fix(copyright): scan bounded YAML credit sections

[mstykow]

Summary

• Parse YAML front matter and explicitly tagged YAML sections without treating the surrounding document as YAML.
• Keep extraction bounded by closing/tag markers and preserve global source line numbers.
• Remove a redundant structured handle prefix only when it exactly matches the declared email local part.

Issues

• Covers: follow-up findings from False-positives when comparing provenant with scancode-toolkit output #1391

Scope and exclusions

• Included: array-valued author/contributor credits in leading YAML front matter and --- yaml tagged sections.
• Explicit exclusions: unmarked code literals that merely resemble metadata and arbitrary recovery from invalid YAML.

How to verify

• Scan Perl 5.38.4 with provenant --copyright --json out.json perl-5.38.4. Relative to fix(copyright): extract structured contributor credits #1408, this adds the source-backed Matt S. Trout credit in the bounded YAML section of cpan/CPAN-Meta-YAML/t/tml-local/yaml-roundtrip/quoting.tml, with no other author changes and no copyright/holder changes. The InfoZIP 3.0 control scan is unchanged in all three fields.

Intentional differences from Python

• Provenant strips mst: because it duplicates the local part of <mst@shadowcatsystems.co.uk>; ScanCode renders the handle as part of the person's name. Unmarked Perl fixture literals remain excluded.

Follow-up work

• Created or intentionally deferred: explicit prose author sections and malformed/template author cleanup remain in later stacked branches.

Expected-output fixture changes

• Files changed: none.
• Why the new expected output is correct: focused tests cover front-matter bounds, tagged-section bounds, source lines, and conservative handle normalization.

---

Generated with Claude Code

[mstykow]

@greptileai review