wrong charset for Polish book plugins

[Halibutt]

GCStar 1.50, Windows (both XP Home and Vista Business checked)

The three Polish book searches have a huge problem with Polish diacritics. In Merlin and Mareno the title search works just fine (although in Mareno barely any field gets filled, no cover picture for instance) - unless the title includes a Polish diacrite (ś, ć, ź, ń and such). In that case the search is completely broken.

Let's take this book for instance: "Życie erotyczne Anais Nin". Merlin finds it fine (http://merlin.pl/Zycie-erotyczne-Anais-Nin/browse/search.html?phrase=%AFycie+erotyczne+Anais+Nin&section=&x=0&y=0&place=0+simple&offer=O&sort=rank&page=1), so does Mareno (http://www.mareno.pl/rezultat.php?tytul=%AFycie+erotyczne&dep=1). Yet when I try to find that book from within GCStar, the search fails. No titles with a Polish diacrite return any searches.

What's more, even the books that are being found have fancy characters instead of Polish diacrites. (æ instead of ć, ± instead of ą, ê instead of ę, ¶ instead of ś and so on). It looks as if GCStar was predefined to use Latin-1 encoding even though the plugins use different encodings. For instance within GCMerlin.pm it is specifically told to use:

    sub getCharset
    {
        my $self = shift;
        #return "UTF-8";
        return "ISO-8859-2";
    }

Yet it seems that the resulting strings are downloaded using latin-1 instead of iso-8859-2.
Cheers

BTW, also filed here: http://forums.gcstar.org/viewtopic.php?pid=5927#p5927

[omvsj (omvsj)]

I'm a Latinamerican user.

In Windows XP, whenever I try to use folders with UTF8 characters, GC can't save them correctly and it use other characters instead:
- ./Imágenes -> ./Imágenes
- ./Películas -> ./Películas

I had to rename my folders to non-UTF8, but I have more than 500 movies and many of them have UTF8 characters. I think this bug is related to global Charset.