medium95%
grep_search Python fallback skips valid UTF-8 files on Windows due to missing encoding parameter in read_text()
When ripgrep is unavailable, the Python fallback in FilesystemFileSearchMiddleware.grep_search reads files using Path.read_text() without specifying an encoding. On Windows, the default encoding is often cp1252, causing UnicodeDecodeError for UTF-8 files. These errors are silently swallowed, leading to incomplete or empty search results even when matching content exists.