RegExp deviation from Perl 5 - By Design or Spec Bug?
2011/1/19 Dave Fugate <dfugate at microsoft.com>:
I’ve come across the following RegExp scenario:
/\s^/m.test("\n")
which results in true, yet the equivalent in Perl 5:
"\n" =~ m/\s^/m
yields false.
It seems that the JavaScript implementations I tested against are correctly implementing ES5. That is, 15.10.2.6, step 4 states that if the character before the current endIndex is a newline then return true. After matching the \s, the previous character is indeed a newline, so the assertion succeeds.
Am I misinterpreting ES5 here and the implementations are exhibiting incorrect behavior or is the spec not matching Perl 5? If it’s the latter, would this be considered a spec defect as ES5’s RegExp feature is roughly a subset of Perl 5?
There are quite a few places where ES differs from perl.
Why is it that perl does not match here?
On Wed, Jan 19, 2011 at 2:13 PM, Erik Corry <erik.corry at gmail.com> wrote:
2011/1/19 Dave Fugate <dfugate at microsoft.com>:
I’ve come across the following RegExp scenario:
/\s^/m.test("\n")which results in true, yet the equivalent in Perl 5:
"\n" =~ m/\s^/m
yields false.
It seems that the JavaScript implementations I tested against are correctly implementing ES5. That is, 15.10.2.6, step 4 states that if the character before the current endIndex is a newline then return true. After matching the \s, the previous character is indeed a newline, so the assertion succeeds.
Am I misinterpreting ES5 here and the implementations are exhibiting incorrect behavior or is the spec not matching Perl 5? If it’s the latter, would this be considered a spec defect as ES5’s RegExp feature is roughly a subset of Perl 5?
There are quite a few places where ES differs from perl.
Why is it that perl does not match here?
If you consider perlre perldoc to be a spec, then Perl5 makes an explicit exemption for newline which is the last character:
You may, however, wish to treat a string as a multi-line buffer, such that
the "^" will match after any newline within the string (except if the newline is the last character in the string), and "$" will match before any newline.
Perl6 deprecates the /m multiline modifier, changes the semantics of ^ and $ and introduces ^^ and $$.
I've come across the following RegExp scenario: /\s^/m.test("\n")
which results in true, yet the equivalent in Perl 5: "\n" =~ m/\s^/m
yields false.
It seems that the JavaScript implementations I tested against are correctly implementing ES5. That is, 15.10.2.6, step 4 states that if the character before the current endIndex is a newline then return true. After matching the \s, the previous character is indeed a newline, so the assertion succeeds.
Am I misinterpreting ES5 here and the implementations are exhibiting incorrect behavior or is the spec not matching Perl 5? If it's the latter, would this be considered a spec defect as ES5's RegExp feature is roughly a subset of Perl 5?
I've come across the following RegExp scenario: /\s^/m.test("\n") which results in true, yet the equivalent in Perl 5: "\n" =~ m/\s^/m yields false. It seems that the JavaScript implementations I tested against are correctly implementing ES5. That is, 15.10.2.6, step 4 states that if the character before the current endIndex is a newline then return true. After matching the \s, the previous character is indeed a newline, so the assertion succeeds. Am I misinterpreting ES5 here and the implementations are exhibiting incorrect behavior *or* is the spec not matching Perl 5? If it's the latter, would this be considered a spec defect as ES5's RegExp feature is roughly a subset of Perl 5? Thanks! Dave -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://mail.mozilla.org/pipermail/es-discuss/attachments/20110119/dfa91a5c/attachment.html>