Update gradio_app.py

gradio_app.py

@@ -60,6 +60,33 @@ def toggle_feedback(request_data, response_data, like_clicked, dislike_clicked):
     else:
         return "Feedback sent successfully!"
 
+# Define the function to preprocess and flatten a specified field in the JSON results
+def preprocess_and_flatten(json_results, nested_field_name, meta_fields=None):
+    # Ensure 'meta_fields' is a list or set default fields
+    if meta_fields is None:
+        meta_fields = ['doc_id', 'details', 'domain']
+
+    # Preprocess to convert the specified nested field to a list
+    for result in json_results:
+        nested_field = result.get(nested_field_name, None)  # Handle missing field
+        if nested_field is None:
+            result[nested_field_name] = []  # Default to empty list if the field is missing
+        elif isinstance(nested_field, dict):
+            result[nested_field_name] = list(nested_field.values())
+        elif not isinstance(nested_field, list):
+            result[nested_field_name] = []  # Default to empty list if it's not a list or dict
+
+    # Flatten the JSON for the specified nested field
+    dataframe_results = pd.json_normalize(
+        json_results,
+        record_path=[nested_field_name],                # Nested array to flatten
+        meta=meta_fields,                               # Top-level fields to retain
+        sep='_',                                        # Separator for nested fields
+        errors='ignore'                                 # To handle cases with missing keys
+    )
+    
+    return dataframe_results
+    
 # Define the functions to handle the inputs and outputs
 def news_analysis(text):
     try:
@@ -77,23 +104,9 @@ def news_analysis(text):
         # Prepare results for JSON output
         json_results = response.json()
 
-        # Preprocess to convert 'claim_objects' from any type to a list
-        for result in json_results:
-            # Ensure 'claim_objects' is a list, or if it's a dict, convert it to a list of its values, else leave it as an empty list
-            claim_objects = result.get('claim_objects')
-            if isinstance(claim_objects, dict):
-                result['claim_objects'] = list(claim_objects.values())
-            elif not isinstance(claim_objects, list):
-                result['claim_objects'] = []  # Default to empty list if it's not a list or dict
-        
-        # Flatten the JSON for 'claim_objects'
-        dataframe_results = pd.json_normalize(
-            json_results,
-            record_path=['claim_objects'],                  # Nested array to flatten
-            meta=['doc_id', 'details', 'domain'],           # Top-level fields to retain
-            sep='_',                                        # Separator for nested fields
-            errors='ignore'                                 # To handle cases with missing keys
-        )
+        # Flatten 'claim_objects' field
+        dataframe_results = preprocess_and_flatten(json_results, 'claim_objects', ['doc_id', 'details', 'domain'])
+
         return json_results, dataframe_results
     except Exception as e:
         results = {'error': str(e)}
@@ -115,37 +128,18 @@ def claim_verification(text):
         # Prepare results for JSON output
         json_results = response.json()
 
-        # Preprocess to convert nested fields from any type to lists
-        for result in json_results:
-            # Ensure 'support' is a list, or if it's a dict, convert it to a list of its values, else leave it as an empty list
-            support = result.get('support')
-            if isinstance(support, dict):
-                result['support'] = list(support.values())
-            elif not isinstance(support, list):
-                result['support'] = []  # Default to empty list if it's not a list or dict
-            
-            # Ensure 'refute' is a list, or if it's a dict, convert it to a list of its values, else leave it as an empty list
-            refute = result.get('refute')
-            if isinstance(refute, dict):
-                result['refute'] = list(refute.values())
-            elif not isinstance(refute, list):
-                result['refute'] = []  # Default to empty list if it's not a list or dict
-            
-            # Ensure 'no_info' is a list, or if it's a dict, convert it to a list of its values, else leave it as an empty list
-            no_info = result.get('no_info')
-            if isinstance(no_info, dict):
-                result['no_info'] = list(no_info.values())
-            elif not isinstance(no_info, list):
-                result['no_info'] = []  # Default to empty list if it's not a list or dict
+        # Flatten 'support' field
+        support_results = preprocess_and_flatten(json_results, 'support', ['doc_id', 'details'])
         
-        # Flatten the JSON
-        dataframe_results = pd.json_normalize(
-            json_results,
-            record_path=['support', 'refute', 'no_info'],   # All possible nested lists
-            meta=['doc_id', 'details'],                     # Top-level fields to retain
-            sep='_',                                        # Separator for nested fields
-            errors='ignore'                                 # To handle cases with missing keys
-        )
+        # Flatten 'refute' field
+        refute_results = preprocess_and_flatten(json_results, 'refute', ['doc_id', 'details'])
+        
+        # Flatten 'no_info' field
+        no_info_results = preprocess_and_flatten(json_results, 'no_info', ['doc_id', 'details'])
+        
+        # concatenate the results if needed
+        dataframe_results = pd.concat([support_results, refute_results, no_info_results], ignore_index=True)
+
         return json_results, dataframe_results
     except Exception as e:
         results = {'error': str(e)}