ci: fix pr-labels job logic

Author: seratch

.github/workflows/pr-labels.yml

@@ -110,6 +110,8 @@ jobs:
         env:
           GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
           PR_NUMBER: ${{ steps.pr.outputs.pr_number }}
+          PR_BASE_SHA: ${{ steps.pr.outputs.base_sha }}
+          PR_HEAD_SHA: ${{ steps.pr.outputs.head_sha }}
           CODEX_OUTPUT_PATH: ${{ steps.codex-output.outputs.output_file }}
         run: |
           python - <<'PY'
@@ -120,10 +122,119 @@ jobs:
           import re
 
           pr_number = os.environ["PR_NUMBER"]
+          pr_base_sha = os.environ.get("PR_BASE_SHA")
+          pr_head_sha = os.environ.get("PR_HEAD_SHA")
           codex_output_path = pathlib.Path(os.environ["CODEX_OUTPUT_PATH"])
           changed_files_path = pathlib.Path(".tmp/pr-labels/changed-files.txt")
           changes_diff_path = pathlib.Path(".tmp/pr-labels/changes.diff")
 
+          def read_file_at(commit, path):
+              if not commit:
+                  return None
+              try:
+                  return subprocess.check_output(
+                      ["git", "show", f"{commit}:{path}"],
+                      text=True,
+                  )
+              except subprocess.CalledProcessError:
+                  return None
+
+          def dependency_lines_for_pyproject(text: str) -> set[int]:
+              dependency_lines: set[int] = set()
+              current_section = None
+              in_project_dependencies = False
+
+              for line_number, raw_line in enumerate(text.splitlines(), start=1):
+                  stripped = raw_line.strip()
+                  if stripped.startswith("[") and stripped.endswith("]"):
+                      if stripped.startswith("[[") and stripped.endswith("]]"):
+                          current_section = stripped[2:-2].strip()
+                      else:
+                          current_section = stripped[1:-1].strip()
+                      in_project_dependencies = False
+                      if current_section in ("project.optional-dependencies", "dependency-groups"):
+                          dependency_lines.add(line_number)
+                      continue
+
+                  if current_section in ("project.optional-dependencies", "dependency-groups"):
+                      dependency_lines.add(line_number)
+                      continue
+
+                  if current_section != "project":
+                      continue
+
+                  if in_project_dependencies:
+                      dependency_lines.add(line_number)
+                      if "]" in stripped:
+                          in_project_dependencies = False
+                      continue
+
+                  if stripped.startswith("dependencies") and "=" in stripped:
+                      dependency_lines.add(line_number)
+                      if "[" in stripped and "]" not in stripped:
+                          in_project_dependencies = True
+
+              return dependency_lines
+
+          def pyproject_dependency_changed(diff_text: str) -> bool:
+              base_text = read_file_at(pr_base_sha, "pyproject.toml")
+              head_text = read_file_at(pr_head_sha, "pyproject.toml")
+              if base_text is None and head_text is None:
+                  return False
+
+              base_dependency_lines = (
+                  dependency_lines_for_pyproject(base_text) if base_text else set()
+              )
+              head_dependency_lines = (
+                  dependency_lines_for_pyproject(head_text) if head_text else set()
+              )
+
+              in_pyproject = False
+              base_line = None
+              head_line = None
+              hunk_re = re.compile(r"@@ -(\d+)(?:,\d+)? \+(\d+)(?:,\d+)? @@")
+
+              for line in diff_text.splitlines():
+                  if line.startswith("+++ b/"):
+                      current_file = line[len("+++ b/") :].strip()
+                      in_pyproject = current_file == "pyproject.toml"
+                      base_line = None
+                      head_line = None
+                      continue
+
+                  if not in_pyproject:
+                      continue
+
+                  if line.startswith("@@ "):
+                      match = hunk_re.match(line)
+                      if not match:
+                          continue
+                      base_line = int(match.group(1))
+                      head_line = int(match.group(2))
+                      continue
+
+                  if base_line is None or head_line is None:
+                      continue
+
+                  if line.startswith(" "):
+                      base_line += 1
+                      head_line += 1
+                      continue
+
+                  if line.startswith("-"):
+                      if base_line in base_dependency_lines:
+                          return True
+                      base_line += 1
+                      continue
+
+                  if line.startswith("+"):
+                      if head_line in head_dependency_lines:
+                          return True
+                      head_line += 1
+                      continue
+
+              return False
+
           changed_files = []
           if changed_files_path.exists():
               changed_files = [
@@ -137,22 +248,15 @@ jobs:
               desired.add("project")
           if any(path.startswith("docs/") for path in changed_files):
               desired.add("documentation")
-          if "uv.lock" in changed_files:
-              desired.add("dependencies")
+          dependencies_allowed = "uv.lock" in changed_files
+          diff_text = None
           if changes_diff_path.exists():
               diff_text = changes_diff_path.read_text()
-              current_file = None
-              for line in diff_text.splitlines():
-                  if line.startswith("+++ b/"):
-                      current_file = line[len("+++ b/") :].strip()
-                      continue
-                  if not current_file or not current_file.startswith(".github/workflows/"):
-                      continue
-                  if not line.startswith(("+", "-")):
-                      continue
-                  if re.search(r"uses:\s*(?!\./)(?!\.\/)[^@\s]+/[^@\s]+@", line):
-                      desired.add("dependencies")
-                      break
+              if "pyproject.toml" in changed_files and pyproject_dependency_changed(diff_text):
+                  dependencies_allowed = True
+
+          if dependencies_allowed:
+              desired.add("dependencies")
 
           allowed = {
               "documentation",
@@ -184,6 +288,8 @@ jobs:
                       pass
 
           for label in codex_labels:
+              if label == "dependencies" and not dependencies_allowed:
+                  continue
               if label in allowed:
                   desired.add(label)