Usage: scws [options] [input] [output]
  -i <file|string> input string or filepath 
                   (default: try to read from <stdin> everyline)
  -o <file>        output filepath (default to <stdout>)
  -c <charset>     set the charset (default: gbk)
                   charset must been same with dictionary & ruleset
  -r <file>        set the ruleset file (default: none)
  -d <file>        set the dictionary file[s] (default: none)
                   if there are multi files, split filepath use ':'
                   if the file suffix is .txt, it will be treated as plain text dict.
  -M <1~15>        use multi child words mode(ÖйúÈË->Öйú+ÈË+ÖйúÈË)
                   1|2|4|8: short|duality|zmain|zall
  -I               ignore the all mark symbol such as ,:
  -A               show the word attribute
  -E               import the xdb dict into xtree(memory)
  -N               don't show time usage and warnings
  -D               debug segment, see the segment detail
  -U               use duality algorithm for single chinese
  -t <NUM>         fetch the top words instead of segment
  -a [~]<attr1,attr2,...>   prefix by ~ means exclude them.
                   For topwords, exclude or include some word attrs
  -v        Show the version.
  -h        Show this page for help.