ci: add manual release workflow

[mohamede1945]

• Replace tag-triggered release with manual release inputs
• Bump gradle.properties and tag the release in workflow
• Reuse the same run to build and publish release artifacts

Testing:

Test 1

See https://github.com/quran/mobile-sync/actions/runs/25288607335.

• Created using command:

gh workflow run release.yml \
  --ref afifi/test-release-workflow \
  -f release_number=0.1.1-test.1 \
  -f target_branch=afifi/test-release-workflow

• It created commit d7495cf
• It created a tag https://github.com/quran/mobile-sync/releases/tag/v0.1.1-test.1
• It created a github release with xcframework attached in https://github.com/quran/mobile-sync/releases/tag/v0.1.1-test.1

Test 2

See https://github.com/quran/mobile-sync/actions/runs/25293771252.

• Created using command:

gh workflow run release.yml \
  --ref afifi/test-release-workflow3 \
  -f release_number=0.1.1-test.2 \
  -f target_branch=afifi/test-release-workflow3

• It created commit bd02d8c
• It created a tag https://github.com/quran/mobile-sync/releases/tag/v0.1.1-test.2
• It created a github release with xcframework attached in https://github.com/quran/mobile-sync/releases/tag/v0.1.1-test.2
• It verified the commit and tag happen after the XCFramework is built and packaged.

[ahmedre]

will this bot have access to write to this repo or do we need to have a token as an env variable?

[mohamede1945]

Yes, there is a test run in the PR's description. It's mainly because of permissions of contents: write we have. See the example here: https://github.com/actions/checkout/blob/main/README.md#push-a-commit-using-the-built-in-token

[ahmedre]

should this happen at the very end? - if anything fails after this point we'll have this commit / tag etc and we didn't release yet

[mohamede1945]

That makes sense, I'll create the commit and tag after creating the xcframework