Since kmhtConvert v0.7.2, the pages in the HTML files coming from EMLs are translated in the user's language, provided it's supported by the EML file itself.
This is done in the following way: each HTML file contains a declaration of the encoding that should be used so that it is properly displayed. This declaration for the Greek language is:
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
kmhtConvert reads this "charset" value and if it matches the KDE "charset" setting, loads and writes to the HTML file, the messages from the program's translation file (kmhtconvert.mo). These messages are the "From:", "To:" etc. fields, the short names of the days and months, etc. If these messages are not translated (in kmhtconvert.po) they will appear in English.
One problem that arises from the way the "charset" value in the HTML file is used, is that different encoding names may be used for the same language. Eg. for the Greek language iso-8859-7 is (almost) equivelant to windows-1253. So kmhtConvert uses the following equivelance table so that it translates the messages properly.
If some of these charset aliases are not correct, or if you know of any more I should add, please tell me about it.