[Po4a-devel]Po4a dependency on libtext-wrapi18n-perl

Ok, I had another look on this one, and it looks to me that we could quite easily drop the dependency. The code of this library itself is very close of what I did in the wrapper part, I think. The first difficulty is to wrap where it's legal, ie after space or after every CJK char. For that, we could use the same trick than libtext-wrapi18n-perl: # Returns 1 for Chinese and Japanese characters. This means that # these characters allow line wrapping after this character even # without whitespaces because these languages don't use whitespaces # between words. # # Character must be given in UCS-4 codepoint value. sub _isCJ { my $u=shift; if ($u >= 0x3000 && $u <= 0x312f) { if ($u == 0x300a || $u == 0x300c || $u == 0x300e || $u == 0x3010 || $u == 0x3014 || $u == 0x3016 || $u == 0x3018 || $u == 0x301a) {return 0;} return 1; } # CJK punctuations, Hiragana, Katakana, Bopomofo if ($u >= 0x31a0 && $u <= 0x31bf) {return 1;} # Bopomofo if ($u >= 0x31f0 && $u <= 0x31ff) {return 1;} # Katakana extension if ($u >= 0x3400 && $u <= 0x9fff) {return 1;} # Han Ideogram if ($u >= 0xf900 && $u <= 0xfaff) {return 1;} # Han Ideogram if ($u >= 0x20000 && $u <= 0x2ffff) {return 1;} # Han Ideogram return 0; } <<<<<<<<<<<<<<<<<<<< Then, the second difficulty is to compute the char width in column right. Google finds this: http://www.cl.cam.ac.uk/~mgk25/ucs/wcwidth.c Reimplementing this in perl seems quite easy. But I'm not sure this source can be trusted, and I feel just too lazy to download the libc source to see how it's done in there... Bye, Mt.