puppeteer web scraper through graphQL with appleData {_, args} ->
return { data: {data}, webdata: boolean }
// hardcoded object as edgecase for not scraper returning null which would make webdata: false;
chatGPT
leverage chatGPT API (graphQL resolver | fetch) to retrieve same candlestick data as puppeteer ->
in puppeteer functions return data: if puppeteerData.webdata === true // scraper returned valid web data
ask chatGPT thru API which data it thinks is more accurate:
it's own retrieved data or puppeteer's data.
again but if puppeteer used hardcoded string don't reprompt chatGPT just use AI stock data rather than hard coded string.
(can hedge against with the return data from puppeteer being previously successful return data, & still uses web data for default)
// notes:
puppeteer once in a {water-drinking-app}: user input data -> scrapes google images data returns BLOB.imgSrc // buggy 50%
never used chatGPT API. last night for first time asked chatGPT for it's API code & looks mostly easy
preferred sources of info as user input data ? or extend feature & ask puppet|API -> to cite sources?
attempting to do:
titled. fetch() candlestock stock data for apple
proposed approach:
puppeteer web scraper through graphQL with appleData {_, args} ->
// hardcoded object as edgecase for not scraper returning null which would make webdata: false;
chatGPT
leverage chatGPT API (graphQL resolver | fetch) to retrieve same candlestick data as puppeteer -> in puppeteer functions return data: if puppeteerData.webdata === true // scraper returned valid web data
ask chatGPT thru API which data it thinks is more accurate:
again but if puppeteer used hardcoded string don't reprompt chatGPT just use AI stock data rather than hard coded string.
(can hedge against with the return data from puppeteer being previously successful return data, & still uses web data for default)
// notes:
preferred sources of info as user input data ? or extend feature & ask puppet|API -> to cite sources?