[clang-tidy] Add an option to exclude files not present in the compile database (#120348)

A change list may include files that are not part of the compile
database, which can cause clang-tidy to fail (e.g., due to missing
included headers). To prevent false negatives, we should allow to skip
processing these files.

Author: wonbinbk

clang-tools-extra/clang-tidy/tool/clang-tidy-diff.py

@@ -35,6 +35,7 @@
 import tempfile
 import threading
 import traceback
+from pathlib import Path
 
 try:
     import yaml
@@ -124,6 +125,23 @@ def merge_replacement_files(tmpdir, mergefile):
         open(mergefile, "w").close()
 
 
+def get_compiling_files(args):
+    """Read a compile_commands.json database and return a set of file paths"""
+    current_dir = Path.cwd()
+    compile_commands_json = (
+        (current_dir / args.build_path) if args.build_path else current_dir
+    )
+    compile_commands_json = compile_commands_json / "compile_commands.json"
+    files = set()
+    with open(compile_commands_json) as db_file:
+        db_json = json.load(db_file)
+        for entry in db_json:
+            if "file" not in entry:
+                continue
+            files.add(Path(entry["file"]))
+    return files
+
+
 def main():
     parser = argparse.ArgumentParser(
         description="Run clang-tidy against changed files, and "
@@ -234,6 +252,13 @@ def main():
         action="store_true",
         help="Allow empty enabled checks.",
     )
+    parser.add_argument(
+        "-only-check-in-db",
+        dest="skip_non_compiling",
+        default=False,
+        action="store_true",
+        help="Only check files in the compilation database",
+    )
 
     clang_tidy_args = []
     argv = sys.argv[1:]
@@ -243,11 +268,13 @@ def main():
 
     args = parser.parse_args(argv)
 
+    compiling_files = get_compiling_files(args) if args.skip_non_compiling else None
+
     # Extract changed lines for each file.
     filename = None
     lines_by_file = {}
     for line in sys.stdin:
-        match = re.search('^\\+\\+\\+\\ "?(.*?/){%s}([^ \t\n"]*)' % args.p, line)
+        match = re.search(r'^\+\+\+\ "?(.*?/){%s}([^ \t\n"]*)' % args.p, line)
         if match:
             filename = match.group(2)
         if filename is None:
@@ -260,6 +287,13 @@ def main():
             if not re.match("^%s$" % args.iregex, filename, re.IGNORECASE):
                 continue
 
+        # Skip any files not in the compiling list
+        if (
+            compiling_files is not None
+            and (Path.cwd() / filename) not in compiling_files
+        ):
+            continue
+
         match = re.search(r"^@@.*\+(\d+)(,(\d+))?", line)
         if match:
             start_line = int(match.group(1))

clang-tools-extra/docs/ReleaseNotes.rst

@@ -108,6 +108,10 @@ Improvements to clang-query
 Improvements to clang-tidy
 --------------------------
 
+- Improved :program:`clang-tidy-diff.py` script. Add the `-only-check-in-db`
+  option to exclude files not present in the compilation database, avoiding
+  false-negative results.
+
 - Improved :program:`run-clang-tidy.py` script. Fixed minor shutdown noise
   happening on certain platforms when interrupting the script.