delimit-cli 2.3.2 → 3.0.0

package/gateway/core/dependency_manifest.py

@@ -0,0 +1,223 @@
+"""
+Delimit Dependency Manifest
+Parses and validates .delimit/dependencies.yaml service dependency declarations.
+
+Per Jamsons Doctrine:
+- Deterministic outputs
+- No credential discovery
+- No telemetry
+"""
+
+import logging
+import re
+from pathlib import Path
+from typing import Any, Dict, List, Optional, Union
+
+logger = logging.getLogger("delimit.dependency_manifest")
+
+# Manifest filename convention
+MANIFEST_FILENAME = "dependencies.yaml"
+MANIFEST_DIR = ".delimit"
+
+# Valid characters for service identifiers: alphanumeric, hyphens, underscores, slashes
+_SERVICE_ID_PATTERN = re.compile(r"^[a-zA-Z0-9][a-zA-Z0-9._/-]*$")
+
+# Required fields in a manifest
+REQUIRED_FIELDS = frozenset(["service"])
+
+
+def _load_yaml():
+    """Lazy import yaml to avoid hard dependency."""
+    try:
+        import yaml
+        return yaml
+    except ImportError:
+        return None
+
+
+def normalize_service_id(service_id: str) -> str:
+    """Normalize a service identifier to lowercase with consistent separators.
+
+    Args:
+        service_id: Raw service identifier string.
+
+    Returns:
+        Normalized lowercase identifier.
+    """
+    return service_id.strip().lower()
+
+
+def validate_service_id(service_id: str) -> Optional[str]:
+    """Validate a service identifier.
+
+    Returns None if valid, or an error message string if invalid.
+    """
+    if not service_id or not isinstance(service_id, str):
+        return "Service identifier must be a non-empty string"
+
+    normalized = normalize_service_id(service_id)
+    if not normalized:
+        return "Service identifier is empty after normalization"
+
+    if not _SERVICE_ID_PATTERN.match(normalized):
+        return (
+            f"Invalid service identifier: {service_id!r}. "
+            "Must start with alphanumeric and contain only "
+            "alphanumeric, hyphens, underscores, dots, or slashes."
+        )
+
+    return None
+
+
+def validate_manifest(data: Dict[str, Any]) -> List[str]:
+    """Validate a parsed dependency manifest.
+
+    Returns a list of error strings. Empty list means valid.
+    """
+    errors: List[str] = []
+
+    if not isinstance(data, dict):
+        return ["Manifest must be a YAML mapping"]
+
+    # Check required fields
+    if "service" not in data:
+        errors.append("Missing required field: 'service'")
+    else:
+        err = validate_service_id(data["service"])
+        if err:
+            errors.append(f"Invalid 'service' field: {err}")
+
+    # Validate consumes list
+    consumes = data.get("consumes")
+    if consumes is not None:
+        if not isinstance(consumes, list):
+            errors.append("Field 'consumes' must be a list")
+        else:
+            for i, item in enumerate(consumes):
+                if not isinstance(item, str):
+                    errors.append(f"consumes[{i}] must be a string, got {type(item).__name__}")
+                else:
+                    err = validate_service_id(item)
+                    if err:
+                        errors.append(f"consumes[{i}]: {err}")
+
+    # Validate produces list
+    produces = data.get("produces")
+    if produces is not None:
+        if not isinstance(produces, list):
+            errors.append("Field 'produces' must be a list")
+        else:
+            for i, item in enumerate(produces):
+                if not isinstance(item, str):
+                    errors.append(f"produces[{i}] must be a string, got {type(item).__name__}")
+                else:
+                    err = validate_service_id(item)
+                    if err:
+                        errors.append(f"produces[{i}]: {err}")
+
+    # Validate optional string fields
+    for field in ("owner", "repository"):
+        val = data.get(field)
+        if val is not None and not isinstance(val, str):
+            errors.append(f"Field '{field}' must be a string")
+
+    return errors
+
+
+def parse_manifest(data: Dict[str, Any]) -> Dict[str, Any]:
+    """Parse and normalize a dependency manifest.
+
+    Args:
+        data: Raw parsed YAML dictionary.
+
+    Returns:
+        Normalized manifest dictionary with deterministic structure.
+
+    Raises:
+        ValueError: If validation fails.
+    """
+    errors = validate_manifest(data)
+    if errors:
+        raise ValueError(f"Manifest validation failed: {'; '.join(errors)}")
+
+    manifest = {
+        "service": normalize_service_id(data["service"]),
+        "consumes": sorted(set(
+            normalize_service_id(s) for s in data.get("consumes", [])
+        )),
+        "produces": sorted(set(
+            normalize_service_id(s) for s in data.get("produces", [])
+        )),
+    }
+
+    if "owner" in data and isinstance(data["owner"], str):
+        manifest["owner"] = data["owner"].strip()
+    if "repository" in data and isinstance(data["repository"], str):
+        manifest["repository"] = data["repository"].strip()
+
+    return manifest
+
+
+def load_manifest_file(path: Union[str, Path]) -> Optional[Dict[str, Any]]:
+    """Load and parse a dependency manifest from a YAML file.
+
+    Args:
+        path: Path to the dependencies.yaml file.
+
+    Returns:
+        Parsed manifest dictionary, or None if the file doesn't exist
+        or can't be parsed.
+    """
+    yaml = _load_yaml()
+    if yaml is None:
+        logger.warning("PyYAML not installed — cannot parse dependency manifests")
+        return None
+
+    filepath = Path(path)
+    if not filepath.exists():
+        return None
+
+    try:
+        with open(filepath, "r", encoding="utf-8") as f:
+            data = yaml.safe_load(f)
+    except Exception as e:
+        logger.warning("Failed to load manifest %s: %s", filepath, e)
+        return None
+
+    if data is None:
+        return None
+
+    try:
+        return parse_manifest(data)
+    except (ValueError, TypeError, AttributeError) as e:
+        logger.warning("Invalid manifest %s: %s", filepath, e)
+        return None
+
+
+def discover_manifests(root_dir: Union[str, Path]) -> List[Dict[str, Any]]:
+    """Discover and load all dependency manifests under a root directory.
+
+    Searches for .delimit/dependencies.yaml files recursively.
+
+    Args:
+        root_dir: Root directory to search.
+
+    Returns:
+        List of parsed manifest dictionaries, sorted by service name.
+    """
+    root = Path(root_dir)
+    manifests = []
+
+    if not root.exists():
+        return manifests
+
+    # Search for manifest files
+    for manifest_path in root.rglob(f"{MANIFEST_DIR}/{MANIFEST_FILENAME}"):
+        manifest = load_manifest_file(manifest_path)
+        if manifest is not None:
+            manifest["_source_path"] = str(manifest_path)
+            manifests.append(manifest)
+
+    # Deterministic ordering
+    manifests.sort(key=lambda m: m["service"])
+    return manifests

package/gateway/core/diff_engine_v2.py

@@ -0,0 +1,477 @@
+"""
+Enhanced OpenAPI diff engine with deep schema comparison.
+Handles nested objects, response schemas, enums, and edge cases.
+"""
+
+from typing import Dict, List, Any, Optional, Set, Tuple
+from dataclasses import dataclass
+from enum import Enum
+
+class ChangeType(Enum):
+    # Breaking changes
+    ENDPOINT_REMOVED = "endpoint_removed"
+    METHOD_REMOVED = "method_removed"
+    REQUIRED_PARAM_ADDED = "required_param_added"
+    PARAM_REMOVED = "param_removed"
+    RESPONSE_REMOVED = "response_removed"
+    REQUIRED_FIELD_ADDED = "required_field_added"
+    FIELD_REMOVED = "field_removed"
+    TYPE_CHANGED = "type_changed"
+    FORMAT_CHANGED = "format_changed"
+    ENUM_VALUE_REMOVED = "enum_value_removed"
+    
+    # Non-breaking changes
+    ENDPOINT_ADDED = "endpoint_added"
+    METHOD_ADDED = "method_added"
+    OPTIONAL_PARAM_ADDED = "optional_param_added"
+    RESPONSE_ADDED = "response_added"
+    OPTIONAL_FIELD_ADDED = "optional_field_added"
+    ENUM_VALUE_ADDED = "enum_value_added"
+    DESCRIPTION_CHANGED = "description_changed"
+
+@dataclass
+class Change:
+    type: ChangeType
+    path: str
+    details: Dict[str, Any]
+    severity: str  # high, medium, low
+    message: str
+    
+    @property
+    def is_breaking(self) -> bool:
+        return self.type in [
+            ChangeType.ENDPOINT_REMOVED,
+            ChangeType.METHOD_REMOVED,
+            ChangeType.REQUIRED_PARAM_ADDED,
+            ChangeType.PARAM_REMOVED,
+            ChangeType.RESPONSE_REMOVED,
+            ChangeType.REQUIRED_FIELD_ADDED,
+            ChangeType.FIELD_REMOVED,
+            ChangeType.TYPE_CHANGED,
+            ChangeType.FORMAT_CHANGED,
+            ChangeType.ENUM_VALUE_REMOVED,
+        ]
+
+class OpenAPIDiffEngine:
+    """Advanced diff engine for OpenAPI specifications."""
+    
+    def __init__(self):
+        self.changes: List[Change] = []
+        self._old_spec: Dict = {}
+        self._new_spec: Dict = {}
+        self._ref_trail: Set[Tuple[str, str]] = set()
+
+    def compare(self, old_spec: Dict, new_spec: Dict) -> List[Change]:
+        """Compare two OpenAPI specifications and return all changes."""
+        self.changes = []
+        self._old_spec = old_spec or {}
+        self._new_spec = new_spec or {}
+        self._ref_trail = set()
+
+        # Compare paths
+        self._compare_paths(self._old_spec.get("paths", {}), self._new_spec.get("paths", {}))
+        
+        # Compare components/schemas
+        self._compare_schemas(
+            self._old_spec.get("components", {}).get("schemas", {}),
+            self._new_spec.get("components", {}).get("schemas", {})
+        )
+        
+        # Compare security schemes
+        self._compare_security(
+            self._old_spec.get("components", {}).get("securitySchemes", {}),
+            self._new_spec.get("components", {}).get("securitySchemes", {})
+        )
+        
+        return self.changes
+    
+    def _compare_paths(self, old_paths: Dict, new_paths: Dict):
+        """Compare API paths/endpoints."""
+        old_set = set(old_paths.keys())
+        new_set = set(new_paths.keys())
+        
+        # Check removed endpoints
+        for path in old_set - new_set:
+            self.changes.append(Change(
+                type=ChangeType.ENDPOINT_REMOVED,
+                path=path,
+                details={"endpoint": path},
+                severity="high",
+                message=f"Endpoint removed: {path}"
+            ))
+        
+        # Check added endpoints
+        for path in new_set - old_set:
+            self.changes.append(Change(
+                type=ChangeType.ENDPOINT_ADDED,
+                path=path,
+                details={"endpoint": path},
+                severity="low",
+                message=f"New endpoint added: {path}"
+            ))
+        
+        # Check modified endpoints
+        for path in old_set & new_set:
+            self._compare_methods(path, old_paths[path], new_paths[path])
+    
+    def _compare_methods(self, path: str, old_methods: Dict, new_methods: Dict):
+        """Compare HTTP methods for an endpoint."""
+        old_set = set(m for m in old_methods.keys() if m in ["get", "post", "put", "delete", "patch", "head", "options"])
+        new_set = set(m for m in new_methods.keys() if m in ["get", "post", "put", "delete", "patch", "head", "options"])
+        
+        # Check removed methods
+        for method in old_set - new_set:
+            self.changes.append(Change(
+                type=ChangeType.METHOD_REMOVED,
+                path=f"{path}:{method.upper()}",
+                details={"endpoint": path, "method": method.upper()},
+                severity="high",
+                message=f"Method removed: {method.upper()} {path}"
+            ))
+        
+        # Check modified methods
+        for method in old_set & new_set:
+            self._compare_operation(
+                f"{path}:{method.upper()}",
+                old_methods[method],
+                new_methods[method]
+            )
+    
+    def _compare_operation(self, operation_id: str, old_op: Dict, new_op: Dict):
+        """Compare operation details (parameters, responses, etc.)."""
+        
+        # Compare parameters
+        old_params = {self._param_key(p): p for p in old_op.get("parameters", [])}
+        new_params = {self._param_key(p): p for p in new_op.get("parameters", [])}
+        
+        # Check removed parameters
+        for param_key in set(old_params.keys()) - set(new_params.keys()):
+            param = old_params[param_key]
+            self.changes.append(Change(
+                type=ChangeType.PARAM_REMOVED,
+                path=operation_id,
+                details={"parameter": param["name"], "in": param["in"]},
+                severity="high",
+                message=f"Parameter removed: {param['name']} from {operation_id}"
+            ))
+        
+        # Check added required parameters
+        for param_key in set(new_params.keys()) - set(old_params.keys()):
+            param = new_params[param_key]
+            if param.get("required", False):
+                self.changes.append(Change(
+                    type=ChangeType.REQUIRED_PARAM_ADDED,
+                    path=operation_id,
+                    details={"parameter": param["name"], "in": param["in"]},
+                    severity="high",
+                    message=f"Required parameter added: {param['name']} to {operation_id}"
+                ))
+        
+        # Check parameter schema changes
+        for param_key in set(old_params.keys()) & set(new_params.keys()):
+            self._compare_parameter_schemas(
+                operation_id,
+                old_params[param_key],
+                new_params[param_key]
+            )
+        
+        # Compare request body
+        if "requestBody" in old_op or "requestBody" in new_op:
+            self._compare_request_body(
+                operation_id,
+                old_op.get("requestBody"),
+                new_op.get("requestBody")
+            )
+        
+        # Compare responses
+        self._compare_responses(
+            operation_id,
+            old_op.get("responses", {}),
+            new_op.get("responses", {})
+        )
+    
+    def _compare_parameter_schemas(self, operation_id: str, old_param: Dict, new_param: Dict):
+        """Compare parameter schemas for type changes."""
+        old_schema = old_param.get("schema", {})
+        new_schema = new_param.get("schema", {})
+
+        # Resolve $ref in parameter schemas
+        if "$ref" in old_schema:
+            old_schema = self._resolve_schema(old_schema, self._old_spec)
+        if "$ref" in new_schema:
+            new_schema = self._resolve_schema(new_schema, self._new_spec)
+        
+        # Check type changes
+        if old_schema.get("type") != new_schema.get("type"):
+            self.changes.append(Change(
+                type=ChangeType.TYPE_CHANGED,
+                path=operation_id,
+                details={
+                    "parameter": old_param["name"],
+                    "old_type": old_schema.get("type"),
+                    "new_type": new_schema.get("type")
+                },
+                severity="high",
+                message=f"Parameter type changed: {old_param['name']} from {old_schema.get('type')} to {new_schema.get('type')}"
+            ))
+        
+        # Check enum changes
+        if "enum" in old_schema or "enum" in new_schema:
+            self._compare_enums(
+                f"{operation_id}:{old_param['name']}",
+                old_schema.get("enum", []),
+                new_schema.get("enum", [])
+            )
+    
+    def _compare_request_body(self, operation_id: str, old_body: Optional[Dict], new_body: Optional[Dict]):
+        """Compare request body schemas."""
+        if old_body and not new_body:
+            self.changes.append(Change(
+                type=ChangeType.FIELD_REMOVED,
+                path=operation_id,
+                details={"field": "request_body"},
+                severity="high",
+                message=f"Request body removed from {operation_id}"
+            ))
+        elif not old_body and new_body and new_body.get("required", False):
+            self.changes.append(Change(
+                type=ChangeType.REQUIRED_FIELD_ADDED,
+                path=operation_id,
+                details={"field": "request_body"},
+                severity="high",
+                message=f"Required request body added to {operation_id}"
+            ))
+        elif old_body and new_body:
+            # Compare content types
+            old_content = old_body.get("content", {})
+            new_content = new_body.get("content", {})
+            
+            for content_type in old_content.keys() & new_content.keys():
+                self._compare_schema_deep(
+                    f"{operation_id}:request",
+                    old_content[content_type].get("schema", {}),
+                    new_content[content_type].get("schema", {})
+                )
+    
+    def _compare_responses(self, operation_id: str, old_responses: Dict, new_responses: Dict):
+        """Compare response definitions."""
+        old_codes = set(old_responses.keys())
+        new_codes = set(new_responses.keys())
+        
+        # Check removed responses
+        for code in old_codes - new_codes:
+            # Only flag 2xx responses as breaking
+            if code.startswith("2"):
+                self.changes.append(Change(
+                    type=ChangeType.RESPONSE_REMOVED,
+                    path=operation_id,
+                    details={"response_code": code},
+                    severity="high",
+                    message=f"Success response {code} removed from {operation_id}"
+                ))
+        
+        # Compare response schemas
+        for code in old_codes & new_codes:
+            old_resp = old_responses[code]
+            new_resp = new_responses[code]
+            
+            if "content" in old_resp or "content" in new_resp:
+                old_content = old_resp.get("content", {})
+                new_content = new_resp.get("content", {})
+                
+                for content_type in old_content.keys() & new_content.keys():
+                    self._compare_schema_deep(
+                        f"{operation_id}:{code}",
+                        old_content[content_type].get("schema", {}),
+                        new_content[content_type].get("schema", {})
+                    )
+    
+    def _resolve_ref(self, ref_string: str, spec: Dict) -> Optional[Dict]:
+        """Resolve a JSON $ref pointer like #/components/schemas/User."""
+        if not ref_string.startswith('#/'):
+            return None
+        parts = ref_string[2:].split('/')
+        current = spec
+        for part in parts:
+            part = part.replace('~1', '/').replace('~0', '~')
+            if isinstance(current, dict) and part in current:
+                current = current[part]
+            else:
+                return None
+        return current if isinstance(current, dict) else None
+
+    def _resolve_schema(self, schema: Dict, spec: Dict, visited: Optional[Set[str]] = None) -> Dict:
+        """Follow $ref chains, detecting circular references."""
+        if visited is None:
+            visited = set()
+        if '$ref' not in schema:
+            return schema
+        ref = schema['$ref']
+        if ref in visited:
+            return schema  # circular — return as-is
+        visited.add(ref)
+        resolved = self._resolve_ref(ref, spec)
+        if resolved is None:
+            return schema  # unresolvable — return as-is
+        if '$ref' in resolved:
+            return self._resolve_schema(resolved, spec, visited)
+        return resolved
+
+    def _compare_schema_deep(self, path: str, old_schema: Dict, new_schema: Dict, required_fields: Optional[Set[str]] = None):
+        """Deep comparison of schemas including nested objects."""
+
+        # Handle references — resolve before comparing
+        if "$ref" in old_schema or "$ref" in new_schema:
+            old_resolved = self._resolve_schema(old_schema, self._old_spec) if "$ref" in old_schema else old_schema
+            new_resolved = self._resolve_schema(new_schema, self._new_spec) if "$ref" in new_schema else new_schema
+            # Track ref pairs to avoid infinite loops on circular schemas
+            ref_key = (old_schema.get("$ref", ""), new_schema.get("$ref", ""))
+            if ref_key in self._ref_trail:
+                return
+            self._ref_trail.add(ref_key)
+            # Compare the resolved schemas
+            self._compare_schema_deep(path, old_resolved, new_resolved, required_fields)
+            return
+        
+        # Compare types
+        old_type = old_schema.get("type")
+        new_type = new_schema.get("type")
+        
+        if old_type != new_type and old_type is not None:
+            self.changes.append(Change(
+                type=ChangeType.TYPE_CHANGED,
+                path=path,
+                details={"old_type": old_type, "new_type": new_type},
+                severity="high",
+                message=f"Type changed from {old_type} to {new_type} at {path}"
+            ))
+            return
+        
+        # Compare object properties
+        if old_type == "object":
+            old_props = old_schema.get("properties", {})
+            new_props = new_schema.get("properties", {})
+            old_required = set(old_schema.get("required", []))
+            new_required = set(new_schema.get("required", []))
+            
+            # Check removed fields
+            for prop in set(old_props.keys()) - set(new_props.keys()):
+                if prop in old_required:
+                    self.changes.append(Change(
+                        type=ChangeType.FIELD_REMOVED,
+                        path=f"{path}.{prop}",
+                        details={"field": prop},
+                        severity="high",
+                        message=f"Required field '{prop}' removed at {path}"
+                    ))
+            
+            # Check new required fields
+            for prop in new_required - old_required:
+                if prop not in old_props:
+                    self.changes.append(Change(
+                        type=ChangeType.REQUIRED_FIELD_ADDED,
+                        path=f"{path}.{prop}",
+                        details={"field": prop},
+                        severity="high",
+                        message=f"New required field '{prop}' added at {path}"
+                    ))
+            
+            # Recursively compare nested properties
+            for prop in set(old_props.keys()) & set(new_props.keys()):
+                self._compare_schema_deep(
+                    f"{path}.{prop}",
+                    old_props[prop],
+                    new_props[prop],
+                    old_required if prop in old_required else None
+                )
+        
+        # Compare arrays
+        elif old_type == "array":
+            if "items" in old_schema and "items" in new_schema:
+                self._compare_schema_deep(
+                    f"{path}[]",
+                    old_schema["items"],
+                    new_schema["items"]
+                )
+        
+        # Compare enums
+        if "enum" in old_schema or "enum" in new_schema:
+            self._compare_enums(path, old_schema.get("enum", []), new_schema.get("enum", []))
+    
+    def _compare_enums(self, path: str, old_enum: List, new_enum: List):
+        """Compare enum values."""
+        old_set = set(old_enum)
+        new_set = set(new_enum)
+        
+        # Removed enum values are breaking
+        for value in old_set - new_set:
+            self.changes.append(Change(
+                type=ChangeType.ENUM_VALUE_REMOVED,
+                path=path,
+                details={"value": value},
+                severity="high",
+                message=f"Enum value '{value}' removed at {path}"
+            ))
+        
+        # Added enum values are non-breaking
+        for value in new_set - old_set:
+            self.changes.append(Change(
+                type=ChangeType.ENUM_VALUE_ADDED,
+                path=path,
+                details={"value": value},
+                severity="low",
+                message=f"Enum value '{value}' added at {path}"
+            ))
+    
+    def _compare_schemas(self, old_schemas: Dict, new_schemas: Dict):
+        """Compare component schemas."""
+        # Schema removal is breaking if referenced
+        for schema_name in set(old_schemas.keys()) - set(new_schemas.keys()):
+            self.changes.append(Change(
+                type=ChangeType.FIELD_REMOVED,
+                path=f"#/components/schemas/{schema_name}",
+                details={"schema": schema_name},
+                severity="medium",
+                message=f"Schema '{schema_name}' removed"
+            ))
+        
+        # Compare existing schemas
+        for schema_name in set(old_schemas.keys()) & set(new_schemas.keys()):
+            self._compare_schema_deep(
+                f"#/components/schemas/{schema_name}",
+                old_schemas[schema_name],
+                new_schemas[schema_name]
+            )
+    
+    def _compare_security(self, old_security: Dict, new_security: Dict):
+        """Compare security schemes."""
+        # Security scheme changes are usually breaking
+        for scheme in set(old_security.keys()) - set(new_security.keys()):
+            self.changes.append(Change(
+                type=ChangeType.FIELD_REMOVED,
+                path=f"#/components/securitySchemes/{scheme}",
+                details={"scheme": scheme},
+                severity="high",
+                message=f"Security scheme '{scheme}' removed"
+            ))
+    
+    def _param_key(self, param: Dict) -> str:
+        """Generate unique key for parameter."""
+        return f"{param.get('in', 'query')}:{param.get('name', '')}"
+    
+    def get_breaking_changes(self) -> List[Change]:
+        """Get only breaking changes."""
+        return [c for c in self.changes if c.is_breaking]
+    
+    def get_summary(self) -> Dict[str, Any]:
+        """Get summary of all changes."""
+        breaking = self.get_breaking_changes()
+        return {
+            "total_changes": len(self.changes),
+            "breaking_changes": len(breaking),
+            "endpoints_removed": len([c for c in breaking if c.type == ChangeType.ENDPOINT_REMOVED]),
+            "methods_removed": len([c for c in breaking if c.type == ChangeType.METHOD_REMOVED]),
+            "parameters_changed": len([c for c in breaking if c.type in [ChangeType.PARAM_REMOVED, ChangeType.REQUIRED_PARAM_ADDED]]),
+            "schemas_changed": len([c for c in breaking if c.type in [ChangeType.FIELD_REMOVED, ChangeType.REQUIRED_FIELD_ADDED, ChangeType.TYPE_CHANGED]]),
+            "is_breaking": len(breaking) > 0
+        }