alita-sdk 0.3.462__py3-none-any.whl → 0.3.627__py3-none-any.whl

alita_sdk/community/inventory/patterns/__init__.py

@@ -0,0 +1,61 @@
+"""
+Cross-file reference pattern detection module.
+
+This module provides extensible pattern matching for detecting cross-file
+relationships in various programming languages and document types.
+
+Pattern categories:
+- imports: Code import/include statements
+- links: Documentation links (Markdown, Wiki, HTML)
+- citations: Text references ("see X", "@see X", etc.)
+- mentions: Entity name mentions in content
+
+Supported approaches:
+1. Regex patterns - Fast, language-agnostic pattern matching
+2. AST parsing - Accurate code analysis using deepwiki parsers (when available)
+
+Each pattern file defines patterns for a specific language or document type.
+"""
+
+from .registry import PatternRegistry, Pattern, PatternCategory, RelationType
+from .loader import (
+    load_all_patterns, 
+    get_patterns_for_file, 
+    get_patterns_for_content_type,
+    extract_references_from_content,
+)
+
+# AST adapter - provides integration with deepwiki parsers
+from .ast_adapter import (
+    is_ast_available,
+    get_supported_ast_languages,
+    parse_file_ast,
+    parse_files_ast,
+    extract_ast_cross_file_relations,
+    get_symbols_for_entity_matching,
+    ASTSymbol,
+    ASTRelation,
+    ASTParseResult,
+)
+
+__all__ = [
+    # Pattern system
+    'PatternRegistry',
+    'Pattern', 
+    'PatternCategory',
+    'RelationType',
+    'load_all_patterns',
+    'get_patterns_for_file',
+    'get_patterns_for_content_type',
+    'extract_references_from_content',
+    # AST adapter
+    'is_ast_available',
+    'get_supported_ast_languages',
+    'parse_file_ast',
+    'parse_files_ast',
+    'extract_ast_cross_file_relations',
+    'get_symbols_for_entity_matching',
+    'ASTSymbol',
+    'ASTRelation',
+    'ASTParseResult',
+]

alita_sdk/community/inventory/patterns/ast_adapter.py

@@ -0,0 +1,380 @@
+"""
+AST Adapter - Bridge between parsers module and patterns module.
+
+This module provides a unified interface for AST-based code analysis,
+integrating with the local parsers module for first-pass extraction.
+"""
+
+from typing import Dict, List, Any, Optional, Set, Tuple
+from pathlib import Path
+import logging
+from dataclasses import dataclass
+
+# Import from local parsers module (self-contained, no external dependencies)
+from ..parsers import (
+    parser_registry,
+    parse_file,
+    parse_files,
+    get_parser_for_file,
+    ParseResult,
+    Symbol,
+    Relationship,
+    RelationshipType,
+    SymbolType,
+)
+
+logger = logging.getLogger(__name__)
+
+# Type aliases for compatibility with patterns module
+ASTSymbol = Symbol
+ASTRelation = Relationship
+ASTParseResult = ParseResult
+
+
+def is_ast_available() -> bool:
+    """
+    Check if AST parsing is available.
+    
+    Since we use the local parsers module with Python's built-in ast
+    and regex patterns, this is always available.
+    
+    Returns:
+        True - AST parsing is always available
+    """
+    return True
+
+
+def get_supported_languages() -> List[str]:
+    """
+    Get list of supported programming languages.
+    
+    Returns:
+        List of language identifiers
+    """
+    return parser_registry.get_supported_languages()
+
+
+def get_supported_ast_languages() -> List[str]:
+    """
+    Get list of supported programming languages for AST parsing.
+    
+    Alias for get_supported_languages() for compatibility with patterns module.
+    
+    Returns:
+        List of language identifiers
+    """
+    return get_supported_languages()
+
+
+def get_supported_extensions() -> List[str]:
+    """
+    Get list of supported file extensions.
+    
+    Returns:
+        List of file extensions (with dot prefix)
+    """
+    return parser_registry.get_supported_extensions()
+
+
+def parse_file_ast(
+    file_path: str,
+    content: Optional[str] = None
+) -> Optional[ParseResult]:
+    """
+    Parse a single file and extract symbols and relationships.
+    
+    Args:
+        file_path: Path to the file to parse
+        content: Optional file content (if not provided, will read from disk)
+        
+    Returns:
+        ParseResult with symbols and relationships, or None if parsing fails
+    """
+    try:
+        return parse_file(file_path, content)
+    except Exception as e:
+        logger.warning(f"Failed to parse {file_path}: {e}")
+        return None
+
+
+def parse_files_ast(
+    file_paths: List[str],
+    file_contents: Optional[Dict[str, str]] = None,
+    resolve_cross_file: bool = True,
+    max_workers: int = 4
+) -> Dict[str, ParseResult]:
+    """
+    Parse multiple files with optional cross-file resolution.
+    
+    Args:
+        file_paths: List of file paths to parse
+        file_contents: Optional dict mapping file paths to content
+        resolve_cross_file: Whether to resolve cross-file references
+        max_workers: Maximum number of parallel workers
+        
+    Returns:
+        Dict mapping file paths to ParseResult objects
+    """
+    results = {}
+    
+    # Group files by language
+    files_by_language: Dict[str, List[Tuple[str, Optional[str]]]] = {}
+    
+    for file_path in file_paths:
+        parser = get_parser_for_file(file_path)
+        if parser:
+            lang = parser.language
+            if lang not in files_by_language:
+                files_by_language[lang] = []
+            content = file_contents.get(file_path) if file_contents else None
+            files_by_language[lang].append((file_path, content))
+    
+    # Parse each language group
+    for lang, files in files_by_language.items():
+        parser = parser_registry.get_parser(lang)
+        if parser and hasattr(parser, 'parse_multiple_files'):
+            try:
+                lang_results = parser.parse_multiple_files(
+                    files,
+                    resolve_cross_file=resolve_cross_file,
+                    max_workers=max_workers
+                )
+                results.update(lang_results)
+            except Exception as e:
+                logger.warning(f"Failed to parse {lang} files: {e}")
+                # Fall back to individual parsing
+                for file_path, content in files:
+                    result = parse_file_ast(file_path, content)
+                    if result:
+                        results[file_path] = result
+        else:
+            # Individual parsing fallback
+            for file_path, content in files:
+                result = parse_file_ast(file_path, content)
+                if result:
+                    results[file_path] = result
+    
+    return results
+
+
+def extract_ast_cross_file_relations(
+    parse_results: Dict[str, ParseResult]
+) -> List[Dict[str, Any]]:
+    """
+    Extract cross-file relationships from parse results.
+    
+    Args:
+        parse_results: Dict mapping file paths to ParseResult objects
+        
+    Returns:
+        List of relationship dictionaries with source_file, target_file, etc.
+    """
+    cross_file_relations = []
+    
+    # Build symbol index
+    symbol_to_file: Dict[str, str] = {}
+    for file_path, result in parse_results.items():
+        for symbol in result.symbols:
+            symbol_to_file[symbol.name] = file_path
+            if symbol.qualified_name:
+                symbol_to_file[symbol.qualified_name] = file_path
+    
+    # Find cross-file relationships
+    for file_path, result in parse_results.items():
+        for rel in result.relationships:
+            target_file = symbol_to_file.get(rel.target)
+            if target_file and target_file != file_path:
+                cross_file_relations.append({
+                    'source_file': file_path,
+                    'source': rel.source,
+                    'target_file': target_file,
+                    'target': rel.target,
+                    'relationship_type': rel.relationship_type.value,
+                    'metadata': rel.metadata
+                })
+    
+    return cross_file_relations
+
+
+def extract_first_pass_entities(
+    file_path: str,
+    content: str
+) -> List[Dict[str, Any]]:
+    """
+    Extract entities (symbols) from a file for first-pass processing.
+    
+    This is a non-LLM based extraction using AST/regex patterns.
+    
+    Args:
+        file_path: Path to the file
+        content: File content
+        
+    Returns:
+        List of entity dictionaries
+    """
+    result = parse_file_ast(file_path, content)
+    if not result:
+        return []
+    
+    entities = []
+    for symbol in result.symbols:
+        entity = {
+            'name': symbol.name,
+            'type': symbol.symbol_type.value,
+            'file_path': file_path,
+            'line_start': symbol.line_start,
+            'line_end': symbol.line_end,
+            'qualified_name': symbol.qualified_name,
+        }
+        if symbol.docstring:
+            entity['docstring'] = symbol.docstring
+        if symbol.metadata:
+            entity['metadata'] = symbol.metadata
+        entities.append(entity)
+    
+    return entities
+
+
+def extract_first_pass_relations(
+    file_path: str,
+    content: str
+) -> List[Dict[str, Any]]:
+    """
+    Extract relationships from a file for first-pass processing.
+    
+    This is a non-LLM based extraction using AST/regex patterns.
+    
+    Args:
+        file_path: Path to the file
+        content: File content
+        
+    Returns:
+        List of relationship dictionaries
+    """
+    result = parse_file_ast(file_path, content)
+    if not result:
+        return []
+    
+    relations = []
+    for rel in result.relationships:
+        relation = {
+            'source': rel.source,
+            'target': rel.target,
+            'relationship_type': rel.relationship_type.value,
+            'file_path': file_path,
+            'line': rel.line,
+        }
+        if rel.metadata:
+            relation['metadata'] = rel.metadata
+        relations.append(relation)
+    
+    return relations
+
+
+def get_file_summary(
+    file_path: str,
+    content: str
+) -> Dict[str, Any]:
+    """
+    Get a summary of a file's structure.
+    
+    Args:
+        file_path: Path to the file
+        content: File content
+        
+    Returns:
+        Summary dictionary with counts and top-level items
+    """
+    result = parse_file_ast(file_path, content)
+    if not result:
+        return {
+            'file_path': file_path,
+            'language': 'unknown',
+            'symbols': [],
+            'relationships': [],
+            'counts': {}
+        }
+    
+    # Count by type
+    symbol_counts: Dict[str, int] = {}
+    for symbol in result.symbols:
+        type_name = symbol.symbol_type.value
+        symbol_counts[type_name] = symbol_counts.get(type_name, 0) + 1
+    
+    rel_counts: Dict[str, int] = {}
+    for rel in result.relationships:
+        type_name = rel.relationship_type.value
+        rel_counts[type_name] = rel_counts.get(type_name, 0) + 1
+    
+    return {
+        'file_path': file_path,
+        'language': result.language,
+        'symbols': [s.name for s in result.symbols],
+        'relationships': len(result.relationships),
+        'counts': {
+            'symbols': symbol_counts,
+            'relationships': rel_counts
+        }
+    }
+
+
+def get_symbols_for_entity_matching(
+    parse_results: Dict[str, ParseResult]
+) -> Dict[str, List[Dict[str, Any]]]:
+    """
+    Get symbols formatted for entity matching.
+    
+    This function extracts symbols from parse results and formats them
+    for use in entity matching algorithms.
+    
+    Args:
+        parse_results: Dict mapping file paths to ParseResult objects
+        
+    Returns:
+        Dict mapping file paths to lists of symbol dictionaries
+    """
+    result = {}
+    for file_path, parse_result in parse_results.items():
+        symbols = []
+        for symbol in parse_result.symbols:
+            symbols.append({
+                'name': symbol.name,
+                'type': symbol.symbol_type.value,
+                'full_name': symbol.full_name or symbol.get_qualified_name(),
+                'file_path': symbol.file_path,
+                'visibility': symbol.visibility,
+                'docstring': symbol.docstring,
+                'range': {
+                    'start_line': symbol.range.start.line,
+                    'end_line': symbol.range.end.line
+                } if symbol.range else None,
+                'metadata': symbol.metadata
+            })
+        result[file_path] = symbols
+    return result
+
+
+# Export key functions and types
+__all__ = [
+    'is_ast_available',
+    'get_supported_languages',
+    'get_supported_ast_languages',
+    'get_supported_extensions',
+    'parse_file_ast',
+    'parse_files_ast',
+    'extract_ast_cross_file_relations',
+    'extract_first_pass_entities',
+    'extract_first_pass_relations',
+    'get_file_summary',
+    'get_symbols_for_entity_matching',
+    # Type aliases
+    'ASTSymbol',
+    'ASTRelation',
+    'ASTParseResult',
+    # Re-export types
+    'ParseResult',
+    'Symbol',
+    'Relationship',
+    'RelationshipType',
+    'SymbolType',
+]