Bug #6723: Global variable scoping problem.

• Status: Closed
• Assignee: drbrain (Eric Hodel)
• Target version: 2.0.0
• ruby -v: -

[iblue (Markus Fenske)]

This issue exists from 1.8 to ruby-head.

See the following example:

def foobar
"Foobar".match(/^(\w*)/)
puts $1.inspect
end

foobar
puts $1.inspect
puts global_variables.inspect

This demonstrates that $1 does not behave like a global variable. However it's listed by Kernel.global_variables as a global variable. And it's named like a global variable.

This leads to some confusion sometimes. See for example http://stackoverflow.com/q/11441782/773690

It may not be a good idea to fix this inconsistent naming, because it may break severaly things. However the fact that $1 and friends are not global variables should be clearly documented. And by the way: Ruby != Perl.

Or maybe make $1 a real global variable?

[ko1 (Koichi Sasada)]

• Target version set to 2.0.0
• Category set to core
• Assignee set to matz (Yukihiro Matsumoto)

[nobu (Nobuyoshi Nakada)]

• Category changed from core to doc
• Status changed from Open to Assigned
• Assignee changed from matz (Yukihiro Matsumoto) to drbrain (Eric Hodel)
• ruby -v changed from ruby 1.9.3p194 to -

[drbrain (Eric Hodel)]

• Status changed from Assigned to Closed
• % Done changed from 0 to 100

This issue was solved with changeset r36526.
Markus, thank you for reporting this issue.
Your contribution to Ruby is greatly appreciated.
May Ruby be with you.

---

• re.c (rb_reg_s_last_match): Update $~ to reference Regexp
  documentation about "special global variables". [Bug #6723]

[drbrain (Eric Hodel)]

These are described as "Special global variables" and are described as "thread-local and method-local" in the Regexp documentation. I updated a reference to them in re.c, but the existing documentation should be sufficient.