Grammalecte  Diff

Differences From Artifact [d137f983fd]:

To Artifact [56d17d9958]:


365
366
367
368
369
370
371
372

373
374
375
376
377
378
379
365
366
367
368
369
370
371

372
373
374
375
376
377
378
379







-
+







            ['‰', "signe pour mille"],
        ]),

    _zPartDemForm: new RegExp("([a-zA-Zà-ö0-9À-Öø-ÿØ-ßĀ-ʯ]+)-(là|ci)$", "i"),
    _aPartDemExceptList: new Set(["celui", "celle", "ceux", "celles", "de", "jusque", "par", "marie-couche-toi"]),
    _zInterroVerb: new RegExp("([a-zA-Zà-ö0-9À-Öø-ÿØ-ßĀ-ʯ]+)(-(?:t-(?:ie?l|elle|on)|je|tu|ie?ls?|elles?|on|[nv]ous))$", "i"),
    _zImperatifVerb: new RegExp("([a-zA-Zà-ö0-9À-Öø-ÿØ-ßĀ-ʯ]+)(-(?:l(?:es?|a)-(?:moi|toi|lui|[nv]ous|leur)|y|en|[mts]['’ʼ‘‛´`′‵՚ꞌꞋ](?:y|en)|les?|la|[mt]oi|leur|lui))$", "i"),
    _zTag: new RegExp("[:;/][a-zA-Z0-9ÑÂĴĈŔÔṼŴ!][^:;/]*", "g"),
    _zTag: new RegExp("[:;/][a-zA-Z0-9É@*!][^:;/]*", "g"),

    split: function (sWord) {
        // returns an arry of strings (prefix, trimed_word, suffix)
        let sPrefix = "";
        let sSuffix = "";
        // préfixe élidé
        let m = /^([ldmtsnjcç]|lorsqu|presqu|jusqu|puisqu|quoiqu|quelqu|qu)['’ʼ‘‛´`′‵՚ꞌꞋ]([a-zA-Zà-öÀ-Ö0-9_ø-ÿØ-ßĀ-ʯfi-st-]+)/i.exec(sWord);
401
402
403
404
405
406
407


408







409
410
411
412

413
414
415

416
417
418
419







420
421
422
423
424







425
426
427
428
429
430
431
401
402
403
404
405
406
407
408
409

410
411
412
413
414
415
416
417
418


419



420




421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446







+
+
-
+
+
+
+
+
+
+


-
-
+
-
-
-
+
-
-
-
-
+
+
+
+
+
+
+





+
+
+
+
+
+
+







    },

    readableMorph: function (sMorph) {
        if (!sMorph) {
            return " mot inconnu";
        }
        let sRes = "";
        let sVType = "";
        if (sMorph.includes(":V")) {
        sMorph = sMorph.replace(/:V([0-3][ea_])[itpqnmr_eaxz]+/, ":V$1");
            sMorph = sMorph.replace(/:V([0-3][ea_])[itpqnmr_eaxz]+/, ":V$1");
        }
        if (sMorph.includes(":Q")) {
            let nVerbTag = sMorph.indexOf(":V")
            sVType = sMorph.slice(nVerbTag, nVerbTag+4);
            sMorph = sMorph.replace(/:V[0123]./, "").replace(/:1[ŝś]/, "");
        }
        let m;
        while ((m = this._zTag.exec(sMorph)) !== null) {
            if (this.dTag.has(m[0])) {
                sRes += this.dTag.get(m[0])[0];
            sRes += this._readableTag(m[0]);
            } else {
                sRes += " [" + m[0] + "]?";
            }
        }
        }
        if (sRes.startsWith(" verbe") && !sRes.includes("infinitif")) {
            sRes += " [" + sMorph.slice(1, sMorph.indexOf("/")) + "]";
        }
        if ((sRes.startsWith(" verbe") && !sRes.includes("infinitif")) || sRes.startsWith(" participe")) {
            if (sVType) {
                sRes += " [" + sMorph.slice(1, sMorph.indexOf("/")) + " : " + this._readableTag(sVType).gl_trimRight(",") + "]";
            } else {
                sRes += " [" + sMorph.slice(1, sMorph.indexOf("/")) + "]";
            }
        }
        if (!sRes) {
            return " [" + sMorph + "]: étiquettes inconnues";
        }
        return sRes.gl_trimRight(",");
    },

    _readableTag: function (sTag) {
        if (this.dTag.has(sTag)) {
            return this.dTag.get(sTag)[0];
        }
        return " [" + sTag + "]?";
    },

    setLabelsOnToken (oToken) {
        // Token: .sType, .sValue, .nStart, .nEnd, .lMorph
        let m = null;
        try {
            switch (oToken.sType) {
                case 'PUNC':