fix(concatjs): update karma to 6.3.2 and fix #2093

[mattsoulanille]

PR Checklist

Please check if your PR fulfills the following requirements:

• Tests for the changes have been added (for bug fixes / features)
  No new tests added, but one was fixed to work with the change.
• Docs have been added / updated (for bug fixes / features)
  Docs shouldn't need to be updated for this change.

PR Type

What kind of change does this PR introduce?

• Bugfix
• Feature (please, look at the "Scope of the project" section in the README.md file)
• Code style update (formatting, local variables)
• Refactoring (no functional changes, no api changes)
• Build related changes
• CI related changes
• Documentation content changes
• Other... Please describe:

What is the current behavior?

Issue Number: #2093

What is the new behavior?

Fixes #2093. Updates karma packages to latest versions.

Does this PR introduce a breaking change?

• Yes
• No

The old version of karma seems to still work with these changes (n = 2 tests), but I haven't extensively tested it.

Other information

[mattsoulanille]

This is somewhat hacky. We're probably not supposed to import from lib

[mattem]

Should we be bumping the peer dep here: https://github.com/bazelbuild/rules_nodejs/blob/stable/packages/concatjs/package.json#L25 ? Or this this API available in 4.0.0 too

[mattsoulanille]

Nice catch. I forgot that there are separate package.json files for the pkg_npm targets.

We might not actually have to bump the peer dependency. The only key missing in the new File object that is present in the current implementation's object is content, which the current one sets to the empty string. Karma 4 seems to work fine without it, at least with our tests in //packages/concatjs/web_test/test/karma_typescript/... and //packages/concatjs/web_test/test/karma/....

We could still bump it to 5.1.0 just in case. What do you think?

[mattem]

As the tests pass I think this shows it's backwards compatible - we could bump the version as part of the upcoming 4.0.0 milestone release if we wanted, then it's a clear breaking point.

[mattsoulanille]

Sounds good to me. Thanks!