[NOTICKET]: add doctypes endpoint & 10MB file size limit

src/api/v1/document.py

@@ -24,6 +24,22 @@ class DocumentResponse(BaseModel):
     created_at: str
  
  
+@router.get("/documents/doctypes")
+@log_execution(logger)
+async def get_document_types():
+    """List supported document types with max file size and status."""
+    return {
+        "status": "success",
+        "data": [
+            {"doc_type": "pdf", "max_size": 10, "status": "active", "message": None},
+            {"doc_type": "docx", "max_size": 10, "status": "active", "message": None},
+            {"doc_type": "txt", "max_size": 10, "status": "active", "message": None},
+            {"doc_type": "csv", "max_size": 10, "status": "active", "message": None},
+            {"doc_type": "xlsx", "max_size": 10, "status": "active", "message": None},
+        ]
+    }
+
+
 @router.get("/documents/{user_id}", response_model=List[DocumentResponse])
 @log_execution(logger)
 async def list_documents(

src/pipeline/document_pipeline/document_pipeline.py

@@ -21,6 +21,13 @@ class DocumentPipeline:
         content = await file.read()
         file_type = file.filename.split(".")[-1].lower() if "." in file.filename else "txt"
 
+        MAX_FILE_SIZE_BYTES = 10 * 1024 * 1024  # 10 MB
+        if len(content) > MAX_FILE_SIZE_BYTES:
+            raise HTTPException(
+                status_code=400,
+                detail="File size exceeds maximum allowed size of 10 MB.",
+            )
+
         if file_type not in SUPPORTED_FILE_TYPES:
             raise HTTPException(
                 status_code=400,