diff options
Diffstat (limited to 'norm-german')
| -rwxr-xr-x | norm-german | 15 |
1 files changed, 7 insertions, 8 deletions
diff --git a/norm-german b/norm-german index 85a39da..5c41f98 100755 --- a/norm-german +++ b/norm-german @@ -1,23 +1,23 @@ #!/usr/bin/env ruby -require 'thread' -require 'optimist' +require "thread" +require "optimist" -STDIN.set_encoding 'utf-8' -STDOUT.set_encoding 'utf-8' +STDIN.set_encoding "utf-8" +STDOUT.set_encoding "utf-8" conf = Optimist::options do banner "norm_german < <file w/ lowercased tokens>" opt :upper, "uppercase", :type => :bool, :default => false - opt :threads, "#threads", :type => :int, :default => 1, :short => '-h' + opt :threads, "#threads", :type => :int, :default => 1, :short => "-h" opt :shard_size, "shard size", :type => :int, :default => 1000 opt :train, "train", :type => :bool opt :apply, "apply", :type => :bool end -pairs_lower = [ ['ß','ss'], ['ue', 'ü'], ['ae','ä'], ['oe', 'ö'] ] -pairs_upper = [ ['Ä', 'Ae'], ['Ö', 'Oe'], ['Ü', 'Ue'] ] +pairs_lower = [ ["ß","ss"], ["ue", "ü"], ["ae","ä"], ["oe", "ö"] ] +pairs_upper = [ ["Ä", "Ae"], ["Ö", "Oe"], ["Ü", "Ue"] ] if conf[:upper] PAIRS = pairs_lower else @@ -84,4 +84,3 @@ token_stock.each { |i| h.merge! build_partial i end } - |