Add count aggregation command tests

[danielfrankcom]

This change adds tests for the count aggregation command.

[documentdb-triage-tool]

🤖 Auto-triaged by documentdb-triage-tool.

Applied: compatibility test, enhancement
Project fields suggested: Component test-coverage · Priority P2 · Effort XL · Status Needs Review
Confidence: 0.85 (mixed)

Reasoning

component from path globs (test-coverage, test-framework); effort from diff stats (4446+43 LOC, 20 files); LLM: Adds new test coverage for the count aggregation command, expanding the compatibility test suite.

If a label is wrong, remove it manually and ping @patty-chow so the rules can be tuned. The bot will not re-label items that already have component labels.

[eerxuan]

I feel "namespace", "maxTimeMS", "comment" and "read_concern" are cross cutting feature outside of the individual command layer. We may just need one test for each in every command folder. The full coverage should be in there own folder. We have query-and-write/read-concern/. I feel "comment" and "maxTimeMS" can belong to /cursor/. As mongosh has helper cursor.comment() and cursor.maxTimeMs(). "namespace" can be under /collections/.
I should have notice this earlier, we can reduce duplication start from new PRs and have a backlog to remove existing duplication later.

Remove the duplication for collation and view in this PR.

[eerxuan]

Can you add a rule in TEST_COVERAGE.md specify these command arguments are cross cutting features and where they belongs to.

[eerxuan]

I feel "namespace", "maxTimeMS", "comment" and "read_concern" are cross cutting feature outside of the individual command layer. We may just need one test for each in every command folder. The full coverage should be in there own folder. We have query-and-write/read-concern/. I feel "comment" and "maxTimeMS" can belong to /cursor/. As mongosh has helper cursor.comment() and cursor.maxTimeMs(). "namespace" can be under /collections/. I should have notice this earlier, we can reduce duplication start from new PRs and have a backlog to remove existing duplication later.

Remove the duplication for collation and view in this PR.

Discussed offline.

• We should remove view, just keep 1 or 2 smoke test here. LIMIT_NOT_POSITIVE error need to verify.
• "collation" and "namespace" only need to keep type_.py, which test each bson type as input + a valid case.
• "comment" and "maxTimeMS" are good as it is now, no need to remove.
• "read_concern" is good as it is now, which covers syntax cases. The semantic cases will be covered under query-and-write/read-concern/. Same for write_concern. Don't need to test options not mentioned in specs.

[danielfrankcom]

Can you add a rule in TEST_COVERAGE.md specify these command arguments are cross cutting features and where they belongs to.

Done in #246