Merge branch 'staging' into backend/add-import-support-python

Author: gkorland

api/analyzers/javascript/__init__.py

@@ -0,0 +1,172 @@
+"""JavaScript analyzer using tree-sitter for code entity extraction."""
+
+from pathlib import Path
+from typing import Optional
+
+from multilspy import SyncLanguageServer
+from ...entities.entity import Entity
+from ...entities.file import File
+from ..analyzer import AbstractAnalyzer
+
+import tree_sitter_javascript as tsjs
+from tree_sitter import Language, Node
+
+import logging
+logger = logging.getLogger('code_graph')
+
+
+class JavaScriptAnalyzer(AbstractAnalyzer):
+    """Analyzer for JavaScript source files using tree-sitter.
+
+    Extracts functions, classes, and methods from JavaScript code.
+    Resolves class inheritance (extends) and function/method call references.
+    """
+
+    def __init__(self) -> None:
+        """Initialize the JavaScript analyzer with the tree-sitter JS grammar."""
+        super().__init__(Language(tsjs.language()))
+
+    def add_dependencies(self, path: Path, files: list[Path]) -> None:
+        """Detect and register JavaScript project dependencies.
+
+        Currently a no-op; npm dependency resolution is not yet implemented.
+        """
+        pass
+
+    def get_entity_label(self, node: Node) -> str:
+        """Return the graph label for a given AST node type.
+
+        Args:
+            node: A tree-sitter AST node representing a JavaScript entity.
+
+        Returns:
+            One of 'Function', 'Class', or 'Method'.
+
+        Raises:
+            ValueError: If the node type is not a recognised entity.
+        """
+        if node.type == 'function_declaration':
+            return "Function"
+        elif node.type == 'class_declaration':
+            return "Class"
+        elif node.type == 'method_definition':
+            return "Method"
+        raise ValueError(f"Unknown entity type: {node.type}")
+
+    def get_entity_name(self, node: Node) -> str:
+        """Extract the declared name from a JavaScript entity node.
+
+        Args:
+            node: A tree-sitter AST node for a function, class, or method.
+
+        Returns:
+            The entity name, or an empty string if no name node is found.
+
+        Raises:
+            ValueError: If the node type is not a recognised entity.
+        """
+        if node.type in ['function_declaration', 'class_declaration', 'method_definition']:
+            name_node = node.child_by_field_name('name')
+            if name_node is None:
+                return ''
+            return name_node.text.decode('utf-8')
+        raise ValueError(f"Unknown entity type: {node.type}")
+
+    def get_entity_docstring(self, node: Node) -> Optional[str]:
+        """Extract a leading comment as a docstring for the entity.
+
+        Looks for a comment node immediately preceding the entity in the AST.
+
+        Args:
+            node: A tree-sitter AST node for a function, class, or method.
+
+        Returns:
+            The comment text, or None if no leading comment exists.
+
+        Raises:
+            ValueError: If the node type is not a recognised entity.
+        """
+        if node.type in ['function_declaration', 'class_declaration', 'method_definition']:
+            if node.prev_sibling and node.prev_sibling.type == 'comment':
+                return node.prev_sibling.text.decode('utf-8')
+            return None
+        raise ValueError(f"Unknown entity type: {node.type}")
+
+    def get_entity_types(self) -> list[str]:
+        """Return the tree-sitter node types recognised as JavaScript entities."""
+        return ['function_declaration', 'class_declaration', 'method_definition']
+
+    def add_symbols(self, entity: Entity) -> None:
+        """Extract symbols (references) from a JavaScript entity.
+
+        For classes: extracts base-class identifiers from ``extends`` clauses.
+        For functions/methods: extracts call-expression references.
+
+        Note:
+            JavaScript parameters are untyped, so they are not captured as
+            symbols — unlike typed languages (Java, Python) where parameter
+            type annotations are meaningful for resolution.
+        """
+        if entity.node.type == 'class_declaration':
+            for child in entity.node.children:
+                if child.type == 'class_heritage':
+                    for heritage_child in child.children:
+                        if heritage_child.type == 'identifier':
+                            entity.add_symbol("base_class", heritage_child)
+        elif entity.node.type in ['function_declaration', 'method_definition']:
+            captures = self._captures("(call_expression) @reference.call", entity.node)
+            if 'reference.call' in captures:
+                for caller in captures['reference.call']:
+                    entity.add_symbol("call", caller)
+
+    def is_dependency(self, file_path: str) -> bool:
+        """Check whether a file path belongs to an external dependency.
+
+        Uses path-segment matching so that directories merely containing
+        'node_modules' in their name (e.g. ``node_modules_utils``) are not
+        treated as dependencies.
+        """
+        return "node_modules" in Path(file_path).parts
+
+    def resolve_path(self, file_path: str, path: Path) -> str:
+        """Resolve an import path relative to the project root."""
+        return file_path
+
+    def resolve_type(self, files: dict[Path, File], lsp: SyncLanguageServer, file_path: Path, path: Path, node: Node) -> list[Entity]:
+        """Resolve a type reference to its class declaration entity."""
+        res = []
+        for file, resolved_node in self.resolve(files, lsp, file_path, path, node):
+            type_dec = self.find_parent(resolved_node, ['class_declaration'])
+            if type_dec in file.entities:
+                res.append(file.entities[type_dec])
+        return res
+
+    def resolve_method(self, files: dict[Path, File], lsp: SyncLanguageServer, file_path: Path, path: Path, node: Node) -> list[Entity]:
+        """Resolve a call expression to the target function or method entity."""
+        res = []
+        if node.type == 'call_expression':
+            func_node = node.child_by_field_name('function')
+            if func_node and func_node.type == 'member_expression':
+                func_node = func_node.child_by_field_name('property')
+            if func_node:
+                node = func_node
+        for file, resolved_node in self.resolve(files, lsp, file_path, path, node):
+            method_dec = self.find_parent(resolved_node, ['function_declaration', 'method_definition', 'class_declaration'])
+            if method_dec and method_dec.type == 'class_declaration':
+                continue
+            if method_dec in file.entities:
+                res.append(file.entities[method_dec])
+        return res
+
+    def resolve_symbol(self, files: dict[Path, File], lsp: SyncLanguageServer, file_path: Path, path: Path, key: str, symbol: Node) -> list[Entity]:
+        """Dispatch symbol resolution based on the symbol category.
+
+        Routes ``base_class`` symbols to type resolution and ``call`` symbols
+        to method resolution.
+        """
+        if key == "base_class":
+            return self.resolve_type(files, lsp, file_path, path, symbol)
+        elif key == "call":
+            return self.resolve_method(files, lsp, file_path, path, symbol)
+        else:
+            raise ValueError(f"Unknown key {key}")

api/analyzers/javascript/analyzer.py

@@ -11,6 +11,7 @@
 from .java.analyzer import JavaAnalyzer
 from .python.analyzer import PythonAnalyzer
 from .csharp.analyzer import CSharpAnalyzer
+from .javascript.analyzer import JavaScriptAnalyzer
 
 from multilspy import SyncLanguageServer
 from multilspy.multilspy_config import MultilspyConfig
@@ -26,7 +27,8 @@
     # '.h': CAnalyzer(),
     '.py': PythonAnalyzer(),
     '.java': JavaAnalyzer(),
-    '.cs': CSharpAnalyzer()}
+    '.cs': CSharpAnalyzer(),
+    '.js': JavaScriptAnalyzer()}
 
 class NullLanguageServer:
     def start_server(self):
@@ -147,9 +149,15 @@ def second_pass(self, graph: Graph, files: list[Path], path: Path) -> None:
             lsps[".cs"] = SyncLanguageServer.create(config, logger, str(path))
         else:
             lsps[".cs"] = NullLanguageServer()
-        with lsps[".java"].start_server(), lsps[".py"].start_server(), lsps[".cs"].start_server():
+        lsps[".js"] = NullLanguageServer()
+        with lsps[".java"].start_server(), lsps[".py"].start_server(), lsps[".cs"].start_server(), lsps[".js"].start_server():
             files_len = len(self.files)
             for i, file_path in enumerate(files):
+                if file_path not in self.files:
+                    continue
+                # Skip symbol resolution when no real LSP is available
+                if isinstance(lsps.get(file_path.suffix), NullLanguageServer):
+                    continue
                 file = self.files[file_path]
                 logging.info(f'Processing file ({i + 1}/{files_len}): {file_path}')
 
@@ -187,7 +195,7 @@ def analyze_files(self, files: list[Path], path: Path, graph: Graph) -> None:
 
     def analyze_sources(self, path: Path, ignore: list[str], graph: Graph) -> None:
         path = path.resolve()
-        files = list(path.rglob("*.java")) + list(path.rglob("*.py")) + list(path.rglob("*.cs"))
+        files = list(path.rglob("*.java")) + list(path.rglob("*.py")) + list(path.rglob("*.cs")) + [f for f in path.rglob("*.js") if "node_modules" not in f.parts]
         # First pass analysis of the source code
         self.first_pass(path, files, ignore, graph)
 

api/analyzers/source_analyzer.py

@@ -13,6 +13,7 @@ dependencies = [
     "tree-sitter-c>=0.24.1,<0.25.0",
     "tree-sitter-python>=0.25.0,<0.26.0",
     "tree-sitter-java>=0.23.5,<0.24.0",
+    "tree-sitter-javascript>=0.23.0,<0.24.0",
     "tree-sitter-c-sharp>=0.23.1,<0.24.0",
     "fastapi>=0.115.0,<1.0.0",
     "uvicorn[standard]>=0.34.0,<1.0.0",

pyproject.toml

@@ -0,0 +1,31 @@
+/**
+ * Base class for shapes
+ */
+class Shape {
+    constructor(name) {
+        this.name = name;
+    }
+
+    area() {
+        return 0;
+    }
+}
+
+class Circle extends Shape {
+    constructor(radius) {
+        super(radius);
+        this.radius = radius;
+    }
+
+    area() {
+        return Math.PI * this.radius * this.radius;
+    }
+}
+
+function calculateTotal(shapes) {
+    let total = 0;
+    for (const shape of shapes) {
+        total += shape.area();
+    }
+    return total;
+}