Fix chat message parsing in Cline

src/codegate/extract_snippets/body_extractor.py

@@ -70,6 +70,25 @@ class ClineBodySnippetExtractor(BodyCodeSnippetExtractor):
     def __init__(self):
         self._snippet_extractor = ClineCodeSnippetExtractor()
 
+    def _extract_from_user_messages(self, data: dict) -> set[str]:
+        """
+        The method extracts the code snippets from the user messages in the data got from Cline.
+
+        It returns a set of filenames extracted from the code snippets.
+        """
+
+        filenames: List[str] = []
+        for msg in data.get("messages", []):
+            if msg.get("role", "") == "user":
+                msgs_content = msg.get("content", [])
+                for msg_content in msgs_content:
+                    if msg_content.get("type", "") == "text":
+                        extracted_snippets = self._snippet_extractor.extract_unique_snippets(
+                            msg_content.get("text")
+                        )
+                        filenames.extend(extracted_snippets.keys())
+        return set(filenames)
+
     def extract_unique_filenames(self, data: dict) -> set[str]:
         return self._extract_from_user_messages(data)
 

src/codegate/muxing/router.py

@@ -52,7 +52,7 @@ async def _get_model_route(
             return model_route
         except Exception as e:
             logger.error(f"Error getting active workspace muxes: {e}")
-            raise HTTPException(str(e), status_code=404)
+            raise HTTPException(detail=str(e), status_code=404)
 
     def _setup_routes(self):
 
@@ -85,7 +85,7 @@ async def route_to_dest_provider(
             model_route = await self._get_model_route(thing_to_match)
             if not model_route:
                 raise HTTPException(
-                    "No matching rule found for the active workspace", status_code=404
+                    detail="No matching rule found for the active workspace", status_code=404
                 )
 
             logger.info(

tests/extract_snippets/test_body_extractor.py

@@ -4,6 +4,7 @@
 
 from codegate.extract_snippets.body_extractor import (
     ClineBodySnippetExtractor,
+    ContinueBodySnippetExtractor,
     OpenInterpreterBodySnippetExtractor,
 )
 
@@ -75,9 +76,13 @@ def test_body_extract_openinterpreter_snippets(test_case: BodyCodeSnippetTest):
         BodyCodeSnippetTest(
             input_body_dict={
                 "messages": [
+                    {"role": "system", "content": "You are Cline, a highly skilled software"},
                     {
                         "role": "user",
-                        "content": '''
+                        "content": [
+                            {
+                                "type": "text",
+                                "text": '''
         [<task>
 now please analyze the folder 'codegate/src/codegate/api/' (see below for folder content)
 </task>
@@ -145,6 +150,8 @@ async def _process_prompt_output_to_partial_qa(
 </file_content>
 </folder_content>
         ''',
+                            }
+                        ],
                     },
                 ]
             },
@@ -157,3 +164,52 @@ def test_body_extract_cline_snippets(test_case: BodyCodeSnippetTest):
     extractor = ClineBodySnippetExtractor()
     filenames = extractor.extract_unique_filenames(test_case.input_body_dict)
     _evaluate_actual_filenames(filenames, test_case)
+
+
+@pytest.mark.parametrize(
+    "test_case",
+    [
+        # Analyze processed snippets from OpenInterpreter
+        BodyCodeSnippetTest(
+            input_body_dict={
+                "messages": [
+                    {
+                        "role": "user",
+                        "content": """
+                        ```file:///Users/user/StacklokRepos/testing_file.py
+import invokehttp
+import fastapi
+from fastapi import FastAPI, Request, Response, HTTPException
+import numpy
+
+GITHUB_TOKEN="ghp_1J9Z3Z2dfg4dfs23dsfsdf232aadfasdfasfasdf32"
+
+def add(a, b):
+        return a + b
+
+def multiply(a, b):
+        return a * b
+
+
+
+def substract(a, b):
+
+```
+
+please analyze testing_file.py
+                        """,
+                    }
+                ],
+                "model": "foo-model-replaced-by-mux",
+                "max_tokens": 4096,
+                "stream": True,
+            },
+            expected_count=1,
+            expected=["testing_file.py"],
+        ),
+    ],
+)
+def test_body_extract_continue_snippets(test_case: BodyCodeSnippetTest):
+    extractor = ContinueBodySnippetExtractor()
+    filenames = extractor.extract_unique_filenames(test_case.input_body_dict)
+    _evaluate_actual_filenames(filenames, test_case)