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 /corpus/support/quote-norm.pl | |
| parent | c2aede0f19b7a5e43581768b8c4fbfae8b92c68c (diff) | |
| parent | db960a8bba81df3217660ec5a96d73e0d6baa01b (diff) | |
Merge branch 'master' of https://github.com/redpony/cdec
Diffstat (limited to 'corpus/support/quote-norm.pl')
| -rwxr-xr-x | corpus/support/quote-norm.pl | 3 |
1 files changed, 3 insertions, 0 deletions
diff --git a/corpus/support/quote-norm.pl b/corpus/support/quote-norm.pl index e4e5055e..b104e73c 100755 --- a/corpus/support/quote-norm.pl +++ b/corpus/support/quote-norm.pl @@ -11,6 +11,8 @@ while(<STDIN>) { s/&\s*squot\s*;/'/gi; s/&\s*quot\s*;/"/gi; s/&\s*amp\s*;/&/gi; + s/&\s*nbsp\s*;/&/gi; + s/&\s*#\s*160\s*;/ /gi; s/ (\d\d): (\d\d)/ $1:$2/g; s/[\x{20a0}]\x{20ac}]/ EUR /g; s/[\x{00A3}]/ GBP /g; @@ -20,6 +22,7 @@ while(<STDIN>) { s/&\s*#45\s*;/--/g; s/�c/--/g; s/ ,,/ "/g; + s/„/"/g; s/``/"/g; s/''/"/g; s/[「」]/"/g; |