add agentws prune

Agents leave their managed worktrees behind, and `agentws rm` takes one path at a time with no idea whether a branch's work is safely upstream, so clearing an accumulation by hand risks destroying unmerged commits. - classify every managed worktree: dirty, PR open, upstream, or unproven - remove only what is safe; delete the local branch only when work is upstream - prove "upstream" with merge-base and git cherry, so squash merges count - match a PR by head.label, which survives the branch deletion a merge does - dry run by default; --yes applies, --keep-branches spares every branch - read the Gitea path from origin's URL rather than assuming the owner
2026-09-09 23:41:09 +10:00
parent 5c0eb1e899
commit 4bbeaae8f0
11 changed files with 1156 additions and 2 deletions
@@ -0,0 +1,456 @@
+package main
+
+import (
+	"bytes"
+	"encoding/json"
+	"net/http"
+	"net/http/httptest"
+	"os"
+	"os/exec"
+	"path/filepath"
+	"strconv"
+	"strings"
+	"testing"
+
+	"git.unkin.net/unkin/agent-tools/internal/agent"
+)
+
+// fixture is a bare origin plus a source checkout named "repo" and a worktree
+// root, wired so managedWorktrees() finds the worktrees created here.
+type fixture struct {
+	root   string
+	bare   string
+	srcDir string
+	wtRoot string
+}
+
+func git(t *testing.T, dir string, args ...string) string {
+	t.Helper()
+	cmd := exec.Command("git", args...)
+	cmd.Dir = dir
+	out, err := cmd.CombinedOutput()
+	if err != nil {
+		t.Fatalf("git %s (in %s): %v: %s", strings.Join(args, " "), dir, err, out)
+	}
+	return strings.TrimSpace(string(out))
+}
+
+func newFixture(t *testing.T) *fixture {
+	t.Helper()
+	root := t.TempDir()
+	f := &fixture{
+		root:   root,
+		bare:   filepath.Join(root, "origin.git"),
+		srcDir: filepath.Join(root, "src", "repo"),
+		wtRoot: filepath.Join(root, "worktrees"),
+	}
+	if err := os.MkdirAll(filepath.Join(root, "src"), 0o755); err != nil {
+		t.Fatal(err)
+	}
+	if err := os.MkdirAll(f.wtRoot, 0o755); err != nil {
+		t.Fatal(err)
+	}
+	git(t, root, "init", "--bare", "-b", "main", f.bare)
+
+	seed := filepath.Join(root, "seed")
+	git(t, root, "init", "-b", "main", seed)
+	identity(t, seed)
+	writeCommit(t, seed, "README.md", "hi\n", "init")
+	git(t, seed, "remote", "add", "origin", f.bare)
+	git(t, seed, "push", "-u", "origin", "main")
+
+	git(t, filepath.Join(root, "src"), "clone", f.bare, f.srcDir)
+	identity(t, f.srcDir)
+
+	t.Setenv("AGENTWS_ROOT", f.wtRoot)
+	t.Setenv("AGENTWS_SRC_ROOT", filepath.Join(root, "src"))
+	t.Setenv("AGENTWS_OWNER", "unkin")
+	return f
+}
+
+func identity(t *testing.T, dir string) {
+	t.Helper()
+	git(t, dir, "config", "user.email", "test@example.com")
+	git(t, dir, "config", "user.name", "Test")
+}
+
+func writeCommit(t *testing.T, dir, name, content, msg string) {
+	t.Helper()
+	if err := os.WriteFile(filepath.Join(dir, name), []byte(content), 0o644); err != nil {
+		t.Fatal(err)
+	}
+	git(t, dir, "add", ".")
+	git(t, dir, "commit", "-m", msg)
+}
+
+// addWorktree creates a managed worktree for branch and returns its path.
+func (f *fixture) addWorktree(t *testing.T, branch string) string {
+	t.Helper()
+	path := filepath.Join(f.wtRoot, agent.WorktreeDirName("repo", branch))
+	git(t, f.srcDir, "worktree", "add", path, "-b", branch, "origin/main")
+	identity(t, path)
+	return path
+}
+
+// landUpstream commits content on origin's main, mimicking a squash merge: the
+// same patch arrives upstream under a different SHA.
+func (f *fixture) landUpstream(t *testing.T, name, content, msg string) {
+	t.Helper()
+	seed := filepath.Join(f.root, "seed")
+	git(t, seed, "pull", "--ff-only", "origin", "main")
+	writeCommit(t, seed, name, content, msg)
+	git(t, seed, "push", "origin", "main")
+}
+
+// fakeGitea serves the pulls listing for unkin/repo with the given PR bodies.
+func fakeGitea(t *testing.T, prs ...map[string]any) *httptest.Server {
+	t.Helper()
+	mux := http.NewServeMux()
+	mux.HandleFunc("/api/v1/repos/unkin/repo/pulls", func(w http.ResponseWriter, r *http.Request) {
+		if r.URL.Query().Get("page") != "1" {
+			_, _ = w.Write([]byte("[]"))
+			return
+		}
+		body, err := json.Marshal(prs)
+		if err != nil {
+			t.Errorf("marshal PRs: %v", err)
+		}
+		_, _ = w.Write(body)
+	})
+	srv := httptest.NewServer(mux)
+	t.Cleanup(srv.Close)
+	return srv
+}
+
+func client(srv *httptest.Server) prLister {
+	return &agent.GiteaClient{BaseURL: srv.URL, HTTP: srv.Client()}
+}
+
+// mergedPR mimics Gitea after a merge: the branch is deleted, so head.ref
+// becomes refs/pull/<n>/head and only head.label still names the branch.
+func mergedPR(number int, branch string) map[string]any {
+	return map[string]any{
+		"number": number,
+		"state":  "closed",
+		"merged": true,
+		"head": map[string]any{
+			"ref":   "refs/pull/" + strconv.Itoa(number) + "/head",
+			"label": branch,
+		},
+	}
+}
+
+func openPR(number int, branch string) map[string]any {
+	return map[string]any{
+		"number": number,
+		"state":  "open",
+		"merged": false,
+		"head":   map[string]any{"ref": branch, "label": branch},
+	}
+}
+
+func closedPR(number int, branch string) map[string]any {
+	return map[string]any{
+		"number": number,
+		"state":  "closed",
+		"merged": false,
+		"head":   map[string]any{"ref": branch, "label": branch},
+	}
+}
+
+func run(t *testing.T, prs prLister, apply, keepBranches bool) string {
+	t.Helper()
+	var out bytes.Buffer
+	if err := runPrune(&out, prs, apply, keepBranches); err != nil {
+		t.Fatalf("runPrune: %v\n%s", err, out.String())
+	}
+	return out.String()
+}
+
+func lineFor(t *testing.T, out, branch string) string {
+	t.Helper()
+	for _, line := range strings.Split(out, "\n") {
+		if strings.Contains(line, " "+branch+" ") {
+			return line
+		}
+	}
+	t.Fatalf("no line for branch %q in:\n%s", branch, out)
+	return ""
+}
+
+func assertVerdict(t *testing.T, out, branch, verdict, reason string) {
+	t.Helper()
+	line := lineFor(t, out, branch)
+	fields := strings.Fields(line)
+	if len(fields) < 3 || fields[2] != verdict {
+		t.Errorf("branch %s: verdict line %q, want verdict %q", branch, line, verdict)
+	}
+	if reason != "" && !strings.Contains(line, reason) {
+		t.Errorf("branch %s: line %q, want reason containing %q", branch, line, reason)
+	}
+}
+
+func exists(path string) bool {
+	_, err := os.Stat(path)
+	return err == nil
+}
+
+// A merged PR's branch is deleted on merge, so its head.ref reads
+// refs/pull/<n>/head; classification must still see the merge (via head.label)
+// and remove the branch, not fall through to "no PR".
+func TestPruneMergedPRWithDeletedBranch(t *testing.T) {
+	f := newFixture(t)
+	wt := f.addWorktree(t, "benvin/merged")
+	writeCommit(t, wt, "m.txt", "m\n", "work")
+
+	srv := fakeGitea(t, mergedPR(3, "benvin/merged"))
+	out := run(t, client(srv), true, false)
+	assertVerdict(t, out, "benvin/merged", verdictRemoveBranch, "PR merged #3")
+
+	if exists(wt) {
+		t.Errorf("worktree %s should have been removed", wt)
+	}
+	if agent.GitBranchExists(f.srcDir, "benvin/merged") {
+		t.Error("branch of a merged PR should be deleted")
+	}
+}
+
+// The same PR without head.label: matching falls back to head.ref, which no
+// longer names the branch, so prune must not guess it is merged — the worktree
+// goes but the branch stays.
+func TestPruneMergedPRWithoutLabelKeepsBranch(t *testing.T) {
+	f := newFixture(t)
+	wt := f.addWorktree(t, "benvin/merged")
+	writeCommit(t, wt, "m.txt", "m\n", "work")
+
+	pr := mergedPR(3, "benvin/merged")
+	pr["head"].(map[string]any)["label"] = ""
+	srv := fakeGitea(t, pr)
+	out := run(t, client(srv), true, false)
+	assertVerdict(t, out, "benvin/merged", verdictRemove, "no PR")
+
+	if !agent.GitBranchExists(f.srcDir, "benvin/merged") {
+		t.Error("branch must survive when the PR could not be matched")
+	}
+}
+
+func TestPruneContainedBranch(t *testing.T) {
+	f := newFixture(t)
+	wt := f.addWorktree(t, "benvin/contained")
+
+	srv := fakeGitea(t)
+	out := run(t, client(srv), true, false)
+	assertVerdict(t, out, "benvin/contained", verdictRemoveBranch, "contained in origin/main")
+
+	if exists(wt) {
+		t.Error("contained worktree should have been removed")
+	}
+	if agent.GitBranchExists(f.srcDir, "benvin/contained") {
+		t.Error("contained branch should be deleted")
+	}
+}
+
+// Squash-merged work keeps a local SHA that is not upstream, so only the
+// patch-equivalence check proves it landed.
+func TestPruneCherryCleanBranch(t *testing.T) {
+	f := newFixture(t)
+	wt := f.addWorktree(t, "benvin/squashed")
+	writeCommit(t, wt, "s.txt", "same\n", "add s")
+	f.landUpstream(t, "s.txt", "same\n", "squashed s")
+
+	srv := fakeGitea(t)
+	out := run(t, client(srv), true, false)
+	assertVerdict(t, out, "benvin/squashed", verdictRemoveBranch, "cherry-clean")
+
+	if agent.GitBranchExists(f.srcDir, "benvin/squashed") {
+		t.Error("cherry-clean branch should be deleted")
+	}
+}
+
+// Uncommitted work outranks every other signal, including a branch that is
+// otherwise fully contained upstream.
+func TestPruneNeverTouchesDirtyWorktree(t *testing.T) {
+	f := newFixture(t)
+	wt := f.addWorktree(t, "benvin/dirty")
+	if err := os.WriteFile(filepath.Join(wt, "wip.txt"), []byte("wip\n"), 0o644); err != nil {
+		t.Fatal(err)
+	}
+
+	srv := fakeGitea(t, mergedPR(4, "benvin/dirty"))
+	out := run(t, client(srv), true, false)
+	assertVerdict(t, out, "benvin/dirty", verdictKeep, "dirty")
+
+	if !exists(wt) {
+		t.Error("dirty worktree must not be removed")
+	}
+	if !agent.GitBranchExists(f.srcDir, "benvin/dirty") {
+		t.Error("dirty worktree's branch must survive")
+	}
+}
+
+// An open PR is kept even when its commits are already upstream.
+func TestPruneKeepsOpenPR(t *testing.T) {
+	f := newFixture(t)
+	wt := f.addWorktree(t, "benvin/open")
+
+	srv := fakeGitea(t, openPR(5, "benvin/open"))
+	out := run(t, client(srv), true, false)
+	assertVerdict(t, out, "benvin/open", verdictKeep, "PR open #5")
+
+	if !exists(wt) {
+		t.Error("worktree with an open PR must not be removed")
+	}
+	if !agent.GitBranchExists(f.srcDir, "benvin/open") {
+		t.Error("branch with an open PR must not be deleted")
+	}
+}
+
+// Closed-unmerged with the branch still on origin: the work is not lost, so the
+// local branch goes too.
+func TestPruneClosedPRWithBranchOnOrigin(t *testing.T) {
+	f := newFixture(t)
+	wt := f.addWorktree(t, "benvin/closed")
+	writeCommit(t, wt, "c.txt", "c\n", "work")
+	git(t, wt, "push", "origin", "benvin/closed")
+
+	srv := fakeGitea(t, closedPR(6, "benvin/closed"))
+	out := run(t, client(srv), true, false)
+	assertVerdict(t, out, "benvin/closed", verdictRemoveBranch, "PR closed #6, branch on origin")
+
+	if exists(wt) {
+		t.Error("worktree should have been removed")
+	}
+	if agent.GitBranchExists(f.srcDir, "benvin/closed") {
+		t.Error("branch should be deleted while origin still has it")
+	}
+}
+
+// Closed-unmerged with nothing on origin: the commits exist only here, so the
+// branch is kept and only the worktree goes.
+func TestPruneClosedPRWithBranchGone(t *testing.T) {
+	f := newFixture(t)
+	wt := f.addWorktree(t, "benvin/orphan")
+	writeCommit(t, wt, "o.txt", "o\n", "work")
+
+	srv := fakeGitea(t, closedPR(7, "benvin/orphan"))
+	out := run(t, client(srv), true, false)
+	assertVerdict(t, out, "benvin/orphan", verdictRemove, "PR closed #7, branch gone")
+
+	if exists(wt) {
+		t.Error("worktree should have been removed")
+	}
+	if !agent.GitBranchExists(f.srcDir, "benvin/orphan") {
+		t.Error("branch must survive when origin does not have the commits")
+	}
+}
+
+func TestPruneNoPRKeepsBranch(t *testing.T) {
+	f := newFixture(t)
+	wt := f.addWorktree(t, "benvin/unpushed")
+	writeCommit(t, wt, "u.txt", "u\n", "work")
+
+	srv := fakeGitea(t)
+	out := run(t, client(srv), true, false)
+	assertVerdict(t, out, "benvin/unpushed", verdictRemove, "no PR")
+
+	if exists(wt) {
+		t.Error("worktree should have been removed")
+	}
+	if !agent.GitBranchExists(f.srcDir, "benvin/unpushed") {
+		t.Error("branch with unproven work must survive")
+	}
+}
+
+// The default run reports and changes nothing.
+func TestPruneDryRunChangesNothing(t *testing.T) {
+	f := newFixture(t)
+	contained := f.addWorktree(t, "benvin/contained")
+	merged := f.addWorktree(t, "benvin/merged")
+	writeCommit(t, merged, "m.txt", "m\n", "work")
+
+	srv := fakeGitea(t, mergedPR(8, "benvin/merged"))
+	out := run(t, client(srv), false, false)
+
+	if !strings.Contains(out, "dry run") {
+		t.Errorf("dry-run output should say so:\n%s", out)
+	}
+	assertVerdict(t, out, "benvin/contained", verdictRemoveBranch, "contained")
+	assertVerdict(t, out, "benvin/merged", verdictRemoveBranch, "PR merged #8")
+	if !exists(contained) || !exists(merged) {
+		t.Error("dry run must not remove worktrees")
+	}
+	if !agent.GitBranchExists(f.srcDir, "benvin/contained") || !agent.GitBranchExists(f.srcDir, "benvin/merged") {
+		t.Error("dry run must not delete branches")
+	}
+}
+
+// --keep-branches removes worktrees but leaves every branch alone.
+func TestPruneKeepBranches(t *testing.T) {
+	f := newFixture(t)
+	wt := f.addWorktree(t, "benvin/contained")
+
+	srv := fakeGitea(t)
+	out := run(t, client(srv), true, true)
+	assertVerdict(t, out, "benvin/contained", verdictRemoveBranch, "contained")
+
+	if exists(wt) {
+		t.Error("worktree should have been removed")
+	}
+	if !agent.GitBranchExists(f.srcDir, "benvin/contained") {
+		t.Error("--keep-branches must not delete the branch")
+	}
+}
+
+// With Gitea unreachable prune falls back to the git signals: provably-upstream
+// work is still cleaned up, and anything unproven keeps its branch.
+func TestPruneDegradesWhenGiteaUnreachable(t *testing.T) {
+	f := newFixture(t)
+	contained := f.addWorktree(t, "benvin/contained")
+	unproven := f.addWorktree(t, "benvin/unproven")
+	writeCommit(t, unproven, "u.txt", "u\n", "work")
+
+	dead := httptest.NewServer(http.NewServeMux())
+	c := &agent.GiteaClient{BaseURL: dead.URL, HTTP: dead.Client()}
+	dead.Close()
+
+	out := run(t, c, true, false)
+	if !strings.Contains(out, "git signals only") {
+		t.Errorf("output should note the Gitea failure:\n%s", out)
+	}
+	assertVerdict(t, out, "benvin/contained", verdictRemoveBranch, "contained")
+	assertVerdict(t, out, "benvin/unproven", verdictRemove, "PR state unknown")
+
+	if exists(contained) || exists(unproven) {
+		t.Error("both worktrees should have been removed")
+	}
+	if agent.GitBranchExists(f.srcDir, "benvin/contained") {
+		t.Error("contained branch is safe to delete without Gitea")
+	}
+	if !agent.GitBranchExists(f.srcDir, "benvin/unproven") {
+		t.Error("unproven branch must survive an unreachable Gitea")
+	}
+}
+
+func TestPruneNoWorktrees(t *testing.T) {
+	newFixture(t)
+	srv := fakeGitea(t)
+	if out := run(t, client(srv), true, false); !strings.Contains(out, "no managed worktrees") {
+		t.Errorf("output = %q", out)
+	}
+}
+
+// Managed repos are not all under AGENTWS_OWNER, so the Gitea path comes from
+// origin's URL; a non-Gitea remote falls back to the configured owner.
+func TestRepoPathFollowsOrigin(t *testing.T) {
+	t.Setenv("AGENTWS_OWNER", "unkin")
+	dir := t.TempDir()
+	git(t, dir, "init", "-b", "main", ".")
+	git(t, dir, "remote", "add", "origin", "https://git.unkin.net/unkinben/dotfiles.git")
+	if got := repoPath(dir, "dotfiles"); got != "unkinben/dotfiles" {
+		t.Errorf("repoPath = %q, want unkinben/dotfiles", got)
+	}
+	git(t, dir, "remote", "set-url", "origin", filepath.Join(dir, "origin.git"))
+	if got := repoPath(dir, "dotfiles"); got != "unkin/dotfiles" {
+		t.Errorf("repoPath for a local remote = %q, want unkin/dotfiles", got)
+	}
+}