Content-type-aware diff: tables (CSV/Excel) and PDF

[gregoryfoster]

Scope

Extend the diff service (src/core/diff/ from #115) with content-type-aware normalizers and renderers for:

1. Tables — CSV, Excel
2. PDF — page-aware

Tables

• Row-key-aware normalization: when the extractor identifies a key column, diff by row-key (row add/remove/modified) instead of line-by-line.
• Render in dashboard as a data-table with adds highlighted green, removes red, per-cell changes yellow with old → new values visible.
• Requires: pandas or equivalent; extractor metadata to carry key-column hints.

PDF

• Page-aware diff: show which pages changed, diff extracted text per-page rather than as one flat blob.
• Integrate with page-level screenshot comparison where available.

Dependencies

• Blocked by Watch Change UI: GitHub-style diff viewer (Phase A + B.1 done; B.2/C/D descoped) #115 Phases A/B (needs src/core/diff/ module, normalization layer, unified-diff core).
• May surface extractor-layer changes (key column detection for tables).

Acceptance

• src/core/diff/normalize.py gains normalize_table and normalize_pdf_text
• Dashboard Change Detail supports content-type-aware rendering for CSV/Excel and PDF watches
• TDD coverage for new normalizers