cat: review-driven cleanup, --no-headers fix and rowskey speedup

[jqnatividad]

Summary

• Fix a cat rowskey --no-headers correctness bug where the second pass reused only the last file's synthetic header for all files, silently truncating trailing columns of any wider file appearing earlier. Per-file synthetic headers are now stored in pass 1 and indexed back in pass 2.
• Remove dead stdin-handling code in cat_rowskey (util::process_input already materializes stdin to a temp file before we run, so Config::is_stdin() was never true here). Drop the now-unused DEFAULT_WTR_BUFFER_CAPACITY import.
• Replace canonicalize().unwrap(), file_name().unwrap(), file_stem().unwrap() and conf_path.unwrap() with ? / graceful fallbacks — consistent with the recent schema/fetch/slice panic-safety passes.
• Warn (wwarn!) when an input column collides with --group-name, in both the headers and --no-headers branches (silent override before).
• Speed up the cat_rowskey hot loop: precompute a Vec<Option<usize>> mapping from the global column set to each file's columns once per file, replacing a per-cell IndexMap probe (with Box<[u8]> hashing) with a flat Vec walk.

Test plan

• cargo build --locked --bin qsv -F all_features
• cargo t cat -F all_features — 43 passed (was 42 before adding regression tests)
• cargo clippy --bin qsv -F all_features --no-deps — clean
• Manual smoke: stdin via -, two-file --group fname, and the wider-first --no-headers case all render correctly
• roborev review datefmt doesn't recognize this timestamp #1692 addressed (LOW: collision warning in --no-headers, mirror narrower-first test) and closed

New regression tests:

• cat_rowskey_no_headers_widerfirst — wider file first, narrower second; would have lost columns under the old code
• cat_rowskey_no_headers_narrowerfirst — symmetric mirror, pins the case that used to "work by accident"
• cat_rowskey_group_name_collision — asserts the warning fires and the documented "file column wins" behavior is preserved

🤖 Generated with Claude Code

[codacy-production]

Up to standards ✅

🟢 Issues 0 issues

Results:
0 new issues

View in Codacy

🟢 Metrics 9 complexity

Metric	Results
Complexity	9

View in Codacy

_{NEW Get contextual insights on your PRs based on Codacy's metrics, along with PR and Jira context, without leaving GitHub. Enable AI reviewer
TIP This summary will be updated as you push new changes.}

[Copilot]

Pull request overview

Fixes correctness and performance issues in cat rowskey, particularly for --no-headers, while improving panic-safety and adding regression coverage.

Changes:

• Fix cat rowskey --no-headers two-pass bug by storing per-file synthetic headers from pass 1 and reusing them in pass 2.
• Improve cat_rowskey throughput by precomputing a global→per-file column index map (Vec<Option<usize>>) to avoid per-cell hashmap probes.
• Add regression tests for wider-first / narrower-first --no-headers cases and --group-name collision warnings.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 3 comments.

File	Description
src/cmd/cat.rs	Implements the --no-headers fix, removes dead stdin code, adds collision warnings, improves panic-safety, and speeds up the rowskey hot loop.
tests/test_cat.rs	Adds regression tests covering the fixed --no-headers behavior and group-name collision warning behavior.