39537701882
100%

Ran 21 Oct 2026 03:18AM UTC

Jobs 1

Files 5

Run time 1min

Badge

Embed ▾

Committed 21 Oct 2026 03:16AM UTC coverage: 99.763%. Remained the same

Build # 39537701882

Build Type

push

github

Committed by

ednolan

Commit Message

Specify code unit adaptor conversion rule in terms of width, not sizeof

Per SG16 feedback, compare the width of the source and target types
([basic.fundamental]: value bits plus sign bit, excluding padding)
rather than sizeof, so that padding bits cannot hide a narrowing
conversion. The implementation computes width from numeric_limits
digits and is_signed, using unsigned char for std::byte.

Update the wording to define width for integer types, character types,
and byte via the underlying type. Replace the design section's single
conversion table with two: one covering all integral type sizes the
standard permits (assuming CHAR_BIT == 8), and one covering typical
x86_64/aarch64 sizes, using "same width" wherever LP64 and LLP64
targets differ.

Also size the exposition-only transcoding buffer as 4, 2, or 1 elements
for char8_t, char16_t, or char32_t respectively, rather than
4 / sizeof(ToType), which under-sizes the buffer on an implementation
whose char16_t has padding bits.

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>

Coverage Stats

421 of 422 relevant lines covered (99.76%)

257.21 hits per line

Jobs

ID	Job ID	Ran	Files	Coverage
1	39537701882.1	21 Oct 2026 03:18AM UTC	5	99.76	GitHub Action Run

bemanproject / utf_view / 39537701882
100%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 39537701882

bemanproject / utf_view / 39537701882 100%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 39537701882

bemanproject / utf_view / 39537701882
100%

README BADGES
x