fix(upstream_triage): force UTF-8 stdout and decode git output as UTF-8 (#544) (#547)

* fix(upstream_triage): force UTF-8 stdout and decode git output as UTF-8 (#544)

* test(tools-utf8): restore assertion in salary converter test and isolate upstream_triage test

CHANGELOG.md

@@ -46,6 +46,14 @@ per-file diff commands.
 
 ### Fixed
 
+- **`upstream_triage` forces UTF-8 stdout and decodes git output as UTF-8** (#544,
+  `tools/upstream_triage.py`, `tests/test_tools_utf8_output.py`) - like the other Python tools,
+  the triage report now calls `_force_utf8_output()` on startup and decodes git subprocess
+  streams with `encoding="utf-8", errors="replace"`. Under Windows default ANSI code pages (such as
+  `cp1252`), upstream commits with Cyrillic, CJK, or emoji characters in their subjects raised
+  `UnicodeEncodeError` when writing stdout or `UnicodeDecodeError` in subprocess pipes, terminating
+  triage runs before the report could be emitted.
+
 - **`jobnet-search` routes external ads to employer `jobAdUrl` rather than broken `find-job` links**
   (#538, `.agents/skills/jobnet-search/cli/src/commands/search.ts`, `.agents/skills/jobnet-search/cli/README.md`,
   `.agents/skills/jobnet-search/cli/tests/search-normalization.test.ts`) - external job postings on
@@ -139,6 +147,7 @@ per-file diff commands.
   says and asserts the shipped file's version still parses out unchanged; it fails on master.
   Found while re-checking #519's class of drift.
 
+
 - **`/reset profile` now targets the headings the shipped skill files actually have**
   (`.claude/commands/reset.md` Step 3, `tests/test_reset_command.py`) - two of its Step 3
   targets had drifted. The `05-cv-templates.md` step anchored on a section beginning

tests/test_tools_utf8_output.py

@@ -149,6 +149,28 @@ class ToolsWriteUtf8(unittest.TestCase):
         ])
         self.assert_clean_utf8(proc, 0, CYRILLIC)
 
+    def test_upstream_triage_handles_non_latin_commit_subjects(self):
+        repo = self.tmp / "repo"
+        repo.mkdir()
+        tools_dir = repo / "tools"
+        tools_dir.mkdir()
+        shutil.copy(TOOLS / "upstream_triage.py", tools_dir / "upstream_triage.py")
+        subprocess.run(["git", "init", "-b", "master"], cwd=repo, check=True, capture_output=True)
+        subprocess.run(["git", "config", "user.name", "Test"], cwd=repo, check=True, capture_output=True)
+        subprocess.run(["git", "config", "user.email", "test@example.com"], cwd=repo, check=True, capture_output=True)
+        subprocess.run(["git", "remote", "add", "upstream", "https://github.com/MadsLorentzen/ai-job-search.git"], cwd=repo, check=True, capture_output=True)
+        (repo / "file.txt").write_text("hello\n", encoding="utf-8")
+        subprocess.run(["git", "add", "file.txt"], cwd=repo, check=True, capture_output=True)
+        subprocess.run(["git", "commit", "-m", "init"], cwd=repo, check=True, capture_output=True)
+        (repo / "file.txt").write_text("change\n", encoding="utf-8")
+        subprocess.run(["git", "add", "file.txt"], cwd=repo, check=True, capture_output=True)
+        msg = f"feat: add support for {CYRILLIC}"
+        subprocess.run(["git", "commit", "-m", msg], cwd=repo, check=True, capture_output=True)
+        subprocess.run(["git", "update-ref", "refs/remotes/upstream/master", "HEAD"], cwd=repo, check=True, capture_output=True)
+        subprocess.run(["git", "reset", "--hard", "HEAD~1"], cwd=repo, check=True, capture_output=True)
+        proc = run_legacy_stdout([repo / "tools" / "upstream_triage.py"], cwd=repo)
+        self.assert_clean_utf8(proc, 0, CYRILLIC)
+
 
 if __name__ == "__main__":
     unittest.main()

tools/upstream_triage.py

@@ -36,9 +36,22 @@ import subprocess
 import sys
 
 
+def _force_utf8_output() -> None:
+    """Write UTF-8 whatever the host's default encoding is.
+
+    A piped stdout on Windows defaults to the ANSI code page (cp1252 on most
+    Western installs), so printing a commit subject or author outside it
+    raised UnicodeEncodeError before the report could be emitted.
+    """
+    for stream in (sys.stdout, sys.stderr):
+        reconfigure = getattr(stream, "reconfigure", None)
+        if reconfigure:
+            reconfigure(encoding="utf-8")
+
+
 def git(*args: str) -> str:
     return subprocess.run(
-        ["git", *args], capture_output=True, text=True, check=True
+        ["git", *args], capture_output=True, text=True, encoding="utf-8", errors="replace", check=True
     ).stdout
 
 
@@ -50,10 +63,10 @@ def rev_list(range_spec: str) -> list[str]:
 def patch_id(sha: str) -> str | None:
     """Stable patch-id for a commit, or None if it has no diff."""
     show = subprocess.run(
-        ["git", "show", sha], capture_output=True, text=True, check=True
+        ["git", "show", sha], capture_output=True, text=True, encoding="utf-8", errors="replace", check=True
     ).stdout
     r = subprocess.run(
-        ["git", "patch-id", "--stable"], input=show, capture_output=True, text=True
+        ["git", "patch-id", "--stable"], input=show, capture_output=True, text=True, encoding="utf-8", errors="replace"
     )
     line = r.stdout.strip()
     return line.split()[0] if line else None
@@ -111,6 +124,7 @@ def commit_cell(short: str, sha: str, slug: str | None) -> str:
 
 
 def main() -> int:
+    _force_utf8_output()
     ap = argparse.ArgumentParser()
     ap.add_argument("--remote", default="upstream")
     ap.add_argument("--branch", default="master")