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>

1 files changed, 46 insertions, 44 deletions

diff --git a/modules/git/repo_commit.go b/modules/git/repo_commit.go
index 8343e34843..a62e0670fe 100644
--- a/modules/git/repo_commit.go
+++ b/modules/git/repo_commit.go
@@ -89,7 +89,7 @@ func (repo *Repository) GetCommitByPath(relpath string) (*Commit, error) {
 
 func (repo *Repository) commitsByRange(id SHA1, page, pageSize int) ([]*Commit, error) {
 	stdout, _, err := NewCommand(repo.Ctx, "log").
-		AddArguments(CmdArg("--skip="+strconv.Itoa((page-1)*pageSize)), CmdArg("--max-count="+strconv.Itoa(pageSize)), prettyLogFormat).
+		AddOptionFormat("--skip=%d", (page-1)*pageSize).AddOptionFormat("--max-count=%d", pageSize).AddArguments(prettyLogFormat).
 		AddDynamicArguments(id.String()).
 		RunStdBytes(&RunOpts{Dir: repo.Path})
 	if err != nil {
@@ -99,32 +99,36 @@ func (repo *Repository) commitsByRange(id SHA1, page, pageSize int) ([]*Commit,
 }
 
 func (repo *Repository) searchCommits(id SHA1, opts SearchCommitsOptions) ([]*Commit, error) {
-	// create new git log command with limit of 100 commis
-	cmd := NewCommand(repo.Ctx, "log", "-100", prettyLogFormat).AddDynamicArguments(id.String())
-	// ignore case
-	args := []CmdArg{"-i"}
+	// add common arguments to git command
+	addCommonSearchArgs := func(c *Command) {
+		// ignore case
+		c.AddArguments("-i")
+
+		// add authors if present in search query
+		if len(opts.Authors) > 0 {
+			for _, v := range opts.Authors {
+				c.AddOptionFormat("--author=%s", v)
+			}
+		}
 
-	// add authors if present in search query
-	if len(opts.Authors) > 0 {
-		for _, v := range opts.Authors {
-			args = append(args, CmdArg("--author="+v))
+		// add committers if present in search query
+		if len(opts.Committers) > 0 {
+			for _, v := range opts.Committers {
+				c.AddOptionFormat("--committer=%s", v)
+			}
 		}
-	}
 
-	// add committers if present in search query
-	if len(opts.Committers) > 0 {
-		for _, v := range opts.Committers {
-			args = append(args, CmdArg("--committer="+v))
+		// add time constraints if present in search query
+		if len(opts.After) > 0 {
+			c.AddOptionFormat("--after=%s", opts.After)
+		}
+		if len(opts.Before) > 0 {
+			c.AddOptionFormat("--before=%s", opts.Before)
 		}
 	}
 
-	// add time constraints if present in search query
-	if len(opts.After) > 0 {
-		args = append(args, CmdArg("--after="+opts.After))
-	}
-	if len(opts.Before) > 0 {
-		args = append(args, CmdArg("--before="+opts.Before))
-	}
+	// create new git log command with limit of 100 commits
+	cmd := NewCommand(repo.Ctx, "log", "-100", prettyLogFormat).AddDynamicArguments(id.String())
 
 	// pretend that all refs along with HEAD were listed on command line as <commis>
 	// https://git-scm.com/docs/git-log#Documentation/git-log.txt---all
@@ -137,12 +141,12 @@ func (repo *Repository) searchCommits(id SHA1, opts SearchCommitsOptions) ([]*Co
 	// note this is done only for command created above
 	if len(opts.Keywords) > 0 {
 		for _, v := range opts.Keywords {
-			cmd.AddArguments(CmdArg("--grep=" + v))
+			cmd.AddOptionFormat("--grep=%s", v)
 		}
 	}
 
 	// search for commits matching given constraints and keywords in commit msg
-	cmd.AddArguments(args...)
+	addCommonSearchArgs(cmd)
 	stdout, _, err := cmd.RunStdBytes(&RunOpts{Dir: repo.Path})
 	if err != nil {
 		return nil, err
@@ -160,7 +164,7 @@ func (repo *Repository) searchCommits(id SHA1, opts SearchCommitsOptions) ([]*Co
 				// create new git log command with 1 commit limit
 				hashCmd := NewCommand(repo.Ctx, "log", "-1", prettyLogFormat)
 				// add previous arguments except for --grep and --all
-				hashCmd.AddArguments(args...)
+				addCommonSearchArgs(hashCmd)
 				// add keyword as <commit>
 				hashCmd.AddDynamicArguments(v)
 
@@ -213,8 +217,8 @@ func (repo *Repository) CommitsByFileAndRange(revision, file string, page int) (
 	go func() {
 		stderr := strings.Builder{}
 		gitCmd := NewCommand(repo.Ctx, "rev-list").
-			AddArguments(CmdArg("--max-count=" + strconv.Itoa(setting.Git.CommitsRangeSize*page))).
-			AddArguments(CmdArg("--skip=" + strconv.Itoa(skip)))
+			AddOptionFormat("--max-count=%d", setting.Git.CommitsRangeSize*page).
+			AddOptionFormat("--skip=%d", skip)
 		gitCmd.AddDynamicArguments(revision)
 		gitCmd.AddDashesAndList(file)
 		err := gitCmd.Run(&RunOpts{
@@ -295,21 +299,21 @@ func (repo *Repository) CommitsBetweenLimit(last, before *Commit, limit, skip in
 	var stdout []byte
 	var err error
 	if before == nil {
-		stdout, _, err = NewCommand(repo.Ctx, "rev-list",
-			"--max-count", CmdArg(strconv.Itoa(limit)),
-			"--skip", CmdArg(strconv.Itoa(skip))).
+		stdout, _, err = NewCommand(repo.Ctx, "rev-list").
+			AddOptionValues("--max-count", strconv.Itoa(limit)).
+			AddOptionValues("--skip", strconv.Itoa(skip)).
 			AddDynamicArguments(last.ID.String()).RunStdBytes(&RunOpts{Dir: repo.Path})
 	} else {
-		stdout, _, err = NewCommand(repo.Ctx, "rev-list",
-			"--max-count", CmdArg(strconv.Itoa(limit)),
-			"--skip", CmdArg(strconv.Itoa(skip))).
+		stdout, _, err = NewCommand(repo.Ctx, "rev-list").
+			AddOptionValues("--max-count", strconv.Itoa(limit)).
+			AddOptionValues("--skip", strconv.Itoa(skip)).
 			AddDynamicArguments(before.ID.String() + ".." + last.ID.String()).RunStdBytes(&RunOpts{Dir: repo.Path})
 		if err != nil && strings.Contains(err.Error(), "no merge base") {
 			// future versions of git >= 2.28 are likely to return an error if before and last have become unrelated.
 			// previously it would return the results of git rev-list --max-count n before last so let's try that...
-			stdout, _, err = NewCommand(repo.Ctx, "rev-list",
-				"--max-count", CmdArg(strconv.Itoa(limit)),
-				"--skip", CmdArg(strconv.Itoa(skip))).
+			stdout, _, err = NewCommand(repo.Ctx, "rev-list").
+				AddOptionValues("--max-count", strconv.Itoa(limit)).
+				AddOptionValues("--skip", strconv.Itoa(skip)).
 				AddDynamicArguments(before.ID.String(), last.ID.String()).RunStdBytes(&RunOpts{Dir: repo.Path})
 		}
 	}
@@ -349,12 +353,11 @@ func (repo *Repository) CommitsCountBetween(start, end string) (int64, error) {
 
 // commitsBefore the limit is depth, not total number of returned commits.
 func (repo *Repository) commitsBefore(id SHA1, limit int) ([]*Commit, error) {
-	cmd := NewCommand(repo.Ctx, "log")
+	cmd := NewCommand(repo.Ctx, "log", prettyLogFormat)
 	if limit > 0 {
-		cmd.AddArguments(CmdArg("-"+strconv.Itoa(limit)), prettyLogFormat).AddDynamicArguments(id.String())
-	} else {
-		cmd.AddArguments(prettyLogFormat).AddDynamicArguments(id.String())
+		cmd.AddOptionFormat("-%d", limit)
 	}
+	cmd.AddDynamicArguments(id.String())
 
 	stdout, _, runErr := cmd.RunStdBytes(&RunOpts{Dir: repo.Path})
 	if runErr != nil {
@@ -393,10 +396,9 @@ func (repo *Repository) getCommitsBeforeLimit(id SHA1, num int) ([]*Commit, erro
 
 func (repo *Repository) getBranches(commit *Commit, limit int) ([]string, error) {
 	if CheckGitVersionAtLeast("2.7.0") == nil {
-		stdout, _, err := NewCommand(repo.Ctx, "for-each-ref",
-			CmdArg("--count="+strconv.Itoa(limit)),
-			"--format=%(refname:strip=2)", "--contains").
-			AddDynamicArguments(commit.ID.String(), BranchPrefix).
+		stdout, _, err := NewCommand(repo.Ctx, "for-each-ref", "--format=%(refname:strip=2)").
+			AddOptionFormat("--count=%d", limit).
+			AddOptionValues("--contains", commit.ID.String(), BranchPrefix).
 			RunStdString(&RunOpts{Dir: repo.Path})
 		if err != nil {
 			return nil, err
@@ -406,7 +408,7 @@ func (repo *Repository) getBranches(commit *Commit, limit int) ([]string, error)
 		return branches, nil
 	}
 
-	stdout, _, err := NewCommand(repo.Ctx, "branch", "--contains").AddDynamicArguments(commit.ID.String()).RunStdString(&RunOpts{Dir: repo.Path})
+	stdout, _, err := NewCommand(repo.Ctx, "branch").AddOptionValues("--contains", commit.ID.String()).RunStdString(&RunOpts{Dir: repo.Path})
 	if err != nil {
 		return nil, err
 	}