INNER CODE UNIT · Python

load_dataframe_to_bamboo_ai_instance

pgalko/BambooAI · web_app/app.py:492

def load_dataframe_to_bamboo_ai_instance(session_id, df=None, file=None, execution_mode='local'):
    new_df_id = generate_dataframe_id()
    prefs = user_preferences.get(session_id, {'planning': False, 'auxiliary_datasets': []})
    prefs['df_id'] = new_df_id
    user_preferences[session_id] = prefs

    if execution_mode == 'api':
        if not file:
            raise ValueError("File is required for API execution mode")
        try:
            # the stream, not the FileStorage: newer werkzeug (3.1+) no longer proxies read() through the
            # FileStorage and requests then cannot serialise it (found on the Mac, 2026-09-15); the stream
            # is what both versions hand to requests unchanged
            file.stream.seek(0)
            files = {'file': (file.filename, file.stream, file.content_type)}
            executor_response = requests.post(
                get_dynamic_executor_urls(get_user_id())['EXECUTOR_API_UPLOAD_URL'],
                files=files,

View source record →

📰 Research Paper
Loading…
⏳ Fetching content…