Grammalecte  Check-in [c55bec5997]

        yield (iStart, m.end())
        iStart = m.end()


def parse (sText, sCountry="${country_default}", bDebug=False, dOptions=None, bContext=False):
    "analyses the paragraph sText and returns list of errors"
    #sText = unicodedata.normalize("NFC", sText)
    aErrors = None
    sRealText = sText
    dPriority = {}  # Key = position; value = priority
    dOpt = _dOptions  if not dOptions  else dOptions
    bShowRuleId = option('idrule')

    # parse paragraph
    try:
        sNew, aErrors = _proofread(None, sText, sRealText, 0, True, dPriority, sCountry, dOpt, bShowRuleId, bDebug, bContext)
        if sNew:
            sText = sNew
    except:
        raise

    # cleanup
    if " " in sText:
        sText = sText.replace(" ", ' ') # nbsp
    if " " in sText:
        sText = sText.replace(" ", ' ') # nnbsp
    if "'" in sText:
        sText = sText.replace("'", "’")
    if "‑" in sText:
        sText = sText.replace("‑", "-") # nobreakdash

    # parse sentences
    for iStart, iEnd in _getSentenceBoundaries(sText):
        if 4 < (iEnd - iStart) < 2000:
            try:
                oSentence = TokenSentence(sText[iStart:iEnd], sRealText[iStart:iEnd], iStart)
                _, errs = _proofread(oSentence, sText[iStart:iEnd], sRealText[iStart:iEnd], iStart, False, dPriority, sCountry, dOpt, bShowRuleId, bDebug, bContext)
                aErrors.update(errs)
            except:
                raise
    return aErrors.values() # this is a view (iterable)


def _proofread (oSentence, s, sx, nOffset, bParagraph, dPriority, sCountry, dOptions, bShowRuleId, bDebug, bContext):
    dErrs = {}
    bParagraphChange = False
    bSentenceChange = False
    dTokenPos = oSentence.dTokenPos if oSentence else {}
    for sOption, lRuleGroup in _getRules(bParagraph):
        if sOption == "@@@@":
            # graph rules

            if not bParagraph and bSentenceChange:
                oSentence.update(s, bDebug)
                bSentenceChange = False
            for sGraphName, sLineId in lRuleGroup:
                if bDebug:
                    print("\n>>>> GRAPH:", sGraphName, sLineId)
                bParagraphChange, s = oSentence.parse(dAllGraph[sGraphName], dPriority, sCountry, dOptions, bShowRuleId, bDebug, bContext)
                dErrs.update(oSentence.dError)
                dTokenPos = oSentence.dTokenPos
        elif not sOption or dOptions.get(sOption, False):
            # regex rules
            for zRegex, bUppercase, sLineId, sRuleId, nPriority, lActions in lRuleGroup:
                if sRuleId not in _aIgnoredRules:
                    for m in zRegex.finditer(s):
                        bCondMemo = None
                        for sFuncCond, cActionType, sWhat, *eAct in lActions:
                            # action in lActions: [ condition, action type, replacement/suggestion/action[, iGroup[, message, URL]] ]
                            try:
                                bCondMemo = not sFuncCond or globals()[sFuncCond](s, sx, m, dTokenPos, sCountry, bCondMemo)
                                if bCondMemo:
                                    if bDebug:
                                        print("RULE:", sLineId)
                                    if cActionType == "-":
                                        # grammar error
                                        nErrorStart = nOffset + m.start(eAct[0])
                                        if nErrorStart not in dErrs or nPriority > dPriority.get(nErrorStart, -1):
                                            dErrs[nErrorStart] = _createError(s, sx, sWhat, nOffset, m, eAct[0], sLineId, sRuleId, bUppercase, eAct[1], eAct[2], bShowRuleId, sOption, bContext)
                                            dPriority[nErrorStart] = nPriority
                                    elif cActionType == "~":
                                        # text processor
                                        s = _rewrite(s, sWhat, eAct[0], m, bUppercase)
                                        bParagraphChange = True
                                        bSentenceChange = True
                                        if bDebug:

                                    else:
                                        echo("# error: unknown action at " + sLineId)
                                elif cActionType == ">":
                                    break
                            except Exception as e:
                                raise Exception(str(e), "# " + sLineId + " # " + sRuleId)
    if bParagraphChange:
        return (s, dErrs)
    return (False, dErrs)


def _createError (s, sx, sRepl, nOffset, m, iGroup, sLineId, sRuleId, bUppercase, sMsg, sURL, bShowRuleId, sOption, bContext):
    nStart = nOffset + m.start(iGroup)
    nEnd = nOffset + m.end(iGroup)
    # suggestions
    if sRepl[0:1] == "=":

                elif dToken["sType"] in sMeta:
                    if bDebug:
                        print("  MATCH: *" + sMeta)
                    yield dGraph[dNode["<meta>"][sMeta]]

    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"
        self.dError = {}
        dPriority = {}  # Key = position; value = priority
        dOpt = _dOptions  if not dOptions  else dOptions
        lPointer = []
        bTagAndRewrite = False
        for dToken in self.lToken:
            if bDebug:
                print("TOKEN:", dToken["sValue"])
            # check arcs for each existing pointer
            lNextPointer = []
            for dPointer in lPointer:
                for dNode in self._getNextMatchingNodes(dToken, dGraph, dPointer["dNode"], bDebug):
                    lNextPointer.append({"iToken": dPointer["iToken"], "dNode": dNode})
            lPointer = lNextPointer
            # check arcs of first nodes
            for dNode in self._getNextMatchingNodes(dToken, dGraph, dGraph[0], bDebug):
                lPointer.append({"iToken": dToken["i"], "dNode": dNode})
            # check if there is rules to check for each pointer
            for dPointer in lPointer:
                #if bDebug:
                #    print("+", dPointer)
                if "<rules>" in dPointer["dNode"]:
                    bChange, dErr = self._executeActions(dGraph, dPointer["dNode"]["<rules>"], dPointer["iToken"]-1, dToken["i"], dPriority, dOpt, sCountry, bShowRuleId, bDebug, bContext)
                    self.dError.update(dErr)
                    if bChange:
                        bTagAndRewrite = True
        if bTagAndRewrite:
            self.rewrite(bDebug)
        if bDebug:
            print(self)
        return (bTagAndRewrite, self.sSentence)

    def _executeActions (self, dGraph, dNode, nTokenOffset, nLastToken, dPriority, dOptions, sCountry, bShowRuleId, bDebug, bContext):
        "execute actions found in the DARG"
        dError = {}
        bChange = False
        for sLineId, nextNodeKey in dNode.items():
            bCondMemo = None
            for sRuleId in dGraph[nextNodeKey]:
                try:
                    if bDebug:
                        print("  TRY:", sRuleId)

                            if cActionType == "-":
                                # grammar error
                                nTokenErrorStart = nTokenOffset + eAct[0]
                                if "bImmune" not in self.lToken[nTokenErrorStart]:
                                    nTokenErrorEnd = (nTokenOffset + eAct[1])  if eAct[1]  else nLastToken
                                    nErrorStart = self.nOffsetWithinParagraph + self.lToken[nTokenErrorStart]["nStart"]
                                    nErrorEnd = self.nOffsetWithinParagraph + self.lToken[nTokenErrorEnd]["nEnd"]
                                    if nErrorStart not in dError or eAct[2] > dPriority.get(nErrorStart, -1):
                                        dError[nErrorStart] = self._createError(sWhat, nTokenOffset, nTokenErrorStart, nErrorStart, nErrorEnd, sLineId, sRuleId, True, eAct[3], eAct[4], bShowRuleId, "notype", bContext)
                                        dPriority[nErrorStart] = eAct[2]
                                        if bDebug:
                                            print("  NEW_ERROR:", dError[nErrorStart], "\n  ", dRule[sRuleId])
                            elif cActionType == "~":
                                # text processor
                                if bDebug:
                                    print("  TAG_PREPARE:\n  ", dRule[sRuleId])
                                nEndToken = (nTokenOffset + eAct[1])  if eAct[1]  else nLastToken
                                self._tagAndPrepareTokenForRewriting(sWhat, nTokenOffset + eAct[0], nEndToken, nTokenOffset, True, bDebug)
                                bChange = True

                                print("# error: unknown action at " + sLineId)
                        elif cActionType == ">":
                            if bDebug:
                                print("  COND_BREAK")
                            break
                except Exception as e:
                    raise Exception(str(e), sLineId, sRuleId, self.sSentence)
        return bChange, dError

    def _createError (self, sSugg, nTokenOffset, iFirstToken, nStart, nEnd, sLineId, sRuleId, bUppercase, sMsg, sURL, bShowRuleId, sOption, bContext):
        # suggestions
        if sSugg[0:1] == "=":
            sSugg = globals()[sSugg[1:]](self.lToken, nTokenOffset)
            lSugg = sSugg.split("|")  if sSugg  else []
        elif sSugg == "_":