Grammalecte  Diff

#!python3

__author__ = "Olivier R."
__license__ = "MPL 2"



import os
import sys
import time
import re
import collections
import zipfile
import math
import argparse
from enum import Enum

# Les dictionnaires
dSUBDIC = { '*': 'Commun',
            'R': 'Réforme1990',
            'M': 'Moderne',
            'C': 'Classique',
            'A': 'Annexe',
            'P': 'Multimots',
            'X': 'Contributeurs' }

dMODERNE = { 'name': 'DICTIONNAIRE ORTHOGRAPHIQUE FRANÇAIS “MODERNE”',
             'shortname': '“Moderne”',
             'asciiName': 'fr-moderne',
             'mozAsciiName': 'fr-FR-modern',
             'subDicts': '*MX',
             'mozId': 'fr-dicollecte-moderne',
             'description': "Dictionnaire français “Moderne”" }

dCLASSIQUE = { 'name': 'DICTIONNAIRE ORTHOGRAPHIQUE FRANÇAIS “CLASSIQUE”',
               'shortname': '“Classique”',
               'asciiName': 'fr-classique',
               'mozAsciiName': 'fr-FR-classic',
               'subDicts': '*MCX',
               'mozId': 'fr-dicollecte-classique',
               'description': "Dictionnaire français “Classique”" }

            'mozId': 'fr-dicollecte',
            'description': "Dictionnaire orthographique de la langue française" }


BUILD_PATH = '_build'
PREFIX_DICT_PATH = 'hunspell-french-dictionaries-v'
EXT_PREFIX_OOO = 'lo-oo-ressources-linguistiques-fr-v'
EXT_PREFIX_MOZ = 'moz-hunspell-fr-dicollecte-v'
LEX_PREFIX = 'lexique-dicollecte-fr-v'
STATS_NAME = 'statistiques-v'

MPLHEADER = "# This Source Code Form is subject to the terms of the Mozilla Public\n" + \
            "# License, v. 2.0. If a copy of the MPL was not distributed with this\n" + \
            "# file, You can obtain one at http://mozilla.org/MPL/2.0/.\n\n"

        "Écrire le fichier des affixes (.aff)"
        echo(' * Dictionnaire >> [ {}.aff ]'.format(dTplVars['asciiName']))
        info = "# This Source Code Form is subject to the terms of the Mozilla Public\n" + \
               "# License, v. 2.0. If a copy of the MPL was not distributed with this\n" + \
               "# file, You can obtain one at http://mozilla.org/MPL/2.0/.\n\n" + \
               "# AFFIXES DU {} v{}\n".format(dTplVars['name'], self.sVersion) + \
               "# par Olivier R. -- licence MPL 2.0\n" + \
               "# Généré le " + time.strftime("%d-%m-%Y à %H:%M") + "\n" \
               "# Pour améliorer le dictionnaire, allez sur https://grammalecte.net/\n\n"

        with open(spDst+'/'+dTplVars['asciiName']+'.aff', 'w', encoding='utf-8', newline="\n") as hDst:
            hDst.write(info)
            hDst.write(self.sSettings + "\n")
            if self.bShortenTags:
                hDst.write("AM {}\n".format(len(self.dAM)))

                    hDst.write(str(dRefW[key]))
                    hDst.write("\n")

    def writeLexicon (self, spfDst, version, oStatsLex):
        echo(' * Lexique >> [ {} ] '.format(spfDst))
        with open(spfDst, 'w', encoding='utf-8', newline="\n") as hDst:
            hDst.write(MPLHEADER)
            hDst.write("# Lexique des formes fléchies du français - Dicollecte v{}\n# Licence : MPL v2.0\n\n".format(version))
            hDst.write(oStatsLex.getInfo())
            hDst.write(Flexion.header(oStatsLex))
            for oFlex in self.lFlexions:
                hDst.write(oFlex.__str__(oStatsLex))

    def writeGrammarCheckerLexicon (self, spfDst, version):
        echo(' * Lexique simplifié >> [ {} ] '.format(spfDst))

    def createMozillaExtensions (self, spBuild, dTplVars, lDictVars, spDestGL=""):
        # Mozilla extension 1
        echo(" * Dictionnaire >> extension pour Mozilla")
        dTplVars['version'] = self.sVersion
        sExtensionName = EXT_PREFIX_MOZ + self.sVersion
        spExt = spBuild + '/' + sExtensionName
        dir_util.mkpath(spExt+'/dictionaries')
        copyTemplate('_templates/moz', spExt, 'install.rdf', dTplVars)
        spDict = spBuild + '/' + PREFIX_DICT_PATH + self.sVersion
        file_util.copy_file(spDict+'/fr-classique.dic', spExt+'/dictionaries/fr-classic.dic')
        file_util.copy_file(spDict+'/fr-classique.aff', spExt+'/dictionaries/fr-classic.aff')
        copyTemplate('orthographe', spExt, 'README_dict_fr.txt', dTplVars)
        createZipFiles(spExt, spBuild, sExtensionName + '.xpi')
        # Grammalecte
        if spDestGL:

                    hDiff.write("{0.iD}\t{0.fq}\n".format(oEntry))
                    hNotes.write("{0.lemma}/{0.flags}\t{0.oldFq} > {0.fq}\n".format(oEntry))

    def createLexiconPackages (self, spBuild, version, oStatsLex, spDestGL=""):
        sLexName = LEX_PREFIX + version
        spLex = spBuild + '/' + sLexName
        dir_util.mkpath(spLex)
        # write Dicollecte lexicon
        self.sortLexiconByFreq()
        self.writeLexicon(spLex + '/' + sLexName + '.txt', version, oStatsLex)
        self.writeGrammarCheckerLexicon(spBuild + '/' + sLexName + '.lex', version)
        copyTemplate('lexique', spLex, 'README_lexique.txt', {'version': version})
        # zip
        createZipFiles(spLex, spBuild, sLexName + '.zip')
        # copy GC lexicon to Grammalecte

        with open(spBuild+'/dictDecl.txt', 'w', encoding='utf-8', newline="\n") as hDst:
            for oEntry in self.lEntry:
                if re.match("[SXFWIA]", oEntry.flags) and (oEntry.po.startswith("nom") or oEntry.po.startswith("adj")):
                    hDst.write(oEntry.getDeclination())
        if spDestGL:
            echo("   Fichier de déclinaison copié dans Grammalecte...")
            file_util.copy_file(spBuild+'/dictDecl.txt', spDestGL)

    def generateSpellVariants (self, nReq, spBuild):
        if nReq < 1: nReq = 1
        if nReq > 2: nReq = 2
        echo(" * Lexique >> variantes par suppression... n = " + str(nReq))
        with open(spBuild+'/dictSpellVariants-'+str(nReq)+'.txt', 'w', encoding='utf-8', newline="\n") as hDst:
            for oFlex in frozenset(self.lFlexions):
                hDst.write(oFlex.sFlexion+"\t_\t_\n")
                if len(oFlex.sFlexion) <= 2:
                    n = 0
                elif len(oFlex.sFlexion) <= 5:
                    n = 1
                else:
                    n = nReq
                #lTup = self._generatePhonetVariants(oFlex.sFlexion)
                lTup = self._generateDeleteVariants(oFlex.sFlexion, oFlex.sFlexion, n)
                for t in lTup:
                    sTag = t[1]  if "\t" in t[1]  else t[1]+"\t_"
                    hDst.write(t[0]+"\t"+sTag+"\n")

    _lTupPhonet = [ ("ph", "f"), ("qu", "k"), ("ss", "c"), ("ss", "ç"), ("ct", "x"),
        ("oe", "œ"), ("ae", "æ"), ("ei", "é"), ("ai", "é"), ("au", "o"), ("eau", "o"),
    ]

    def _generatePhonetVariants (self, s):
        l = []
        for torep, rep in self._lTupPhonet():
            for m in torep.finditer(s):
                l.append( (s[:m.start(0)] + rep + s[m.end(0):], str(m.start(0))+":"+str(m.start(0)+len(rep))+">"+torep) )
        return l

    def _generateDeleteVariants (self, sWord0, sWordCur, n):
        "renvoie une liste de tuples : (forme dégradée de sWord, code de genèse de sWord)"
        # caution: recursive function
        if n == 0:
            return []
        lTup = []
        for i in range(len(sWordCur)):
            sNew = sWordCur[0:i]+sWordCur[i+1:]
            lTup.append( ( sNew, self._generateAddCode(sWord0, sNew) ) )
            lTup += self._generateDeleteVariants(sWord0, sNew, n-1)
        return lTup

    def _generateAddCode (self, sWord, sCrippled):
        "returns addCode to generate sWord from sCrippled"
        sAdd = ""
        for i in range(len(sWord)):
            if sWord[i] != sCrippled[i:i+1]:
                sCrippled = sCrippled[:i] + sWord[i] + sCrippled[i:]
                if sAdd:
                    sAdd += "\t"
                sAdd += str(i)+"+"+sWord[i]
        return sAdd  if sAdd  else "0"



class Entree:
    def __init__ (self, sLine):
        self.lemma = ''
        self.flags = ''
        # champs morphologiques Hunspell

            sErr += 'lemme vide'
        if re.match(r"^\s", self.lemma):
            sErr += 'premier caractère un espace dans <' + self.lemma + '>'
        if re.search(r"\s$", self.lemma):
            sErr += 'espace en fin de lemme'
        if re.match(r"v[0123]", self.po) and not re.match(r"[eas_][ix_][tx_][nx_][pqreuvx_][mx_][ex_z][ax_z]\b", self.po[2:]):
            sErr += 'verbe inconnu: ' + self.po
        if (re.match(r"S[*.]", self.flags) and re.search("[sxz]$", self.lemma)) or (re.match(r"X[*.]", self.flags) and not re.search("[ul]$", self.lemma)):
            sErr += 'drapeau inutile'
        if self.iz == '' and re.match(r"[SXAI](?!=)", self.flags) and self.po:
            sErr += '[is] vide'
        if re.match(r"pl|sg|inv", self.iz):
            sErr += '[is] incomplet'
        if re.match(r"[FW]", self.flags) and re.search(r"epi|mas|fem|inv|sg|pl", self.iz):
            sErr += '[is] incohérent'
        if re.match(r".\*", self.flags) and re.match(r"[bcdfgjklmnpqrstvwxz]", self.lemma):
            sErr += 'drapeau pour lemme commençant par une voyelle'
        if re.search(r"pl|sg|inv", self.iz) and re.match(r"[SXAIFW](?!=)", self.flags):
            sErr += '[is] incohérent'
        if re.search(r"nom|adj", self.po) and re.match(r"(?i)[aâàäáeéèêëiîïíìoôöóòuûüúù]", self.lemma) and re.match("[SFWXAI][.]", self.flags) \
           and "pel" not in self.lx:
            sErr += 'le drapeau derait finir avec *'
        if self.iz.endswith(("mas", "fem", "epi")) and (not self.flags or not self.flags.startswith(("S", "X", "F", "W", "A", "I", "U"))):
            sErr += '[is] incomplet'
        if self.flags.startswith(("a", "b", "c", "d")) and not self.lemma.endswith("er"):
            sErr += "drapeau pour verbe du 1ᵉʳ groupe sur un lemme non conforme"
        if self.flags.startswith("f") and not self.lemma.endswith(("ir", "ïr")):
            sErr += "drapeau pour verbe du 2ᵉ groupe sur un lemme non conforme"
        if sErr:
            echo('   error -  id: ' + self.iD, end = "")
            echo('  ' + sErr + '  in  ' + self.__str__())

            morph = self.lexMorph()
        lFlexions = [(self.lemma, morph, self.di)]  if iPR == 0 and not self.flags.endswith('()')  else []
        lFlexPrefix = []
        lFlexSuffix = []
        for sFlag in makeLongFlags(self.flags):
            if sFlag not in dFlags:
                if sFlag not in ['**', '()', '||', '--']:
                    lFlexions.append( (self.lemma, '[unknown flag: {}]'.format(sFlag)) )
                    echo("ERROR: "  + self.lemma + ' - unknown flag: ' + sFlag)
            else:
                oFlag = dFlags[sFlag]
                if not oFlag.bSfx:
                    # cas des préfixes
                    for oRule in oFlag.lRules:
                        if oRule.motif.search(self.lemma):

    def getGrammarCheckerRepr (self):
        return "{0.sFlexion}\t{0.oEntry.lemma}\t{1}\n".format(self, self._getSimpleTags())

    _dTagReplacement = {
        # POS
        "nom": ":N", "adj": ":A", "adv": ":W", "negadv": ":X", "mg": ":G", "nb": ":B", "nbro": ":Br",
        "loc.nom": ":Ñ", "loc.adj": ":Â", "loc.adv": ":Ŵ", "loc.verb": ":Ṽ",
        "interj": ":J", "loc.interj": ":Ĵ", "titr": ":T",
        "mas": ":m", "fem": ":f", "epi": ":e", "sg": ":s", "pl": ":p", "inv": ":i",
        "infi": ":Y",
        "ppre": ":P", "ppas": ":Q",
        "ipre": ":Ip", "iimp": ":Iq", "ipsi": ":Is", "ifut": ":If",
        "spre": ":Sp", "simp": ":Sq", "cond": ":K", "impe": ":E",
        "1sg": ":1s", "1isg": ":1ś", "1jsg": ":1ŝ", "2sg": ":2s", "3sg": ":3s", "1pl": ":1p", "2pl": ":2p", "3pl": ":3p", "3pl!": ":3p!",
        "prepv": ":Rv", "prep": ":R", "loc.prep": ":Ŕ", "loc.prepv": "Ŕv",
        "detpos": ":Dp", "detdem": ":Dd", "detind": ":Di", "detneg": ":Dn", "detex": ":De", "det": ":D",
        "advint": ":U",
        "prodem": ":Od", "proind": ":Oi", "proint": ":Ot", "proneg": ":On", "prorel": ":Or", "proadv": ":Ow",
        "properobj": ":Oo", "propersuj": ":Os", "1pe": ":O1", "2pe": ":O2", "3pe": ":O3", "preverb": ":Ov",
        "cjco": ":Cc", "cjsub": ":Cs", "cj": ":C", "loc.cj": ":Ĉ", "loc.cjsub": ":Ĉs",
        "prn": ":M1", "patr": ":M2", "loc.patr": ":Ḿ2", "npr": ":MP", "nompr": ":NM",
        "pfx": ":Zp", "sfx": ":Zs",
        "div": ":H",
        "err": ":F",
        "ponc": ":@p", "sign": ":@s",
        # LEX
        "symb": ";S"
    }

    def _getSimpleTags (self):
        s = ""
        # POS
        for sTag in self.sMorph.split():
            if sTag.startswith("v"):

def main ():
    xParser = argparse.ArgumentParser()
    xParser.add_argument("-v", "--verdic", help="set dictionary version, i.e. 5.4", type=str, default="X.Y.z")
    xParser.add_argument("-m", "--mode", help="0: no tags,  1: Hunspell tags (default),  2: All tags", type=int, choices=[0, 1, 2], default=1)
    xParser.add_argument("-u", "--uncompress", help="do not use Hunspell compression", action="store_true")
    xParser.add_argument("-s", "--simplify", help="no virtual lemmas", action="store_true")
    xParser.add_argument("-sv", "--spellvariants", help="generate spell variants", action="store_true")
    xParser.add_argument("-gl", "--grammalecte", help="copy generated files to Grammalecte folders", action="store_true")
    xArgs = xParser.parse_args()

    if xArgs.simplify:
        xArgs.mode = 0
        xArgs.uncompress = True

    spBuild = BUILD_PATH + '/' + xArgs.verdic
    dir_util.mkpath(spBuild)

    ### Lecture des fichiers et création du dictionnaire
    oFrenchDict = Dictionnaire(xArgs.verdic, "French dictionary")
    for sFile in ['orthographe/FRANCAIS.dic']:
        oFrenchDict.readDictionary(sFile)
    oFrenchDict.readAffixes('orthographe/FRANCAIS_5.aff')

    ### Contrôle
    oFrenchDict.sortEntriesNatural()
    oFrenchDict.checkEntries()

    ### Lexique
    oFrenchDict.generateFlexions()
    oFrenchDict.calcMetagraphe()
    oFrenchDict.calcMetaphone2()

    #oFrenchDict.createNgrams(spBuild, 3)
    if xArgs.spellvariants:
        oFrenchDict.generateSpellVariants(1, spBuild)

    ### Statistiques
    spfStats = spBuild+'/'+STATS_NAME+xArgs.verdic+'.txt'
    oStatsLex = StatsLex(oFrenchDict)
    oStatsLex.addLexFromFile('lexique/corpus_data/stats_google_ngram_1.txt', 'G', 'Google 1-grams')
    oStatsLex.addLexFromFile('lexique/corpus_data/stats_frwiki.txt', 'W', 'Wikipédia')
    oStatsLex.addLexFromFile('lexique/corpus_data/stats_frwikisource.txt', 'S', 'Wikisource')