fix(tutor): creates summariy in the required language (#104)

Author: sandragjacinto

src/app/api/api_v1/endpoints/tutor.py

@@ -58,14 +58,15 @@ async def extract_files_content(
     files: Annotated[list[UploadFile], File()],
     response: Response,
     chatfactory=Depends(get_chat_service),
+    lang: str = "en",
 ) -> ExtractorOutputList | None:
     files_content = await get_files_content(files)
     files_content_str = ("__DOCUMENT_SEPARATOR__").join(files_content)
 
     messages = [
         {
             "role": "system",
-            "content": extractor_system_prompt,
+            "content": extractor_system_prompt + lang,
         },
         {
             "role": "user",

src/app/services/tutor/prompts.py

@@ -3,6 +3,7 @@
 backstory="You are specialised in analysing documents, summarizing them and extracting the main themes. You value precision and clarity.",
 goal="Analyse each document, summarize it and extract the main themes, explaining why each theme was identified.",
 expected_output="You must follow the following JSON schema: {extracts: [{'original_document': 'Document', 'summary': 'Summary', 'themes': [{'theme': 'Theme 1', 'reason': 'Reason for Theme 1'}, {'theme': 'Theme 2', 'reason': 'Reason for Theme 2'}, ...]}, {'original_document': 'Document', 'summary': 'Sumamry', 'themes': [{'theme': 'Theme 1', 'reason': 'Reason for Theme 1'}, {'theme': 'Theme 2', 'reason': 'Reason for Theme 2'}, ...]}] } an entry by document make sure to use double quotes when formating the JSON",
+The extracts must be written in lang:
 """
 
 extractor_user_prompt = """