diff options
author | Paul Baltescu <pauldb89@gmail.com> | 2013-04-24 17:18:10 +0100 |
---|---|---|
committer | Paul Baltescu <pauldb89@gmail.com> | 2013-04-24 17:18:10 +0100 |
commit | ba206aaac1d95e76126443c9e7ccc5941e879849 (patch) | |
tree | 13a918da3f3983fd8e4cb74e7cdc3f5e1fc01cd1 /word-aligner | |
parent | c2aede0f19b7a5e43581768b8c4fbfae8b92c68c (diff) | |
parent | db960a8bba81df3217660ec5a96d73e0d6baa01b (diff) |
Merge branch 'master' of https://github.com/redpony/cdec
Diffstat (limited to 'word-aligner')
-rwxr-xr-x | word-aligner/ortho-norm/ru.pl | 44 |
1 files changed, 44 insertions, 0 deletions
diff --git a/word-aligner/ortho-norm/ru.pl b/word-aligner/ortho-norm/ru.pl new file mode 100755 index 00000000..34452d06 --- /dev/null +++ b/word-aligner/ortho-norm/ru.pl @@ -0,0 +1,44 @@ +#!/usr/bin/perl -w +use strict; +use utf8; +binmode(STDIN,":utf8"); +binmode(STDOUT,":utf8"); +while(<STDIN>) { + $_ = uc $_; + s/А/a/g; + s/І/i/g; + s/Б/b/g; + s/В/v/g; + s/Г/g/g; + s/Д/d/g; + s/Е/e/g; + s/Ж/zh/g; + s/З/z/g; + s/И/i/g; + s/Й/i/g; + s/К/k/g; + s/Л/l/g; + s/М/m/g; + s/Н/n/g; + s/О/o/g; + s/П/p/g; + s/Р/r/g; + s/С/s/g; + s/Т/t/g; + s/У/u/g; + s/Ф/f/g; + s/Х/kh/g; + s/Ц/c/g; + s/Ч/ch/g; + s/Ш/sh/g; + s/Щ/shch/g; + s/Ъ//g; + s/Ы//g; + s/Ь//g; + s/Э/e/g; + s/Ю/yo/g; + s/Я/ya/g; + $_ = lc $_; + print; +} + |