[6247] unicode string_width implementation

[Gavin D. Smith]

Revision: 6247
          http://svn.sv.gnu.org/viewvc/?view=rev&root=texinfo&revision=6247
Author:   gavin
Date:     2015-05-04 22:06:03 +0000 (Mon, 04 May 2015)
Log Message:
-----------
unicode string_width implementation

Modified Paths:
--------------
    trunk/ChangeLog
    trunk/tp/Texinfo/Convert/Unicode.pm

Modified: trunk/ChangeLog
===================================================================
--- trunk/ChangeLog     2015-05-04 18:13:48 UTC (rev 6246)
+++ trunk/ChangeLog     2015-05-04 22:06:03 UTC (rev 6247)
@@ -1,3 +1,8 @@
+2015-05-03  Gavin Smith  <address@hidden>
+
+       * tp/Texinfo/Convert/Unicode.pm (string_width): Calculate string
+       width using "unpack" function.
+
 2015-05-03  Karl Berry  <address@hidden>
 
        * pretest 5.9.91.

Modified: trunk/tp/Texinfo/Convert/Unicode.pm
===================================================================
--- trunk/tp/Texinfo/Convert/Unicode.pm 2015-05-04 18:13:48 UTC (rev 6246)
+++ trunk/tp/Texinfo/Convert/Unicode.pm 2015-05-04 22:06:03 UTC (rev 6247)
@@ -1459,6 +1459,17 @@
 {
   my $string = shift;
 
+  $string =~ s/\p{InFullwidth}/\x{02}/g;
+  $string =~ s/\pM/\x{00}/g;
+  $string =~ s/\p{IsPrint}/\x{01}/g;
+  $string =~ s/\p{IsSpace}/\x{01}/g;
+  $string =~ s/[^\x{01}\x{02}]/\x{00}/g;
+
+  # This sums up the byte values of the bytes in $string, which now are
+  # all either 0, 1 or 2.  This is faster.  The original, more readable
+  # version is below.
+  return unpack("U0%32A*", $string);
+
   if (! defined($string)) {
     Carp::cluck();
   }