fix message extraaction

Author: yrobla

src/codegate/pipeline/codegate_context_retriever/codegate.py

@@ -36,7 +36,7 @@ def generate_context_str(self, objects: list[object], context: PipelineContext)
         matched_packages = []
         for obj in objects:
             # The object is already a dictionary with 'properties'
-            package_obj = obj["properties"]
+            package_obj = obj["properties"]  # type: ignore
             matched_packages.append(f"{package_obj['name']} ({package_obj['type']})")
             # Add one alert for each package found
             context.add_alert(
@@ -97,7 +97,7 @@ async def process(
         user_messages = re.sub(r"⋮...*?⋮...\n\n", "", user_messages, flags=re.DOTALL)
 
         # split messages into double newlines, to avoid passing so many content in the search
-        split_messages = user_messages.split("\n\n")
+        split_messages = re.split(r'</?task>|(\n\n)', user_messages)
         collected_bad_packages = []
         for item_message in split_messages:
             # Vector search to find bad packages

src/codegate/pipeline/extract_snippets/extract_snippets.py

@@ -125,7 +125,8 @@ def extract_snippets(message: str) -> List[CodeSnippet]:
 
         #  just correct the typescript exception
         lang_map = {"typescript": "javascript"}
-        lang = lang_map.get(lang, lang)
+        if lang:
+            lang = lang_map.get(lang, lang)
         snippets.append(CodeSnippet(filepath=filename, code=content, language=lang))
 
     return snippets

src/codegate/storage/storage_engine.py

@@ -1,7 +1,7 @@
 import os
 import re
 import sqlite3
-from typing import List
+from typing import List, Optional
 
 import numpy as np
 import sqlite_vec_sl_tmp
@@ -51,9 +51,13 @@ def __init__(self, data_path="./sqlite_data"):
         )
 
         self.inference_engine = LlamaCppInferenceEngine()
-        self.model_path = (
-            f"{Config.get_config().model_base_path}/{Config.get_config().embedding_model}"
-        )
+        conf = Config.get_config()
+        if conf and conf.model_base_path and conf.embedding_model:
+            self.model_path = (
+                f"{conf.model_base_path}/{conf.embedding_model}"
+            )
+        else:
+            self.model_path = ""
 
         self.conn = self._get_connection()
         self._setup_schema()
@@ -131,10 +135,10 @@ async def search_by_property(self, name: str, properties: List[str]) -> list[dic
 
     async def search(
         self,
-        query: str = None,
-        language: str = None,
-        ecosystem: str = None,
-        packages: List[str] = None,
+        query: Optional[str] = None,
+        language: Optional[str] = None,
+        ecosystem: Optional[str] = None,
+        packages: Optional[List[str]] = None,
         limit: int = 50,
         distance: float = 0.3,
     ) -> list[object]: