represents the repository format version,
+// as per defined at:
+//
+//	https://git-scm.com/docs/repository-version
+type RepositoryFormatVersion string
+
+const (
+	// Version_0 is the format defined by the initial version of git,
+	// including but not limited to the format of the repository
+	// directory, the repository configuration file, and the object
+	// and ref storage.
+	//
+	// Specifying the complete behavior of git is beyond the scope
+	// of this document.
+	Version_0 = "0"
+
+	// Version_1 is identical to version 0, with the following exceptions:
+	//
+	//   1. When reading the core.repositoryformatversion variable, a git
+	//		implementation which supports version 1 MUST also read any
+	//		configuration keys found in the extensions section of the
+	//		configuration file.
+	//
+	//	 2. If a version-1 repository specifies any extensions.* keys that
+	//		the running git has not implemented, the operation MUST NOT proceed.
+	//		Similarly, if the value of any known key is not understood by the
+	//		implementation, the operation MUST NOT proceed.
+	//
+	// Note that if no extensions are specified in the config file, then
+	// core.repositoryformatversion SHOULD be set to 0 (setting it to 1 provides
+	// no benefit, and makes the repository incompatible with older
+	// implementations of git).
+	Version_1 = "1"
+
+	// DefaultRepositoryFormatVersion holds the default repository format version.
+	DefaultRepositoryFormatVersion = Version_0
+)
+
+// ObjectFormat defines the object format.
+type ObjectFormat string
+
+const (
+	// SHA1 represents the object format used for SHA1.
+	SHA1 ObjectFormat = "sha1"
+
+	// SHA256 represents the object format used for SHA256.
+	SHA256 ObjectFormat = "sha256"
+
+	// DefaultObjectFormat holds the default object format.
+	DefaultObjectFormat = SHA1
+)

plumbing/format/idxfile/decoder.go

@@ -6,6 +6,7 @@ import (
 	"errors"
 	"io"
 
+	"github.com/go-git/go-git/v5/plumbing/hash"
 	"github.com/go-git/go-git/v5/utils/binary"
 )
 
@@ -19,7 +20,7 @@ var (
 
 const (
 	fanout         = 256
-	objectIDLength = 20
+	objectIDLength = hash.Size
 )
 
 // Decoder reads and decodes idx files from an input stream.

plumbing/format/idxfile/encoder.go

@@ -1,7 +1,6 @@
 package idxfile
 
 import (
-	"crypto"
 	"io"
 
 	"github.com/go-git/go-git/v5/plumbing/hash"
@@ -16,7 +15,7 @@ type Encoder struct {
 
 // NewEncoder returns a new stream encoder that writes to w.
 func NewEncoder(w io.Writer) *Encoder {
-	h := hash.New(crypto.SHA1)
+	h := hash.New(hash.CryptoType)
 	mw := io.MultiWriter(w, h)
 	return &Encoder{mw, h}
 }
@@ -133,10 +132,10 @@ func (e *Encoder) encodeChecksums(idx *MemoryIndex) (int, error) {
 		return 0, err
 	}
 
-	copy(idx.IdxChecksum[:], e.hash.Sum(nil)[:20])
+	copy(idx.IdxChecksum[:], e.hash.Sum(nil)[:hash.Size])
 	if _, err := e.Write(idx.IdxChecksum[:]); err != nil {
 		return 0, err
 	}
 
-	return 40, nil
+	return hash.HexSize, nil
 }

plumbing/format/idxfile/idxfile.go

@@ -8,6 +8,7 @@ import (
 	encbin "encoding/binary"
 
 	"github.com/go-git/go-git/v5/plumbing"
+	"github.com/go-git/go-git/v5/plumbing/hash"
 )
 
 const (
@@ -53,8 +54,8 @@ type MemoryIndex struct {
 	Offset32         [][]byte
 	CRC32            [][]byte
 	Offset64         []byte
-	PackfileChecksum [20]byte
-	IdxChecksum      [20]byte
+	PackfileChecksum [hash.Size]byte
+	IdxChecksum      [hash.Size]byte
 
 	offsetHash       map[int64]plumbing.Hash
 	offsetHashIsFull bool

plumbing/format/index/decoder.go

@@ -3,7 +3,6 @@ package index
 import (
 	"bufio"
 	"bytes"
-	"crypto"
 	"errors"
 	"io"
 	"io/ioutil"
@@ -49,7 +48,7 @@ type Decoder struct {
 
 // NewDecoder returns a new decoder that reads from r.
 func NewDecoder(r io.Reader) *Decoder {
-	h := hash.New(crypto.SHA1)
+	h := hash.New(hash.CryptoType)
 	return &Decoder{
 		r:         io.TeeReader(r, h),
 		hash:      h,