Grammalecte  Check-in [31ddfb6e1a]

    "analyses the paragraph sText and returns list of errors"
    #sText = unicodedata.normalize("NFC", sText)
    aErrors = None
    sRealText = sText
    dDA = {}        # Disambiguisator. Key = position; value = list of morphologies
    dPriority = {}  # Key = position; value = priority
    dOpt = _dOptions  if not dOptions  else dOptions
    bShowRuleId = option('idrule')

    # parse paragraph
    try:
        sNew, aErrors = _proofread(sText, sRealText, 0, True, dDA, dPriority, sCountry, dOpt, bShowRuleId, bDebug, bContext)
        if sNew:
            sText = sNew
    except:
        raise

    # cleanup
    if " " in sText:

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


def _getSentenceBoundaries (sText):
    iStart = _zBeginOfParagraph.match(sText).end()
    for m in _zEndOfSentence.finditer(sText):
        yield (iStart, m.end())
        iStart = m.end()


def _proofread (s, sx, nOffset, bParagraph, dDA, dPriority, sCountry, dOptions, bShowRuleId, bDebug, bContext):
    dErrs = {}
    bChange = False


    for sOption, lRuleGroup in _getRules(bParagraph):
        if not sOption or dOptions.get(sOption, False):
            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, dDA, sCountry, bCondMemo)
                                if bCondMemo:
                                    if cActionType == "-":
                                        # grammar error
                                        nErrorStart = nOffset + m.start(eAct[0])
                                        if nErrorStart not in dErrs or nPriority > dPriority[nErrorStart]:
                                            dErrs[nErrorStart] = _createRegexError(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)
                                        bChange = True
                                        if bDebug:
                                            echo("~ " + s + "  -- " + m.group(eAct[0]) + "  # " + sLineId)

                            except Exception as e:
                                raise Exception(str(e), "# " + sLineId + " # " + sRuleId)
    if bChange:
        return (s, dErrs)
    return (False, dErrs)


def _createRegexWriterError (s, sx, sRepl, nOffset, m, iGroup, sLineId, sRuleId, bUppercase, sMsg, sURL, bShowRuleId, sOption, bContext):
    "error for Writer (LO/OO)"
    xErr = SingleProofreadingError()
    #xErr = uno.createUnoStruct( "com.sun.star.linguistic2.SingleProofreadingError" )
    xErr.nErrorStart = nOffset + m.start(iGroup)
    xErr.nErrorLength = m.end(iGroup) - m.start(iGroup)
    xErr.nErrorType = PROOFREADING
    xErr.aRuleIdentifier = sRuleId

    # Message
    if sMsg[0:1] == "=":
        sMessage = globals()[sMsg[1:]](s, m)
    else:
        sMessage = m.expand(sMsg)
    xErr.aShortComment = sMessage   # sMessage.split("|")[0]     # in context menu
    xErr.aFullComment = sMessage   # sMessage.split("|")[-1]    # in dialog
    if bShowRuleId:
        xErr.aShortComment += "  # " + sLineId + " # " + sRuleId
    # URL
    if sURL:
        p = PropertyValue()
        p.Name = "FullCommentURL"
        p.Value = sURL
        xErr.aProperties = (p,)
    else:
        xErr.aProperties = ()
    return xErr


def _createRegexDictError (s, sx, sRepl, nOffset, m, iGroup, sLineId, sRuleId, bUppercase, sMsg, sURL, bShowRuleId, sOption, bContext):
    "error as a dictionary"
    dErr = {}
    dErr["nStart"] = nOffset + m.start(iGroup)
    dErr["nEnd"] = nOffset + m.end(iGroup)
    dErr["sLineId"] = sLineId
    dErr["sRuleId"] = sRuleId
    dErr["sType"] = sOption  if sOption  else "notype"

            dErr["aSuggestions"] = m.expand(sRepl).split("|")
    # Message
    if sMsg[0:1] == "=":
        sMessage = globals()[sMsg[1:]](s, m)
    else:
        sMessage = m.expand(sMsg)
    dErr["sMessage"] = sMessage
    if bShowRuleId:
        dErr["sMessage"] += "  # " + sLineId + " # " + sRuleId
    # URL
    dErr["URL"] = sURL  if sURL  else ""
    # Context
    if bContext:
        dErr['sUnderlined'] = sx[m.start(iGroup):m.end(iGroup)]
        dErr['sBefore'] = sx[max(0,m.start(iGroup)-80):m.start(iGroup)]
        dErr['sAfter'] = sx[m.end(iGroup):m.end(iGroup)+80]
    return dErr


def _createTokenWriterError (lToken, sSentence, sSentence0, sRepl, iFirstToken, nStart, nEnd, sLineId, sRuleId, bUppercase, sMsg, sURL, bShowRuleId, sOption, bContext):
    "error for Writer (LO/OO)"
    xErr = SingleProofreadingError()
    #xErr = uno.createUnoStruct( "com.sun.star.linguistic2.SingleProofreadingError" )
    xErr.nErrorStart = nStart
    xErr.nErrorLength = nEnd - nStart
    xErr.nErrorType = PROOFREADING
    xErr.aRuleIdentifier = sRuleId

    # Message
    if sMsg[0:1] == "=":
        sMessage = globals()[sMsg[1:]](lToken)
    else:
        sMessage = sMsg
    xErr.aShortComment = sMessage   # sMessage.split("|")[0]     # in context menu
    xErr.aFullComment = sMessage   # sMessage.split("|")[-1]    # in dialog
    if bShowRuleId:
        xErr.aShortComment += "  " + sLineId + " # " + sRuleId
    # URL
    if sURL:
        p = PropertyValue()
        p.Name = "FullCommentURL"
        p.Value = sURL
        xErr.aProperties = (p,)
    else:
        xErr.aProperties = ()
    return xErr

                                                         
def _createTokenDictError (lToken, sSentence, sSentence0, sRepl, iFirstToken, nStart, nEnd, sLineId, sRuleId, bUppercase, sMsg, sURL, bShowRuleId, sOption, bContext):
    "error as a dictionary"
    dErr = {}
    dErr["nStart"] = nStart
    dErr["nEnd"] = nEnd
    dErr["sLineId"] = sLineId
    dErr["sRuleId"] = sRuleId
    dErr["sType"] = sOption  if sOption  else "notype"

            dErr["aSuggestions"] = sRepl.split("|")
    # Message
    if sMsg[0:1] == "=":
        sMessage = globals()[sMsg[1:]](lToken)
    else:
        sMessage = sMsg
    dErr["sMessage"] = sMessage
    if bShowRuleId:
        dErr["sMessage"] += "  " + sLineId + " # " + sRuleId
    # URL
    dErr["URL"] = sURL  if sURL  else ""
    # Context
    if bContext:
        dErr['sUnderlined'] = sSentence0[dErr["nStart"]:dErr["nEnd"]]
        dErr['sBefore'] = sSentence0[max(0,dErr["nStart"]-80):dErr["nStart"]]

        if "<re_morph>" in dNode:
            for sRegex in dNode["<re_morph>"]:
                for sMorph in _oSpellChecker.getMorph(dToken["sValue"]):
                    if re.search(sRegex, sMorph):
                        #print("morph regex matching: ", sRegex)
                        yield dGraph[dNode["<re_morph>"][sRegex]]

    def parse (self, dPriority, sCountry="${country_default}", dOptions=None, bShowRuleId=False, bDebug=False, bContext=False):
        dErr = {}
        dPriority = {}  # Key = position; value = priority
        dOpt = _dOptions  if not dOptions  else dOptions
        lPointer = []

        bChange = False
        for dToken in self.lToken:
            # check arcs for each existing pointer
            lNewPointer = []
            for i, dPointer in enumerate(lPointer):
                bValid = False
                bFirst = True

            lPointer.extend(lNewPointer)
            # check arcs of first nodes
            for dNode in self._getNextMatchingNodes(dToken, dGraph[0]):
                lPointer.append({"nOffset": dToken["i"], "dNode": dNode})
            # check if there is rules to check for each pointer
            for dPointer in lPointer:
                if "<rules>" in dPointer["dNode"]:
                    bHasChanged, errs = self._executeActions(dPointer["dNode"]["<rules>"], dPointer["nOffset"], dPriority, dOpt, bShowRuleId, bContext)
                    dErr.update(errs)
                    if bHasChanged:
                        bChange = True
        if dErr:
            print(dErr)
        return (bChange, dErr)

    def _executeActions (self, dNode, nTokenOffset, dPriority, dOpt, bShowRuleId, bContext):
        #print(locals())
        dErrs = {}
        bChange = False
        for sLineId, nextNodeKey in dNode.items():
            for sRuleId in dGraph[nextNodeKey]:
                print(sRuleId)
                bCondMemo = None
                sFuncCond, cActionType, sWhat, *eAct = dRule[sRuleId]
                # action in lActions: [ condition, action type, replacement/suggestion/action[, iTokenStart, iTokenEnd[, nPriority, message, URL]] ]
                try:
                    bCondMemo = not sFuncCond or globals()[sFuncCond](self, sCountry, bCondMemo)
                    if bCondMemo:
                        if cActionType == "-":
                            # grammar error
                            print("-")
                            nTokenErrorStart = nTokenOffset + eAct[0]
                            nTokenErrorEnd = nTokenOffset + eAct[1]
                            nErrorStart = self.iStart + self.lToken[nTokenErrorStart]["nStart"]
                            nErrorEnd = self.iStart + self.lToken[nTokenErrorEnd]["nEnd"]
                            if nErrorStart not in dErrs or eAct[2] > dPriority[nErrorStart]:
                                dErrs[nErrorStart] = _createTokenError(self.lToken, self.sSentence, self.sSentence0, sWhat, nTokenErrorStart, nErrorStart, nErrorEnd, sLineId, sRuleId, True, eAct[3], eAct[4], bShowRuleId, "notype", bContext)
                                dPriority[nErrorStart] = eAct[2]
                        elif cActionType == "~":
                            # text processor
                            print("~")
                            self._rewrite(sWhat, nErrorStart, nErrorEnd)
                        elif cActionType == "@":
                            # jump