Re: [Vimcdoc-translate] help with iconv

[Yi-an H. <yia...@gm...>]

Hi lang2,

For your list of problematic files, there are a few different cases.

Case 1, digraph, arabic, hebrew, mbyte, farsi, usr_24: foreign alphabets

utf-8 are designed to incorporate characters from different languages,
enc-cn/cp936/gb2312/gb18030 is not, so in a few special cases some
characters cannot be converted correctly and there are unfortunately no eas=
y
fix. In my opinion, utf-8 is the better form for translations for at least
those documents that discuss digraphs and a few foreign languages. Likely,
people who only read gb encodings will not care about these documents
anyway, but we never know for sure...

Add "-c" parameter to iconv can work around the problem and iconv would
succeed by not translating these characters. But it is better for us to
leave unfixable characters with question marks. So for gb translations some
manual work is needed, but i would still argue that the utf-8 versions
should be the primary ones.

Case 2, pi_netrw, pattern: small nuisance

pi_netrw.txt: the name of a French author near the end of the file
pattern.txt:  "a" with a hat on top of it (actually reverting to the
previous version fixes the problem. but I wonder whether people reading
utf-8 may feel the current way is cleaner)

Case 3, various, index: incorrect characters in use and they should be
fixed.

netbean used to have this problem too with two unusual dashes (-), a recent
update fixed it, thanks.

I can upload the fixes of case 2 and 3.

Willlis