feat: add library field to v1 and v2 public channel APIs, gate v1 ChannelVersion lookup

[rtibblesbot]

Plan: Add a library enum field ("KOLIBRI", "COMMUNITY", or null) to the v1 and v2 public channel API endpoints, and gate v1 ChannelVersion token resolution behind a channel_versions=true query parameter.

• Bump le-utils to 0.2.17
• Write the failing gate test
• Implement the gate in _get_channel_list_v1
• Update the three existing channel-version tests
• Write failing tests for V1 library field
• Implement library in _serialize_channel_version()
• Add library field to PublicChannelSerializer
• Write failing tests for V2 library field
• Implement library in V2 consolidate()

Summary

Adds a library field to the v1 and v2 public channel API endpoints, enabling Kolibri to distinguish Kolibri Library, Community Library, and other channels. Also gates v1 ChannelVersion token resolution behind ?channel_versions=true to prevent older Kolibri clients from accidentally retrieving data they cannot parse.

References

• Closes Add library enum field to v1 and v2 public channel API responses #5821
• Related: Add support for channel version tokens on v1 public channel endpoint #5772 (channel version token support added to v1)

Reviewer guidance

Risky areas worth careful review:

1. views_v1.py gate — ?channel_versions=true is a deliberate breaking change to the v1 API contract. Three existing channel-version tests were updated to include the parameter; without it, channel-version token lookups now return 404.
2. _get_channel_version_library() status filter — must include both STATUS_APPROVED and STATUS_LIVE; any other status (including STATUS_PENDING) correctly returns null.
3. V2 consolidate() mapping — non-public channels map to "COMMUNITY" (not null), which differs from v1 behaviour where only channel-version tokens with a qualifying submission return "COMMUNITY".

To test: pytest contentcuration/kolibri_public/tests/ -v

AI usage

Implemented using Claude Code following a pre-approved TDD plan (red-green-refactor). Tests were written before each implementation step; the full test suite passed before each commit.

---

@rtibblesbot's comments are generated by an LLM, and should be evaluated accordingly

How was this generated?

• Ran pre-flight CI checks (lint, format, tests) and verified all pass
• Rebased onto the target branch and resolved any conflicts
• Reorganized commit history into clean, logical commits
• Audited the diff to ensure only issue-relevant files are changed
• Built PR body from the repository's PR template with evidence blocks

[rtibbles]

Implemented to spec, but just one area missing testing coverage.

[rtibbles]

Really expected to see a test confirming that without the channel_versions query parameter, the ChannelVersion token would not produce any result. Please add.

[rtibblesbot]

Added test_channel_version_token_without_param_returns_404 — a focused standalone test that uses a channel-version token without ?channel_versions=true and asserts a 404 response. Kept it separate from the combined gate test so the negative case is unambiguously visible on its own.

[rtibbles]

Changes are good - test coverage gives confidence.