Spaces:
Runtime error
Runtime error
| from AssistantService import GPTAssistant | |
| from openai.error import AuthenticationError | |
| import streamlit as st | |
| import configparser | |
| import os | |
| config = configparser.ConfigParser() | |
| config.read('config.ini') | |
| if 'DEFAULT' in config: | |
| assistant_api_key = config['DEFAULT'].get('API-KEY', '') | |
| os.environ["LANGCHAIN_TRACING_V2"]="true" | |
| os.environ["LANGCHAIN_ENDPOINT"]="https://api.smith.langchain.com" | |
| os.environ["LANGCHAIN_API_KEY"]=st.secrets["LANGCHAIN_API_KEY"] | |
| os.environ["LANGCHAIN_PROJECT"]=st.secrets["LANGCHAIN_PROJECT"] | |
| st.write("This app helps you to extract data from HTML code using web scraping. It uses GPT-3.5-turbo to generate the code for you. \n *Contribute to this project on [GitHub](https://github.com/CognitiveLabs/GPT-auto-webscraping)*") | |
| st.write("") | |
| with st.expander(label="Check out the video demo"): | |
| yt_video = st.video("https://www.youtube.com/watch?v=_zeCun4OlCc") | |
| info_text = """ | |
| **Quick start** \n | |
| Fill the input with the HTML code you want to extract data from | |
| Example below: | |
| """ | |
| st.write(info_text) | |
| st.image("https://j.gifs.com/gpqvPl.gif") | |
| if assistant_api_key == '': | |
| assistant_api_key = st.text_input("Paste your API key here:") | |
| if assistant_api_key: | |
| gpt_assistant = GPTAssistant(assistant_api_key) | |
| else: | |
| gpt_assistant = GPTAssistant(assistant_api_key) | |
| html_content = st.text_input("Paste your piece of HTML here:") | |
| extract_button = st.button("Extract data format") | |
| if html_content and extract_button: | |
| try: | |
| output = gpt_assistant.chain_response_format(html_content) | |
| st.session_state['output_format'] = output | |
| except NameError: | |
| st.write("Complete the API key field") | |
| except AuthenticationError: | |
| st.write("Invalid API key") | |
| if 'output_format' in st.session_state: | |
| output_format = st.code(st.session_state['output_format'], language="json") | |
| if st.button("Generate the code"): | |
| try: | |
| python_code = gpt_assistant.chain_code_generator(st.session_state['output_format'], html_content) | |
| st.session_state['code_generated'] = python_code | |
| st.session_state['code_generated_exec'] = python_code + "\nresult = extract_info(html_data)" | |
| except NameError: | |
| st.write("Complete the API key field") | |
| except AuthenticationError: | |
| st.write("Invalid API key") | |
| if 'code_generated' in st.session_state: | |
| python_function_label = st.write("Here is your python function:") | |
| code_generated = st.code(st.session_state['code_generated'],language="python") | |
| full_content = st.text_input("Paste your complete HTML here:") | |
| test_code = st.button("Test the code") | |
| if full_content and test_code: | |
| html_data = full_content | |
| result = None | |
| exec(st.session_state['code_generated_exec'], globals()) | |
| if result: | |
| st.write("data extracted successfully") | |
| # show data in table | |
| st.table(result) | |
| else: | |
| st.write("error extracting data") | |