From: Steve Hancock <perltidy@users.sourceforge.net>
Date: Tue, 15 Aug 2023 05:15:45 +0000 (-0700)
Subject: optimize pretoken search order
X-Git-Tag: 20230701.03~16
X-Git-Url: https://git.donarmstrong.com/?a=commitdiff_plain;h=c2cb582a916fbe78fb034d474c9d15058d291c70;p=perltidy.git

optimize pretoken search order

Benchmarking shows that this simple change speeds up perltidy about 0.5%
---

diff --git a/lib/Perl/Tidy/Tokenizer.pm b/lib/Perl/Tidy/Tokenizer.pm
index d46e9b20..9e38e794 100644
--- a/lib/Perl/Tidy/Tokenizer.pm
+++ b/lib/Perl/Tidy/Tokenizer.pm
@@ -9968,19 +9968,18 @@ sub pre_tokenize {
 
     do {
 
-        # whitespace
+        # whitespace - this must come before \W
         if ( $str =~ /\G(\s+)/gc ) { push @type, 'b'; }
 
-        # numbers
-        # note that this must come before words!
+        # non-whitespace single-character punctuation
+        elsif ( $str =~ /\G(\W)/gc ) { push @type, $1; }
+
+        # sequence of digits - this must come before \w
         elsif ( $str =~ /\G(\d+)/gc ) { push @type, 'd'; }
 
-        # words
+        # words not starting with a digit
         elsif ( $str =~ /\G(\w+)/gc ) { push @type, 'w'; }
 
-        # single-character punctuation
-        elsif ( $str =~ /\G(\W)/gc ) { push @type, $1; }
-
         # that's all..
         else {
             return ( \@tokens, \@token_map, \@type );