Grammalecte: tokenizer.py at [8cc845ffb8]

File gc_core/py/tokenizer.py artifact 27b6fefad2 part of check-in 8cc845ffb8

# Very simple tokenizer

import re

_PATTERNS = {
    "default":
        (
            r'(?P<PUNC>[.,?!:;…«»“”"()/·]+)',
            r'(?P<LINK>(?:https?://|www[.]|\w+[@.]\w+[@.])[\w./?&!%=+*"\'@$#-]+)',
            r'(?P<HASHTAG>[#@][\w-]+)',
            r'(?P<HTML><\w+.*?>|</\w+ *>)',
            r'(?P<PSEUDOHTML>\[/?\w+\])',
            r'(?P<HOUR>\d\d?h\d\d\b)',
            r'(?P<NUM>-?\d+(?:[.,]\d+))',
            r"(?P<WORD>\w+(?:[’'`-]\w+)*)"
        ),
    "fr":
        (
            r'(?P<PUNC>[.,?!:;…«»“”"()/·]+)',
            r'(?P<LINK>(?:https?://|www[.]|\w+[@.]\w+[@.])[\w./?&!%=+*"\'@$#-]+)',
            r'(?P<HASHTAG>[#@][\w-]+)',
            r'(?P<HTML><\w+.*?>|</\w+ *>)',
            r'(?P<PSEUDOHTML>\[/?\w+\])',
            r"(?P<ELPFX>(?:l|d|n|m|t|s|j|c|ç|lorsqu|puisqu|jusqu|quoiqu|qu)['’`])",
            r'(?P<ORDINAL>\d+(?:er|nd|e|de|ième|ème|eme)\b)',
            r'(?P<HOUR>\d\d?h\d\d\b)',
            r'(?P<NUM>-?\d+(?:[.,]\d+|))',
            r"(?P<WORD>\w+(?:[’'`-]\w+)*)"
        )
}


class Tokenizer:

    def __init__ (self, sLang):
        self.sLang = sLang
        if sLang not in _PATTERNS:
            self.sLang = "default"
        self.zToken = re.compile( "(?i)" + '|'.join(sRegex for sRegex in _PATTERNS[sLang]) )

    def genTokens (self, sText):
        for m in self.zToken.finditer(sText):
            yield { "sType": m.lastgroup, "sValue": m.group(), "nStart": m.start(), "nEnd": m.end() }

Grammalecte tokenizer.py at [8cc845ffb8]

File gc_core/py/tokenizer.py artifact 27b6fefad2 part of check-in 8cc845ffb8