added the chat endpoint

Author: swilly22

code_graph/__init__.py

@@ -1,4 +1,5 @@
 from .info import *
+from .llm import ask
 from .graph import *
 from .project import *
 from .entities import *

code_graph/analyzers/python/analyzer.py

@@ -75,7 +75,7 @@ def process_class_definition(self, node: Node, path: Path) -> tuple[Class, list[
 
         return (c, inherited_classes)
 
-    def process_function_definition(self, node: Node, path: Path) -> Function:
+    def process_function_definition(self, node: Node, path: Path, source_code: str) -> Function:
         """
         Processes a function definition node from the syntax tree and extracts relevant information.
 
@@ -154,7 +154,8 @@ def process_function_definition(self, node: Node, path: Path) -> Function:
         ret_type = return_type.text.decode('utf-8') if return_type else None
 
         # Create Function object
-        f = Function(str(path), function_name, docstring, ret_type, '', start_line, end_line)
+        src = source_code[node.start_byte:node.end_byte]
+        f = Function(str(path), function_name, docstring, ret_type, src, start_line, end_line)
 
         # Add arguments to Function object
         for arg in args:
@@ -163,7 +164,7 @@ def process_function_definition(self, node: Node, path: Path) -> Function:
         return f
 
     def first_pass_traverse(self, parent: Union[File,Class,Function], node: Node,
-                            path: Path, graph: Graph) -> None:
+                            path: Path, graph: Graph, source_code: str) -> None:
         """
         Recursively traverses a syntax tree node, processes class and function definitions,
         and connects them in a graph representation.
@@ -198,7 +199,7 @@ def first_pass_traverse(self, parent: Union[File,Class,Function], node: Node,
             graph.add_class(entity)
 
         elif node.type == "function_definition":
-            entity = self.process_function_definition(node, path)
+            entity = self.process_function_definition(node, path, source_code)
             # Add Function object to the graph
             graph.add_function(entity)
 
@@ -209,7 +210,7 @@ def first_pass_traverse(self, parent: Union[File,Class,Function], node: Node,
 
         # Recursivly visit child nodes
         for child in node.children:
-            self.first_pass_traverse(parent, child, path, graph)
+            self.first_pass_traverse(parent, child, path, graph, source_code)
 
     def first_pass(self, path: Path, f: io.TextIOWrapper, graph:Graph) -> None:
         """
@@ -231,11 +232,16 @@ def first_pass(self, path: Path, f: io.TextIOWrapper, graph:Graph) -> None:
         graph.add_file(file)
 
         # Parse file
-        content = f.read()
-        tree = self.parser.parse(content)
+        source_code = f.read()
+        tree = self.parser.parse(source_code)
+        try:
+            source_code = source_code.decode('utf-8')
+        except Exception as e:
+            logger.error(f"Failed decoding source code: {e}")
+            source_code = ''
 
         # Walk thought the AST
-        self.first_pass_traverse(file, tree.root_node, path, graph)
+        self.first_pass_traverse(file, tree.root_node, path, graph, source_code)
 
     def process_function_call(self, node) -> Optional[str]:
         """
@@ -324,7 +330,7 @@ def process_inheritance(self, cls: Class, super_classes: list[str],
             graph.connect_entities('INHERITS', cls.id, _super_class.id)
 
     def second_pass_traverse(self, parent: Union[File, Class, Function],
-                             node: Node, path: Path, graph: Graph) -> None:
+                             node: Node, path: Path, graph: Graph, source_code: str) -> None:
         """
         Traverse the AST nodes during the second pass and process each node accordingly.
 
@@ -341,7 +347,9 @@ def second_pass_traverse(self, parent: Union[File, Class, Function],
             parent = cls
 
         elif node.type == "function_definition":
-            func = self.process_function_definition(node, path)
+            # TODO: simply extract function name, no need to parse entire function
+            # see C analyzer
+            func = self.process_function_definition(node, path, source_code)
             parent = graph.get_function_by_name(func.name)
         elif node.type == "call":
             callee = self.process_function_call(node)
@@ -350,7 +358,7 @@ def second_pass_traverse(self, parent: Union[File, Class, Function],
 
         # Recursivly visit child nodes
         for child in node.children:
-            self.second_pass_traverse(parent, child, path, graph)
+            self.second_pass_traverse(parent, child, path, graph, source_code)
 
     def second_pass(self, path: Path, f: io.TextIOWrapper, graph: Graph) -> None:
         """
@@ -375,10 +383,10 @@ def second_pass(self, path: Path, f: io.TextIOWrapper, graph: Graph) -> None:
 
         try:
             # Parse file
-            content = f.read()
-            tree = self.parser.parse(content)
+            source_code = f.read()
+            tree = self.parser.parse(source_code)
 
             # Walk thought the AST
-            self.second_pass_traverse(file, tree.root_node, path, graph)
+            self.second_pass_traverse(file, tree.root_node, path, graph, source_code)
         except Exception as e:
             logger.error(f"Failed to process file {path}: {e}")

code_graph/analyzers/source_analyzer.py

@@ -11,7 +11,8 @@
 from .python.analyzer import PythonAnalyzer
 
 import logging
-logger = logging.getLogger('code_graph')
+# Configure logging
+logging.basicConfig(level=logging.DEBUG, format='%(filename)s - %(asctime)s - %(levelname)s - %(message)s')
 
 # List of available analyzers
 analyzers = {'.c': CAnalyzer(),
@@ -41,11 +42,11 @@ def first_pass(self, ignore: List[str], executor: concurrent.futures.Executor) -
             if dirpath in ignore:
                 # in-place clear dirnames to prevent os.walk from recursing into
                 # any of the nested directories
-                logger.info(f'ignoring directory: {dirpath}')
+                logging.info(f'ignoring directory: {dirpath}')
                 dirnames[:] = []
                 continue
 
-            logger.info(f'Processing directory: {dirpath}')
+            logging.info(f'Processing directory: {dirpath}')
 
             # Process each file in the current directory
             for filename in filenames:
@@ -54,11 +55,10 @@ def first_pass(self, ignore: List[str], executor: concurrent.futures.Executor) -
                 # Skip none supported files
                 ext = file_path.suffix
                 if ext not in analyzers:
-                    logger.info(f"Skipping none supported file {file_path}")
+                    logging.info(f"Skipping none supported file {file_path}")
                     continue
 
-                logger.info(f'Processing file: {file_path}')
-                print(f'Processing file: {file_path}')
+                logging.info(f'Processing file: {file_path}')
 
                 def process_file(path: Path) -> None:
                     with open(path, 'rb') as f:
@@ -89,11 +89,11 @@ def second_pass(self, ignore: List[str], executor: concurrent.futures.Executor)
             if dirpath in ignore:
                 # in-place clear dirnames to prevent os.walk from recursing into
                 # any of the nested directories
-                logger.info(f'ignoring directory: {dirpath}')
+                logging.info(f'ignoring directory: {dirpath}')
                 dirnames[:] = []
                 continue
 
-            logger.info(f'Processing directory: {dirpath}')
+            logging.info(f'Processing directory: {dirpath}')
 
             # Process each file in the current directory
             for filename in filenames:
@@ -104,7 +104,7 @@ def second_pass(self, ignore: List[str], executor: concurrent.futures.Executor)
                 if ext not in analyzers:
                     continue
 
-                logger.info(f'Processing file: {file_path}')
+                logging.info(f'Processing file: {file_path}')
 
                 def process_file(path: Path) -> None:
                     with open(path, 'rb') as f:
@@ -119,8 +119,8 @@ def process_file(path: Path) -> None:
 
     def analyze_file(self, path: Path, graph: Graph) -> None:
         ext = path.suffix
-        print(f"analyze_file: path: {path}")
-        print(f"analyze_file: ext: {ext}")
+        logging.info(f"analyze_file: path: {path}")
+        logging.info(f"analyze_file: ext: {ext}")
         if ext not in analyzers:
             return
 
@@ -145,6 +145,9 @@ def analyze(self, path: str, g: Graph, ignore: Optional[List[str]] = []) -> None
             ignore (List(str)): List of paths to skip
         """
 
+        # Save original working directory for later restore
+        original_dir = Path.cwd()
+
         # change working directory to path
         os.chdir(path)
 
@@ -154,7 +157,10 @@ def analyze(self, path: str, g: Graph, ignore: Optional[List[str]] = []) -> None
         # Analyze source files
         self.analyze_sources(ignore)
 
-        logger.info("Done analyzing path")
+        logging.info("Done analyzing path")
+
+        # Restore original working dir
+        os.chdir(original_dir)
 
     def analyze_local_repository(self, path: str, ignore: Optional[List[str]] = []) -> Graph:
         """

code_graph/app.py

@@ -387,8 +387,6 @@ def repo_info():
             'info': stats
         }
 
-        print(f"response: {response}")
-
         return jsonify(response), 200
 
     @app.route('/find_paths', methods=['POST'])
@@ -479,6 +477,29 @@ def unreachable_entities():
 
         return jsonify(response), 200
 
+    @app.route('/chat', methods=['POST'])
+    def chat():
+        # Get JSON data from the request
+        data = request.get_json()
+
+        # Validate 'repo' parameter
+        repo = data.get('repo')
+        if repo is None:
+            return jsonify({'status': f'Missing mandatory parameter "repo"'}), 400
+
+        # Get optional 'label' and 'relation' parameters
+        msg = data.get('msg')
+        if msg is None:
+            return jsonify({'status': f'Missing mandatory parameter "msg"'}), 400
+
+        answer = ask(repo, msg)
+
+        # Create and return a successful response
+        response = { 'status': 'success', 'response': answer }
+
+        return jsonify(response), 200
+
+
     return app
 
 if __name__ == '__main__':

code_graph/git_utils/git_graph.py

@@ -1,7 +1,11 @@
 import os
+import logging
 from falkordb import FalkorDB, Node
 from typing import List, Optional
 
+# Configure logging
+logging.basicConfig(level=logging.DEBUG, format='%(filename)s - %(asctime)s - %(levelname)s - %(message)s')
+
 class GitGraph():
     """
     Represents a git commit graph
@@ -40,6 +44,9 @@ def add_commit(self, commit_hash: str, author: str, message: str, date: int) ->
         """
             Add a new commit to the graph
         """
+
+        logging.info(f"Adding commit {commit_hash}: {message}")
+
         q = "MERGE (c:Commit {hash: $hash, author: $author, message: $message, date: $date})"
         params = {'hash': commit_hash, 'author': author, 'message': message, 'date': date}
         self.g.query(q, params)
@@ -55,6 +62,8 @@ def list_commits(self) -> List[Node]:
         return [self._commit_from_node(row[0]) for row in result_set]
 
     def get_commits(self, hashes: List[str]) -> List[dict]:
+        logging.info(f"Searching for commits {hashes}")
+
         q = """MATCH (c:Commit)
                WHERE c.hash IN $hashes
                RETURN c"""
@@ -67,13 +76,17 @@ def get_commits(self, hashes: List[str]) -> List[dict]:
             commit = self._commit_from_node(row[0])
             commits.append(commit)
 
+        logging.info(f"retrived commits: {commits}")
         return commits
 
     def connect_commits(self, child: str, parent: str) -> None:
         """
             connect commits via both PARENT and CHILD edges
         """
 
+        logging.info(f"Connecting commits {child} -PARENT-> {parent}")
+        logging.info(f"Connecting commits {parent} -CHILD-> {child}")
+
         q = """MATCH (child :Commit {hash: $child_hash}), (parent :Commit {hash: $parent_hash})
                MERGE (child)-[:PARENT]->(parent)
                MERGE (parent)-[:CHILD]->(child)"""