attune-ai 2.0.0__py3-none-any.whl

attune/workflows/keyboard_shortcuts/parsers.py

@@ -0,0 +1,416 @@
+"""Feature discovery parsers for keyboard shortcut generation.
+
+Supports multiple input sources:
+- VSCode extension package.json
+- Python pyproject.toml entry points
+- Custom features.yaml manifest
+- LLM-based code analysis
+"""
+
+import json
+from abc import ABC, abstractmethod
+from pathlib import Path
+
+import yaml
+
+from .schema import Category, Feature, FeatureManifest, FrequencyTier
+
+
+class FeatureParser(ABC):
+    """Base class for feature parsers."""
+
+    @abstractmethod
+    def parse(self, path: Path) -> list[Feature]:
+        """Extract features from source file."""
+
+    @abstractmethod
+    def can_parse(self, path: Path) -> bool:
+        """Check if this parser can handle the given path."""
+
+
+class VSCodeCommandParser(FeatureParser):
+    """Parse VSCode extension package.json for commands."""
+
+    def can_parse(self, path: Path) -> bool:
+        return path.name == "package.json" and path.exists()
+
+    def parse(self, path: Path) -> list[Feature]:
+        """Extract commands from VSCode package.json."""
+        try:
+            pkg = json.loads(path.read_text())
+        except (json.JSONDecodeError, OSError):
+            return []
+
+        commands = pkg.get("contributes", {}).get("commands", [])
+        features = []
+
+        for cmd in commands:
+            command_id = cmd.get("command", "")
+            title = cmd.get("title", "")
+
+            # Skip internal/hidden commands
+            if not title or command_id.startswith("_"):
+                continue
+
+            # Extract category from title (e.g., "Empathy: Quick → Morning")
+            category = self._extract_category(title)
+            name = self._extract_name(title)
+
+            # Determine frequency tier based on category
+            tier = self._infer_tier(category)
+
+            features.append(
+                Feature(
+                    id=command_id.split(".")[-1],
+                    name=name,
+                    description=f"VSCode command: {command_id}",
+                    command=command_id,
+                    icon=cmd.get("icon", "$(symbol-misc)"),
+                    frequency=tier,
+                ),
+            )
+
+        return features
+
+    def _extract_category(self, title: str) -> str:
+        """Extract category from title like 'Empathy: Quick → Morning'."""
+        if "→" in title:
+            parts = title.split("→")
+            if len(parts) >= 2:
+                # Get the part before the arrow
+                prefix = parts[0].strip()
+                if ":" in prefix:
+                    return prefix.split(":")[-1].strip()
+                return prefix
+        if ":" in title:
+            return title.split(":")[0].strip()
+        return "General"
+
+    def _extract_name(self, title: str) -> str:
+        """Extract feature name from title like 'Empathy: Quick → Morning'."""
+        if "→" in title:
+            return title.split("→")[-1].strip()
+        if ":" in title:
+            return title.split(":")[-1].strip()
+        return title.strip()
+
+    def _infer_tier(self, category: str) -> FrequencyTier:
+        """Infer frequency tier from category name."""
+        category_lower = category.lower()
+        if "quick" in category_lower or "daily" in category_lower:
+            return FrequencyTier.DAILY
+        if "view" in category_lower or "workflow" in category_lower:
+            return FrequencyTier.FREQUENT
+        return FrequencyTier.ADVANCED
+
+
+class PyProjectParser(FeatureParser):
+    """Parse Python pyproject.toml for CLI entry points."""
+
+    def can_parse(self, path: Path) -> bool:
+        return path.name == "pyproject.toml" and path.exists()
+
+    def parse(self, path: Path) -> list[Feature]:
+        """Extract CLI scripts from pyproject.toml."""
+        try:
+            # Python 3.11+ has tomllib built-in
+            import tomllib
+
+            data = tomllib.loads(path.read_text())
+        except ImportError:
+            # Fallback for older Python
+            try:
+                import tomli as tomllib
+
+                data = tomllib.loads(path.read_text())
+            except ImportError:
+                return []
+            except (OSError, Exception):
+                return []
+        except (OSError, Exception):
+            return []
+
+        features = []
+
+        # Parse [project.scripts] section
+        scripts = data.get("project", {}).get("scripts", {})
+        for name, _entry_point in scripts.items():
+            features.append(
+                Feature(
+                    id=name.replace("-", "_"),
+                    name=name.replace("-", " ").title(),
+                    description=f"CLI command: {name}",
+                    command=f"cli.{name}",
+                    cli_alias=name,
+                    frequency=FrequencyTier.FREQUENT,
+                ),
+            )
+
+        # Parse [project.entry-points] section
+        entry_points = data.get("project", {}).get("entry-points", {})
+        for group, entries in entry_points.items():
+            if isinstance(entries, dict):
+                for name, entry_point in entries.items():
+                    features.append(
+                        Feature(
+                            id=f"{group}_{name}".replace("-", "_"),
+                            name=name.replace("-", " ").title(),
+                            description=f"Entry point: {group}.{name}",
+                            command=entry_point,
+                            frequency=FrequencyTier.ADVANCED,
+                        ),
+                    )
+
+        return features
+
+
+class YAMLManifestParser(FeatureParser):
+    """Parse custom features.yaml manifest."""
+
+    def can_parse(self, path: Path) -> bool:
+        return path.name in ("features.yaml", "features.yml") and path.exists()
+
+    def parse(self, path: Path) -> list[Feature]:
+        """Extract features from custom YAML manifest."""
+        try:
+            manifest = yaml.safe_load(path.read_text())
+        except (yaml.YAMLError, OSError):
+            return []
+
+        if not manifest or not isinstance(manifest, dict):
+            return []
+
+        features = []
+        categories = manifest.get("categories", [])
+
+        for category in categories:
+            if not isinstance(category, dict):
+                continue
+
+            category.get("name", "General")
+            tier_str = category.get("tier", "frequent")
+            tier = (
+                FrequencyTier(tier_str)
+                if tier_str in FrequencyTier._value2member_map_
+                else FrequencyTier.FREQUENT
+            )
+
+            for feat in category.get("features", []):
+                if not isinstance(feat, dict):
+                    continue
+
+                freq_str = feat.get("frequency", tier_str)
+                freq = (
+                    FrequencyTier(freq_str)
+                    if freq_str in FrequencyTier._value2member_map_
+                    else tier
+                )
+
+                features.append(
+                    Feature(
+                        id=feat.get("id", feat.get("name", "").lower().replace(" ", "_")),
+                        name=feat.get("name", ""),
+                        description=feat.get("description", ""),
+                        command=feat.get("command", ""),
+                        cli_alias=feat.get("cli_alias", ""),
+                        frequency=freq,
+                        context=feat.get("context", "global"),
+                        icon=feat.get("icon", "$(symbol-misc)"),
+                    ),
+                )
+
+        return features
+
+    def parse_full_manifest(self, path: Path) -> FeatureManifest | None:
+        """Parse complete manifest including project metadata."""
+        try:
+            data = yaml.safe_load(path.read_text())
+        except (yaml.YAMLError, OSError):
+            return None
+
+        if not data:
+            return None
+
+        project = data.get("project", {})
+        categories = []
+
+        for cat_data in data.get("categories", []):
+            features = []
+            tier_str = cat_data.get("tier", "frequent")
+            tier = (
+                FrequencyTier(tier_str)
+                if tier_str in FrequencyTier._value2member_map_
+                else FrequencyTier.FREQUENT
+            )
+
+            for feat_data in cat_data.get("features", []):
+                freq_str = feat_data.get("frequency", tier_str)
+                freq = (
+                    FrequencyTier(freq_str)
+                    if freq_str in FrequencyTier._value2member_map_
+                    else tier
+                )
+
+                features.append(
+                    Feature(
+                        id=feat_data.get("id", ""),
+                        name=feat_data.get("name", ""),
+                        description=feat_data.get("description", ""),
+                        command=feat_data.get("command", ""),
+                        cli_alias=feat_data.get("cli_alias", ""),
+                        frequency=freq,
+                        context=feat_data.get("context", "global"),
+                        icon=feat_data.get("icon", "$(symbol-misc)"),
+                    ),
+                )
+
+            categories.append(
+                Category(
+                    name=cat_data.get("name", "General"),
+                    icon=cat_data.get("icon", "$(folder)"),
+                    tier=tier,
+                    features=features,
+                ),
+            )
+
+        return FeatureManifest(
+            project_name=project.get("name", "project"),
+            project_type=project.get("type", "custom"),
+            prefix=project.get("prefix", "ctrl+shift+e"),
+            categories=categories,
+        )
+
+
+class LLMFeatureAnalyzer:
+    """Use LLM to analyze codebase and discover features."""
+
+    def __init__(self, llm_client=None):
+        """Initialize with optional LLM client."""
+        self.llm_client = llm_client
+
+    async def analyze_codebase(self, project_path: Path) -> list[Feature]:
+        """Analyze codebase using LLM to discover features.
+
+        This is a placeholder for LLM-based feature discovery.
+        The actual implementation would:
+        1. Scan for function signatures, CLI decorators, etc.
+        2. Send code snippets to LLM for analysis
+        3. Extract feature descriptions and usage patterns
+        """
+        # TODO: Implement LLM-based feature discovery
+        # For now, return empty list
+        return []
+
+
+class CompositeParser:
+    """Combines multiple parsers for comprehensive feature discovery."""
+
+    def __init__(self):
+        self.parsers: list[FeatureParser] = [
+            VSCodeCommandParser(),
+            PyProjectParser(),
+            YAMLManifestParser(),
+        ]
+        self.llm_analyzer = LLMFeatureAnalyzer()
+
+    def discover_features(self, project_path: Path) -> FeatureManifest:
+        """Discover all features from a project.
+
+        Scans for:
+        - VSCode package.json commands
+        - Python pyproject.toml entry points
+        - Custom features.yaml manifest
+        """
+        all_features: list[Feature] = []
+
+        # Try each parser
+        for parser in self.parsers:
+            for file_pattern in self._get_patterns(parser):
+                for found_path in project_path.rglob(file_pattern):
+                    if parser.can_parse(found_path):
+                        features = parser.parse(found_path)
+                        all_features.extend(features)
+
+        # Check for custom manifest with full metadata
+        manifest_path = project_path / "features.yaml"
+        if manifest_path.exists():
+            yaml_parser = YAMLManifestParser()
+            full_manifest = yaml_parser.parse_full_manifest(manifest_path)
+            if full_manifest:
+                return full_manifest
+
+        # Create manifest from discovered features
+        return self._create_manifest_from_features(project_path, all_features)
+
+    def _get_patterns(self, parser: FeatureParser) -> list[str]:
+        """Get file patterns for a parser."""
+        if isinstance(parser, VSCodeCommandParser):
+            return ["package.json"]
+        if isinstance(parser, PyProjectParser):
+            return ["pyproject.toml"]
+        if isinstance(parser, YAMLManifestParser):
+            return ["features.yaml", "features.yml"]
+        return []
+
+    def _create_manifest_from_features(
+        self,
+        project_path: Path,
+        features: list[Feature],
+    ) -> FeatureManifest:
+        """Create a manifest from discovered features."""
+        # Group features by inferred category
+        categories_dict: dict[str, list[Feature]] = {}
+        for feature in features:
+            # Try to infer category from feature properties
+            category = "General"
+            if "quick" in feature.id.lower() or feature.frequency == FrequencyTier.DAILY:
+                category = "Quick Actions"
+            elif "view" in feature.id.lower() or "dashboard" in feature.id.lower():
+                category = "Views"
+            elif "workflow" in feature.id.lower() or "run" in feature.id.lower():
+                category = "Workflows"
+            elif "security" in feature.id.lower() or "audit" in feature.id.lower():
+                category = "Security"
+
+            if category not in categories_dict:
+                categories_dict[category] = []
+            categories_dict[category].append(feature)
+
+        # Convert to Category objects
+        categories = []
+        for name, feats in categories_dict.items():
+            # Determine tier based on most common tier in category
+            tier = FrequencyTier.FREQUENT
+            if feats:
+                tier_counts: dict[FrequencyTier, int] = {}
+                for f in feats:
+                    tier_counts[f.frequency] = tier_counts.get(f.frequency, 0) + 1
+                tier = max(tier_counts.keys(), key=lambda t: tier_counts[t])
+
+            categories.append(
+                Category(
+                    name=name,
+                    tier=tier,
+                    features=feats,
+                ),
+            )
+
+        # Infer project name
+        project_name = project_path.name
+
+        # Infer project type - use cast for Literal typing
+        from typing import Literal, cast
+
+        project_type_str = "custom"
+        if (project_path / "package.json").exists():
+            project_type_str = "vscode-extension"
+        elif (project_path / "pyproject.toml").exists():
+            project_type_str = "python-cli"
+
+        project_type = cast("Literal['vscode-extension', 'python-cli', 'custom']", project_type_str)
+
+        return FeatureManifest(
+            project_name=project_name,
+            project_type=project_type,
+            categories=categories,
+        )