如何抓取 ChatGPT：输入提示，获取回复

				
					from camoufox.sync_api import Camoufox
from markdownify import markdownify
import logging

query = "What are the top three dog breeds?"

output_file = "output_md.txt"

proxy_server = "XXX"
proxy_uname = "YYY"
proxy_pass = "ZZZ"

timeout = 60_000 

url = "https://chatgpt.com/"

selector_textarea = "div#prompt-textarea"
selector_submit = "button#composer-submit-button"
selector_response = "div.markdown"

logging.basicConfig(filename="chatgpt_scraper.log", level=logging.DEBUG)
logger = logging.getLogger(__name__)

def write_output(response_md: str) -> None:
    try:
        with open (output_file, "w") as f:
            f.write(response_md)
    except Exception as e:
        logger.error(f"Write failed. Error: {str(e)}", exc_info=True)

def scrape(url: str, page: Camoufox) -> None:
    try:
        response = page.goto(
            url, 
            wait_until="domcontentloaded",
            timeout=timeout)
        if response and not response.ok:
            logger.warning(f"Page returned status code: {response.status}")
        
        textarea_elem = page.locator(selector_textarea)

        if not textarea_elem:
            raise Exception("Textarea element not found.")
        
        textarea_elem.fill(query)
        
        submit_button = page.locator(selector_submit)
        
        if not submit_button:
            raise Exception("Submit button not found")
        
        submit_button.click()

        page.wait_for_load_state("networkidle", timeout=timeout)

        response_elem = page.locator(selector_response)
        if not response_elem:
            raise Exception(f"Response element with selector '{selector_response}' not found")

        response_text = response_elem.text_content()

        print ("Output: ")
        print (response_text)

        response_md = markdownify(response_elem.inner_html())
        print ("Markdown:")
        print (response_md)

        if not response_text:
            logger.warning("Response element is empty")

        write_output(response_md)

    except Exception as e:
        logger.error(f"Scrape failed: {str(e)}", exc_info=True)


def main() -> None:
    # Setting up basic camoufox, no need for anything fancy to unblock ChatGPT yet.
    # https://camoufox.com/python/usage/
    # Setting it up with a rotating proxy
    with Camoufox(geoip=True,
        proxy={
            "server": proxy_server,
            "username": proxy_uname,
            "password": proxy_pass,
        },
        headless = True, #False to see the browser in operation.
    ) as browser:
        page = browser.new_page()
        # Beggining the scrape in the scrape() function
        scrape(url, page)
        # Closing the browser
        browser.close()


if __name__ == "__main__":
    main()