fix(pipeline): preserve dotted TypeScript import basenames

[pcristin]

What does this PR do?

Preserve extensionless dotted JavaScript/TypeScript module basenames such as
featureX.engine during relative-import resolution.

• Strip only recognized JS/TS module file extensions instead of treating every
  final dot suffix as an extension.
• Avoid applying generic module-extension stripping a second time after the
  relative resolver has normalized the path.
• Add a five-file TypeScript regression reproducing TypeScript: CALLS edges silently missing for some directly-imported functions — trace_path reports callers_total: 0 (v0.10.5) #1682 and assert that the
  helperB and SomeType IMPORTS edges both target featureX.engine.ts.

Refs #1682.

Verification

• make -f Makefile.cbm test-focused TEST_SUITES='fqn edge_imports' — 143 passed
• make -f Makefile.cbm test-par — 7,577 passed, 0 failed, 8 skipped across 139 suites
• git diff --check
• Changed files pass clang-format --dry-run --Werror
• Diff-scoped clang-tidy passes

The repository-wide lint-ci baseline is not clean in untouched files: cppcheck
reports existing style findings in src/store/store.c and src/cypher/cypher.c,
and the formatter reports existing findings in src/mcp/mcp.c and
src/pipeline/pipeline_incremental.c. No unrelated lint cleanup is included here.

Checklist

• Every commit is signed off (git commit -s) — required, CI rejects
  unsigned commits (DCO, see CONTRIBUTING.md)
• Tests pass locally (make -f Makefile.cbm test)
• Lint passes (make -f Makefile.cbm lint-ci)
• New behavior is covered by a test (reproduce-first for bug fixes)

[github-actions]

Thanks for opening this — it has been seen, and it is queued.

This note is automated, but it is not a brush-off: it exists so you know where your PR stands instead of having to guess from silence.

Current review status: working through a backlog. 0.9.1-rc.1 is out, so the release freeze that held reviews is over — but it left a large queue of open pull requests behind it, and we are reading through them oldest-first. The background is in discussion #1144.

What that means for this PR, concretely:

• It will not be closed for inactivity. No stale bot touches pull requests here.
• It may still sit a while before a human reads it. That is on us, not on you.
• Older PRs are read first, so a recent one is not being skipped — it is behind a queue.

Things that will genuinely speed it up whenever review does happen:

• Keep it rebased on main — the tree is moving quickly right now, and a conflicting branch cannot be reviewed as the diff you intended.
• Get CI green, or say which failures you believe are pre-existing.
• Keep the change to one claim. Bundled features and refactors get split before they get merged, which costs you a round trip.
• Every commit needs a sign-off (git commit -s) — CI enforces DCO.

If this fixes a bug, a reproduction we can run is worth more than a description of the symptom.

Thanks for contributing, and sorry in advance for the wait.

[pcristin]

CI triage: the only substantive red job is test / test-windows-guards, where section_cold_storm lost one of six concurrent clients while they raced to spawn the ephemeral daemon. All dedicated Windows test shards, sanitizers, lint, security, and DCO jobs passed.

This PR changes only FQN/import resolution and its regression tests; it does not touch daemon or Windows-guard code. The same cold-storm timeout occurred on the unrelated PR #1245 workflow run, while surrounding runs of this guard passed, so this matches an
existing intermittent guard failure rather than this patch.

I tried to rerun only the failed jobs, but GitHub rejected it because contributors do not have repository Actions rerun permission. Could you @DeusData rerun the failed jobs? I have deliberately not added an unrelated code change or empty commit just to retrigger CI.

[DeusData]

Thank you for the careful CI attribution. I checked the diff and current main: this PR is limited to FQN and import resolution plus its regression tests, while the red job is the Windows daemon guard. That failure is outside this change path and matches the existing startup-flake family you linked.

You are also correct that contributors cannot rerun repository jobs. This should be rerun from the maintainer side, and there is no need to create an unrelated commit. Thank you for documenting the evidence instead of treating the red aggregate as proof against the patch.