Scraper graphic elements failed

Hello,
I am unable to scrape the graphic elements within this table to then display them in a CSV file. Can you help me?

Thank you in advance for your assistance.

Hello,

Unfortunately we did not receive any links to the table in question, could you please resend?

Looking forward to your reply.
Victoria

Customer Support - bardeen.ai
Knowledge Base https://support.bardeen.ai/hc/en-us
Explore | @bardeenai | Bardeen Community

1 Like

You are right, here is what you requested. Let me give a brief recap.

Currently, I am exploring deep scraping for product pages on the website accessible at the following link:

The goal is to extract certain sections of the document, such as: ā€˜vantaggiā€™ and ā€˜svantaggiā€™. Initially, I used CSS selectors but, encountering limitations, I developed an XPath script with the help of ChatGPT, available here: MigliorRobot - Scrape Single page šŸ„‡ | Bardeen Community Playbook.
The script works correctly, however, it does not preserve the original text formatting. Is it possible to modify the script to maintain a bulleted list, or to insert appropriate line breaks that facilitate reading the extracted points?

Furthermore, I am encountering difficulties in extracting the initial table on the page.

I cannot extract it, neither as text nor as an image. I also add that during the scraping the ā€˜green checkā€™ symbol is not recognized, an element that seems particularly difficult to isolate.

Finally, I take this opportunity to ask if and how I can configure a command that allows ChatGPT to automatically process the data once extracted. I have tried to create a new command but it seems not possible to make ChatGPT view the CSV obtained from scraping and insert a specific prompt to ask ChatGPT what you want.

Thank you in advance for your support and I look forward to your valuable guidance.

@Bardeeni can you help me? Thank you so much.

Hi,

It looks like you are building a very complex scraper. First of all, I would like to mention that it is not possible to preserve initial formatting for scraped text. Our engineers are working on adding this feature and I added request from you to the list as well. We will let you know when this becomes available.

As for the table, scraper cannot convert it to an image, you should scrape individual table elements separately. Otherwise, if you select the table as single element, you may end up with unformatted plain text. ā€œGreen checkmarksā€ should be scraped as images, you will not get them inside of text, but at least you will be able to mark field as checked. For example when you createfield Mappatura, hereā€™s how you can scrapea value for it:

For the ChatGPT step, I suggest you use ā€œGet custom prompt to OpenAIā€ action. You do not have to send a CSV to it. When you add an action, you can directly connect it to scraping action, here is an example:

Here is our general guide that I suggest you to look through: https://support.bardeen.ai/hc/en-us/articles/23925200748173-How-to-use-Bardeen-scraper Also here are tutorial videos: https://www.youtube.com/playlist?list=PLcXybhQ4q2HndLcnFJj2AvesiJdtfkRdv

If youā€™re interested in having our team build a custom playbook for you, we offer this as part of our Business plan. Just let us know, and weā€™d be happy to connect you with our Sales Team to discuss details.

Best regards,
Victoria

Customer Support - bardeen.ai
Knowledge Base https://support.bardeen.ai/hc/en-us
Explore | @bardeenai | Bardeen Community