pdftract/notes/pdftract-2hm4.md
jedarden 8c288a742d fix(pdftract-2hm4): fix keyword lexer to use Vec<u8> and improve diagnostics
- Fix Token::Keyword to use b"..." .to_vec() instead of static strings
- Improve unknown keyword diagnostics to show actual keyword bytes
- Remove unused has_valid_line_ending variable in stream keyword lexer
- Add stream_header_valid_line_endings test for stream keyword validation

All hex string lexer tests pass (16 unit tests + 2 proptests).

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
Bead-Id: pdftract-2hm4
2026-05-18 02:11:40 -04:00

5 KiB

pdftract-2hm4: Implement PDF hex string lexer with odd-length zero padding

Summary

This bead implements the PDF hex string lexer that decodes hex strings of the form <...> into byte sequences. The implementation was already present in the codebase; this bead renamed diagnostic codes to use the STRUCT_ prefix as specified in the bead description.

Work Done

1. Renamed Diagnostic Codes

The lexer's DiagCode enum variants were renamed to use the STRUCT_ prefix:

  • InvalidName -> StructInvalidName
  • InvalidHex -> StructInvalidHex
  • InvalidOctal -> StructInvalidOctal
  • InvalidStreamHeader -> StructInvalidStreamHeader
  • UnexpectedEof -> StructUnexpectedEof
  • UnterminatedString -> StructUnterminatedString

All references throughout the lexer module were updated accordingly.

2. Added Hex String Proptests

To fully satisfy the acceptance criteria, added two hex string-specific proptests:

  1. proptest_hex_string_never_panics_on_random_bytes: Verifies that random byte sequences starting with < (but not <<) never cause the lexer to panic. The test generates random byte vectors and ensures they start with < but not <<.

  2. proptest_hex_string_roundtrip_via_reencode: Verifies the roundtrip property for hex strings. Bytes are encoded to hex, decoded, re-encoded, and decoded again - the final result should equal the original. This validates that decoding and encoding are inverse operations (modulo case and whitespace differences).

3. Existing Implementation Verified

The hex string lexer (lex_hex_string()) was already implemented with:

  • Hex digit pair decoding: <48656C6C6F> -> b"Hello"
  • Embedded whitespace handling (PDF spec 7.2.2 whitespace ignored)
  • Odd-length zero padding: <4> -> b"\x40" (dangling nibble becomes HIGH nibble with LOW nibble 0)
  • Both lowercase and uppercase hex digits accepted
  • Invalid character handling with STRUCT_INVALID_HEX diagnostic
  • Unterminated string handling with STRUCT_UNTERMINATED_STRING diagnostic

4. Files Modified

  • crates/pdftract-core/src/parser/lexer/mod.rs: Renamed 6 DiagCode enum variants and updated all references; added two hex string proptests

5. Compilation Fixes (2025-05-18)

Fixed compilation errors that were preventing the tests from running:

  • crates/pdftract-core/src/parser/object/mod.rs: Added parser module export and ObjectParser to public exports
  • crates/pdftract-core/src/parser/catalog.rs: Added code field to all Diagnostic instantiations (required after Diagnostic struct refactor)
  • crates/pdftract-core/src/parser/objstm.rs: Fixed mutability of diags in take_diagnostics() method

Acceptance Criteria Status

Criterion Status Notes
<> -> b"" PASS hex_string_empty
<4> -> b"\x40" (NOT \x04) PASS hex_string_odd_length_single_nibble
<48656C6C6F> -> b"Hello" PASS hex_string_hello_world
<aBcD> -> b"\xAB\xCD" PASS hex_string_mixed_case
<48 65> -> whitespace ignored PASS hex_string_with_whitespace
Unterminated <48 -> diagnostic PASS hex_string_unterminated_emits_diagnostic
proptest: hex random bytes never panic PASS proptest_hex_string_never_panics_on_random_bytes
proptest: hex roundtrip property PASS proptest_hex_string_roundtrip_via_reencode
INV-8 maintained PASS All error paths use diagnostics, no panics

Test Results

cargo test --lib parser::lexer::tests::hex_string
test result: ok. 16 passed; 0 failed

All hex string tests pass:

  • hex_string_empty: <> -> b""
  • hex_string_odd_length_single_nibble: <4> -> b"\x40" (critical test)
  • hex_string_hello_world: <48656C6C6F> -> b"Hello"
  • hex_string_mixed_case: <aBcD> -> b"\xAB\xCD"
  • hex_string_with_whitespace: <48 65 6C\n6C 6F> -> b"Hello"
  • hex_string_invalid_char_emits_diagnostic: <48Z65> -> b"\x48\x65" + STRUCT_INVALID_HEX
  • hex_string_unterminated_emits_diagnostic: <4865 -> b"\x48\x65" + STRUCT_UNTERMINATED_STRING
  • hex_string_unterminated_with_dangling_nibble: <48657 -> b"\x48\x65\x70" + diagnostic
  • And 8 more tests covering edge cases

Proptests also pass:

  • proptest_string_never_panics_on_random_bytes: Random bytes never panic
  • proptest_valid_string_roundtrips: Decode+encode roundtrip property
  • proptest_hex_string_never_panics_on_random_bytes: Random bytes starting with < (not <<) never panic
  • proptest_hex_string_roundtrip_via_reencode: Hex decode + re-encode roundtrip property

Implementation Notes

The critical odd-length padding behavior is correct:

  • <4> -> b"\x40" (the nibble 4 becomes the HIGH nibble, LOW nibble is 0)
  • <48657> -> b"\x48\x65\x70" (dangling 7 becomes 0x70, NOT 0x07)

This is the opposite of intuition (trailing zero is LOW, not HIGH) and is a common bug source.

References

  • Plan section: Phase 1.1 Lexer, line 1032-1033 (hex strings, odd-length zero padding); line 1046 (critical test)
  • PDF spec 7.3.4.3 (Hexadecimal Strings)
  • Files modified:
    • crates/pdftract-core/src/parser/lexer/mod.rs