THRIFT-5967: Parallelize static code analysis run on GitHub

[kpumuk]

Currently SCA job runs as a single monolythic workflow, with each job checking different libraries separately. Jobs are allowed to fail, and there is a final summary job that check the status of each other job and fails if any failed. This makes it a little bit hard to read, makes the job slower, and somewhat hard to add new languages.

This pull request follows the similar structure we have in our build.yml - each language gets its own job, that depends on compiler build (used to generate thrift and verify generated code to follow the same rules as the library itself).

The time went down from 7m4s to 6m45s, which is not very significant, but that's pretty much a cap - compiler + cppcheck for C++ library, and the rest of the languages will finish in parallel before that, so it will never take longer. The old workflow would keep increasing the total time.

Additionally, replaces sloccount with scc, as it does not know the modern PHP syntax and was spitting lots of noise in the output:

Warning! Unclosed PHP file /home/runner/work/thrift/thrift/lib/php/test/Unit/Lib/Transport/TransportErrorScenariosTest.php, mode=1
Warning! Unclosed PHP file /home/runner/work/thrift/thrift/lib/php/test/Unit/Lib/Serializer/TBinarySerializerTest.php, mode=1

Nice and pretty now with https://github.com/apache/thrift/actions/runs/25285416874/job/74129027301

• Did you create an Apache Jira ticket? THRIFT-5967
• If a ticket exists: Does your pull request title follow the pattern "THRIFT-NNNN: describe my issue"?
• Did you squash your changes to a single commit? (not required, but preferred)
• Did you do your best to avoid breaking changes? If one was needed, did you label the Jira ticket with "Breaking-Change"?
• If your change does not involve any code, include [skip ci] anywhere in the commit message to free up build resources.

[kpumuk]

Created an umbrella ticket https://issues.apache.org/jira/browse/THRIFT-5966 to add SCA for other libraries.