Is there a way to extract thumbnails from hyperlink cards? Or better yet, scrape some info from these hyperlinked pages

1

I have this huge table with almost 3k rows, with lots of links from shutterstock. We use it in an editorial process where we try to reuse images we’ve bought in the past (before going and buying new ones). Each row on this table is an image we’ve bought and I turned those links into cards so we can see the thumbnail for each one.

When my team needs a new image (for example, a “vector illustration of a hand”), they look at this table and try to find it by looking at those thumbnails one-by-one. We even started manually adding tags to these images but gave up because there are so many.

I’m trying to nudge towards a better solution, and would be grateful for any ideas.

  • It would help if the thumbnails were IN the table instead of being displayed as link cards, because those link cards take some time to load and seem to unload when scrolled out of view.
  • Better yet, shutterstock has some tags for each image. If I could scrape them into a table cell in coda, then we’d have a better text search. Even retrieving the link title would help.

Ultimately I’d like to plug an AI and just ask it for what i need and it would give me the most relevant results from my table ! :grimacing:

Any ideas?

image
image1910×1298 255 KB

1 Like
2

Hi @CleberSantana ,

Great question, you can look at adding the OpenAi Pack to your doc and see if you can work with the AI functionally to find scrub the data you are looking for.

As for the downloading and adding the actual image to the table, with 3k rows you will see some doc performance slowing with the amount of data that is looking to load, but you will have direct access to the image in question.

I think one of the other options you could do, is pull the direct download link and have it directly download rather than take them to the webpage and then have to download, as it sounds like that is what the current process it, but I could be completely wrong!

Check out the AI pack and see if that will help you out first with some of the elements you are trying to scrub.

Best,
Dan

3

I think you can somewhat achieve what you’re looking for with a few formulas.

image
image1092×1107 198 KB

The Image column is a Image URL column type with this formula below. Change the ‘Link’ text to your link column.

If(
  thisRow.Link.Split("/").Last().Split("-").Count() >= 7,
  Concatenate(
    thisRow.Link
      .RegexReplace(
        thisRow.Link.Split("/").Last().Split("-").Nth(6), "600w"
      ),
    ".jpg
    "
  ),
  Concatenate(
    thisRow.Link
      .RegexReplace(
        thisRow.Link.Split("/").Last().Split("-").Last(), "600w-"
      ),
    thisRow.Link.Split("/").Last().Split("-").Last(),
    ".jpg
    "
  )
)

For the tags column, use this formula, again replacing the “Link” with your link column.

thisRow.Link.Split("/").Last().Split("-")

Then on the tags column, go to filter, canvas control, create control. So whatever keyword you type in, it’ll filter by that!

1 Like
4

Nice idea! That already helps a bunch!

5

Actually, that won’t work :grimacing:. At least it won’t be so easy. The URL I have isn’t like yours.

What I originally had was a spreadsheet with each image’s ID on shutterstock. Like this:
1573275691

Then I used these IDs to build this URL that points to the image page on Shutterstock. Like this:
https://www.shutterstock.com/pt/search/1573275691

And finally I’ve set the column formating to CARD so that I can SEE the thumbnail and page title. Kinda hacky but worked and that was an improvement compared to what i had but it is also very frustrating. I can see the info, but can work with it.

6

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.