Align foundational Go conversion capabilities

[shps951023]

Summary

This draft advances Go parity through nine independently validated stages:

• bound OOXML ZIP loading and expose ErrInvalidPackage
• add Reader/Writer conversion and complete font registry lifecycle APIs
• embed registered TTF fonts as Type0/CID fonts with ToUnicode maps
• add CLI --fonts support
• fix one-byte PDF content stream length errors
• remove synthetic XLSX sheet labels and paginate wide worksheets horizontally
• add Flate-compressed PDF content streams and CLI --compress
• parse DOCX section margins and support validated margin overrides
• add XLSX row/column limits and portrait/landscape overrides in API and CLI

Validation

• go vet ./...
• go test ./...
• go test -race ./...
• real DOCX CLI conversion validated with pdfinfo and pdftotext
• Go classic XLSX five-case visual benchmark: 0.8093 baseline to 0.9975
• classic05_wide_table: 0.6867 at 1/3 pages to 0.9953 at 3/3 pages
• PDF structure errors: present in all five baseline candidates, zero after the stream fix
• Go classic DOCX five-case benchmark: 0.9872 baseline to 0.9889 after native margin parsing

Dependency

Adds golang.org/x/image v0.24.0 for SFNT parsing. Registered fonts are currently embedded in full; subsetting, TTC support, system font discovery, and complex-script shaping remain explicit follow-up work.

Remaining Go parity work

This PR remains draft because the matrix still tracks rendering and API gaps, including XLSX sheet selection, fit/scale, styles/merges/images, DOCX tables/images/styles, PPTX shapes/images, and complete benchmark evidence.

[coderabbitai]

Important

Draft PR not reviewed

Draft PRs are not automatically reviewed by default.

• Trigger a manual review

To automatically review draft PRs, update your CodeRabbit configuration:

reviews:
  auto_review:
    drafts: true

---

_{Comment @coderabbitai help to get the list of available commands.}