Grammalecte  Diff

from itertools import chain

from ..graphspell.spellchecker import SpellChecker
from ..graphspell.echo import echo
from . import gc_options

from ..graphspell.tokenizer import Tokenizer
from .gc_rules_graph import dGraph, dRule


__all__ = [ "lang", "locales", "pkg", "name", "version", "author", \
            "load", "parse", "getSpellChecker", \
            "setOption", "setOptions", "getOptions", "getDefaultOptions", "getOptionsLabels", "resetOptions", "displayOptions", \
            "ignoreRule", "resetIgnoreRules", "reactivateRule", "listRules", "displayRules" ]

        self.iStart = iStart
        self.lToken = list(_oTokenizer.genTokens(sSentence))

    def parse (self):
        dErr = {}
        lPointer = []
        for dToken in self.lToken:
            # check arcs for each existing pointer
            lNewPointer = []
            for i, dPointer in enumerate(lPointer):
                bValid = False
                bFirst = True
                for dNode in self._getNextMatchingNodes(dToken, dPointer["dNode"]):
                    if bFirst:
                        dPointer["nOffset"] = dToken["i"]
                        dPointer["dNode"] = dNode
                    else:
                        lNewPointer.append({"nOffset": dPointer["nOffset"], "dNode": dNode})
                    bFirst = False
                    bValid = True
                if not bValid:
                    del lPointer[i]
            lPointer.extend(lNewPointer)
            # check arcs of first nodes
            for dNode in self._getNextMatchingNodes(dToken, dGraph[0]):
                lPointer.append({"nOffset": 0, "dNode": dNode})
            # check if there is rules to check for each pointer
            for dPointer in lPointer:
                if "<rules>" in dPointer["dNode"]:

                    dErr = self._executeActions(dPointer["dNode"]["<rules>"], dPointer["nOffset"])
        if dErr:
            print(dErr)
        return dErr

    def _getNextMatchingNodes (self, dToken, dNode):
        # token value
        if dToken["sValue"] in dNode:
            print("value found: ", dToken["sValue"])
            yield dGraph[dNode[dToken["sValue"]]]
        # token lemmas
        if "<lemmas>" in dNode:
            for sLemma in _oSpellChecker.getLemma(dToken["sValue"]):
                if sLemma in dNode["<lemmas>"]:
                    print("lemma found: ", sLemma)
                    yield dGraph[dNode["<lemmas>"][sLemma]]
        # universal arc
        if "*" in dNode:
            print("generic arc")
            yield dGraph[dNode["*"]]
        # regex value arcs
        if "<re_value>" in dNode:
            for sRegex in dNode["<re_value>"]:
                if re.search(sRegex, dToken["sValue"]):
                    print("value regex matching: ", sRegex)
                    yield dGraph[dNode["<re_value>"][sRegex]]
        # regex morph arcs
        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 _executeActions (self, dNode, nOffset):
        dErrs = {}
        for sLineId, nextNodeKey in dNode.items():
            for sArc in dGraph[nextNodeKey]:
                print(sArc)
                bCondMemo = None
                sFuncCond, cActionType, sWhat, *eAct = dRule[sArc]
                # action in lActions: [ condition, action type, replacement/suggestion/action[, iGroupStart, iGroupEnd[, message, URL]] ]
                try:
                    bCondMemo = not sFuncCond or globals()[sFuncCond](self, sCountry, bCondMemo)
                    if bCondMemo:
                        if cActionType == "-":
                            # grammar error
                            print("-")
                            nErrorStart = self.iStart + self.lToken[eAct[0]]["nStart"]
                            nErrorEnd = self.iStart + self.lToken[eAct[1]]["nEnd"]
                            if nErrorStart not in dErrs or nPriority > dPriority[nErrorStart]:
                                dErrs[nErrorStart] = _createError(self, sWhat, nErrorStart, nErrorEnd, sLineId, bUppercase, eAct[2], eAct[3], bIdRule, sOption, bContext)
                                dPriority[nErrorStart] = nPriority
                        elif cActionType == "~":
                            # text processor
                            print("~")
                            self._rewrite(sWhat, nErrorStart, nErrorEnd)

                            print(">")
                            pass
                        else:
                            print("# error: unknown action at " + sLineId)
                    elif cActionType == ">":
                        break
                except Exception as e:
                    raise Exception(str(e), sLineId)



        return dErrs





    def _rewrite (self, sWhat, nErrorStart, nErrorEnd):
        "text processor: rewrite tokens between <nErrorStart> and <nErrorEnd> position"
        lTokenValue = sWhat.split("|")
        if len(lTokenValue) != (nErrorEnd - nErrorStart + 1):
            print("Error. Text processor: number of replacements != number of tokens.")
            return