hermes-agent/tests/tools/test_working_diff.py
teknium1 0fa5e41c86 feat(diff): cross-surface /diff with staged/all/session modes
Widen the cherry-picked /diff base (#4839 by @SHL0MS) into one
cross-surface implementation, folding in the review feedback and the
best ideas from the two sibling PRs (#22703, #53527):

- tools/working_diff.py: shared git collection layer — unstaged
  (default), staged, and all (vs HEAD) modes; untracked files folded in
  via `git diff --no-index` so new files appear as additions (Codex
  /diff parity); shlex-split arguments preserve quoted paths.
- CLI: handler moved to hermes_cli/cli_commands_mixin.py per the
  current god-file decomposition (dispatch stays in cli.py), renders
  through the rich console with a 400-line terminal-flood guard.
- Gateway: _handle_diff_command in gateway/slash_commands.py + dispatch
  in gateway/run.py; fenced ```diff output truncated to 60 lines /
  3000 chars before the platform senders apply their own per-platform
  message clamps (tool-progress-style layered truncation). Localized
  strings in all 17 locale catalogs.
- /diff session (from #53527): cumulative checkpoint-baseline diff of
  everything Hermes changed, via new CheckpointManager.session_diff();
  docstring records the retained-baseline approximation caveat from
  review. Works on both surfaces; degrades with an actionable message
  when checkpoints are off.
- Slack: /diff routed via /hermes diff (50-slash cap; keeps
  telegram-parity test green and /version native).
- Registry: cross-surface CommandDef with staged|all|session
  subcommands; docs: slash-commands reference (CLI + gateway tables +
  both-surfaces list) and hermes-agent skill reference.
- Tests: tests/tools/test_working_diff.py (real git repos),
  tests/hermes_cli/test_diff_command.py (real git + stubbed checkpoint
  manager), tests/gateway/test_diff_command.py (end-to-end handler,
  real checkpoint store), TestSessionDiff in
  tests/tools/test_checkpoint_manager.py.

Salvaged from the /diff PR cluster #4839 + #22703 + #53527.

Co-authored-by: Ninso112 <ninso112@proton.me>
Co-authored-by: Harshkamdar67 <harshkamdar67@gmail.com>
2026-07-26 18:28:20 -07:00

111 lines
3.5 KiB
Python

"""Tests for tools.working_diff.collect_working_diff — the git collection
layer shared by the CLI and gateway ``/diff`` command.
Runs against real temporary git repositories (no mocks) so the staged /
unstaged / untracked semantics are proven against actual git behaviour.
"""
import shutil
import subprocess
import pytest
from tools.working_diff import collect_working_diff
pytestmark = pytest.mark.skipif(
shutil.which("git") is None, reason="git required for working-diff tests"
)
def _git(repo, *args):
subprocess.run(
["git", *args], cwd=repo, check=True, capture_output=True,
env={"HOME": str(repo), "GIT_AUTHOR_NAME": "t", "GIT_AUTHOR_EMAIL": "t@t",
"GIT_COMMITTER_NAME": "t", "GIT_COMMITTER_EMAIL": "t@t",
"PATH": __import__("os").environ["PATH"]},
)
@pytest.fixture()
def repo(tmp_path):
d = tmp_path / "repo"
d.mkdir()
_git(d, "init", "-q")
(d / "tracked.py").write_text("print('hello')\n")
_git(d, "add", "-A")
_git(d, "commit", "-q", "-m", "init")
return d
def test_clean_repo_reports_empty(repo):
result = collect_working_diff(str(repo))
assert result["success"] is True
assert result.get("empty") is True
assert result["diff"] == ""
def test_unstaged_change_appears_in_default_mode(repo):
(repo / "tracked.py").write_text("print('changed')\n")
result = collect_working_diff(str(repo))
assert result["success"] is True
assert "-print('hello')" in result["diff"]
assert "+print('changed')" in result["diff"]
assert "tracked.py" in result["stat"]
def test_untracked_file_appears_as_addition(repo):
(repo / "brand_new.py").write_text("x = 1\n")
result = collect_working_diff(str(repo))
assert result["success"] is True
assert "brand_new.py" in result["untracked"]
assert "+x = 1" in result["diff"]
assert not result.get("empty")
def test_staged_mode_shows_only_staged(repo):
(repo / "tracked.py").write_text("print('staged')\n")
_git(repo, "add", "tracked.py")
(repo / "unstaged.py").write_text("y = 2\n") # untracked, must not appear
result = collect_working_diff(str(repo), mode="staged")
assert result["success"] is True
assert "+print('staged')" in result["diff"]
assert "unstaged.py" not in result["diff"]
def test_all_mode_spans_staged_unstaged_and_untracked(repo):
(repo / "tracked.py").write_text("print('staged')\n")
_git(repo, "add", "tracked.py")
(repo / "extra.py").write_text("z = 3\n")
result = collect_working_diff(str(repo), mode="all")
assert result["success"] is True
assert "+print('staged')" in result["diff"]
assert "+z = 3" in result["diff"]
def test_paths_filter_restricts_output(repo):
(repo / "tracked.py").write_text("print('changed')\n")
(repo / "other.py").write_text("o = 1\n")
_git(repo, "add", "other.py")
_git(repo, "commit", "-q", "-m", "add other")
(repo / "other.py").write_text("o = 2\n")
result = collect_working_diff(str(repo), paths=["tracked.py"])
assert result["success"] is True
assert "tracked.py" in result["diff"]
assert "other.py" not in result["diff"]
def test_non_git_directory_fails_cleanly(tmp_path):
plain = tmp_path / "plain"
plain.mkdir()
result = collect_working_diff(str(plain))
assert result["success"] is False
assert "not a git repository" in result["error"].lower()
def test_unknown_mode_rejected(repo):
result = collect_working_diff(str(repo), mode="bogus")
assert result["success"] is False
assert "bogus" in result["error"]