Added a small script to help when creating charsets code.

[squirrelmail.git] / po / charsetconvert.pl

diff --git a/po/charsetconvert.pl b/po/charsetconvert.pl
new file mode 100755 (executable)
index 0000000..7443b6d
--- /dev/null
+++ b/po/charsetconvert.pl
@@ -0,0 +1,29 @@
+#! /usr/bin/perl
+
+# This script takes an Unicode character map as input and outputs
+# conversion code for SquirrelMail charset support.
+
+# This code is placed in the Public Domain. Written by Gustav Foseid,
+# gustavf@squirrelmail.org
+
+$min = 160;
+
+while (<>) {
+  chop;
+
+  unless (/^\#/) {
+
+    ($orig_text, $unicode_text, $dummy, $name) = split /\t/;
+       
+    # oct does not only do what it's name suggest. If a string starts 
+    # with 0x it is interpreted as a hexadecimal value.
+    $orig = oct $orig_text;
+    $unicode = oct $unicode_text;
+       
+    if ($orig >= $min) {
+      print "      # $name\n";
+      printf ('      $string = str_replace("\%o", "&#%d"), $string);'."\n",
+             $orig, $unicode);
+    }
+  } 
+}