Grammalecte  Diff

    global _dOptions
    _dOptions = getDefaultOptions()


#### Parsing

def parse (sText, sCountry="${country_default}", bDebug=False, dOptions=None, bContext=False, bFullInfo=False):
    "init point to analyse <sText> and returns an iterable of errors or (with option <bFullInfo>) paragraphs errors and sentences with tokens and errors"
    oText = TextParser(sText)
    return oText.parse(sCountry, bDebug, dOptions, bContext, bFullInfo)


#### TEXT PARSER

class TextParser:

                s += "\t" + str(dToken["aTags"])
            s += "\n"
        #for nPos, dToken in self.dTokenPos.items():
        #    s += "{}\t{}\n".format(nPos, dToken)
        return s

    def parse (self, sCountry="${country_default}", bDebug=False, dOptions=None, bContext=False, bFullInfo=False):
        "analyses <sText> and returns an iterable of errors or (with option <bFullInfo>) paragraphs errors and sentences with tokens and errors"
        #sText = unicodedata.normalize("NFC", sText)
        dOpt = dOptions or _dOptions
        bShowRuleId = option('idrule')
        # parse paragraph
        try:
            self.parseText(self.sText, self.sText0, True, 0, sCountry, dOpt, bShowRuleId, bDebug, bContext)
        except:
            raise
        if bFullInfo:
            lParagraphErrors = list(self.dError.values())
            lSentences = []
            self.dSentenceError.clear()
        # parse sentences
        sText = self._getCleanText()

        for iStart, iEnd in text.getSentenceBoundaries(sText):
            if 4 < (iEnd - iStart) < 2000:
                try:
                    self.sSentence = sText[iStart:iEnd]
                    self.sSentence0 = self.sText0[iStart:iEnd]
                    self.nOffsetWithinParagraph = iStart
                    self.lToken = list(_oTokenizer.genTokens(self.sSentence, True))
                    self.dTokenPos = { dToken["nStart"]: dToken  for dToken in self.lToken  if dToken["sType"] != "INFO" }
                    if bFullInfo:
                        dSentence = { "nStart": iStart, "nEnd": iEnd, "sSentence": self.sSentence, "lToken": list(self.lToken) }
                        for dToken in dSentence["lToken"]:
                            if dToken["sType"] == "WORD":
                                dToken["bValidToken"] = _oSpellChecker.isValidToken(dToken["sValue"])
                        # the list of tokens is duplicated, to keep all tokens from being deleted when analysis
                    self.parseText(self.sSentence, self.sSentence0, False, iStart, sCountry, dOpt, bShowRuleId, bDebug, bContext)
                    if bFullInfo:
                        dSentence["lGrammarErrors"] = list(self.dSentenceError.values())
                        lSentences.append(dSentence)
                        self.dSentenceError.clear()
                except:
                    raise
        if bFullInfo:
            # Grammar checking and sentence analysis
            return lParagraphErrors, lSentences
        else:
            # Grammar checking only
            return self.dError.values() # this is a view (iterable)

    def _getCleanText (self):
        sText = self.sText
        if " " in sText: