Refactor git command package to improve security and maintainability (#22678)

This PR follows #21535 (and replace #22592)

## Review without space diff

https://github.com/go-gitea/gitea/pull/22678/files?diff=split&w=1

## Purpose of this PR

1. Make git module command completely safe (risky user inputs won't be
passed as argument option anymore)
2. Avoid low-level mistakes like
https://github.com/go-gitea/gitea/pull/22098#discussion_r1045234918
3. Remove deprecated and dirty `CmdArgCheck` function, hide the `CmdArg`
type
4. Simplify code when using git command

## The main idea of this PR

* Move the `git.CmdArg` to the `internal` package, then no other package
except `git` could use it. Then developers could never do
`AddArguments(git.CmdArg(userInput))` any more.
* Introduce `git.ToTrustedCmdArgs`, it's for user-provided and already
trusted arguments. It's only used in a few cases, for example: use git
arguments from config file, help unit test with some arguments.
* Introduce `AddOptionValues` and `AddOptionFormat`, they make code more
clear and simple:
    * Before: `AddArguments("-m").AddDynamicArguments(message)`
    * After: `AddOptionValues("-m", message)`
    * -
* Before: `AddArguments(git.CmdArg(fmt.Sprintf("--author='%s <%s>'",
sig.Name, sig.Email)))`
* After: `AddOptionFormat("--author='%s <%s>'", sig.Name, sig.Email)`

## FAQ

### Why these changes were not done in #21535 ?

#21535 is mainly a search&replace, it did its best to not change too
much logic.

Making the framework better needs a lot of changes, so this separate PR
is needed as the second step.


### The naming of `AddOptionXxx`

According to git's manual, the `--xxx` part is called `option`.

### How can it guarantee that `internal.CmdArg` won't be not misused?

Go's specification guarantees that. Trying to access other package's
internal package causes compilation error.

And, `golangci-lint` also denies the git/internal package. Only the
`git/command.go` can use it carefully.

### There is still a `ToTrustedCmdArgs`, will it still allow developers
to make mistakes and pass untrusted arguments?

Generally speaking, no. Because when using `ToTrustedCmdArgs`, the code
will be very complex (see the changes for examples). Then developers and
reviewers can know that something might be unreasonable.

### Why there was a `CmdArgCheck` and why it's removed?

At the moment of #21535, to reduce unnecessary changes, `CmdArgCheck`
was introduced as a hacky patch. Now, almost all code could be written
as `cmd := NewCommand(); cmd.AddXxx(...)`, then there is no need for
`CmdArgCheck` anymore.


### Why many codes for `signArg == ""` is deleted?

Because in the old code, `signArg` could never be empty string, it's
either `-S[key-id]` or `--no-gpg-sign`. So the `signArg == ""` is just
dead code.

---------

Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>

5 files changed, 17 insertions, 22 deletions

diff --git a/services/repository/check.go b/services/repository/check.go
index e9d65aea4a..3a1f0b7f30 100644
--- a/services/repository/check.go
+++ b/services/repository/check.go
@@ -23,7 +23,7 @@ import (
 )
 
 // GitFsckRepos calls 'git fsck' to check repository health.
-func GitFsckRepos(ctx context.Context, timeout time.Duration, args []git.CmdArg) error {
+func GitFsckRepos(ctx context.Context, timeout time.Duration, args git.TrustedCmdArgs) error {
 	log.Trace("Doing: GitFsck")
 
 	if err := db.Iterate(
@@ -47,10 +47,10 @@ func GitFsckRepos(ctx context.Context, timeout time.Duration, args []git.CmdArg)
 }
 
 // GitFsckRepo calls 'git fsck' to check an individual repository's health.
-func GitFsckRepo(ctx context.Context, repo *repo_model.Repository, timeout time.Duration, args []git.CmdArg) error {
+func GitFsckRepo(ctx context.Context, repo *repo_model.Repository, timeout time.Duration, args git.TrustedCmdArgs) error {
 	log.Trace("Running health check on repository %-v", repo)
 	repoPath := repo.RepoPath()
-	if err := git.Fsck(ctx, repoPath, timeout, args...); err != nil {
+	if err := git.Fsck(ctx, repoPath, timeout, args); err != nil {
 		log.Warn("Failed to health check repository (%-v): %v", repo, err)
 		if err = system_model.CreateRepositoryNotice("Failed to health check repository (%s): %v", repo.FullName(), err); err != nil {
 			log.Error("CreateRepositoryNotice: %v", err)
@@ -60,9 +60,8 @@ func GitFsckRepo(ctx context.Context, repo *repo_model.Repository, timeout time.
 }
 
 // GitGcRepos calls 'git gc' to remove unnecessary files and optimize the local repository
-func GitGcRepos(ctx context.Context, timeout time.Duration, args ...git.CmdArg) error {
+func GitGcRepos(ctx context.Context, timeout time.Duration, args git.TrustedCmdArgs) error {
 	log.Trace("Doing: GitGcRepos")
-	args = append([]git.CmdArg{"gc"}, args...)
 
 	if err := db.Iterate(
 		ctx,
@@ -86,9 +85,9 @@ func GitGcRepos(ctx context.Context, timeout time.Duration, args ...git.CmdArg)
 }
 
 // GitGcRepo calls 'git gc' to remove unnecessary files and optimize the local repository
-func GitGcRepo(ctx context.Context, repo *repo_model.Repository, timeout time.Duration, args []git.CmdArg) error {
+func GitGcRepo(ctx context.Context, repo *repo_model.Repository, timeout time.Duration, args git.TrustedCmdArgs) error {
 	log.Trace("Running git gc on %-v", repo)
-	command := git.NewCommand(ctx, args...).
+	command := git.NewCommand(ctx, "gc").AddArguments(args...).
 		SetDescription(fmt.Sprintf("Repository Garbage Collection: %s", repo.FullName()))
 	var stdout string
 	var err error
diff --git a/services/repository/files/patch.go b/services/repository/files/patch.go
index 73ee0fa815..f65199cfcb 100644
--- a/services/repository/files/patch.go
+++ b/services/repository/files/patch.go
@@ -141,14 +141,12 @@ func ApplyDiffPatch(ctx context.Context, repo *repo_model.Repository, doer *user
 	stdout := &strings.Builder{}
 	stderr := &strings.Builder{}
 
-	args := []git.CmdArg{"apply", "--index", "--recount", "--cached", "--ignore-whitespace", "--whitespace=fix", "--binary"}
-
+	cmdApply := git.NewCommand(ctx, "apply", "--index", "--recount", "--cached", "--ignore-whitespace", "--whitespace=fix", "--binary")
 	if git.CheckGitVersionAtLeast("2.32") == nil {
-		args = append(args, "-3")
+		cmdApply.AddArguments("-3")
 	}
 
-	cmd := git.NewCommand(ctx, args...)
-	if err := cmd.Run(&git.RunOpts{
+	if err := cmdApply.Run(&git.RunOpts{
 		Dir:    t.basePath,
 		Stdout: stdout,
 		Stderr: stderr,
diff --git a/services/repository/files/temp_repo.go b/services/repository/files/temp_repo.go
index 1f3375cdcc..a086d15a4f 100644
--- a/services/repository/files/temp_repo.go
+++ b/services/repository/files/temp_repo.go
@@ -233,11 +233,9 @@ func (t *TemporaryUploadRepository) CommitTreeWithDate(parent string, author, co
 	_, _ = messageBytes.WriteString(message)
 	_, _ = messageBytes.WriteString("\n")
 
-	var args []git.CmdArg
+	cmdCommitTree := git.NewCommand(t.ctx, "commit-tree").AddDynamicArguments(treeHash)
 	if parent != "" {
-		args = []git.CmdArg{"commit-tree", git.CmdArgCheck(treeHash), "-p", git.CmdArgCheck(parent)}
-	} else {
-		args = []git.CmdArg{"commit-tree", git.CmdArgCheck(treeHash)}
+		cmdCommitTree.AddOptionValues("-p", parent)
 	}
 
 	var sign bool
@@ -249,7 +247,7 @@ func (t *TemporaryUploadRepository) CommitTreeWithDate(parent string, author, co
 		sign, keyID, signer, _ = asymkey_service.SignInitialCommit(t.ctx, t.repo.RepoPath(), author)
 	}
 	if sign {
-		args = append(args, git.CmdArg("-S"+keyID))
+		cmdCommitTree.AddOptionFormat("-S%s", keyID)
 		if t.repo.GetTrustModel() == repo_model.CommitterTrustModel || t.repo.GetTrustModel() == repo_model.CollaboratorCommitterTrustModel {
 			if committerSig.Name != authorSig.Name || committerSig.Email != authorSig.Email {
 				// Add trailers
@@ -264,7 +262,7 @@ func (t *TemporaryUploadRepository) CommitTreeWithDate(parent string, author, co
 			committerSig = signer
 		}
 	} else {
-		args = append(args, "--no-gpg-sign")
+		cmdCommitTree.AddArguments("--no-gpg-sign")
 	}
 
 	if signoff {
@@ -281,7 +279,7 @@ func (t *TemporaryUploadRepository) CommitTreeWithDate(parent string, author, co
 
 	stdout := new(bytes.Buffer)
 	stderr := new(bytes.Buffer)
-	if err := git.NewCommand(t.ctx, args...).
+	if err := cmdCommitTree.
 		Run(&git.RunOpts{
 			Env:    env,
 			Dir:    t.basePath,
@@ -364,7 +362,7 @@ func (t *TemporaryUploadRepository) DiffIndex() (*gitdiff.Diff, error) {
 			t.repo.FullName(), err, stderr)
 	}
 
-	diff.NumFiles, diff.TotalAddition, diff.TotalDeletion, err = git.GetDiffShortStat(t.ctx, t.basePath, "--cached", "HEAD")
+	diff.NumFiles, diff.TotalAddition, diff.TotalDeletion, err = git.GetDiffShortStat(t.ctx, t.basePath, git.TrustedCmdArgs{"--cached"}, "HEAD")
 	if err != nil {
 		return nil, err
 	}
diff --git a/services/repository/files/update.go b/services/repository/files/update.go
index 58b7a5e082..45a4692396 100644
--- a/services/repository/files/update.go
+++ b/services/repository/files/update.go
@@ -370,7 +370,7 @@ func CreateOrUpdateRepoFile(ctx context.Context, repo *repo_model.Repository, do
 	if setting.LFS.StartServer && hasOldBranch {
 		// Check there is no way this can return multiple infos
 		filename2attribute2info, err := t.gitRepo.CheckAttribute(git.CheckAttributeOpts{
-			Attributes: []git.CmdArg{"filter"},
+			Attributes: []string{"filter"},
 			Filenames:  []string{treePath},
 			CachedOnly: true,
 		})
diff --git a/services/repository/files/upload.go b/services/repository/files/upload.go
index e7289dd60d..cf2f7019b6 100644
--- a/services/repository/files/upload.go
+++ b/services/repository/files/upload.go
@@ -96,7 +96,7 @@ func UploadRepoFiles(ctx context.Context, repo *repo_model.Repository, doer *use
 	var filename2attribute2info map[string]map[string]string
 	if setting.LFS.StartServer {
 		filename2attribute2info, err = t.gitRepo.CheckAttribute(git.CheckAttributeOpts{
-			Attributes: []git.CmdArg{"filter"},
+			Attributes: []string{"filter"},
 			Filenames:  names,
 			CachedOnly: true,
 		})