Bug #8920: Exception when using a variable defined in end-of-line if statement

• Status: Rejected
• Assignee: -
• Target version: -
• ruby -v: ruby 2.0.0p247
• Backport: 1.9.3: UNKNOWN, 2.0.0: UNKNOWN

[jfredett (Joe Fredette)]

=begin
The following code raises an exception on each MRI version (exact ((%ruby -v%)) output attached at the end of this post):

class Test
def foo
puts "No Exception!"
end
end

begin
t.foo if t = Test.new
rescue Exception => e
STDERR.puts "Exception raised! ``#{e.class}: #{e}''"
exit 1
end

exit 0

The exception raised is:

NameError: undefined local variable or method `t' for main:Object

The output I expected was "No Exception."

This script was run on the following system:

$ rvm --version
#=> rvm 1.21.17 (stable) by Wayne E. Seguin [e-mail], Michal Papis [e-mail] [https://rvm.io/]

$ uname -a
#=> Darwin JGF.local 12.4.0 Darwin Kernel Version 12.4.0: Wed May 1 17:57:12 PDT 2013; root:xnu-2050.24.15~1/RELEASE_X86_64 x86_64 i386 MacBookPro8,2 Darwin

(For reference, this is a mid-2011 Macbook Pro)

and raises the unexpected exception on the following ruby versions:

ruby 1.8.7 (2012-02-08 patchlevel 358) [universal-darwin12.0]
ruby 1.9.3p448 (2013-06-27 revision 41675) [x86_64-darwin12.3.0]
ruby 2.0.0p247 (2013-06-27 revision 41674) [x86_64-darwin12.3.0]

It also appears to raise in JRuby, but not in Rubinius, though that may or may not be relevant for this bugtracker.

Please let me know if I need to provide any other details. I put 2.0.0 in the ((%ruby -v%)) field, but it is reproducible on each of the MRI versions I listed above.

=end

• Status changed from Open to Rejected

This is intended behaviour. Variables are defined in order of their appearance in the source.

I believe matz has rejected making variables defined in the condition of a postfix if before, saying he didn't want to introduce inconsistent behaviour just for this.

[nobu (Nobuyoshi Nakada)]

• Description updated (diff)

[jfredett (Joe Fredette)]

-- accidental doublepost --

[jfredett (Joe Fredette)]

Excellent, thanks charliesome for the fast response, and nobu for fixing my formatting. I'll report back to the rubinius project that this is, indeed, inconsistent behavior.