Fix pull request update showing too many commits with multiple branches (#22856)

When the base repository contains multiple branches with the same
commits as the base branch, pull requests can show a long list of
commits already in the base branch as having been added.

What this is supposed to do is exclude commits already in the base
branch. But the mechansim to do so assumed a commit only exists in a
single branch. Now use `git rev-list A B --not branchName` instead of
filtering commits afterwards.

The logic to detect if there was a force push also was wrong for
multiple branches. If the old commit existed in any branch in the base
repository it would assume there was no force push. Instead check if the
old commit is an ancestor of the new commit.

modules/git/commit.go

@@ -218,6 +218,19 @@ func (c *Commit) HasPreviousCommit(commitHash SHA1) (bool, error) {
 	return false, err
 }
 
+// IsForcePush returns true if a push from oldCommitHash to this is a force push
+func (c *Commit) IsForcePush(oldCommitID string) (bool, error) {
+	if oldCommitID == EmptySHA {
+		return false, nil
+	}
+	oldCommit, err := c.repo.GetCommit(oldCommitID)
+	if err != nil {
+		return false, err
+	}
+	hasPreviousCommit, err := c.HasPreviousCommit(oldCommit.ID)
+	return !hasPreviousCommit, err
+}
+
 // CommitsBeforeLimit returns num commits before current revision
 func (c *Commit) CommitsBeforeLimit(num int) ([]*Commit, error) {
 	return c.repo.getCommitsBeforeLimit(c.ID, num)

modules/git/repo_commit.go

@@ -323,6 +323,27 @@ func (repo *Repository) CommitsBetweenLimit(last, before *Commit, limit, skip in
 	return repo.parsePrettyFormatLogToList(bytes.TrimSpace(stdout))
 }
 
+// CommitsBetweenNotBase returns a list that contains commits between [before, last), excluding commits in baseBranch.
+// If before is detached (removed by reset + push) it is not included.
+func (repo *Repository) CommitsBetweenNotBase(last, before *Commit, baseBranch string) ([]*Commit, error) {
+	var stdout []byte
+	var err error
+	if before == nil {
+		stdout, _, err = NewCommand(repo.Ctx, "rev-list").AddDynamicArguments(last.ID.String()).AddOptionValues("--not", baseBranch).RunStdBytes(&RunOpts{Dir: repo.Path})
+	} else {
+		stdout, _, err = NewCommand(repo.Ctx, "rev-list").AddDynamicArguments(before.ID.String()+".."+last.ID.String()).AddOptionValues("--not", baseBranch).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 before last so let's try that...
+			stdout, _, err = NewCommand(repo.Ctx, "rev-list").AddDynamicArguments(before.ID.String(), last.ID.String()).AddOptionValues("--not", baseBranch).RunStdBytes(&RunOpts{Dir: repo.Path})
+		}
+	}
+	if err != nil {
+		return nil, err
+	}
+	return repo.parsePrettyFormatLogToList(bytes.TrimSpace(stdout))
+}
+
 // CommitsBetweenIDs return commits between twoe commits
 func (repo *Repository) CommitsBetweenIDs(last, before string) ([]*Commit, error) {
 	lastCommit, err := repo.GetCommit(last)

modules/repository/push.go

@@ -4,10 +4,8 @@
 package repository
 
 import (
-	"context"
 	"strings"
 
-	repo_model "code.gitea.io/gitea/models/repo"
 	"code.gitea.io/gitea/modules/git"
 )
 
@@ -96,19 +94,3 @@ func (opts *PushUpdateOptions) RefName() string {
 func (opts *PushUpdateOptions) RepoFullName() string {
 	return opts.RepoUserName + "/" + opts.RepoName
 }
-
-// IsForcePush detect if a push is a force push
-func IsForcePush(ctx context.Context, opts *PushUpdateOptions) (bool, error) {
-	if !opts.IsUpdateBranch() {
-		return false, nil
-	}
-
-	output, _, err := git.NewCommand(ctx, "rev-list", "--max-count=1").AddDynamicArguments(opts.OldCommitID, "^"+opts.NewCommitID).
-		RunStdString(&git.RunOpts{Dir: repo_model.RepoPath(opts.RepoUserName, opts.RepoName)})
-	if err != nil {
-		return false, err
-	} else if len(output) > 0 {
-		return true, nil
-	}
-	return false, nil
-}