Grammalecte  Check-in [043d1fdd77]

                dToken["lMorph"] = self.dTokenPos[dToken["nStart"]]["lMorph"]
        self.lToken = lNewToken
        self.dTokenPos = { dToken["nStart"]: dToken  for dToken in self.lToken  if dToken["sType"] != "INFO" }
        if bDebug:
            print("UPDATE:")
            print(self)

    def _getNextMatchingNodes (self, dToken, dGraph, iNode1, dNode, bDebug=False):
    def _getNextPointers (self, dToken, dGraph, iNode1, dNode, bKeep=False, bDebug=False):
        "generator: return nodes where <dToken> “values” match <dNode> arcs"
        bTokenFound = False
        # token value
        if dToken["sValue"] in dNode:
            if bDebug:
                print("  MATCH:", dToken["sValue"])
            yield { "iNode1": iNode1, "dNode": dGraph[dNode[dToken["sValue"]]] }
            bTokenFound = True
        if dToken["sValue"][0:2].istitle(): # we test only 2 first chars, to make valid words such as "Laissez-les", "Passe-partout".
            sValue = dToken["sValue"].lower()
            if sValue in dNode:
                if bDebug:
                    print("  MATCH:", sValue)
                yield { "iNode1": iNode1, "dNode": dGraph[dNode[sValue]] }
                bTokenFound = True
        elif dToken["sValue"].isupper():
            sValue = dToken["sValue"].lower()
            if sValue in dNode:
                if bDebug:
                    print("  MATCH:", sValue)
                yield { "iNode1": iNode1, "dNode": dGraph[dNode[sValue]] }
                bTokenFound = True
            sValue = dToken["sValue"].capitalize()
            if sValue in dNode:
                if bDebug:
                    print("  MATCH:", sValue)
                yield { "iNode1": iNode1, "dNode": dGraph[dNode[sValue]] }
                bTokenFound = True
        # regex value arcs
        if "<re_value>" in dNode:
            for sRegex in dNode["<re_value>"]:
                if "¬" not in sRegex:
                    # no anti-pattern
                    if re.search(sRegex, dToken["sValue"]):
                        if bDebug:
                            print("  MATCH: ~" + sRegex)
                        yield { "iNode1": iNode1, "dNode": dGraph[dNode["<re_value>"][sRegex]] }
                        bTokenFound = True
                else:
                    # there is an anti-pattern
                    sPattern, sNegPattern = sRegex.split("¬", 1)
                    if sNegPattern and re.search(sNegPattern, dToken["sValue"]):
                        continue
                    if not sPattern or re.search(sPattern, dToken["sValue"]):
                        if bDebug:
                            print("  MATCH: ~" + sRegex)
                        yield { "iNode1": iNode1, "dNode": dGraph[dNode["<re_value>"][sRegex]] }
                        bTokenFound = True
        # analysable tokens
        if dToken["sType"][0:4] == "WORD":
            # token lemmas
            if "<lemmas>" in dNode:
                for sLemma in _oSpellChecker.getLemma(dToken["sValue"]):
                    if sLemma in dNode["<lemmas>"]:
                        if bDebug:
                            print("  MATCH: >" + sLemma)
                        yield { "iNode1": iNode1, "dNode": dGraph[dNode["<lemmas>"][sLemma]] }
                        bTokenFound = True
            # regex morph arcs
            if "<re_morph>" in dNode:
                for sRegex in dNode["<re_morph>"]:
                    if "¬" not in sRegex:
                        # no anti-pattern
                        lMorph = dToken.get("lMorph", _oSpellChecker.getMorph(dToken["sValue"]))
                        if any(re.search(sRegex, sMorph)  for sMorph in lMorph):
                            if bDebug:
                                print("  MATCH: @" + sRegex)
                            yield { "iNode1": iNode1, "dNode": dGraph[dNode["<re_morph>"][sRegex]] }
                            bTokenFound = True
                    else:
                        # there is an anti-pattern
                        sPattern, sNegPattern = sRegex.split("¬", 1)
                        if sNegPattern == "*":
                            # all morphologies must match with <sPattern>
                            if sPattern:
                                lMorph = dToken.get("lMorph", _oSpellChecker.getMorph(dToken["sValue"]))
                                if lMorph and all(re.search(sPattern, sMorph)  for sMorph in lMorph):
                                    if bDebug:
                                        print("  MATCH: @" + sRegex)
                                    yield { "iNode1": iNode1, "dNode": dGraph[dNode["<re_morph>"][sRegex]] }
                                    bTokenFound = True
                        else:
                            lMorph = dToken.get("lMorph", _oSpellChecker.getMorph(dToken["sValue"]))
                            if sNegPattern and any(re.search(sNegPattern, sMorph)  for sMorph in lMorph):
                                continue
                            if not sPattern or any(re.search(sPattern, sMorph)  for sMorph in lMorph):
                                if bDebug:
                                    print("  MATCH: @" + sRegex)
                                yield { "iNode1": iNode1, "dNode": dGraph[dNode["<re_morph>"][sRegex]] }
                                bTokenFound = True
        # token tags
        if "tags" in dToken and "<tags>" in dNode:
            for sTag in dToken["tags"]:
                if sTag in dNode["<tags>"]:
                    if bDebug:
                        print("  MATCH: /" + sTag)
                    yield { "iNode1": iNode1, "dNode": dGraph[dNode["<tags>"][sTag]] }
                    bTokenFound = True
        # meta arc (for token type)
        if "<meta>" in dNode:
            for sMeta in dNode["<meta>"]:
                # no regex here, we just search if <dNode["sType"]> exists within <sMeta>
                if sMeta == "*":
                    if bDebug:
                        print("  MATCH: *" + sMeta)
                    yield { "iNode1": iNode1, "dNode": dGraph[dNode["<meta>"]["*"]] }
                    bTokenFound = True
                elif "¬" in sMeta:
                    if dToken["sType"] not in sMeta:
                        if bDebug:
                            print("  MATCH: *" + sMeta)
                        yield { "iNode1": iNode1, "dNode": dGraph[dNode["<meta>"][sMeta]] }
                        bTokenFound = True
                elif dToken["sType"] in sMeta:
                    if bDebug:
                        print("  MATCH: *" + sMeta)
                    yield { "iNode1": iNode1, "dNode": dGraph[dNode["<meta>"][sMeta]] }
                    bTokenFound = True
        if bKeep and not bTokenFound:
            yield { "iNode1": iNode1, "dNode": dNode, "bKeep": True }
        # JUMP
        # Warning! Recurssion!
        if "<>" in dNode:
            yield from self._getNextPointers(self, dToken, dGraph, iNode1, dGraph[dNode["<>"]], True, bDebug)


    def parse (self, dGraph, dPriority, sCountry="${country_default}", dOptions=None, bShowRuleId=False, bDebug=False, bContext=False):
        "parse tokens from the text and execute actions encountered"
        dOpt = _dOptions  if not dOptions  else dOptions
        lPointer = []
        bTagAndRewrite = False
        for iToken, dToken in enumerate(self.lToken):
            if bDebug:
                print("TOKEN:", dToken["sValue"])
            # check arcs for each existing pointer
            lNextPointer = []
            for dPointer in lPointer:
                lNextPointer.extend(self._getNextPointers(dToken, dGraph, dPointer["iNode1"], dPointer["dNode"], bDebug))
                lNextPointer.extend(self._getNextPointers(dToken, dGraph, dPointer["iNode1"], dPointer["dNode"], dPointer.get("bKeep", False), bDebug))
            lPointer = lNextPointer
            # check arcs of first nodes
            lPointer.extend(self._getNextPointers(dToken, dGraph, iToken, dGraph[0], bDebug))
            lPointer.extend(self._getNextPointers(dToken, dGraph, iToken, dGraph[0], False, bDebug))
            # check if there is rules to check for each pointer
            for dPointer in lPointer:
                #if bDebug:
                #    print("+", dPointer)
                if "<rules>" in dPointer["dNode"]:
                    bChange = self._executeActions(dGraph, dPointer["dNode"]["<rules>"], dPointer["iNode1"]-1, iToken, dPriority, dOpt, sCountry, bShowRuleId, bDebug, bContext)
                    if bChange:

        <<- /conf/ -1>> =\1.replace("escell", "écel").replace("essell", "écel")
        # Confusion probable si ce mot se rapporte à “\3 \4”. Desceller signifie briser un sceau, un cachet… Desseller signifie ôter une selle.|http://fr.wiktionary.org/wiki/déceler

    [>desseller|>desceller] [l’|@:[BD]] [>acidité|>activité|>allergie|>anévrisme|>anomalie|>arnaque|>appendicite|>atrophie|>baisse|>bébé|>blessure|>bug|>bogue|>carie|>cancer|>cause|>changement|>complot|>comète|>concentration|>corrélation|>croissance|>défaut|>défaillance|>demande|>dépression|>diabète|>différence|>diminution|>effluve|>épilepsie|>erreur|>essai|>existence|>grossesse|>grosseur|>faille|>faute|>fuite|>fraude|>grippe|>handicap|>hausse|>hémorragie|>hostilité|>hypertrophie|>incompatibilité|>incohérence|>infection|>infraction|>indice|>infidélité|>insuffisance|>intrigue|>irrégularité|>leucémie|>lésion|>lueur|>lumière|>maladie|>malformation|>manœuvre|>manipulation|>molécule|>mensonge|>mutation|>once|>perturbation|>personnalité|>piste|>perte|>planète|>exoplanète|>présence|>qualité|>odeur|>opportunité|>otite|>problème|>surdité|>talent|>tendance|>tentative|>tumeur|>utilisation|hoax|>variation|>vie|virus]
        <<- /conf/ -1>> =\1.replace("escell", "écel").replace("essell", "écel")
        # Confusion probable si ce mot se rapporte à “\3”. Desceller signifie briser un sceau, un cachet… Desseller signifie ôter une selle.|http://fr.wiktionary.org/wiki/déceler

TEST: il faut régler les problèmes {{descellés}} dans le code de l’analyseur syntaxique.
TEST: il y a une erreur qu’on peut {{desceller}} dans ses analyses.
TEST: elle a {{dessellé}} une forte hostilité dans ses propos.
TEST: elle a {{dessellé}} deux problèmes.



@@@@