chore: archive google-ai and google-vertex integrations, migrate to google-genai

[vblagoje]

Why:

Archives google-ai-haystack and google-vertex-haystack integrations to consolidate Google offerings under the unified google-genai-haystack integration, reducing maintenance burden and providing users with a single, actively maintained Google integration.

• fixes Archive google-ai-haystack, google-vertex-haystack. Point to google-genai-haystack. #2348

What:

• Added deprecation notices to google_ai and google_vertex integration READMEs directing users to google-genai-haystack
• Updated main README to mark both integrations as Archived in the inventory table
• Extended CI workflows (CI_pypi_release.yml, CI_docusaurus_sync.yml) to support PEP 440 post-release versions (.postN)
• Enhanced validate_version.py to recognize and validate post-release versions against PyPI

Manual PyPI steps after this PR merge:

1. Tag post version for these two integrations
2. Release them via our pypi actions
3. Update PyPI project descriptions with deprecation notices
4. Mark packages as archived via PyPI Settings

How did you test it:

• Verified regex patterns match post-release version formats
• Validated CI workflow tag patterns support .postN suffix
• Tested version validation logic handles both regular and post-release versions

Notes for the reviewer:

• Post-release versions (.post0) are PEP 440 compliant and appropriate for final archival releases
• Manual PyPI steps documented but cannot be automated (no API support)
• CI workflow changes are generic and will support post-releases for any future integration

[vblagoje]

Perfect @anakin87 - I wouldn't know anyone else more suitable for this one than you. Anything I missed?

[anakin87]

Looks good.