support product can be str or list

app/schemas/requests.py

@@ -1,4 +1,4 @@
-from typing import Any, Dict, List, Optional
+from typing import Any, Dict, List, Optional, Union
 
 from pydantic import BaseModel
 
@@ -18,7 +18,7 @@ class ExtractionRequest(BaseModel):
     img_urls: Optional[List[str]] = None
     product_taxonomy: str
     request_meta: Optional[Dict[str, str]] = None
-    product_data: Optional[Dict[str, str]] = None
+    product_data: Optional[Dict[str, Union[str, List]]] = None
     ai_model: str = settings.DEFAULT_MODEL  # type: ignore
     max_attempts: int = settings.DEFAULT_MAX_ATTEMPTS  # type: ignore
 

app/services/base.py

@@ -1,5 +1,5 @@
 from abc import ABC, abstractmethod
-from typing import Any, Dict, List, Type
+from typing import Any, Dict, List, Type, Union
 
 from pydantic import BaseModel
 
@@ -34,7 +34,7 @@ class BaseAttributionService(ABC):
         ai_model: str,
         img_urls: List[str],
         product_taxonomy: str,
-        product_data: Dict[str, str],
+        product_data: Dict[str, Union[str, List]],
         pil_images: List[Any] = None,
         img_paths: List[str] = None,
     ) -> Dict[str, Any]:

app/services/service_anthropic.py

@@ -1,6 +1,6 @@
 import json
 import os
-from typing import Any, Dict, List, Type
+from typing import Any, Dict, List, Type, Union
 
 import anthropic
 import weave
@@ -41,7 +41,7 @@ class AnthropicService(BaseAttributionService):
         ai_model: str,
         img_urls: List[str],
         product_taxonomy: str,
-        product_data: Dict[str, str],
+        product_data: Dict[str, Union[str, List]],
         pil_images: List[Any] = None,  # do not remove, this is for weave
         img_paths: List[str] = None,
     ) -> Dict[str, Any]:

app/services/service_openai.py

@@ -1,6 +1,6 @@
 import json
 import os
-from typing import Any, Dict, List, Type
+from typing import Any, Dict, List, Type, Union
 
 import openai
 import weave
@@ -58,7 +58,7 @@ class OpenAIService(BaseAttributionService):
         ai_model: str,
         img_urls: List[str],
         product_taxonomy: str,
-        product_data: Dict[str, str],
+        product_data: Dict[str, Union[str, List]],
         pil_images: List[Any] = None,  # do not remove, this is for weave
         img_paths: List[str] = None,
     ) -> Dict[str, Any]:

app/utils/converter.py

@@ -11,4 +11,29 @@ def product_data_to_str(product_data: dict[str, any]) -> str:
     if product_data is None:
         return ""
 
-    return "\n".join([f"{k}: {v}" for k, v in product_data.items()])
+    data_list = []
+    for k, v in product_data.items():
+        data_line = None
+        if isinstance(v, list):
+            delimiter = ","
+            for sub_v in v:
+                if "," in sub_v:
+                    delimiter = (
+                        ";"  # Use semicolon as delimiter if comma is found in the value
+                    )
+
+            for i, sub_v in enumerate(v):
+                if ";" in sub_v:
+                    v[i] = f'"{sub_v}"'
+
+            data_line = f"{k}: {f'{delimiter} '.join(v)}"
+        elif isinstance(v, str):
+            data_line = f"{k}: {v}"
+        else:
+            raise ValueError(
+                f"Unsupported data type for value of product data: {type(v)}. Only list and str are supported."
+            )
+
+        data_list.append(data_line)
+
+    return "\n".join(data_list)

clean_for_gradio.sh

@@ -1,4 +1,4 @@
 rm -rf app/api
 rm -rf app/aws
 rm app/main.py
-rm worker.py
+rm app/worker.py