Changeset 9125

Timestamp:

2005-02-22T10:21:35+13:00 (19 years ago)

Author:

mdewsnip

Message:

Added a substr function to unicode.pm that should work correctly on Unicode strings. It might be slow, but it should be more reliable than "use utf8; substr(...)", which has never worked for me and causes problems on Windows with Perl 5.6.

Location:

trunk/gsdl/perllib

Files:

• plugins/HTMLPlug.pm (modified) (1 diff)
• unicode.pm (modified) (1 diff)

trunk/gsdl/perllib/plugins/HTMLPlug.pm

@@ -630 +630 @@
     $tmptext =~ s/\s+$//;
     $tmptext =~ s/\s+/ /gs;
-    # with perl 5.6 at least, substr might segment at a multi-byte char...
-    use utf8;
-    $tmptext = substr ($tmptext, 0, $size);
+    $tmptext = &unicode::substr ($tmptext, 0, $size);
     $tmptext =~ s/\s\S*$/…/; # adds an ellipse (...)
     $doc_obj->add_utf8_metadata ($thissection, "First$size", $tmptext);

trunk/gsdl/perllib/unicode.pm

@@ -527 +527 @@
 }
 
+
+sub substr
+{
+    my ($utf8_string, $offset, $length) = @_;
+
+    my @unicode_string = @{&utf82unicode($utf8_string)};
+    my $unicode_string_length = scalar(@unicode_string);
+
+    my $substr_start = $offset;
+    if ($substr_start >= $unicode_string_length) {
+    return "";
+    }
+
+    my $substr_end = $offset + $length - 1;
+    if ($substr_end >= $unicode_string_length) {
+    $substr_end = $unicode_string_length - 1;
+    }
+
+    my @unicode_substring = @unicode_string[$substr_start..$substr_end];
+    return &unicode2utf8(\@unicode_substring);
+}
+
+
 1;