When I tried to re-create
aspell-bn (bn-BD) i.e. of Bengali(BD) language spellchecker with help of this link for aspell6 by compiling like the following
$ cvs -z3 -d:pserver:firstname.lastname@example.org:/sources/aspell co aspell-lang $ cd aspell-lang $ ./pre bn u-beng $ cd bn $ ./proc create
This creates configure script and Makefile.pre and finally copying the word list file to working directory and named it as bn.wl Word list file which contains list of words separated by new lines. And then
$ ./configure $ make
(/usr/sbin/prezip-bin -d < bn.cwl | /usr/sbin/aspell --lang=bn create master ./bn.rws)
But when I run "make", it prints many error messages and warnings like the following 2 types (The Unicode code points U+09CE and U+FEFF are unsupported), though according to u-beng.txt U+09CE is clearly within the defined range.
Warning: The string "হৃৎযন্ত্র" is invalid. The Unicode code point U+09CE is unsupported. Skipping string. Warning: The string "ক্ল্যারিজ" is invalid. The Unicode code point U+FEFF is unsupported. Skipping string.
After all of these, I do get a bn.rws file but it does not include all the words listed in bn.wl.
I am using parabola-arch x86-64
pacman -Qi aspell returns
Name : aspell Version : 0.60.6.1-5 Description : A spell checker designed to eventually replace Ispell Architecture : x86_64 URL : http://aspell.net/ Licenses : LGPL Groups : None Provides : None Depends On : gcc-libs ncurses Optional Deps : perl: to import old dictionaries [installed]
If you can help me correct my mistakes then please. I am also including the error.log.
And a pull request.