feat: now use batches but seem to write less rows, not sure

new_write.py

@@ -26,6 +26,10 @@ json_files = glob.glob('./sec_data/companyfacts/*.json')
 file_count = len(json_files)
 current_file = 1
 
+# Batch size configuration - process `batch_size` files at a time
+batch_size = 50  # Adjust this number to your preference or based on system resources
+rows = []
+
 # Data type mapping for the DataFrame to SQL conversion
 dtype_map = {
     'entity_cik': Integer,
@@ -46,69 +50,71 @@ dtype_map = {
     'frame': String(255)
 }
 
-for json_file in json_files:
+# Iterate through the JSON files in batches
-    try:
+for i in range(0, file_count, batch_size):
-        # Load the JSON data
+    batch_files = json_files[i:i+batch_size]
-        with open(json_file) as f:
-            data = json.load(f)
     
-        # Informing the user about the current file being processed
+    for json_file in batch_files:
-        print(f"Processing file {current_file}/{file_count}: {json_file}")
+        try:
-        current_file += 1
+            # Load the JSON data
+            with open(json_file) as f:
+                data = json.load(f)
 
-        # Check if the JSON has the keys we're interested in
+            # Informing the user about the current file being processed
-        cik = data.get('cik', None)
+            print(f"Processing file {current_file}/{file_count}: {json_file}")
-        entity_name = data.get('entityName', None)
+            current_file += 1
-        facts = data.get('facts', {})
 
-        # Initialize a list to hold rows
+            # Check if the JSON has the keys we're interested in
-        rows = []
+            cik = data.get('cik', None)
+            entity_name = data.get('entityName', None)
+            facts = data.get('facts', {})
 
-        # Skip files that don't have facts to process
+            # Skip files that don't have facts to process
-        if not facts:
+            if not facts:
-            print(f"File {json_file} has no facts to process. Skipping...")
+                print(f"File {json_file} has no facts to process. Skipping...")
-            continue
+                continue
 
-        # Process the facts dynamically
+            # Process the facts dynamically
-        for taxonomy, fact_items in facts.items(): 
+            for taxonomy, fact_items in facts.items(): 
-            for fact_id, fact_data in fact_items.items():
+                for fact_id, fact_data in fact_items.items():
-                label = fact_data.get('label', None)
+                    label = fact_data.get('label', None)
-                description = fact_data.get('description', None)
+                    description = fact_data.get('description', None)
-                units = fact_data.get('units', {})
+                    units = fact_data.get('units', {})
 
-                for unit, details_list in units.items():
+                    for unit, details_list in units.items():
-                    for details in details_list:
+                        for details in details_list:
-                        # Generate row dictionary dynamically, only updating non-None values
+                            # Generate row dictionary dynamically, only updating non-None values
-                        row = {
+                            row = {
-                            'entity_cik': cik,
+                                'entity_cik': cik,
-                            'entity_name': entity_name,
+                                'entity_name': entity_name,
-                            'fact_id': fact_id,
+                                'fact_id': fact_id,
-                            'fact_taxonomy': taxonomy,
+                                'fact_taxonomy': taxonomy,
-                            'fact_label': label,
+                                'fact_label': label,
-                            'fact_description': description,
+                                'fact_description': description,
-                            'fact_unit': unit
+                                'fact_unit': unit
-                        }
+                            }
 
-                        # Add the remaining keys in details to row
+                            # Add the remaining keys in details to row
-                        for key, value in details.items():
+                            for key, value in details.items():
-                            row[key] = value
+                                row[key] = value
                             
-                        # Append the row to rows list
+                            # Append the row to rows list
-                        rows.append(row)
+                            rows.append(row)
 
-        # Create DataFrame only if there are rows
+        except json.JSONDecodeError as e:
-        if rows:
+            print(f"Failed to process file {json_file}: Invalid JSON format. Error: {e}")
-            df = pd.DataFrame(rows)
 
-            # Write DataFrame to the 'data' table, appending if table exists
+        except Exception as e:
-            df.to_sql('data', con=engine, if_exists='append', index=False, dtype=dtype_map)
+            print(f"An error occurred while processing file {json_file}: {e}")
-        else:
-            print(f"No data rows were created for file {json_file}. Skipping insert...")
 
-    except json.JSONDecodeError as e:
+    # After processing batch_files, insert accumulated rows into the database
-        print(f"Failed to process file {json_file}: Invalid JSON format. Error: {e}")
+    if rows:
+        df = pd.DataFrame(rows)
 
-    except Exception as e:
+        # Write DataFrame to the 'data' table, appending if table exists
-        print(f"An error occurred while processing file {json_file}: {e}")
+        df.to_sql('data', con=engine, if_exists='append', index=False, dtype=dtype_map)
+
+        # Clear the list of rows for the next batch
+        rows.clear()
 
 print("Processing complete. All files have been handled.")