Merge pull request #107 from FalkorDB/csharp

Add C# language support

Author: gkorland

README.md

@@ -56,7 +56,7 @@ a graph representation of your source code, the graph name should be the same as
 the name of the folder you've requested to analyze, for the example above a graph named:
 "GraphRAG-SDK".
 
-At the moment only the Python and C languages are supported, we do intend to support additional languages.
+At the moment Python, Java, and C# languages are supported, we do intend to support additional languages.
 
 At this point you can explore and query your source code using various tools
 Here are several options:

api/analyzers/analyzer.py

@@ -127,7 +127,7 @@ def add_symbols(self, entity: Entity) -> None:
         pass
 
     @abstractmethod
-    def resolve_symbol(self, files: dict[Path, File], lsp: SyncLanguageServer, file_path: Path, path: Path, key: str, symbol: Node) -> Entity:
+    def resolve_symbol(self, files: dict[Path, File], lsp: SyncLanguageServer, file_path: Path, path: Path, key: str, symbol: Node) -> list[Entity]:
         """
         Resolve a symbol to an entity.
 
@@ -138,7 +138,7 @@ def resolve_symbol(self, files: dict[Path, File], lsp: SyncLanguageServer, file_
             symbol (Node): The symbol node.
 
         Returns:
-            Entity: The entity.
+            list[Entity]: The resolved entities.
         """
 
         pass

api/analyzers/csharp/__init__.py

@@ -0,0 +1,144 @@
+import subprocess
+from pathlib import Path
+
+from multilspy import SyncLanguageServer
+from ...entities.entity import Entity
+from ...entities.file import File
+from typing import Optional
+from ..analyzer import AbstractAnalyzer
+
+import tree_sitter_c_sharp as tscsharp
+from tree_sitter import Language, Node, QueryCursor
+
+import logging
+logger = logging.getLogger('code_graph')
+
+class CSharpAnalyzer(AbstractAnalyzer):
+    def __init__(self) -> None:
+        super().__init__(Language(tscsharp.language()))
+
+    def _captures(self, pattern: str, node: Node) -> dict:
+        """Run a tree-sitter query and return captures dict."""
+        query = self.language.query(pattern)
+        cursor = QueryCursor(query)
+        return cursor.captures(node)
+
+    def add_dependencies(self, path: Path, files: list[Path]):
+        if Path(f"{path}/temp_deps_cs").is_dir():
+            return
+        if any(Path(f"{path}").glob("*.csproj")) or any(Path(f"{path}").glob("*.sln")):
+            subprocess.run(["dotnet", "restore"], cwd=str(path))
+
+    def get_entity_label(self, node: Node) -> str:
+        if node.type == 'class_declaration':
+            return "Class"
+        elif node.type == 'interface_declaration':
+            return "Interface"
+        elif node.type == 'enum_declaration':
+            return "Enum"
+        elif node.type == 'struct_declaration':
+            return "Struct"
+        elif node.type == 'method_declaration':
+            return "Method"
+        elif node.type == 'constructor_declaration':
+            return "Constructor"
+        raise ValueError(f"Unknown entity type: {node.type}")
+
+    def get_entity_name(self, node: Node) -> str:
+        if node.type in ['class_declaration', 'interface_declaration', 'enum_declaration',
+                         'struct_declaration', 'method_declaration', 'constructor_declaration']:
+            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]:
+        if node.type in ['class_declaration', 'interface_declaration', 'enum_declaration',
+                         'struct_declaration', 'method_declaration', 'constructor_declaration']:
+            # Walk back through contiguous comment siblings to collect
+            # multi-line XML doc comments (each /// line is a separate node)
+            lines = []
+            sibling = node.prev_sibling
+            while sibling and sibling.type == "comment":
+                lines.insert(0, sibling.text.decode('utf-8'))
+                sibling = sibling.prev_sibling
+            return '\n'.join(lines) if lines else None
+        raise ValueError(f"Unknown entity type: {node.type}")
+
+    def get_entity_types(self) -> list[str]:
+        return ['class_declaration', 'interface_declaration', 'enum_declaration',
+                'struct_declaration', 'method_declaration', 'constructor_declaration']
+
+    def add_symbols(self, entity: Entity) -> None:
+        if entity.node.type in ['class_declaration', 'struct_declaration']:
+            base_list_captures = self._captures("(base_list (_) @base_type)", entity.node)
+            if 'base_type' in base_list_captures:
+                first = True
+                for base_type in base_list_captures['base_type']:
+                    if first and entity.node.type == 'class_declaration':
+                        # NOTE: Without semantic analysis, we cannot distinguish a base
+                        # class from an interface in C# base_list. By convention, the
+                        # base class is listed first; if a class only implements
+                        # interfaces, this will produce a spurious base_class edge that
+                        # the LSP resolution in second_pass can correct.
+                        entity.add_symbol("base_class", base_type)
+                        first = False
+                    else:
+                        entity.add_symbol("implement_interface", base_type)
+        elif entity.node.type == 'interface_declaration':
+            base_list_captures = self._captures("(base_list (_) @base_type)", entity.node)
+            if 'base_type' in base_list_captures:
+                for base_type in base_list_captures['base_type']:
+                    entity.add_symbol("extend_interface", base_type)
+        elif entity.node.type in ['method_declaration', 'constructor_declaration']:
+            captures = self._captures("(invocation_expression) @reference.call", entity.node)
+            if 'reference.call' in captures:
+                for caller in captures['reference.call']:
+                    entity.add_symbol("call", caller)
+            captures = self._captures("(parameter_list (parameter type: (_) @parameter))", entity.node)
+            if 'parameter' in captures:
+                for parameter in captures['parameter']:
+                    entity.add_symbol("parameters", parameter)
+            if entity.node.type == 'method_declaration':
+                return_type = entity.node.child_by_field_name('type')
+                if return_type:
+                    entity.add_symbol("return_type", return_type)
+
+    def is_dependency(self, file_path: str) -> bool:
+        return "temp_deps_cs" in file_path
+
+    def resolve_path(self, file_path: str, path: Path) -> str:
+        return file_path
+
+    def resolve_type(self, files: dict[Path, File], lsp: SyncLanguageServer, file_path: Path, path: Path, node: Node) -> list[Entity]:
+        res = []
+        for file, resolved_node in self.resolve(files, lsp, file_path, path, node):
+            type_dec = self.find_parent(resolved_node, ['class_declaration', 'interface_declaration', 'enum_declaration', 'struct_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]:
+        res = []
+        if node.type == 'invocation_expression':
+            func_node = node.child_by_field_name('function')
+            if func_node and func_node.type == 'member_access_expression':
+                func_node = func_node.child_by_field_name('name')
+            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, ['method_declaration', 'constructor_declaration', 'class_declaration', 'interface_declaration', 'enum_declaration', 'struct_declaration'])
+            if method_dec and method_dec.type in ['class_declaration', 'interface_declaration', 'enum_declaration', 'struct_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]:
+        if key in ["implement_interface", "base_class", "extend_interface", "parameters", "return_type"]:
+            return self.resolve_type(files, lsp, file_path, path, symbol)
+        elif key in ["call"]:
+            return self.resolve_method(files, lsp, file_path, path, symbol)
+        else:
+            raise ValueError(f"Unknown key {key}")

api/analyzers/csharp/analyzer.py

@@ -125,7 +125,7 @@ def resolve_method(self, files: dict[Path, File], lsp: SyncLanguageServer, file_
                 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) -> Entity:
+    def resolve_symbol(self, files: dict[Path, File], lsp: SyncLanguageServer, file_path: Path, path: Path, key: str, symbol: Node) -> list[Entity]:
         if key in ["implement_interface", "base_class", "extend_interface", "parameters", "return_type"]:
             return self.resolve_type(files, lsp, file_path, path, symbol)
         elif key in ["call"]:

api/analyzers/java/analyzer.py

@@ -114,7 +114,7 @@ def resolve_method(self, files: dict[Path, File], lsp: SyncLanguageServer, file_
                 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) -> Entity:
+    def resolve_symbol(self, files: dict[Path, File], lsp: SyncLanguageServer, file_path: Path, path: Path, key: str, symbol: Node) -> list[Entity]:
         if key in ["base_class", "parameters", "return_type"]:
             return self.resolve_type(files, lsp, file_path, path, symbol)
         elif key in ["call"]:

api/analyzers/python/analyzer.py

@@ -10,6 +10,7 @@
 # from .c.analyzer import CAnalyzer
 from .java.analyzer import JavaAnalyzer
 from .python.analyzer import PythonAnalyzer
+from .csharp.analyzer import CSharpAnalyzer
 
 from multilspy import SyncLanguageServer
 from multilspy.multilspy_config import MultilspyConfig
@@ -24,7 +25,8 @@
     # '.c': CAnalyzer(),
     # '.h': CAnalyzer(),
     '.py': PythonAnalyzer(),
-    '.java': JavaAnalyzer()}
+    '.java': JavaAnalyzer(),
+    '.cs': CSharpAnalyzer()}
 
 class NullLanguageServer:
     def start_server(self):
@@ -136,7 +138,12 @@ def second_pass(self, graph: Graph, files: list[Path], path: Path) -> None:
             lsps[".py"] = SyncLanguageServer.create(config, logger, str(path))
         else:
             lsps[".py"] = NullLanguageServer()
-        with lsps[".java"].start_server(), lsps[".py"].start_server():
+        if any(path.rglob('*.cs')):
+            config = MultilspyConfig.from_dict({"code_language": "csharp"})
+            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():
             files_len = len(self.files)
             for i, file_path in enumerate(files):
                 file = self.files[file_path]
@@ -166,7 +173,8 @@ def analyze_files(self, files: list[Path], path: Path, graph: Graph) -> None:
         self.second_pass(graph, files, path)
 
     def analyze_sources(self, path: Path, ignore: list[str], graph: Graph) -> None:
-        files = list(path.rglob("*.java")) + list(path.rglob("*.py"))
+        path = path.resolve()
+        files = list(path.rglob("*.java")) + list(path.rglob("*.py")) + list(path.rglob("*.cs"))
         # 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-c-sharp>=0.23.1,<0.24.0",
     "flask>=3.1.0,<4.0.0",
     "python-dotenv>=1.0.1,<2.0.0",
     "multilspy @ git+https://github.com/AviAvni/multilspy.git@python-init-params",

pyproject.toml

@@ -0,0 +1,48 @@
+using System;
+
+namespace TestProject
+{
+    public interface ILogger
+    {
+        void Log(string message);
+    }
+
+    public class ConsoleLogger : ILogger
+    {
+        public void Log(string message)
+        {
+            Console.WriteLine(message);
+        }
+    }
+
+    /// <summary>
+    /// Represents a task to be executed.
+    /// </summary>
+    public class Task
+    {
+        public string Name { get; set; }
+        public int Duration { get; set; }
+
+        private ILogger _logger;
+
+        public Task(string name, int duration, ILogger logger)
+        {
+            Name = name;
+            Duration = duration;
+            _logger = logger;
+            _logger.Log("Task created: " + name);
+        }
+
+        public bool Execute()
+        {
+            _logger.Log("Executing: " + Name);
+            return true;
+        }
+
+        public void Abort(float delay)
+        {
+            _logger.Log("Aborting: " + Name);
+            Execute();
+        }
+    }
+}

tests/source_files/csharp/Program.cs

@@ -0,0 +1,69 @@
+import os
+import unittest
+
+from api import SourceAnalyzer, Graph
+
+
+class Test_CSharp_Analyzer(unittest.TestCase):
+    def setUp(self):
+        self.g = Graph("csharp")
+
+    def tearDown(self):
+        self.g.delete()
+
+    def test_analyzer(self):
+        analyzer = SourceAnalyzer()
+
+        # Get the current file path
+        current_file_path = os.path.abspath(__file__)
+
+        # Get the directory of the current file
+        current_dir = os.path.dirname(current_file_path)
+
+        # Append 'source_files/csharp' to the current directory
+        path = os.path.join(current_dir, 'source_files')
+        path = os.path.join(path, 'csharp')
+        path = str(path)
+
+        analyzer.analyze_local_folder(path, self.g)
+
+        # Verify ILogger interface was detected
+        q = "MATCH (n:Interface {name: 'ILogger'}) RETURN n LIMIT 1"
+        res = self.g._query(q).result_set
+        self.assertEqual(len(res), 1)
+
+        # Verify ConsoleLogger class was detected
+        q = "MATCH (n:Class {name: 'ConsoleLogger'}) RETURN n LIMIT 1"
+        res = self.g._query(q).result_set
+        self.assertEqual(len(res), 1)
+
+        # Verify Task class was detected
+        q = "MATCH (n:Class {name: 'Task'}) RETURN n LIMIT 1"
+        res = self.g._query(q).result_set
+        self.assertEqual(len(res), 1)
+
+        # Verify methods were detected
+        for method_name in ['Log', 'Execute', 'Abort']:
+            q = "MATCH (n {name: $name}) RETURN n LIMIT 1"
+            res = self.g._query(q, {'name': method_name}).result_set
+            self.assertGreaterEqual(len(res), 1, f"Method {method_name} not found")
+
+        # Verify Constructor was detected
+        q = "MATCH (n:Constructor {name: 'Task'}) RETURN n LIMIT 1"
+        res = self.g._query(q).result_set
+        self.assertEqual(len(res), 1)
+
+        # Verify DEFINES relationships exist (File -> Class/Interface)
+        q = "MATCH (f:File)-[:DEFINES]->(n) RETURN count(n)"
+        res = self.g._query(q).result_set
+        self.assertGreater(res[0][0], 0)
+
+        # Verify class defines methods
+        q = "MATCH (c:Class {name: 'Task'})-[:DEFINES]->(m) RETURN count(m)"
+        res = self.g._query(q).result_set
+        self.assertGreater(res[0][0], 0)
+
+        # Verify ConsoleLogger implements ILogger
+        q = "MATCH (c:Class {name: 'ConsoleLogger'})-[:IMPLEMENTS]->(i:Interface {name: 'ILogger'}) RETURN c, i LIMIT 1"
+        res = self.g._query(q).result_set
+        self.assertEqual(len(res), 1)