claude-code-workflow 6.2.5 → 6.2.6

package/codex-lens/src/codexlens/parsers/encoding.py

@@ -0,0 +1,202 @@
+"""Optional encoding detection module for CodexLens.
+
+Provides automatic encoding detection with graceful fallback to UTF-8.
+Install with: pip install codexlens[encoding]
+"""
+
+from __future__ import annotations
+
+import logging
+from pathlib import Path
+from typing import Tuple, Optional
+
+log = logging.getLogger(__name__)
+
+# Feature flag for encoding detection availability
+ENCODING_DETECTION_AVAILABLE = False
+_import_error: Optional[str] = None
+
+
+def _detect_chardet_backend() -> Tuple[bool, Optional[str]]:
+    """Detect if chardet or charset-normalizer is available."""
+    try:
+        import chardet
+        return True, None
+    except ImportError:
+        pass
+
+    try:
+        from charset_normalizer import from_bytes
+        return True, None
+    except ImportError:
+        pass
+
+    return False, "chardet not available. Install with: pip install codexlens[encoding]"
+
+
+# Initialize on module load
+ENCODING_DETECTION_AVAILABLE, _import_error = _detect_chardet_backend()
+
+
+def check_encoding_available() -> Tuple[bool, Optional[str]]:
+    """Check if encoding detection dependencies are available.
+
+    Returns:
+        Tuple of (available, error_message)
+    """
+    return ENCODING_DETECTION_AVAILABLE, _import_error
+
+
+def detect_encoding(content_bytes: bytes, confidence_threshold: float = 0.7) -> str:
+    """Detect encoding from file content bytes.
+
+    Uses chardet or charset-normalizer with configurable confidence threshold.
+    Falls back to UTF-8 if confidence is too low or detection unavailable.
+
+    Args:
+        content_bytes: Raw file content as bytes
+        confidence_threshold: Minimum confidence (0.0-1.0) to accept detection
+
+    Returns:
+        Detected encoding name (e.g., 'utf-8', 'iso-8859-1', 'gbk')
+        Returns 'utf-8' as fallback if detection fails or confidence too low
+    """
+    if not ENCODING_DETECTION_AVAILABLE:
+        log.debug("Encoding detection not available, using UTF-8 fallback")
+        return "utf-8"
+
+    if not content_bytes:
+        return "utf-8"
+
+    try:
+        # Try chardet first
+        try:
+            import chardet
+            result = chardet.detect(content_bytes)
+            encoding = result.get("encoding")
+            confidence = result.get("confidence", 0.0)
+
+            if encoding and confidence >= confidence_threshold:
+                log.debug(f"Detected encoding: {encoding} (confidence: {confidence:.2f})")
+                # Normalize encoding name: replace underscores with hyphens
+                return encoding.lower().replace('_', '-')
+            else:
+                log.debug(
+                    f"Low confidence encoding detection: {encoding} "
+                    f"(confidence: {confidence:.2f}), using UTF-8 fallback"
+                )
+                return "utf-8"
+        except ImportError:
+            pass
+
+        # Fallback to charset-normalizer
+        try:
+            from charset_normalizer import from_bytes
+            results = from_bytes(content_bytes)
+            if results:
+                best = results.best()
+                if best and best.encoding:
+                    log.debug(f"Detected encoding via charset-normalizer: {best.encoding}")
+                    # Normalize encoding name: replace underscores with hyphens
+                    return best.encoding.lower().replace('_', '-')
+        except ImportError:
+            pass
+
+    except Exception as e:
+        log.warning(f"Encoding detection failed: {e}, using UTF-8 fallback")
+
+    return "utf-8"
+
+
+def read_file_safe(
+    path: Path | str,
+    confidence_threshold: float = 0.7,
+    max_detection_bytes: int = 100_000
+) -> Tuple[str, str]:
+    """Read file with automatic encoding detection and safe decoding.
+
+    Reads file bytes, detects encoding, and decodes with error replacement
+    to preserve file structure even with encoding issues.
+
+    Args:
+        path: Path to file to read
+        confidence_threshold: Minimum confidence for encoding detection
+        max_detection_bytes: Maximum bytes to use for encoding detection (default 100KB)
+
+    Returns:
+        Tuple of (content, detected_encoding)
+        - content: Decoded file content (with � for unmappable bytes)
+        - detected_encoding: Detected encoding name
+
+    Raises:
+        OSError: If file cannot be read
+        IsADirectoryError: If path is a directory
+    """
+    file_path = Path(path) if isinstance(path, str) else path
+
+    # Read file bytes
+    try:
+        content_bytes = file_path.read_bytes()
+    except Exception as e:
+        log.error(f"Failed to read file {file_path}: {e}")
+        raise
+
+    # Detect encoding from first N bytes for performance
+    detection_sample = content_bytes[:max_detection_bytes] if len(content_bytes) > max_detection_bytes else content_bytes
+    encoding = detect_encoding(detection_sample, confidence_threshold)
+
+    # Decode with error replacement to preserve structure
+    try:
+        content = content_bytes.decode(encoding, errors='replace')
+        log.debug(f"Successfully decoded {file_path} using {encoding}")
+        return content, encoding
+    except Exception as e:
+        # Final fallback to UTF-8 with replacement
+        log.warning(f"Failed to decode {file_path} with {encoding}, using UTF-8: {e}")
+        content = content_bytes.decode('utf-8', errors='replace')
+        return content, 'utf-8'
+
+
+def is_binary_file(path: Path | str, sample_size: int = 8192) -> bool:
+    """Check if file is likely binary by sampling first bytes.
+
+    Uses heuristic: if >30% of sample bytes are null or non-text, consider binary.
+
+    Args:
+        path: Path to file to check
+        sample_size: Number of bytes to sample (default 8KB)
+
+    Returns:
+        True if file appears to be binary, False otherwise
+    """
+    file_path = Path(path) if isinstance(path, str) else path
+
+    try:
+        with file_path.open('rb') as f:
+            sample = f.read(sample_size)
+
+        if not sample:
+            return False
+
+        # Count null bytes and non-printable characters
+        null_count = sample.count(b'\x00')
+        non_text_count = sum(1 for byte in sample if byte < 0x20 and byte not in (0x09, 0x0a, 0x0d))
+
+        # If >30% null bytes or >50% non-text, consider binary
+        null_ratio = null_count / len(sample)
+        non_text_ratio = non_text_count / len(sample)
+
+        return null_ratio > 0.3 or non_text_ratio > 0.5
+
+    except Exception as e:
+        log.debug(f"Binary check failed for {file_path}: {e}, assuming text")
+        return False
+
+
+__all__ = [
+    "ENCODING_DETECTION_AVAILABLE",
+    "check_encoding_available",
+    "detect_encoding",
+    "read_file_safe",
+    "is_binary_file",
+]

package/codex-lens/src/codexlens/parsers/factory.py

@@ -0,0 +1,256 @@
+"""Parser factory for CodexLens.
+
+Python and JavaScript/TypeScript parsing use Tree-Sitter grammars when
+available. Regex fallbacks are retained to preserve the existing parser
+interface and behavior in minimal environments.
+"""
+
+from __future__ import annotations
+
+import re
+from dataclasses import dataclass
+from pathlib import Path
+from typing import Dict, List, Optional, Protocol
+
+from codexlens.config import Config
+from codexlens.entities import IndexedFile, Symbol
+from codexlens.parsers.treesitter_parser import TreeSitterSymbolParser
+
+
+class Parser(Protocol):
+    def parse(self, text: str, path: Path) -> IndexedFile: ...
+
+
+@dataclass
+class SimpleRegexParser:
+    language_id: str
+
+    def parse(self, text: str, path: Path) -> IndexedFile:
+        # Try tree-sitter first for supported languages
+        if self.language_id in {"python", "javascript", "typescript"}:
+            ts_parser = TreeSitterSymbolParser(self.language_id, path)
+            if ts_parser.is_available():
+                symbols = ts_parser.parse_symbols(text)
+                if symbols is not None:
+                    return IndexedFile(
+                        path=str(path.resolve()),
+                        language=self.language_id,
+                        symbols=symbols,
+                        chunks=[],
+                    )
+
+        # Fallback to regex parsing
+        if self.language_id == "python":
+            symbols = _parse_python_symbols_regex(text)
+        elif self.language_id in {"javascript", "typescript"}:
+            symbols = _parse_js_ts_symbols_regex(text)
+        elif self.language_id == "java":
+            symbols = _parse_java_symbols(text)
+        elif self.language_id == "go":
+            symbols = _parse_go_symbols(text)
+        elif self.language_id == "markdown":
+            symbols = _parse_markdown_symbols(text)
+        elif self.language_id == "text":
+            symbols = _parse_text_symbols(text)
+        else:
+            symbols = _parse_generic_symbols(text)
+
+        return IndexedFile(
+            path=str(path.resolve()),
+            language=self.language_id,
+            symbols=symbols,
+            chunks=[],
+        )
+
+
+class ParserFactory:
+    def __init__(self, config: Config) -> None:
+        self.config = config
+        self._parsers: Dict[str, Parser] = {}
+
+    def get_parser(self, language_id: str) -> Parser:
+        if language_id not in self._parsers:
+            self._parsers[language_id] = SimpleRegexParser(language_id)
+        return self._parsers[language_id]
+
+
+# Regex-based fallback parsers
+_PY_CLASS_RE = re.compile(r"^\s*class\s+([A-Za-z_]\w*)\b")
+_PY_DEF_RE = re.compile(r"^\s*(?:async\s+)?def\s+([A-Za-z_]\w*)\s*\(")
+
+
+
+
+def _parse_python_symbols(text: str) -> List[Symbol]:
+    """Parse Python symbols, using tree-sitter if available, regex fallback."""
+    ts_parser = TreeSitterSymbolParser("python")
+    if ts_parser.is_available():
+        symbols = ts_parser.parse_symbols(text)
+        if symbols is not None:
+            return symbols
+    return _parse_python_symbols_regex(text)
+
+
+def _parse_js_ts_symbols(
+    text: str,
+    language_id: str = "javascript",
+    path: Optional[Path] = None,
+) -> List[Symbol]:
+    """Parse JS/TS symbols, using tree-sitter if available, regex fallback."""
+    ts_parser = TreeSitterSymbolParser(language_id, path)
+    if ts_parser.is_available():
+        symbols = ts_parser.parse_symbols(text)
+        if symbols is not None:
+            return symbols
+    return _parse_js_ts_symbols_regex(text)
+
+
+def _parse_python_symbols_regex(text: str) -> List[Symbol]:
+    symbols: List[Symbol] = []
+    current_class_indent: Optional[int] = None
+    for i, line in enumerate(text.splitlines(), start=1):
+        class_match = _PY_CLASS_RE.match(line)
+        if class_match:
+            current_class_indent = len(line) - len(line.lstrip(" "))
+            symbols.append(Symbol(name=class_match.group(1), kind="class", range=(i, i)))
+            continue
+        def_match = _PY_DEF_RE.match(line)
+        if def_match:
+            indent = len(line) - len(line.lstrip(" "))
+            kind = "method" if current_class_indent is not None and indent > current_class_indent else "function"
+            symbols.append(Symbol(name=def_match.group(1), kind=kind, range=(i, i)))
+            continue
+        if current_class_indent is not None:
+            indent = len(line) - len(line.lstrip(" "))
+            if line.strip() and indent <= current_class_indent:
+                current_class_indent = None
+    return symbols
+
+
+_JS_FUNC_RE = re.compile(r"^\s*(?:export\s+)?(?:async\s+)?function\s+([A-Za-z_$][\w$]*)\s*\(")
+_JS_CLASS_RE = re.compile(r"^\s*(?:export\s+)?class\s+([A-Za-z_$][\w$]*)\b")
+_JS_ARROW_RE = re.compile(
+    r"^\s*(?:export\s+)?(?:const|let|var)\s+([A-Za-z_$][\w$]*)\s*=\s*(?:async\s*)?\(?[^)]*\)?\s*=>"
+)
+_JS_METHOD_RE = re.compile(r"^\s+(?:async\s+)?([A-Za-z_$][\w$]*)\s*\([^)]*\)\s*\{")
+
+
+def _parse_js_ts_symbols_regex(text: str) -> List[Symbol]:
+    symbols: List[Symbol] = []
+    in_class = False
+    class_brace_depth = 0
+    brace_depth = 0
+
+    for i, line in enumerate(text.splitlines(), start=1):
+        brace_depth += line.count("{") - line.count("}")
+
+        class_match = _JS_CLASS_RE.match(line)
+        if class_match:
+            symbols.append(Symbol(name=class_match.group(1), kind="class", range=(i, i)))
+            in_class = True
+            class_brace_depth = brace_depth
+            continue
+
+        if in_class and brace_depth < class_brace_depth:
+            in_class = False
+
+        func_match = _JS_FUNC_RE.match(line)
+        if func_match:
+            symbols.append(Symbol(name=func_match.group(1), kind="function", range=(i, i)))
+            continue
+
+        arrow_match = _JS_ARROW_RE.match(line)
+        if arrow_match:
+            symbols.append(Symbol(name=arrow_match.group(1), kind="function", range=(i, i)))
+            continue
+
+        if in_class:
+            method_match = _JS_METHOD_RE.match(line)
+            if method_match:
+                name = method_match.group(1)
+                if name != "constructor":
+                    symbols.append(Symbol(name=name, kind="method", range=(i, i)))
+
+    return symbols
+
+
+_JAVA_CLASS_RE = re.compile(r"^\s*(?:public\s+)?class\s+([A-Za-z_]\w*)\b")
+_JAVA_METHOD_RE = re.compile(
+    r"^\s*(?:public|private|protected|static|\s)+[\w<>\[\]]+\s+([A-Za-z_]\w*)\s*\("
+)
+
+
+def _parse_java_symbols(text: str) -> List[Symbol]:
+    symbols: List[Symbol] = []
+    for i, line in enumerate(text.splitlines(), start=1):
+        class_match = _JAVA_CLASS_RE.match(line)
+        if class_match:
+            symbols.append(Symbol(name=class_match.group(1), kind="class", range=(i, i)))
+            continue
+        method_match = _JAVA_METHOD_RE.match(line)
+        if method_match:
+            symbols.append(Symbol(name=method_match.group(1), kind="method", range=(i, i)))
+    return symbols
+
+
+_GO_FUNC_RE = re.compile(r"^\s*func\s+(?:\([^)]+\)\s+)?([A-Za-z_]\w*)\s*\(")
+_GO_TYPE_RE = re.compile(r"^\s*type\s+([A-Za-z_]\w*)\s+(?:struct|interface)\b")
+
+
+def _parse_go_symbols(text: str) -> List[Symbol]:
+    symbols: List[Symbol] = []
+    for i, line in enumerate(text.splitlines(), start=1):
+        type_match = _GO_TYPE_RE.match(line)
+        if type_match:
+            symbols.append(Symbol(name=type_match.group(1), kind="class", range=(i, i)))
+            continue
+        func_match = _GO_FUNC_RE.match(line)
+        if func_match:
+            symbols.append(Symbol(name=func_match.group(1), kind="function", range=(i, i)))
+    return symbols
+
+
+_GENERIC_DEF_RE = re.compile(r"^\s*(?:def|function|func)\s+([A-Za-z_]\w*)\b")
+_GENERIC_CLASS_RE = re.compile(r"^\s*(?:class|struct|interface)\s+([A-Za-z_]\w*)\b")
+
+
+def _parse_generic_symbols(text: str) -> List[Symbol]:
+    symbols: List[Symbol] = []
+    for i, line in enumerate(text.splitlines(), start=1):
+        class_match = _GENERIC_CLASS_RE.match(line)
+        if class_match:
+            symbols.append(Symbol(name=class_match.group(1), kind="class", range=(i, i)))
+            continue
+        def_match = _GENERIC_DEF_RE.match(line)
+        if def_match:
+            symbols.append(Symbol(name=def_match.group(1), kind="function", range=(i, i)))
+    return symbols
+
+
+# Markdown heading regex: # Heading, ## Heading, etc.
+_MD_HEADING_RE = re.compile(r"^(#{1,6})\s+(.+)$")
+
+
+def _parse_markdown_symbols(text: str) -> List[Symbol]:
+    """Parse Markdown headings as symbols.
+    
+    Extracts # headings as 'section' symbols with heading level as kind suffix.
+    """
+    symbols: List[Symbol] = []
+    for i, line in enumerate(text.splitlines(), start=1):
+        heading_match = _MD_HEADING_RE.match(line)
+        if heading_match:
+            level = len(heading_match.group(1))
+            title = heading_match.group(2).strip()
+            # Use 'section' kind with level indicator
+            kind = f"h{level}"
+            symbols.append(Symbol(name=title, kind=kind, range=(i, i)))
+    return symbols
+
+
+def _parse_text_symbols(text: str) -> List[Symbol]:
+    """Parse plain text files - no symbols, just index content."""
+    # Text files don't have structured symbols, return empty list
+    # The file content will still be indexed for FTS search
+    return []
+

package/codex-lens/src/codexlens/parsers/tokenizer.py

@@ -0,0 +1,98 @@
+"""Token counting utilities for CodexLens.
+
+Provides accurate token counting using tiktoken with character count fallback.
+"""
+
+from __future__ import annotations
+
+from typing import Optional
+
+try:
+    import tiktoken
+    TIKTOKEN_AVAILABLE = True
+except ImportError:
+    TIKTOKEN_AVAILABLE = False
+
+
+class Tokenizer:
+    """Token counter with tiktoken primary and character count fallback."""
+
+    def __init__(self, encoding_name: str = "cl100k_base") -> None:
+        """Initialize tokenizer.
+
+        Args:
+            encoding_name: Tiktoken encoding name (default: cl100k_base for GPT-4)
+        """
+        self._encoding: Optional[object] = None
+        self._encoding_name = encoding_name
+
+        if TIKTOKEN_AVAILABLE:
+            try:
+                self._encoding = tiktoken.get_encoding(encoding_name)
+            except Exception:
+                # Fallback to character counting if encoding fails
+                self._encoding = None
+
+    def count_tokens(self, text: str) -> int:
+        """Count tokens in text.
+
+        Uses tiktoken if available, otherwise falls back to character count / 4.
+
+        Args:
+            text: Text to count tokens for
+
+        Returns:
+            Estimated token count
+        """
+        if not text:
+            return 0
+
+        if self._encoding is not None:
+            try:
+                return len(self._encoding.encode(text))  # type: ignore[attr-defined]
+            except Exception:
+                # Fall through to character count fallback
+                pass
+
+        # Fallback: rough estimate using character count
+        # Average of ~4 characters per token for English text
+        return max(1, len(text) // 4)
+
+    def is_using_tiktoken(self) -> bool:
+        """Check if tiktoken is being used.
+
+        Returns:
+            True if tiktoken is available and initialized
+        """
+        return self._encoding is not None
+
+
+# Global default tokenizer instance
+_default_tokenizer: Optional[Tokenizer] = None
+
+
+def get_default_tokenizer() -> Tokenizer:
+    """Get the global default tokenizer instance.
+
+    Returns:
+        Shared Tokenizer instance
+    """
+    global _default_tokenizer
+    if _default_tokenizer is None:
+        _default_tokenizer = Tokenizer()
+    return _default_tokenizer
+
+
+def count_tokens(text: str, tokenizer: Optional[Tokenizer] = None) -> int:
+    """Count tokens in text using default or provided tokenizer.
+
+    Args:
+        text: Text to count tokens for
+        tokenizer: Optional tokenizer instance (uses default if None)
+
+    Returns:
+        Estimated token count
+    """
+    if tokenizer is None:
+        tokenizer = get_default_tokenizer()
+    return tokenizer.count_tokens(text)