hanzo-mcp 0.6.13__py3-none-any.whl → 0.7.1__py3-none-any.whl

hanzo_mcp/tools/search/find_tool.py

@@ -0,0 +1,581 @@
+"""Fast file finding tool using ffind and intelligent caching."""
+
+import os
+import time
+import fnmatch
+from typing import List, Optional, Dict, Any, Set
+from pathlib import Path
+from dataclasses import dataclass
+import subprocess
+import json
+from datetime import datetime
+
+from hanzo_mcp.tools.common.base import BaseTool
+from hanzo_mcp.tools.common.paginated_response import AutoPaginatedResponse
+from hanzo_mcp.tools.common.decorators import with_context_normalization
+from hanzo_mcp.types import MCPResourceDocument
+
+try:
+    import ffind
+    FFIND_AVAILABLE = True
+except ImportError:
+    FFIND_AVAILABLE = False
+
+
+@dataclass
+class FileMatch:
+    """Represents a found file."""
+    path: str
+    name: str
+    size: int
+    modified: float
+    is_dir: bool
+    extension: str
+    depth: int
+    
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "path": self.path,
+            "name": self.name,
+            "size": self.size,
+            "modified": datetime.fromtimestamp(self.modified).isoformat(),
+            "is_dir": self.is_dir,
+            "extension": self.extension,
+            "depth": self.depth
+        }
+
+
+class FindTool(BaseTool):
+    """Fast file and directory finding tool.
+    
+    This tool is optimized for quickly finding files and directories by name,
+    pattern, or attributes. It uses ffind (when available) for blazing fast
+    performance and falls back to optimized Python implementation.
+    
+    Key features:
+    - Lightning fast file discovery
+    - Smart pattern matching (glob, regex, fuzzy)
+    - File attribute filtering (size, date, type)
+    - Intelligent result ranking
+    - Built-in caching for repeated searches
+    - Respects .gitignore by default
+    """
+    
+    name = "find"
+    description = """Find files and directories by name, pattern, or attributes.
+    
+    Examples:
+    - find("*.py") - Find all Python files
+    - find("test_", type="file") - Find files starting with test_
+    - find("src", type="dir") - Find directories named src
+    - find("TODO", in_content=True) - Find files containing TODO
+    - find("large", min_size="10MB") - Find files larger than 10MB
+    - find("recent", modified_after="1 day ago") - Recently modified files
+    
+    This is the primary tool for discovering files in a project. Use it before
+    reading or searching within files.
+    """
+    
+    def __init__(self):
+        super().__init__()
+        self._cache = {}
+        self._gitignore_cache = {}
+        
+    def _parse_size(self, size_str: str) -> int:
+        """Parse human-readable size to bytes."""
+        # Order matters - check longer units first
+        units = [
+            ('TB', 1024**4), ('GB', 1024**3), ('MB', 1024**2), ('KB', 1024),
+            ('T', 1024**4), ('G', 1024**3), ('M', 1024**2), ('K', 1024), ('B', 1)
+        ]
+        
+        size_str = size_str.upper().strip()
+        for unit, multiplier in units:
+            if size_str.endswith(unit):
+                num_str = size_str[:-len(unit)].strip()
+                if num_str:
+                    try:
+                        return int(float(num_str) * multiplier)
+                    except ValueError:
+                        return 0
+        
+        try:
+            return int(size_str)
+        except ValueError:
+            return 0
+    
+    def _parse_time(self, time_str: str) -> float:
+        """Parse human-readable time to timestamp."""
+        import re
+        from datetime import datetime, timedelta
+        
+        # Handle relative times like "1 day ago", "2 hours ago"
+        match = re.match(r'(\d+)\s*(second|minute|hour|day|week|month|year)s?\s*ago', time_str.lower())
+        if match:
+            amount = int(match.group(1))
+            unit = match.group(2)
+            
+            if unit == 'second':
+                delta = timedelta(seconds=amount)
+            elif unit == 'minute':
+                delta = timedelta(minutes=amount)
+            elif unit == 'hour':
+                delta = timedelta(hours=amount)
+            elif unit == 'day':
+                delta = timedelta(days=amount)
+            elif unit == 'week':
+                delta = timedelta(weeks=amount)
+            elif unit == 'month':
+                delta = timedelta(days=amount * 30)  # Approximate
+            elif unit == 'year':
+                delta = timedelta(days=amount * 365)  # Approximate
+            
+            return (datetime.now() - delta).timestamp()
+        
+        # Try parsing as date
+        try:
+            return datetime.fromisoformat(time_str).timestamp()
+        except:
+            return datetime.now().timestamp()
+    
+    def _load_gitignore(self, root: str) -> Set[str]:
+        """Load and parse .gitignore patterns."""
+        if root in self._gitignore_cache:
+            return self._gitignore_cache[root]
+        
+        patterns = set()
+        gitignore_path = Path(root) / '.gitignore'
+        
+        if gitignore_path.exists():
+            try:
+                with open(gitignore_path, 'r') as f:
+                    for line in f:
+                        line = line.strip()
+                        if line and not line.startswith('#'):
+                            patterns.add(line)
+            except:
+                pass
+        
+        # Add common ignore patterns
+        patterns.update([
+            '*.pyc', '__pycache__', '.git', '.svn', '.hg',
+            'node_modules', '.env', '.venv', 'venv',
+            '*.swp', '*.swo', '.DS_Store', 'Thumbs.db'
+        ])
+        
+        self._gitignore_cache[root] = patterns
+        return patterns
+    
+    def _should_ignore(self, path: str, ignore_patterns: Set[str]) -> bool:
+        """Check if path should be ignored."""
+        path_obj = Path(path)
+        
+        for pattern in ignore_patterns:
+            # Check against full path and basename
+            if fnmatch.fnmatch(path_obj.name, pattern):
+                return True
+            if fnmatch.fnmatch(str(path_obj), pattern):
+                return True
+            
+            # Check if any parent directory matches
+            for parent in path_obj.parents:
+                if fnmatch.fnmatch(parent.name, pattern):
+                    return True
+        
+        return False
+    
+    async def run(self,
+                  pattern: str = "*",
+                  path: str = ".",
+                  type: Optional[str] = None,  # "file", "dir", "any"
+                  min_size: Optional[str] = None,
+                  max_size: Optional[str] = None,
+                  modified_after: Optional[str] = None,
+                  modified_before: Optional[str] = None,
+                  max_depth: Optional[int] = None,
+                  case_sensitive: bool = False,
+                  regex: bool = False,
+                  fuzzy: bool = False,
+                  in_content: bool = False,
+                  follow_symlinks: bool = False,
+                  respect_gitignore: bool = True,
+                  max_results: int = 1000,
+                  sort_by: str = "path",  # "path", "name", "size", "modified"
+                  reverse: bool = False,
+                  page_size: int = 100,
+                  page: int = 1,
+                  **kwargs) -> MCPResourceDocument:
+        """Find files and directories.
+        
+        Args:
+            pattern: Search pattern (glob by default, regex if regex=True)
+            path: Root directory to search from
+            type: Filter by type ("file", "dir", "any")
+            min_size: Minimum file size (e.g., "1MB", "500K")
+            max_size: Maximum file size
+            modified_after: Find files modified after this time
+            modified_before: Find files modified before this time
+            max_depth: Maximum directory depth to search
+            case_sensitive: Case-sensitive matching
+            regex: Treat pattern as regex instead of glob
+            fuzzy: Use fuzzy matching for pattern
+            in_content: Search for pattern inside files (slower)
+            follow_symlinks: Follow symbolic links
+            respect_gitignore: Respect .gitignore patterns
+            max_results: Maximum results to return
+            sort_by: Sort results by attribute
+            reverse: Reverse sort order
+            page_size: Results per page
+            page: Page number
+        """
+        
+        start_time = time.time()
+        
+        # Resolve path
+        root_path = Path(path).resolve()
+        if not root_path.exists():
+            return MCPResourceDocument(data={
+                "error": f"Path does not exist: {path}",
+                "results": []
+            })
+        
+        # Get ignore patterns
+        ignore_patterns = set()
+        if respect_gitignore:
+            ignore_patterns = self._load_gitignore(str(root_path))
+        
+        # Parse filters
+        min_size_bytes = self._parse_size(min_size) if min_size else None
+        max_size_bytes = self._parse_size(max_size) if max_size else None
+        modified_after_ts = self._parse_time(modified_after) if modified_after else None
+        modified_before_ts = self._parse_time(modified_before) if modified_before else None
+        
+        # Collect matches
+        matches = []
+        
+        if FFIND_AVAILABLE and not in_content:
+            # Use ffind for fast file discovery
+            matches = await self._find_with_ffind(
+                pattern, root_path, type, case_sensitive, regex,
+                max_depth, follow_symlinks, ignore_patterns
+            )
+        else:
+            # Fall back to Python implementation
+            matches = await self._find_with_python(
+                pattern, root_path, type, case_sensitive, regex, fuzzy,
+                in_content, max_depth, follow_symlinks, respect_gitignore, ignore_patterns
+            )
+        
+        # Apply filters
+        filtered_matches = []
+        for match in matches:
+            # Size filters
+            if min_size_bytes and match.size < min_size_bytes:
+                continue
+            if max_size_bytes and match.size > max_size_bytes:
+                continue
+            
+            # Time filters
+            if modified_after_ts and match.modified < modified_after_ts:
+                continue
+            if modified_before_ts and match.modified > modified_before_ts:
+                continue
+            
+            filtered_matches.append(match)
+            
+            if len(filtered_matches) >= max_results:
+                break
+        
+        # Sort results
+        if sort_by == "name":
+            filtered_matches.sort(key=lambda m: m.name, reverse=reverse)
+        elif sort_by == "size":
+            filtered_matches.sort(key=lambda m: m.size, reverse=reverse)
+        elif sort_by == "modified":
+            filtered_matches.sort(key=lambda m: m.modified, reverse=reverse)
+        else:  # path
+            filtered_matches.sort(key=lambda m: m.path, reverse=reverse)
+        
+        # Paginate
+        total_results = len(filtered_matches)
+        start_idx = (page - 1) * page_size
+        end_idx = start_idx + page_size
+        page_results = filtered_matches[start_idx:end_idx]
+        
+        # Format results
+        formatted_results = [match.to_dict() for match in page_results]
+        
+        # Statistics
+        stats = {
+            "total_found": total_results,
+            "search_time_ms": int((time.time() - start_time) * 1000),
+            "search_method": "ffind" if FFIND_AVAILABLE and not in_content else "python",
+            "root_path": str(root_path),
+            "filters_applied": {
+                "pattern": pattern,
+                "type": type,
+                "size": {"min": min_size, "max": max_size} if min_size or max_size else None,
+                "modified": {"after": modified_after, "before": modified_before} if modified_after or modified_before else None,
+                "max_depth": max_depth,
+                "gitignore": respect_gitignore
+            }
+        }
+        
+        return MCPResourceDocument(data={
+            "results": formatted_results,
+            "pagination": {
+                "page": page,
+                "page_size": page_size,
+                "total_results": total_results,
+                "total_pages": (total_results + page_size - 1) // page_size,
+                "has_next": end_idx < total_results,
+                "has_prev": page > 1
+            },
+            "statistics": stats
+        })
+    
+    async def call(self, **kwargs) -> str:
+        """Tool interface for MCP - converts result to JSON string."""
+        result = await self.run(**kwargs)
+        return result.to_json_string()
+    
+    def register(self, mcp_server) -> None:
+        """Register tool with MCP server."""
+        from mcp.server import FastMCP
+        
+        @mcp_server.tool(name=self.name, description=self.description)
+        async def find_handler(
+            pattern: str,
+            path: str = ".",
+            type: Optional[str] = None,
+            max_results: int = 100,
+            max_depth: Optional[int] = None,
+            case_sensitive: bool = False,
+            regex: bool = False,
+            fuzzy: bool = False,
+            min_size: Optional[str] = None,
+            max_size: Optional[str] = None,
+            modified_after: Optional[str] = None,
+            modified_before: Optional[str] = None,
+            follow_symlinks: bool = True,
+            respect_gitignore: bool = True,
+            sort_by: str = "name",
+            reverse: bool = False,
+            page_size: int = 50,
+            page: int = 1,
+        ) -> str:
+            """Execute file finding."""
+            return await self.call(
+                pattern=pattern,
+                path=path,
+                type=type,
+                max_results=max_results,
+                max_depth=max_depth,
+                case_sensitive=case_sensitive,
+                regex=regex,
+                fuzzy=fuzzy,
+                min_size=min_size,
+                max_size=max_size,
+                modified_after=modified_after,
+                modified_before=modified_before,
+                follow_symlinks=follow_symlinks,
+                respect_gitignore=respect_gitignore,
+                sort_by=sort_by,
+                reverse=reverse,
+                page_size=page_size,
+                page=page,
+            )
+    
+    async def _find_with_ffind(self,
+                              pattern: str,
+                              root: Path,
+                              file_type: Optional[str],
+                              case_sensitive: bool,
+                              regex: bool,
+                              max_depth: Optional[int],
+                              follow_symlinks: bool,
+                              ignore_patterns: Set[str]) -> List[FileMatch]:
+        """Use ffind for fast file discovery."""
+        matches = []
+        
+        # Configure ffind
+        ffind_args = {
+            'path': str(root),
+            'pattern': pattern,
+            'regex': regex,
+            'case_sensitive': case_sensitive,
+            'follow_symlinks': follow_symlinks,
+        }
+        
+        if max_depth:
+            ffind_args['max_depth'] = max_depth
+        
+        try:
+            # Run ffind
+            results = ffind.find(**ffind_args)
+            
+            for result in results:
+                path = result['path']
+                
+                # Check ignore patterns
+                if self._should_ignore(path, ignore_patterns):
+                    continue
+                
+                # Get file info
+                try:
+                    stat = os.stat(path)
+                    is_dir = os.path.isdir(path)
+                    
+                    # Apply type filter
+                    if file_type == "file" and is_dir:
+                        continue
+                    if file_type == "dir" and not is_dir:
+                        continue
+                    
+                    match = FileMatch(
+                        path=path,
+                        name=os.path.basename(path),
+                        size=stat.st_size,
+                        modified=stat.st_mtime,
+                        is_dir=is_dir,
+                        extension=Path(path).suffix,
+                        depth=len(Path(path).relative_to(root).parts)
+                    )
+                    matches.append(match)
+                    
+                except OSError:
+                    continue
+                    
+        except Exception as e:
+            # Fall back to Python implementation
+            return await self._find_with_python(
+                pattern, root, file_type, case_sensitive, regex, False,
+                False, max_depth, follow_symlinks, respect_gitignore, ignore_patterns
+            )
+        
+        return matches
+    
+    async def _find_with_python(self,
+                               pattern: str,
+                               root: Path,
+                               file_type: Optional[str],
+                               case_sensitive: bool,
+                               regex: bool,
+                               fuzzy: bool,
+                               in_content: bool,
+                               max_depth: Optional[int],
+                               follow_symlinks: bool,
+                               respect_gitignore: bool,
+                               ignore_patterns: Set[str]) -> List[FileMatch]:
+        """Python implementation of file finding."""
+        matches = []
+        
+        import re
+        from difflib import SequenceMatcher
+        
+        # Prepare pattern matcher
+        if regex:
+            flags = 0 if case_sensitive else re.IGNORECASE
+            try:
+                pattern_re = re.compile(pattern, flags)
+                matcher = lambda name: pattern_re.search(name) is not None
+            except re.error:
+                matcher = lambda name: pattern in name
+        elif fuzzy:
+            pattern_lower = pattern.lower() if not case_sensitive else pattern
+            matcher = lambda name: SequenceMatcher(None, pattern_lower, 
+                                                  name.lower() if not case_sensitive else name).ratio() > 0.6
+        else:
+            # Glob pattern
+            if not case_sensitive:
+                pattern = pattern.lower()
+                matcher = lambda name: fnmatch.fnmatch(name.lower(), pattern)
+            else:
+                matcher = lambda name: fnmatch.fnmatch(name, pattern)
+        
+        # Walk directory tree
+        for dirpath, dirnames, filenames in os.walk(str(root), followlinks=follow_symlinks):
+            # Check depth
+            if max_depth is not None:
+                depth = len(Path(dirpath).relative_to(root).parts)
+                if depth > max_depth:
+                    dirnames.clear()  # Don't recurse deeper
+                    continue
+            
+            # Filter directories to skip
+            if respect_gitignore:
+                dirnames[:] = [d for d in dirnames if not self._should_ignore(os.path.join(dirpath, d), ignore_patterns)]
+            
+            # Check directories
+            if file_type != "file":
+                for dirname in dirnames:
+                    if matcher(dirname):
+                        full_path = os.path.join(dirpath, dirname)
+                        if not self._should_ignore(full_path, ignore_patterns):
+                            try:
+                                stat = os.stat(full_path)
+                                match = FileMatch(
+                                    path=full_path,
+                                    name=dirname,
+                                    size=0,  # Directories don't have size
+                                    modified=stat.st_mtime,
+                                    is_dir=True,
+                                    extension="",
+                                    depth=len(Path(full_path).relative_to(root).parts)
+                                )
+                                matches.append(match)
+                            except OSError:
+                                continue
+            
+            # Check files
+            if file_type != "dir":
+                for filename in filenames:
+                    full_path = os.path.join(dirpath, filename)
+                    
+                    if self._should_ignore(full_path, ignore_patterns):
+                        continue
+                    
+                    # Match against filename
+                    if matcher(filename):
+                        match_found = True
+                    elif in_content:
+                        # Search in file content
+                        match_found = await self._search_in_file(full_path, pattern, case_sensitive)
+                    else:
+                        match_found = False
+                    
+                    if match_found:
+                        try:
+                            stat = os.stat(full_path)
+                            match = FileMatch(
+                                path=full_path,
+                                name=filename,
+                                size=stat.st_size,
+                                modified=stat.st_mtime,
+                                is_dir=False,
+                                extension=Path(filename).suffix,
+                                depth=len(Path(full_path).relative_to(root).parts)
+                            )
+                            matches.append(match)
+                        except OSError:
+                            continue
+        
+        return matches
+    
+    async def _search_in_file(self, file_path: str, pattern: str, case_sensitive: bool) -> bool:
+        """Search for pattern in file content."""
+        try:
+            with open(file_path, 'r', encoding='utf-8', errors='ignore') as f:
+                content = f.read()
+                if not case_sensitive:
+                    return pattern.lower() in content.lower()
+                else:
+                    return pattern in content
+        except:
+            return False
+
+
+# Tool registration
+def create_find_tool():
+    """Factory function to create find tool."""
+    return FindTool()