haspirater

detect aspirated 'h' in French words (local mirror of https://gitlab.com/a3nm/haspirater)
git clone https://a3nm.net/git/haspirater/
Log | Files | Refs | README | LICENSE

make.sh (458B)

      1 #!/bin/bash
      2 
      3 # From a French text input and an exceptions dictionary, prepare the
      4 # trie.
      5 
      6 ./prepare.sh | # reformat the text
      7   ./detect.pl | # identify and label occurrences
      8   cat - $* | # add in exceptions
      9   sed 's/ h/ /' | # we don't keep the useless leading 'h' in the trie
     10   ../haspirater/buildtrie.py  | # prepare the trie
     11   ../haspirater/compresstrie.py | # compress the trie
     12   ../haspirater/majoritytrie.py # keep only the most frequent information
     13