| Summary: | dev-lang/ruby-1.8.7: securerandom with UTF-8 produces output incompatible to 1.9 | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Diego Elio Pettenò (RETIRED) <flameeyes> |
| Component: | New packages | Assignee: | Gentoo Ruby Team <ruby> |
| Status: | RESOLVED WORKSFORME | ||
| Severity: | normal | ||
| Priority: | High | ||
| Version: | unspecified | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
This is not yet fixed in ruby 1.8.7_p173 Looking at this again I wonder if this isn't a bug in uuidtools: perhaps it should check for the $KCODE stuff and change its code accordingly. After all you've requested all string to be UTF-8 with the -K option. In any case 1.8.7_p302 and ree18 1.8.7.2010.02-r1 both still have this behaviour. jruby runs out of heap space. Thinking about this a bit more it seems to be that this is by design due to ruby 1.8's lack of proper support for UTF-8. |
% irb18 -KU irb(main):001:0> require 'securerandom' => true irb(main):003:0> SecureRandom.random_bytes(32).split("").size => 22 Without UTF-8 strings, or with Ruby 1.9, the problem is not there and it always returns 32. This actually causes uuidtools to break.