Apply this patch to String::Similarity 0.02 from CPAN to turn it into
String::Unicode::Similarity, required by the MusicBrainz server.

Only in String-Similarity-0.02: blib
diff -r -u t/String-Similarity-0.02/Changes String-Similarity-0.02/Changes
--- t/String-Similarity-0.02/Changes	2000-11-16 00:30:35.000000000 +0000
+++ String-Similarity-0.02/Changes	2003-01-07 19:41:04.000000000 +0000
@@ -1,4 +1,4 @@
-Revision history for Perl extension String::Similarity
+Revision history for Perl extension String::Unicode::Similarity
 
 0.02  Thu Nov 16 01:30:16 CET 2000
 	- added minimum_similarity limit to speed up some searches
diff -r -u t/String-Similarity-0.02/fstrcmp.c String-Similarity-0.02/fstrcmp.c
--- t/String-Similarity-0.02/fstrcmp.c	2000-11-16 00:35:28.000000000 +0000
+++ String-Similarity-0.02/fstrcmp.c	2003-01-07 19:33:22.000000000 +0000
@@ -50,7 +50,7 @@
 struct string_data
 {
   /* The string to be compared. */
-  const char *data;
+  const INT32 *data;
 
   /* The length of the string to be compared. */
   int data_length;
@@ -161,8 +161,8 @@
 {
   int *const fd = fdiag;	/* Give the compiler a chance. */
   int *const bd = bdiag;	/* Additional help for the compiler. */
-  const char *const xv = string[0].data;	/* Still more help for the compiler. */
-  const char *const yv = string[1].data;	/* And more and more . . . */
+  const INT32 *const xv = string[0].data;	/* Still more help for the compiler. */
+  const INT32 *const yv = string[1].data;	/* And more and more . . . */
   const int dmin = xoff - ylim;	/* Minimum valid diagonal. */
   const int dmax = xlim - yoff;	/* Maximum valid diagonal. */
   const int fmid = xoff - yoff;	/* Center diagonal of top-down search. */
@@ -485,8 +485,8 @@
      int ylim;
      int minimal;
 {
-  const char *const xv = string[0].data;	/* Help the compiler.  */
-  const char *const yv = string[1].data;
+  const INT32 *const xv = string[0].data;	/* Help the compiler.  */
+  const INT32 *const yv = string[1].data;
 
   if (string[1].edit_count + string[0].edit_count > max_edits)
     return;
@@ -560,7 +560,7 @@
 	fstrcmp - fuzzy string compare
 
    SYNOPSIS
-	double fstrcmp(const char *, const char *, double);
+	double fstrcmp(const INT32 *, const INT32 *, int, int, double);
 
    DESCRIPTION
 	The fstrcmp function may be used to compare two string for
@@ -574,7 +574,7 @@
 	similar.  */
 
 double
-fstrcmp (const char *string1, const char *string2, double minimum)
+fstrcmp (const INT32 *string1, const INT32 *string2, int length1, int length2, double minimum)
 {
   int i;
 
@@ -584,9 +584,9 @@
 
   /* set the info for each string.  */
   string[0].data = string1;
-  string[0].data_length = strlen (string1);
+  string[0].data_length = length1;
   string[1].data = string2;
-  string[1].data_length = strlen (string2);
+  string[1].data_length = length2;
 
   /* short-circuit obvious comparisons */
   if (string[0].data_length == 0 && string[1].data_length == 0)
diff -r -u t/String-Similarity-0.02/fstrcmp.h String-Similarity-0.02/fstrcmp.h
--- t/String-Similarity-0.02/fstrcmp.h	2000-11-16 00:28:38.000000000 +0000
+++ String-Similarity-0.02/fstrcmp.h	2003-01-07 19:32:10.000000000 +0000
@@ -22,6 +22,8 @@
 
 #define PARAMS(proto) proto
 
-double fstrcmp (const char *__s1, const char *__s2, double __minimum);
+#define INT32 long
+
+double fstrcmp (const INT32 *__s1, const INT32 *__s2, int __l1, int __l2, double __minimum);
 
 #endif
Only in String-Similarity-0.02: Makefile
diff -r -u t/String-Similarity-0.02/Makefile.PL String-Similarity-0.02/Makefile.PL
--- t/String-Similarity-0.02/Makefile.PL	2000-04-07 07:09:10.000000000 +0100
+++ String-Similarity-0.02/Makefile.PL	2003-01-10 23:21:58.000000000 +0000
@@ -6,7 +6,20 @@
 	            COMPRESS	=> 'gzip -9v',
 	            SUFFIX	=> '.gz',
 	           },
-    NAME => "String::Similarity",
+    NAME => "String::Unicode::Similarity",
     VERSION_FROM => "Similarity.pm",
 );
 
+print <<EOF;
+
+Please check that the type named in this '#define' occupies exactly four bytes
+on your system:
+EOF
+print "\t";
+system "grep '#define INT32' fstrcmp.h";
+print <<EOF;
+If it's not correct, please edit 'fstrcmp.h' and fix it before continuing with
+'make'.
+
+EOF
+
Only in String-Similarity-0.02: pm_to_blib
diff -r -u t/String-Similarity-0.02/README String-Similarity-0.02/README
--- t/String-Similarity-0.02/README	2000-11-16 00:51:11.000000000 +0000
+++ String-Similarity-0.02/README	2003-01-07 19:41:04.000000000 +0000
@@ -1,8 +1,8 @@
 NAME
-    String::Similarity - calculate the similarity of two strings
+    String::Unicode::Similarity - calculate the similarity of two strings
 
 SYNOPSIS
-     use String::Similarity;
+     use String::Unicode::Similarity;
 
      $similarity = similarity $string1, $string2;
      $similarity = similarity $string1, $string2, $limit;
Only in String-Similarity-0.02: Similarity.bs
Only in String-Similarity-0.02: Similarity.c
Only in String-Similarity-0.02: Similarity.o
diff -r -u t/String-Similarity-0.02/Similarity.pm String-Similarity-0.02/Similarity.pm
--- t/String-Similarity-0.02/Similarity.pm	2000-11-16 00:40:06.000000000 +0000
+++ String-Similarity-0.02/Similarity.pm	2003-01-10 23:24:13.000000000 +0000
@@ -1,10 +1,10 @@
 =head1 NAME
 
-String::Similarity - calculate the similarity of two strings
+String::Unicode::Similarity - calculate the similarity of two strings
 
 =head1 SYNOPSIS
 
- use String::Similarity;
+ use String::Unicode::Similarity;
 
  $similarity = similarity $string1, $string2;
  $similarity = similarity $string1, $string2, $limit;
@@ -15,16 +15,17 @@
 
 =cut
 
-package String::Similarity;
+package String::Unicode::Similarity;
 
 require DynaLoader;
 
+use 5.008;
 $VERSION = 0.02;
 @ISA = qw/Exporter DynaLoader/;
 @EXPORT = qw(similarity);
-@EXPORT_OK = qw(fstrcmp);
+#@EXPORT_OK = qw(fstrcmp);
 
-bootstrap String::Similarity $VERSION;
+bootstrap String::Unicode::Similarity $VERSION;
 
 =item $factor = similarity $string1, $string2, [$limit]
 
@@ -48,7 +49,21 @@
 =cut
 
 # out of historical reasons, I prefer "fstrcmp" as the original name.
-*similarity = *fstrcmp;
+#*similarity = *fstrcmp;
+
+sub similarity($$)
+{
+	my ($s1, $s2) = @_;
+	
+	use Encode qw( encode );
+
+	fstrcmp(
+		encode("UCS-4BE", $s1),
+		encode("UCS-4BE", $s2),
+		length($s1),
+		length($s2),
+	);
+}
 
 1;
 
@@ -74,5 +89,6 @@
  modified by Peter Miller <pmiller@agso.gov.au> and Marc Lehmann
  <pcg@goof.com>).
 
+ Unicode (UCS-4) patch by Dave Evans <http://rudolf.org.uk/>
 
 
diff -r -u t/String-Similarity-0.02/Similarity.xs String-Similarity-0.02/Similarity.xs
--- t/String-Similarity-0.02/Similarity.xs	2000-11-16 00:28:17.000000000 +0000
+++ String-Similarity-0.02/Similarity.xs	2003-01-07 19:51:13.000000000 +0000
@@ -5,12 +5,13 @@
 #include "fstrcmp.h"
 #include "fstrcmp.c"
 
-MODULE = String::Similarity		PACKAGE = String::Similarity
+MODULE = String::Unicode::Similarity		PACKAGE = String::Unicode::Similarity
 
 double
-fstrcmp(s1, s2, minimum_similarity = 0)
+fstrcmp(s1, s2, l1, l2, minimum_similarity = 0)
 	char *	s1
         char *	s2
+	int l1
+	int l2
         double	minimum_similarity
-        PROTOTYPE: @
 
diff -r -u t/String-Similarity-0.02/t/00_load.t String-Similarity-0.02/t/00_load.t
--- t/String-Similarity-0.02/t/00_load.t	2000-04-07 07:12:01.000000000 +0100
+++ String-Similarity-0.02/t/00_load.t	2003-01-07 19:52:08.000000000 +0000
@@ -1,6 +1,6 @@
-BEGIN { $| = 1; print "1..4\n"; }
+BEGIN { $| = 1; print "1..5\n"; }
 END {print "not ok 1\n" unless $loaded;}
-use String::Similarity;
+use String::Unicode::Similarity;
 $loaded = 1;
 print "ok 1\n";
 
@@ -8,4 +8,5 @@
 my $s = similarity("this should be same the", "this should be the same");
 print $s > 0.825 && $s < 0.827 ? "" : "not ", "ok 3\n";
 print similarity("A", "B") == 0 ? "" : "not ", "ok 4\n";
+print similarity(chr(0x100), chr(0x101)) == 0 ? "" : "not ", "ok 5\n";