Crystal: Regression in `BigDecimal#to_big_i`

Created on 12 Feb 2020  路  8Comments  路  Source: crystal-lang/crystal

Regression introduced in Crystal 0.31.0 (#8120)

require "big"

"-100.000".to_big_d.to_big_i # => -100000

Before: https://carc.in/#/r/8jrq
After: https://carc.in/#/r/8jro

/cc @bcardiff

bug topicnumeric
Source
picture Sija
👍2

Most helpful comment

Other broken case: "-1.1".to_big_d.to_big_i # => -10. #8790 doesn't fix this.

picture straight-shoota on 12 Feb 2020
👍3 👀1

All 8 comments

Other broken case: "-1.1".to_big_d.to_big_i # => -10. #8790 doesn't fix this.

straight-shoota picture straight-shoota on 12 Feb 2020
👍3 👀1

@straight-shoota Force-pushed amended commit with the fix.

Sija picture Sija on 12 Feb 2020

I scheduled the PR for 0.34.0. It could not review it yet.

bcardiff picture bcardiff on 14 Feb 2020
😕1

Would've been nice to include it in 0.33.0 since it's already fixed. But I guess it's fine.

straight-shoota picture straight-shoota on 14 Feb 2020
👍1

@bcardiff It's been reviewed already and there are tests, so why hold it back...? :/

Sija picture Sija on 14 Feb 2020

@straight-shoota Thanks for merging! I guess it should have the 0.34.0 milestone added too.

Sija picture Sija on 18 Feb 2020

Usually only PRs are milestoned (see previous ones) to keep the noise down. Both PR and issue relate to the same bug or feature and we only need to have one representation in the milestone.

straight-shoota picture straight-shoota on 18 Feb 2020
👍1

@straight-shoota Good to know, makes sense! I got mistaken by this:

image

Sija picture Sija on 18 Feb 2020
Was this page helpful?
0 / 5 - 0 ratings

Related issues

oprypin picture oprypin  路  3Comments
RX14 picture RX14  路  3Comments
Papierkorb picture Papierkorb  路  3Comments
jhass picture jhass  路  3Comments
relonger picture relonger  路  3Comments