tools: Add various improvements for clang_format.py/githelper.py

1 files changed, 27 insertions, 9 deletions

diff --git a/tools/githelper.py b/tools/githelper.py
index b7eb1656b1..29e9ee0a4e 100644
--- a/tools/githelper.py
+++ b/tools/githelper.py
@@ -26,41 +26,52 @@ def get_toplevel_path() -> str:
 def get_changed_lines(
     from_ref=None, filter_lines=None, include_files=None
 ) -> typing.Iterable[Line]:
-    """Inspect `git diff` output, yields changed lines."""
+    """Inspect `git diff-index` output, yields changed lines."""
 
     logger = logging.getLogger(__name__)
 
-    cmd = ["git", "diff", "--unified=0", from_ref if from_ref else "HEAD"]
+    cmd = [
+        "git",
+        "diff-index",
+        "--patch",
+        "--unified=0",
+        from_ref if from_ref else "HEAD",
+    ]
     if include_files:
         cmd.extend(["--", *include_files])
     logger.debug("Executing: %r", cmd)
     proc = subprocess.run(cmd, capture_output=True)
     proc.check_returncode()
     current_file = None
-    lines_left = 0
+    hunk_lines_left = 0
     for line in proc.stdout.decode(errors="replace").splitlines():
         match_file = re.match(r"^\+\+\+ b/(.*)$", line)
         if match_file:
+            # Current line contains a diff filename
+            assert hunk_lines_left == 0
             current_file = match_file.group(1)
-            lines_left = 0
             continue
 
         match_lineno = re.match(
             r"^@@ -(\d+(?:,\d+)?) \+([0-9]+(?:,[0-9]+)?) @@", line
         )
         if match_lineno:
+            # Current line contains a hunk header
+            assert current_file is not None
+            assert hunk_lines_left == 0
             start_removed, _, length_removed = match_lineno.group(1).partition(
                 ","
             )
             start_added, _, length_added = match_lineno.group(2).partition(",")
             lineno_removed = int(start_removed)
             lineno_added = int(start_added)
-            lines_left = int(length_removed) if length_removed else 1
-            lines_left += int(length_added) if length_added else 1
+            hunk_lines_left = int(length_removed) if length_removed else 1
+            hunk_lines_left += int(length_added) if length_added else 1
             continue
 
-        if lines_left and line:
-            lines_left -= 1
+        if hunk_lines_left and line:
+            # Current line contains an added/removed line
+            hunk_lines_left -= 1
 
             assert line[0] in ("+", "-")
             if line.startswith("+"):
@@ -80,11 +91,18 @@ def get_changed_lines(
             if filter_lines is None or filter_lines(lineobj):
                 yield lineobj
 
+        # If we reach this part, the line does not contain a diff filename or a
+        # hunk header and does not belog to a hunk. This means that this line
+        # will be ignored implicitly.
+
+    # Make sure we really parsed all lines from the last hunk
+    assert hunk_lines_left == 0
+
 
 def get_changed_lines_grouped(
     from_ref=None, filter_lines=None, include_files=None
 ) -> typing.Iterable[FileLines]:
-    """Inspect `git diff` output, yields changed lines grouped by file."""
+    """Inspect `git diff-index` output, yields lines grouped by file."""
 
     lines = get_changed_lines(from_ref, filter_lines, include_files)
     for filename, file_lines in itertools.groupby(