Front page | perl.perl5.porters |
Postings from February 2012
"Eric Brine via RT" <perlbug-followup@perl.org> wrote
on Sat, 04 Feb 2012 16:19:41 PST:
> Also note that "e" is not mentioned in perlre. The regex flags
> are (as listed in perlre): m, s, i, x, p, g, c, a, d, l, u.
That's incorrect, and potentially misleading.
With blank lines removed for brevity, these are
the only pattern modifiers:
=head2 Pattern Modifiers
=begin table picture Regular expression modifiers
=headrow
=row
=cell Modifier
=cell Meaning
=bodyrows
=row
=cell C</i>
=cell Ignore alphabetic case distinctions (case insensitive).
=row
=cell C</s>
=cell Let C<.> also match newline.
=row
=cell C</m>
=cell Let C<^> and C<$> also match next to embedded C<\n>.
=row
=cell C</x>
=cell Ignore (most) whitespace and permit comments in pattern.
=row
=cell
=row
=cell C</o>
=cell Compile pattern once only.
=row
=cell C</p>
=cell Preserve C<${^PREMATCH}>, C<${^MATCH}>, and C<${^POSTMATCH}> variables.
=row
=cell
=row
=cell C</d>
=cell Dual ASCII–Unicode mode charset behavior (old default)
=row
=cell C</a>
=cell ASCII charset behavior
=row
=cell C</u>
=cell Unicode charset behavior (new default)
=row
=cell C</l>
=cell the run-time locale’s charset behavior (default under R<use locale>)
=end table
That's all. Those apply to patterns.
These flags, however, apply to the match operator, which is different:
=head2 The m// Operator (Matching)
=begin table picture m// Modifiers
=headrow
=row
=cell Modifier
=cell Meaning
=bodyrows
=row
=cell C</i>
=cell Ignore alphabetic case.
=row
=cell C</m>
=cell Let C<^> and C<$> also match next to embedded C<\n>.
=row
=cell C</s>
=cell Let C<.> also match newline.
=row
=cell C</x>
=cell Ignore (most) whitespace and permit comments in pattern.
=row
=cell C</o>
=cell Compile pattern once only.
=row
=cell C</p>
=cell Preserve the matched string.
=row
=cell
=row
=cell C</d>
=cell Dual ASCII–Unicode mode charset behavior (old default).
=row
=cell C</u>
=cell Unicode charset behavior (new default).
=row
=cell C</a>
=cell ASCII charset behavior
=row
=cell C</l>
=cell The run-time locale’s charset behavior (default under R<use locale>).
=row
=cell
=row
=cell C</g>
=cell Globally find all matches.
=row
=cell C</cg>
=cell Allow continued search after failed C</g> match.
=end table
Here are the s/// flags:
=head2 The s/// Operator (Substitution)
=begin table picture s/// Modifiers
=headrow
=row
=cell Modifier
=cell Meaning
=bodyrows
=row
=cell C</i>
=cell Ignore alphabetic case (when matching).
=row
=cell C</m>
=cell Let C<^> and C<$> also match next to embedded C<\n>.
=row
=cell C</s>
=cell Let C<.> also match newline.
=row
=cell C</x>
=cell Ignore (most) whitespace and permit comments in pattern.
=row
=cell C</o>
=cell Compile pattern once only.
=row
=cell C</p>
=cell Preserve the matched string.
=row
=cell
=row
=cell C</d>
=cell Dual ASCII–Unicode mode charset behavior (old default).
=row
=cell C</u>
=cell Unicode charset behavior (new default).
=row
=cell C</a>
=cell ASCII charset behavior.
=row
=cell C</l>
=cell The run-time locale’s charset behavior (default under R<use locale>).
=row
=cell
=row
=cell C</g>
=cell Replace globally, that is, all occurrences.
=row
=cell C</r>
=cell Return substitution and leave the original string untouched.
=row
=cell C</e>
=cell Evaluate the right side as an expression.
=end table
And tr/// has its own as well:
=begin table picture tr/// Modifiers
=headrow
=row
=cell Modifier
=cell Meaning
=bodyrows
=row
=cell C</c>
=cell Complement R<SEARCHLIST>.X</c pattern modifier:c5 pattern>
=row
=cell C</d>
=cell Delete found but unreplaced characters.X</d pattern modifier:d pattern>
=row
=cell C</s>
=cell Squash duplicate replaced characters.X</s pattern modifier:s pattern>
=row
=cell C</r>
=cell Return transliteration and leave the original string untouched.
=end table
See? Four distinct sets of flags.
> (I find this very odd that p, g and c are regex flags instead
> of match substitute operator flags, but they are are.)
You should find it odd, because they aren't.
$ perl -E 'use re "/e"; say "ok"'
Unknown regular expression flag "e" at -e line 1
ok
$ perl -E 'use re "/g"; say "ok"'
Unknown regular expression flag "g" at -e line 1
ok
$ perl -E 'use re "/c"; say "ok"'
Unknown regular expression flag "c" at -e line 1
ok
$ perl -E 'use re "/p"; say "ok"'
ok
What's going on with the first three is that it isn't properly dying.
$ perl -Mwarnings=FATAL,all -E 'use re "/g"; say "ok"'
Unknown regular expression flag "g" at -e line 1
ok
That's a bug.
What's going on with the last one is more subtle. /p can be
embedded in a pattern with (?p), but it cannot be turned off
once turned on.
Anyway, probably perlre should be clearer (read: more correct)
about the s/uper/man/mxyzptlk flags, and that jazz.
--tom
Thread Previous
|
Thread Next