The following patch fixes krakozyabry (garbled characters due to incorrect charset detection) in the title of the document in translation into XML DocBook format. Unfortunately, this is a hack - after applying this patch, the conversion problem of conversion moves to the documents with titles in the Western European encoding (instead for ones in the Cyrillic).
--- chartrans.c.orig 2005-01-11 22:44:46.000000000 +0300
+++ ../chartrans.c 2010-10-05 22:51:30.000000000 +0400
@@ -359,7 +359,7 @@
break;
case encoding_latin_1:
default:
- usCharSet = usCp1252;
+ usCharSet = usCp1251;
break;
}
}
@@ -367,7 +367,7 @@
if (usChar >= 0x80 && usChar <= 0x9f) {
/* Translate implementation defined characters */
usChar = usCharSet[usChar - 0x80];
- } else if (iWordVersion < 8 && usChar >= 0xa0 && usChar <= 0xff) {
+ } else if (iWordVersion <= 8 && usChar >= 0xa0 && usChar <= 0xff) {
/* Translate old character set to Unixcode */
usChar = usCharSet[usChar - 0x80];
}