I’m using an GET node to do an http get from
https://api.deezer.com/2.0/track/isrc:GBF088500674 to verify if it is a valid ISRC number… (I’ll feed this from a list of numbers)
If the ISRC number if correct then a JSON object is returned with information about the music track and I use the JSON Path node to get to the data I want. Works fine. When the ISRC number is invalid, then a different JSON structure is returned;
“message”: “no data”,
This is expected. In this case, I want to be able to detect that the ISRC number is ni valid. What the best practice on this? Is there an example workflow?
You can use the Row Splitter node with the following dialog configuration:
Include (Exclude) rows by attribute value
Column to test: the column containing your JSON response
Use pattern matching with “contains wild cards” option selected and
This way you will have different responses split and can treat them in accordance with your needs afterwards.
Sounds like you may have a use for the Palladian and/or Selenium nodes. Here’s a post from
@qqilihq a while back that talks about them some - he may be able to provide some more recent info.
Hi Fabio, afair there are some examples on the KNIME example server. Also you'll find some workflows on the Palladian KNIME
community pages. However, if you're targeting modern websites and web apps, I would strongly recommend to have a look at the Selenium Nodes. These have dedicated functionality which closer resembles a web browser (in contrast to the relatively "dumb" HttpRetriever/HtmlParser). This allows you to scrape practically any website, as you're running an actual web browser…