[#285]:svarga:test, raise library line coverage above 95%

[steven-varga]

Closes part of #285 (see status note); >95% completion gated on #286.

What

Raises reachable h5cpp library line coverage 82.3% → 94.0% (functions 96.3%), 56/56 tests pass.

• New tests for the H5cout handle pretty-printers (37→94%), recursive dfs/bfs traversal (55→75%), the rank-2 chunked pipeline write path, the synchronous high_throughput write/open/read dispatch (H5Dread 79→90%, H5Dopen 60→80%), and attribute / auto-chunk-create / fletcher32 edges.
• Excludes the unreachable H5Zpipeline_pool.hpp from the Codecov pipeline (180 dead lines — see fix, carry per-file worker pool in a fileno-keyed singleton registry (H5Fget_access_plist strips FAPL props) #286), with an inline comment to remove the exclusion once activation is fixed.

Why not ≥95% in this PR

The remaining uncovered lines are dominated by the #286 dead worker-pool branches inside H5Dwrite/H5Dread/H5Dappend (~44 lines). Coverage excludes are file-granular, so they can't be dropped without dropping the tested files. Fixing #286 makes ~224 lines live and clears 95% over the full denominator with no exclusions — tracked as the follow-up.

Related

• CDash-side coverage scope fix (thirdparty/test/examples) is in ci, off-CI CDash submissions from any architecture #278.
• Worker-pool activation defect + two more high_throughput defects: fix, carry per-file worker pool in a fileno-keyed singleton registry (H5Fget_access_plist strips FAPL props) #286.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!