Fix crash in match when passing a frozen value as app_identifiers

[AliSoftware]

Checklist

• I've run bundle exec rspec from the root directory to see all new and existing tests pass
• I've followed the fastlane code style and run bundle exec rubocop -a to ensure the code style is valid
• I've read the Contribution Guidelines
• I've updated the documentation if necessary.

Motivation and Context

When calling match with a frozen array (our_app_ids.freeze) for the app_identifiers parameter, we got a crash because match is trying to mutate that input value in-place. See wordpress-mobile/WordPress-iOS@cd585ae for some more details.

This PR replaces the mutation in-place with a new assignment instead, to avoid modifying the original value that the call site passed to match.

Description

This crash was due to this patching PR (made back in 2017 by @taquitos to quick-fix another issue), which calls app_identifiers.flatten!, mutating that parameter in place.

• My quick fix simply consists of changing this to app_identifiers = app_identifiers.flatten, to avoid modifying the original in-place and thus allow frozen values to be passed at call site.
• But tbh I'm not sure this fix is still required nowadays in 2022, as I expect FastlaneCore::ConfigItem#auto_convert_value to already take care of this nowadays, right?
• Though I'm not sure if that auto_convert_value appeared after the patch from 2017 (meaning that that auto_convert_value of comma-separated strings to Array did not exist back then and was added later, rendering this .split(/\s*,\s*/).uniq in match now obsolete) or was already present back then (meaning that the bug in match that we tried to patch was due to something else and more complicated)… 🤔
• Also worth mentioning the following related past PRs/Issues where we seem to have done some back and forth on this:
  • Match could not find App ID #11243
  • Updated handling of Arrays that are passed as Strings #11407
  • Fix not auto-converting values after user input #11423
  • Revert 'Updated handling of Arrays that are passed as Strings' #11424 (reverting 11407)

Testing Steps

I thought about writing a spec for this but couldn't think of a reliable way of how to test just this part of the Match::Runner without stubbing basically everything, only to test that line in isolation.

Another option could have been to extract lines L79-L87 into a dedicated private method, and write a test for it. But:

• I'm not sure how the fastlane team regards "testing private methods" (quite a controversial topic)
• I preferred to gather feedback on the fix implementation (i.e. will this patch do for now, or do we consider we can get rid of all the L79-L87 logic and consider that ConfigItem#auto_convert_value should already cover all this nowadays) first

(\cc @mokagio FYI, as the discoverer of that bug in our tooling)

[taquitos]

Ya, ooops 🤦‍♂️
I test private methods in languages that make it easy to do if the effort is low and potential impact of a bug is high. In this case, I'd say... nahhh.

🎈🐐

[fastlane-bot]

Hey @AliSoftware 👋

Thank you for your contribution to fastlane and congrats on getting this pull request merged 🎉
The code change now lives in the master branch, however it wasn't released to RubyGems yet.
We usually ship about once a week, and your PR will be included in the next one.

Please let us know if this change requires an immediate release by adding a comment here 👍
We'll notify you once we shipped a new release with your changes 🚀

[mokagio]

❤️ y'all

[taquitos]

❤️ y'all

(°◡°♡) .:｡

[fastlane-bot]

Congratulations! 🎉 This was released as part of fastlane 2.205.2 🚀