Grammalecte  Check-in [9f88778630]

    resetOptions: function () {
        _dOptions = gc_options.getOptions(_sAppContext).gl_shallowCopy();
    },

    //// Parsing

    parse: function (sText, sCountry="${country_default}", bDebug=false, dOptions=null, bContext=false, bFullInfo=false) {
        // init point to analyse <sText> and returns an iterable of errors or (with option <bFullInfo>) a list of sentences with tokens and errors
        let oText = new TextParser(sText);
        return oText.parse(sCountry, bDebug, dOptions, bContext, bFullInfo);
    }
};


class TextParser {

    constructor (sText) {
        this.sText = sText;
        this.sText0 = sText;
        this.sSentence = "";
        this.sSentence0 = "";
        this.nOffsetWithinParagraph = 0;
        this.lToken = [];
        this.dTokenPos = new Map();         // {position: token}
        this.dTags = new Map();             // {position: tags}
        this.dError = new Map();            // {position: error}
        this.dSentenceError = new Map();    // {position: error} (for the current sentence only)
        this.dErrorPriority = new Map();    // {position: priority of the current error}
    }

    asString () {
        let s = "===== TEXT =====\n";
        s += "sentence: " + this.sSentence0 + "\n";
        s += "now:      " + this.sSentence  + "\n";
        for (let dToken of this.lToken) {
            s += `#${dToken["i"]}\t${dToken["nStart"]}:${dToken["nEnd"]}\t${dToken["sValue"]}\t${dToken["sType"]}`;
            if (dToken.hasOwnProperty("lMorph")) {
                s += "\t" + dToken["lMorph"].toString();
            }
            if (dToken.hasOwnProperty("aTags")) {
                s += "\t" + dToken["aTags"].toString();
            }
            s += "\n";
        }
        return s;
    }

    parse (sCountry="${country_default}", bDebug=false, dOptions=null, bContext=false, bFullInfo=false) {
        // analyses <sText> and returns an iterable of errors or (with option <bFullInfo>) a list of sentences with tokens and errors
        let dOpt = dOptions || _dOptions;
        let bShowRuleId = option('idrule');
        // parse paragraph
        try {
            this.parseText(this.sText, this.sText0, true, 0, sCountry, dOpt, bShowRuleId, bDebug, bContext);
        }
        catch (e) {
            console.error(e);
        }
        // parse sentence














        let sText = this._getCleanText();

        let lSentences = [];
        let oSentence = null;
        for (let [iStart, iEnd] of text.getSentenceBoundaries(sText)) {
            try {
                this.sSentence = sText.slice(iStart, iEnd);
                this.sSentence0 = this.sText0.slice(iStart, iEnd);
                this.nOffsetWithinParagraph = iStart;
                this.lToken = Array.from(_oTokenizer.genTokens(this.sSentence, true));
                this.dTokenPos.clear();
                for (let dToken of this.lToken) {
                    if (dToken["sType"] != "INFO") {
                        this.dTokenPos.set(dToken["nStart"], dToken);
                    }
                }
                if (bFullInfo) {
                    oSentence = { "nStart": iStart, "nEnd": iEnd, "sSentence": this.sSentence, "lToken": Array.from(this.lToken) };
                    // the list of tokens is duplicated, to keep all tokens from being deleted when analysis
                }
                this.parseText(this.sSentence, this.sSentence0, false, iStart, sCountry, dOpt, bShowRuleId, bDebug, bContext);
                if (bFullInfo) {
                    oSentence["aGrammarErrors"] = Array.from(this.dSentenceError.values());
                    lSentences.push(oSentence);
                    this.dSentenceError.clear();
                }
            }
            catch (e) {
                console.error(e);
            }
        }
        if (bFullInfo) {
            // Grammar checking and sentence analysis
            return lSentences;
        } else {
            // Grammar checking only
            return Array.from(this.dError.values());
        }
    }

    _getCleanText () {
        let sText = this.sText;
        if (sText.includes(" ")) {
            sText = sText.replace(/ /g, ' '); // nbsp
        }
        if (sText.includes(" ")) {
            sText = sText.replace(/ /g, ' '); // snbsp
        }
        if (sText.includes("'")) {
            sText = sText.replace(/'/g, "’");
        }
        if (sText.includes("‑")) {
            sText = sText.replace(/‑/g, "-"); // nobreakdash
        }
        if (sText.includes("@@")) {
            sText = sText.replace(/@@+/g, "");
        }
        return sText;
    }

    parseText (sText, sText0, bParagraph, nOffset, sCountry, dOptions, bShowRuleId, bDebug, bContext) {
        let bChange = false;
        let m;

        for (let [sOption, lRuleGroup] of gc_engine.getRules(bParagraph)) {

                                            case "-":
                                                // grammar error
                                                //console.log("-> error detected in " + sLineId + "\nzRegex: " + zRegex.source);
                                                let nErrorStart = nOffset + m.start[eAct[0]];
                                                if (!this.dError.has(nErrorStart) || nPriority > this.dErrorPriority.get(nErrorStart)) {
                                                    this.dError.set(nErrorStart, this._createErrorFromRegex(sText, sText0, sWhat, nOffset, m, eAct[0], sLineId, sRuleId, bUppercase, eAct[1], eAct[2], bShowRuleId, sOption, bContext));
                                                    this.dErrorPriority.set(nErrorStart, nPriority);
                                                    this.dSentenceError.set(nErrorStart, this.dError.get(nErrorStart));
                                                }
                                                break;
                                            case "~":
                                                // text processor
                                                //console.log("-> text processor by " + sLineId + "\nzRegex: " + zRegex.source);
                                                sText = this.rewriteText(sText, sWhat, eAct[0], m, bUppercase);
                                                bChange = true;

                                if (!this.lToken[nTokenErrorStart].hasOwnProperty("bImmune")) {
                                    let nTokenErrorEnd = (iTokenEnd > 0) ? nTokenOffset + iTokenEnd : nLastToken + iTokenEnd;
                                    let nErrorStart = this.nOffsetWithinParagraph + ((cStartLimit == "<") ? this.lToken[nTokenErrorStart]["nStart"] : this.lToken[nTokenErrorStart]["nEnd"]);
                                    let nErrorEnd = this.nOffsetWithinParagraph + ((cEndLimit == ">") ? this.lToken[nTokenErrorEnd]["nEnd"] : this.lToken[nTokenErrorEnd]["nStart"]);
                                    if (!this.dError.has(nErrorStart) || nPriority > this.dErrorPriority.gl_get(nErrorStart, -1)) {
                                        this.dError.set(nErrorStart, this._createErrorFromTokens(sWhat, nTokenOffset, nLastToken, nTokenErrorStart, nErrorStart, nErrorEnd, sLineId, sRuleId, bCaseSvty, sMessage, sURL, bShowRuleId, sOption, bContext));
                                        this.dErrorPriority.set(nErrorStart, nPriority);
                                        this.dSentenceError.set(nErrorStart, this.dError.get(nErrorStart));
                                        if (bDebug) {
                                            console.log("    NEW_ERROR: ",  this.dError.get(nErrorStart));
                                        }
                                    }
                                }
                            }
                            else if (cActionType == "~") {