Commit graph

1 commit

Author SHA1 Message Date
jedarden
1dfaf73aa4 feat(pdftract-3g6ne): implement CMap codespace range parser
Some checks are pending
Schema Generation Validation / Validate JSON Schema (push) Waiting to run
Schema Generation Validation / Validate JSON Syntax (push) Waiting to run
This commit adds the codespace range parser for CMap streams. The parser
extracts the begincodespacerange / endcodespacerange blocks that define
legal byte-width boundaries for character codes in a CMap.

## Implementation

- CodespaceRange: Single range with lo/hi bounds (stored as [u8; 4]) and width (1-4 bytes)
- CodespaceRanges: Collection with SmallVec<[CodespaceRange; 8]>
- CodespaceParser: PostScript-style tokenizer for begincodespacerange blocks

## Acceptance Criteria (all PASS)

- Parse <00> <7F> → 1 range, width=1 
- Parse <00> <7F> <8000> <FFFF> in one block → 2 ranges 
- Width inference: 2-char hex → width=1; 4-char hex → width=2 
- Case-insensitive hex (<C0> and <c0> equivalent) 
- Malformed range (width mismatch) → diagnostic + skipped 
- Empty CMap → empty ranges 
- JIS range <8140> <FEFE> → 2-byte CJK 
- 3-byte and 4-byte range support 

Also adds encrypted fixture provenance entries to PROVENANCE.md.

Co-Authored-By: Claude Code <noreply@anthropic.com>
2026-05-28 05:47:07 -04:00