C wide string literals L"..." / wchar_t

• Type: feature (clexer + crtl <wchar.h>). Track C. Low priority.
• Found: 2026-07-06 c-testsuite run.

Failing test

• 00220: wchar_t s[] = L"hello$$你好¢¢世界€€world"; — UTF-8 source decoded to wchar_t (4-byte on Linux) array; test prints code points %04X. We output empty line (literal decays to nothing). Needs: L-prefix lexing, UTF-8 → UCS-4 decode, wchar_t array init, crtl wchar.h with wchar_t typedef.

Gate

Drop 00220.c from test/c-conformance/pxx.skip; runner green.

Triage 2026-07-07

Foundations present: wchar_t = int (4-byte) in lib/crtl/include/wchar.h, and 00220 COMPILES (so wchar_t/wchar.h resolve). It just produces empty output — L"..." is not decoded. Needs a coordinated add across clexer + cparser:

1. clexer: detect an L immediately followed by " as a wide-string prefix (before the identifier path), lex the string, and UTF-8-DECODE its bytes to Unicode code points. The token model stores byte strings (SVal:AnsiString), so the code points must be stored as 4-byte-LE in SVal (SLen = 4*count) with a wide flag, or in a side table.
2. cparser: wchar_t s[] = L"..." must initialize a 4-byte-element (int32) array from the wide token's code points + a wide NUL, sized count+1 (mirror the char buf[]="lit" per-element init path but 4-byte). Multi-subsystem; ~60-80 lines across two files with UTF-8 edge cases. Focused session. (00220's chars are all BMP, so 1-3 byte UTF-8 -> <=U+FFFF.)

Log

• 2026-07-08 — resolved, commit 658d284c.