Support for non-JSON responses

[humble-barnacle001]

Could you please add support for responses which may be as text/plain by checking the response headers before parsing to JSON and saving the data in a txt file instead?

[JamesIves]

Seems reasonable. Do you have an API handy I can test this with?

[humble-barnacle001]

Here is link but headers are absent: https://hello-world-api-sbn2k60ruxvs.runkit.sh/

[humble-barnacle001]

Here is one with headers I created: https://fetch-api-data-action-test-scvrfk4481as.runkit.sh/

[JamesIves]

Thanks, I'll get supported added to this later next week.

[humble-barnacle001]

Thanks @JamesIves for the super quick response

[JamesIves]

I've got a test version of this feature ready if you'd like to give it a try. Instead of checking the headers the action will now only parse JSON if it's valid, otherwise it will save the raw response. You can modify the file format directly using the new format property, in your case you can set this to txt to generate a file of that type.

- name: Fetch API Data 📦
   uses: JamesIves/fetch-api-data-action@releases/v2
   with:
      endpoint: https://fetch-api-data-action-test-scvrfk4481as.runkit.sh
      debug: true
      format: txt

Let me know if this works for you! If it does I'll promote the build.

[JamesIves]

Released as part of 2.1.0.

[humble-barnacle001]

I've got a test version of this feature ready if you'd like to give it a try. Instead of checking the headers the action will now only parse JSON if it's valid, otherwise it will save the raw response. You can modify the file format directly using the new format property, in your case you can set this to txt to generate a file of that type.

- name: Fetch API Data 📦
   uses: JamesIves/fetch-api-data-action@releases/v2
   with:
      endpoint: https://fetch-api-data-action-test-scvrfk4481as.runkit.sh
      debug: true
      format: txt

Let me know if this works for you! If it does I'll promote the build.

Hey @JamesIves thanks for the release. I will surely check this out.

[humble-barnacle001]

I see after the release #468 (reply in thread) and the release source in commit 570f45b it is actually writing the data as text into the file encoded as utf-8.

Could you please also possibly allow an option to set the character encoding as desired?

[JamesIves]

Go for it! I'd start by searching for format and taking a look at how that property gets passed to the file generation function.

[humble-barnacle001]

Thanks for the help. I shall use it as my starting point.

[humble-barnacle001]

Hey I did some changes, and while running the tests, I get this error:

I guess this is due to different Node.js versions, isn't it?

[JamesIves]

Most likely yes. Your node version should match the one found in the nvmrc file. The contributing guide has details on how to see this up.

[JamesIves]

This has been published and released as part of v2.2.0. Thanks for your contribution!