From 15f34d3043a06b8fa541cc7d37ce9c563215fd32 Mon Sep 17 00:00:00 2001
From: GalaxyGamingFR <153150716+GalaxyGamingFR@users.noreply.github.com>
Date: Sun, 11 Oct 2026 02:38:52 -0700
Subject: [PATCH] fix(tools): decode pr_overlap git output with errors=replace

One PR diff containing a byte that is not valid UTF-8 made git() and
merge() raise, which aborted the hourly run and dropped the overlap
comment on every PR. Both subprocess calls now pass errors="replace",
matching check_conventions.py. A test commits a Latin-1 byte and checks
the export is still found.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01V9UU71SP1UKGR51Sowbgmq
---
 tools/pr_overlap.py                     |  4 ++--
 tools/tests/test_pr_overlap_encoding.py | 20 ++++++++++++++++++++
 2 files changed, 22 insertions(+), 2 deletions(-)

diff --git a/tools/pr_overlap.py b/tools/pr_overlap.py
index 40a5efb9d..d42e24592 100644
--- a/tools/pr_overlap.py
+++ b/tools/pr_overlap.py
@@ -16,7 +16,7 @@ MARKER = "<!-- pr-overlap -->"
 
 
 def git(*args):
-    return subprocess.run(["git", *args], capture_output=True, text=True, encoding="utf-8", check=True).stdout
+    return subprocess.run(["git", *args], capture_output=True, text=True, encoding="utf-8", errors="replace", check=True).stdout
 
 
 def gh(*args):
@@ -49,7 +49,7 @@ def fetch(base, prs):
 
 def merge(*args):
     result = subprocess.run(["git", "merge-tree", "--write-tree", "--name-only", "--no-messages", *args],
-                            capture_output=True, text=True, encoding="utf-8")
+                            capture_output=True, text=True, encoding="utf-8", errors="replace")
     lines = result.stdout.splitlines()
     if result.returncode not in (0, 1) or not lines:
         raise RuntimeError(f"git merge-tree {' '.join(args)} failed: {result.stderr.strip()}")
diff --git a/tools/tests/test_pr_overlap_encoding.py b/tools/tests/test_pr_overlap_encoding.py
index 9a4c3f449..2d2b92c42 100644
--- a/tools/tests/test_pr_overlap_encoding.py
+++ b/tools/tests/test_pr_overlap_encoding.py
@@ -1,5 +1,6 @@
 from contextlib import contextmanager
 import json
+import os
 from pathlib import Path
 import subprocess
 import sys
@@ -74,6 +75,25 @@ class EncodingTests(unittest.TestCase):
             with self.assertRaisesRegex(RuntimeError, UNICODE):
                 pr_overlap.merge("main", "topic")
 
+    def test_git_diff_survives_invalid_utf8(self):
+        with TemporaryDirectory() as directory:
+            repo = Path(directory)
+            source = repo / "core" / "libs" / "prx" / "x" / "a.cpp"
+            source.parent.mkdir(parents=True)
+            ident = ["-c", "user.name=t", "-c", "user.email=t@example.test"]
+            for args in (["init", "-q"], ["commit", "-q", "--allow-empty", "-m", "base"]):
+                subprocess.run(["git", *ident, *args], cwd=repo, check=True, capture_output=True)
+            source.write_bytes(b"int APS5_VABI f() {} // caf\xe9\n")
+            for args in (["add", "."], ["commit", "-q", "-m", "add"]):
+                subprocess.run(["git", *ident, *args], cwd=repo, check=True, capture_output=True)
+            old = os.getcwd()
+            os.chdir(repo)
+            try:
+                diff = pr_overlap.git("diff", "-U0", "--no-color", "HEAD~1", "HEAD", "--", "core/libs/prx/*.cpp")
+            finally:
+                os.chdir(old)
+            self.assertEqual(pr_overlap.exports(diff), {"f"})
+
     def test_report_is_written_as_utf8(self):
         with TemporaryDirectory() as directory:
             path = Path(directory) / "1.md"