6885
89%

Ran 28 Sep 2026 08:21AM UTC

Jobs 1

Files 454

Run time 2min

Badge

Embed ▾

Committed 28 Sep 2026 08:17AM UTC coverage: 89.433% (-0.006%) from 89.439%

Build # 6885

Build Type

push

github

Committed by

bobzhang

Commit Message

refactor(builtin,float): deprecate Double::until and Float::until

Both build each element as the previous one plus step, so rounding error
accumulates (0.0.until(1.0, step=0.1) yields 11 elements). Once adding
step no longer changes the element (a sub-precision step, an infinite
start, or infinity reached with an inclusive infinite end), it is
yielded forever. Rather than fix them, point users at @range.iter, which
computes element k as start + k * step and stops when the next element
would not advance.

The functions move unchanged to deprecated.mbt. Their docs say what is
wrong, how @range.iter's results differ (accumulated error, repeated
elements, and a zero step with from == to), and how to hit decimal end
points exactly. The float README now uses @range.iter. Existing tests
keep pinning the deprecated behaviour under #warnings("-deprecated").

Int::until and Int64::until stay: integer steps are exact.

Codex review (2 rounds): round 1 found the migration notes missed the
infinity cases that repeat forever; round 2 approved.

Signed-off-by: Codex CLI <codex@openai.com>
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>

Coverage Stats

16673 of 18643 relevant lines covered (89.43%)

331721.09 hits per line

Jobs

ID	Job ID	Ran	Files	Coverage
1	6885.1	28 Sep 2026 08:21AM UTC	454	89.43	GitHub Action Run

moonbitlang / core / 6885
89%

README BADGES
x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 6885

moonbitlang / core / 6885 89%

README BADGES x

Markdown

Textile

RDoc

HTML

Rst

Jobs

Source Files on build 6885

moonbitlang / core / 6885
89%

README BADGES
x