This plugin allows you to extract metadata from any site on demand.
The metadata of a link is for example the title, description and image that are displayed when posting its URL on social networks.
Usage
This plugin adds an API Call (Metascraper) type used to represent the extracted metadata.
Metadata extraction is done through the action Plugins > Metascraper - Extract a list of metadata, which takes as parameter :
URLs to extract: the list of URLs to be analyzed.
Metadata Type: the type API Call (Metascraper) provided by the plugin
Types of metadata
In this first version, the extracted metadata are the following parameters:
Author
Date
Description
Image
Logo (clearbit)
Publisher
Title
URL
This plugin uses the https://metascraper.js.org library and tries to extract these parameters by taking a maximum of standards into account (Open Graph, Twitter, …) for greater reliability.
Ugh, could I have some help getting this to work? I have been wasting hours trying. I even emailed Bubble support - who said I needed to provide the plugin with a LIST of urls (evidently just a single line of text won’t do). So I did… and now it gives me a pop-up error that says invalid url. However I checked… and the url is fine. I checked multiple ways - checked the database to be sure… and even had it display the very field after creating it JUST before I have metascraper go. I am seeing a list of text, with one url in it, and the url looks fine.
I even tried the same url on your demo page and it worked fine. So I don’t know if maybe the most recent version of the Bubble editor isn’t compatible with the plugin now, or what. It’s driving me crazy. I just need some scraping
Hey @CramerTech ! Sorry to hear you have problem with the plugin !
Could you provide me a project reproducing the error (with, at least, an “everyone can view” setting) ? We can also plan a call if it is simpler for you
For others who experience issues scraping a webpage, be sure to include https:// before the rest of the URL. This missing part was causing errors to be thrown for me today.
The plugin is expecting a “list of texts” as URLs to extract input parameter.
In your screenshot, Domain seems to be a single text or a table’s row, and not a “list of texts”.
You probably have access to the :converted to list operator on ShopsBeta or Domain object, which allows you to “cast” a single item as a “list”.
You can also go through a custom state of type “list of texts”.
I could add an option in the plugin to allow the use of a single URL input, but I’m overbooked right now, so not before 1 or 2 weeks.
The results are, then, available in the following workflow steps, as result of step N.
