Grammalecte  Changes On Branch 2bfe79e9aa31c8ac

Hide Diff Patch Unified Diff

Changes In Branch bdic_opt Through [2bfe79e9aa] Excluding Merge-Ins

This is equivalent to a diff from 4a19028115 to 2bfe79e9aa

2020-09-15
16:57
[graphspell][py] suggestion: use Jaro-Winkler check-in: 7db21c89fd user: olr tags: graphspell, bdic_opt
14:09
[graphspell][js] str_transform: fix cleanWord() check-in: 2bfe79e9aa user: olr tags: graphspell, bdic_opt
14:01
[graphspell][js] remove specific trick in cleanWord() check-in: 6569849b49 user: olr tags: graphspell, bdic_opt
2020-09-14
07:55
[fr] ajustements check-in: cd8c458e7b user: olr tags: trunk, fr
2020-09-11
19:20
merge trunk check-in: 43afb8b856 user: olr tags: bdic_opt
19:18
[fr] tests: spellchecker.suggest() check-in: 4a19028115 user: olr tags: trunk, fr
14:21
[fr] faux positif check-in: 86f302f4ef user: olr tags: trunk, fr

Modified gc_lang/fr/modules/tests_modules.py from [1156e862bc] to [8a9c45807b].

︙︙
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32

def timeblock (label, hDst=None):
    "performance counter (contextmanager)"
    start = time.perf_counter()
    try:
        yield
    finally:
        end = time.perf_counter()
        print('{} : {}'.format(label, end - start))
        if hDst:
            hDst.write("{:<12.6}".format(end-start))


class TestDictionary (unittest.TestCase):
    "Test du correcteur orthographique"









|








18
19
20
21
22
23
24
25
26
27
28
29
30
31
32

def timeblock (label, hDst=None):
    "performance counter (contextmanager)"
    start = time.perf_counter()
    try:
        yield
    finally:
        end = time.perf_counter()
        print('{:<20} : {}'.format(label, end - start))
        if hDst:
            hDst.write("{:<12.6}".format(end-start))


class TestDictionary (unittest.TestCase):
    "Test du correcteur orthographique"
︙︙
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61

            self.assertTrue(self.oDic.isValid(sWord), sWord)

    def test_isvalid_failed (self):
        for sWord in ["BranchE", "BRanche", "BRAnCHE", "émilie", "éMILIE", "émiLie"]:
            self.assertFalse(self.oDic.isValid(sWord), sWord)

    def test_suggest (self):
        for sWord in ["déelirranttesss", "vallidasion", "Emilie", "exibission"]:
            with timeblock(sWord):
                self.assertNotEqual(0, self.oDic.suggest(sWord))


class TestConjugation (unittest.TestCase):
    "Tests des conjugaisons"









|








47
48
49
50
51
52
53
54
55
56
57
58
59
60
61

            self.assertTrue(self.oDic.isValid(sWord), sWord)

    def test_isvalid_failed (self):
        for sWord in ["BranchE", "BRanche", "BRAnCHE", "émilie", "éMILIE", "émiLie"]:
            self.assertFalse(self.oDic.isValid(sWord), sWord)

    def test_suggest (self):
        for sWord in ["déelirranttesss", "vallidasion", "Emilie", "exibission", "ditirembique", "jai", "email"]:
            with timeblock(sWord):
                self.assertNotEqual(0, self.oDic.suggest(sWord))


class TestConjugation (unittest.TestCase):
    "Tests des conjugaisons"
︙︙

Modified gc_lang/fr/perf_memo.text from [6a0d81df00] to [ad156793c1].

︙︙
26
27
28
29
30
31
32
33
34


0.6.2       2018.02.19 19:06    5.51302     1.29359     0.874157    0.260415    0.271596    0.290641    0.684754    0.376905    0.0815201   0.00919633  (spelling normalization)
1.0         2018.11.23 10:59    2.88577     0.702486    0.485648    0.139897    0.14079     0.148125    0.348751    0.201061    0.0360297   0.0043535   (x2, with new GC engine)
1.1         2019.05.16 09:42    1.50743     0.360923    0.261113    0.0749272   0.0763827   0.0771537   0.180504    0.102942    0.0182762   0.0021925   (×2, but new processor: AMD Ryzen 7 2700X)
1.2.1       2019.08.06 20:57    1.42886     0.358425    0.247356    0.0704405   0.0754886   0.0765604   0.177197    0.0988517   0.0188103   0.0020243
1.6.0       2020.01.03 20:22    1.38847     0.346214    0.240242    0.0709539   0.0737499   0.0748733   0.176477    0.0969171   0.0187857   0.0025143   (nouveau dictionnaire avec lemmes masculins)
1.9.0       2020.04.20 19:57    1.51183     0.369546    0.25681     0.0734314   0.0764396   0.0785668   0.183922    0.103674    0.0185812   0.002099    (NFC normalization)
1.9.2       2020.05.12 08:43    1.62465     0.398831    0.273012    0.0810811   0.080937    0.0845885   0.204133    0.114146    0.0212864   0.0029547
1.12.2      2020.09.09 13:34    1.50568     0.374504    0.233108    0.0798712   0.0804466   0.0769674   0.171519    0.0945132   0.0165344   0.0019474   
1.12.2      2020.09.09 13:35    1.41094     0.359093    0.236443    0.06968     0.0734418   0.0738087   0.169371    0.0946279   0.0167106   0.0019773   









|
|
>

26
27
28
29
30
31
32
33
34
35

0.6.2       2018.02.19 19:06    5.51302     1.29359     0.874157    0.260415    0.271596    0.290641    0.684754    0.376905    0.0815201   0.00919633  (spelling normalization)
1.0         2018.11.23 10:59    2.88577     0.702486    0.485648    0.139897    0.14079     0.148125    0.348751    0.201061    0.0360297   0.0043535   (x2, with new GC engine)
1.1         2019.05.16 09:42    1.50743     0.360923    0.261113    0.0749272   0.0763827   0.0771537   0.180504    0.102942    0.0182762   0.0021925   (×2, but new processor: AMD Ryzen 7 2700X)
1.2.1       2019.08.06 20:57    1.42886     0.358425    0.247356    0.0704405   0.0754886   0.0765604   0.177197    0.0988517   0.0188103   0.0020243
1.6.0       2020.01.03 20:22    1.38847     0.346214    0.240242    0.0709539   0.0737499   0.0748733   0.176477    0.0969171   0.0187857   0.0025143   (nouveau dictionnaire avec lemmes masculins)
1.9.0       2020.04.20 19:57    1.51183     0.369546    0.25681     0.0734314   0.0764396   0.0785668   0.183922    0.103674    0.0185812   0.002099    (NFC normalization)
1.9.2       2020.05.12 08:43    1.62465     0.398831    0.273012    0.0810811   0.080937    0.0845885   0.204133    0.114146    0.0212864   0.0029547
1.12.2      2020.09.09 13:34    1.50568     0.374504    0.233108    0.0798712   0.0804466   0.0769674   0.171519    0.0945132   0.0165344   0.0019474
1.12.2      2020.09.09 13:35    1.41094     0.359093    0.236443    0.06968     0.0734418   0.0738087   0.169371    0.0946279   0.0167106   0.0019773
1.12.2      2020.09.11 19:16    1.35297     0.330545    0.221731    0.0666998   0.0692539   0.0701707   0.160564    0.0891676   0.015807    0.0045998

Modified gc_lang/fr/webext/gce_worker.js from [c94d68b34b] to [d11e3267b8].

︙︙
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195

            oTokenizer = new Tokenizer("fr");
            if (dOptions !== null) {
                if (!(dOptions instanceof Map)) {
                    dOptions = helpers.objectToMap(dOptions);
                }
                gc_engine.setOptions(dOptions);
            }
            //tests();
            bInitDone = true;
        } else {
            console.log("[Worker] Already initialized…")
        }
        // we always retrieve options from the gc_engine, for setOptions filters obsolete options
        dOptions = helpers.mapToObject(gc_engine.getOptions());
        postMessage(createResponse("init", dOptions, oInfo, true));








|








181
182
183
184
185
186
187
188
189
190
191
192
193
194
195

            oTokenizer = new Tokenizer("fr");
            if (dOptions !== null) {
                if (!(dOptions instanceof Map)) {
                    dOptions = helpers.objectToMap(dOptions);
                }
                gc_engine.setOptions(dOptions);
            }
            tests();
            bInitDone = true;
        } else {
            console.log("[Worker] Already initialized…")
        }
        // we always retrieve options from the gc_engine, for setOptions filters obsolete options
        dOptions = helpers.mapToObject(gc_engine.getOptions());
        postMessage(createResponse("init", dOptions, oInfo, true));
︙︙
288
289
290
291
292
293
294
295
296
297
298
299
300
301









302
303
304
305
306
307
308

function resetOptions (oInfo={}) {
    gc_engine.resetOptions();
    let dOptions = helpers.mapToObject(gc_engine.getOptions());
    postMessage(createResponse("resetOptions", dOptions, oInfo, true));
}

function tests () {
    console.log(conj.getConj("devenir", ":E", ":2s"));
    console.log(mfsp.getMasForm("emmerdeuse", true));
    console.log(mfsp.getMasForm("pointilleuse", false));
    console.log(phonet.getSimil("est"));
    let aRes = gc_engine.parse("Je suit...");
    for (let oErr of aRes) {
        console.log(text.getReadableError(oErr));









    }
}

function textToTest (sText, sCountry, bDebug, bContext, oInfo={}) {
    if (!gc_engine) {
        postMessage(createResponse("textToTest", "# Grammar checker not loaded.", oInfo, true));
        return;








|






>
>
>
>
>
>
>
>
>








288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317

function resetOptions (oInfo={}) {
    gc_engine.resetOptions();
    let dOptions = helpers.mapToObject(gc_engine.getOptions());
    postMessage(createResponse("resetOptions", dOptions, oInfo, true));
}

function tests () {
    /*console.log(conj.getConj("devenir", ":E", ":2s"));
    console.log(mfsp.getMasForm("emmerdeuse", true));
    console.log(mfsp.getMasForm("pointilleuse", false));
    console.log(phonet.getSimil("est"));
    let aRes = gc_engine.parse("Je suit...");
    for (let oErr of aRes) {
        console.log(text.getReadableError(oErr));
    }*/
    for (let sWord of ["fatiqué", "coeur", "trèèèèèèèèès", "vraaaaiiiimeeeeennnt", "apele", "email", "Co2", "emmppâiiiller", "testt", "apelaion", "exsepttion", "sintaxik", "ebriete", "ennormmement"]) {
        console.time("Suggestions for " + sWord);
        for (let aSugg of oSpellChecker.suggest(sWord)) {
            if (aSugg.length) {
                console.log(sWord + " -> ", aSugg.join(" "));
            }
        }
        console.timeEnd("Suggestions for " + sWord);
    }
}

function textToTest (sText, sCountry, bDebug, bContext, oInfo={}) {
    if (!gc_engine) {
        postMessage(createResponse("textToTest", "# Grammar checker not loaded.", oInfo, true));
        return;
︙︙

Modified graphspell-js/char_player.js from [0602ec129b] to [dab3d08473]. 

1
2
3
4
5
6
7
8
9
10
11




12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46

// list of similar chars
// useful for suggestion mechanism

/* jshint esversion:6 */
/* jslint esversion:6 */

${map}


var char_player = {





    oDistanceBetweenChars: {
        "a": {},
        "e": {"é": 0.5},
        "é": {"e": 0.5},
        "i": {"y": 0.2},
        "o": {},
        "u": {},
        "y": {"i": 0.3},
        "b": {"d": 0.8, "h": 0.9},
        "c": {"ç": 0.1, "k": 0.5, "q": 0.5, "s": 0.5, "x": 0.5, "z": 0.8},
        "d": {"b": 0.8},
        "f": {"v": 0.8},
        "g": {"j": 0.5},
        "h": {"b": 0.9},
        "j": {"g": 0.5, "i": 0.9},
        "k": {"c": 0.5, "q": 0.1, "x": 0.5},
        "l": {"i": 0.9},
        "m": {"n": 0.8},
        "n": {"m": 0.8, "r": 0.9},
        "p": {"q": 0.9},
        "q": {"c": 0.5, "k": 0.1, "p": 0.9},
        "r": {"n": 0.9, "j": 0.9},
        "s": {"c": 0.5, "ç": 0.1, "x": 0.5, "z": 0.5},
        "t": {"d": 0.9},
        "v": {"f": 0.8, "w": 0.1},
        "w": {"v": 0.1},
        "x": {"c": 0.5, "k": 0.5, "q": 0.5, "s": 0.5},
        "z": {"s": 0.5}
    },

    distanceBetweenChars: function (c1, c2) {
        if (c1 == c2) {
            return 0;
        }
        if (this.oDistanceBetweenChars.hasOwnProperty(c1) && this.oDistanceBetweenChars[c1].hasOwnProperty(c2)) {











|
>
>
>
>

|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|








1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50

// list of similar chars
// useful for suggestion mechanism

/* jshint esversion:6 */
/* jslint esversion:6 */

${map}


var char_player = {
    /*
        oDistanceBetweenChars:
            - with Jaro-Winkler, values between 1 and 10
            - with Damerau-Levenshtein, values / 10 (between 0 and 1: 0.1, 0.2 ... 0.9)
    */
    oDistanceBetweenChars: {
        //"a": {},
        "e": {"é": 5},
        //"é": {"e": 5},
        "i": {"y": 2},
        //"o": {},
        //"u": {},
        "y": {"i": 3},
        "b": {"d": 8, "h": 9},
        "c": {"ç": 1, "k": 5, "q": 5, "s": 5, "x": 5, "z": 8},
        "d": {"b": 8},
        "f": {"v": 8},
        "g": {"j": 5},
        "h": {"b": 9},
        "j": {"g": 5, "i": 9},
        "k": {"c": 5, "q": 1, "x": 5},
        "l": {"i": 9},
        "m": {"n": 8},
        "n": {"m": 8, "r": 9},
        "p": {"q": 9},
        "q": {"c": 5, "k": 1, "p": 9},
        "r": {"n": 9, "j": 9},
        "s": {"c": 5, "ç": 1, "x": 5, "z": 5},
        "t": {"d": 9},
        "v": {"f": 8, "w": 1},
        "w": {"v": 1},
        "x": {"c": 5, "k": 5, "q": 5, "s": 5},
        "z": {"s": 5}
    },

    distanceBetweenChars: function (c1, c2) {
        if (c1 == c2) {
            return 0;
        }
        if (this.oDistanceBetweenChars.hasOwnProperty(c1) && this.oDistanceBetweenChars[c1].hasOwnProperty(c2)) {
︙︙
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338


339
340
341
342
343
344
345

        ["UN", ["EIN",]],
    ]),

    // End of word
    dFinal1: new Map([
        ["a", ["as", "at", "ant", "ah"]],
        ["A", ["AS", "AT", "ANT", "AH"]],
        ["c", ["ch",]],
        ["C", ["CH",]],
        ["e", ["et", "er", "ets", "ée", "ez", "ai", "ais", "ait", "ent", "eh"]],
        ["E", ["ET", "ER", "ETS", "ÉE", "EZ", "AI", "AIS", "AIT", "ENT", "EH"]],
        ["é", ["et", "er", "ets", "ée", "ez", "ai", "ais", "ait"]],
        ["É", ["ET", "ER", "ETS", "ÉE", "EZ", "AI", "AIS", "AIT"]],
        ["è", ["et", "er", "ets", "ée", "ez", "ai", "ais", "ait"]],
        ["È", ["ET", "ER", "ETS", "ÉE", "EZ", "AI", "AIS", "AIT"]],
        ["ê", ["et", "er", "ets", "ée", "ez", "ai", "ais", "ait"]],
        ["Ê", ["ET", "ER", "ETS", "ÉE", "EZ", "AI", "AIS", "AIT"]],
        ["ë", ["et", "er", "ets", "ée", "ez", "ai", "ais", "ait"]],
        ["Ë", ["ET", "ER", "ETS", "ÉE", "EZ", "AI", "AIS", "AIT"]],
        ["g", ["gh",]],
        ["G", ["GH",]],
        ["i", ["is", "it", "ie", "in"]],
        ["I", ["IS", "IT", "IE", "IN"]],


        ["n", ["nt", "nd", "ns", "nh"]],
        ["N", ["NT", "ND", "NS", "NH"]],
        ["o", ["aut", "ot", "os"]],
        ["O", ["AUT", "OT", "OS"]],
        ["ô", ["aut", "ot", "os"]],
        ["Ô", ["AUT", "OT", "OS"]],
        ["ö", ["aut", "ot", "os"]],








|
|














>
>








320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351

        ["UN", ["EIN",]],
    ]),

    // End of word
    dFinal1: new Map([
        ["a", ["as", "at", "ant", "ah"]],
        ["A", ["AS", "AT", "ANT", "AH"]],
        ["c", ["ch", "que"]],
        ["C", ["CH", "QUE"]],
        ["e", ["et", "er", "ets", "ée", "ez", "ai", "ais", "ait", "ent", "eh"]],
        ["E", ["ET", "ER", "ETS", "ÉE", "EZ", "AI", "AIS", "AIT", "ENT", "EH"]],
        ["é", ["et", "er", "ets", "ée", "ez", "ai", "ais", "ait"]],
        ["É", ["ET", "ER", "ETS", "ÉE", "EZ", "AI", "AIS", "AIT"]],
        ["è", ["et", "er", "ets", "ée", "ez", "ai", "ais", "ait"]],
        ["È", ["ET", "ER", "ETS", "ÉE", "EZ", "AI", "AIS", "AIT"]],
        ["ê", ["et", "er", "ets", "ée", "ez", "ai", "ais", "ait"]],
        ["Ê", ["ET", "ER", "ETS", "ÉE", "EZ", "AI", "AIS", "AIT"]],
        ["ë", ["et", "er", "ets", "ée", "ez", "ai", "ais", "ait"]],
        ["Ë", ["ET", "ER", "ETS", "ÉE", "EZ", "AI", "AIS", "AIT"]],
        ["g", ["gh",]],
        ["G", ["GH",]],
        ["i", ["is", "it", "ie", "in"]],
        ["I", ["IS", "IT", "IE", "IN"]],
        ["k", ["que"]],
        ["K", ["QUE"]],
        ["n", ["nt", "nd", "ns", "nh"]],
        ["N", ["NT", "ND", "NS", "NH"]],
        ["o", ["aut", "ot", "os"]],
        ["O", ["AUT", "OT", "OS"]],
        ["ô", ["aut", "ot", "os"]],
        ["Ô", ["AUT", "OT", "OS"]],
        ["ö", ["aut", "ot", "os"]],
︙︙

Modified graphspell-js/ibdawg.js from [69d7490b82] to [fa91a09215].

︙︙
18
19
20
21
22
23
24
25
26
27
28
29

30
31
32





33
34
35
36
37
38
39
40
41

42
43
44
45


46
47
48


49
50
51

52
53
54
55
56





57
58
59
60
61
62

63
64
65
66


67

68



69
70





71



72
73
74
75
76

77
78
79
80
81
82
83
84
85
86
87
88
89
90

91
92
93
94
95
96
97

    var char_player = require("./char_player.js");
}


class SuggResult {
    // Structure for storing, classifying and filtering suggestions

    constructor (sWord, nDistLimit=-1) {
        this.sWord = sWord;
        this.sSimplifiedWord = str_transform.simplifyWord(sWord);
        this.nDistLimit = (nDistLimit >= 0) ? nDistLimit :  Math.floor(sWord.length / 3) + 1;
        this.nMinDist = 1000;

        this.aSugg = new Set();
        this.dSugg = new Map([ [0, []],  [1, []],  [2, []] ]);
        this.aAllSugg = new Set();      // all found words even those refused





    }

    addSugg (sSugg, nDeep=0) {
        // add a suggestion
        if (this.aAllSugg.has(sSugg)) {
            return;
        }
        this.aAllSugg.add(sSugg);
        if (!this.aSugg.has(sSugg)) {

            let nDist = Math.floor(str_transform.distanceDamerauLevenshtein(this.sSimplifiedWord, str_transform.simplifyWord(sSugg)));
            if (nDist <= this.nDistLimit) {
                if (sSugg.includes(" ")) { // add 1 to distance for split suggestions
                    nDist += 1;


                }
                if (!this.dSugg.has(nDist)) {
                    this.dSugg.set(nDist, []);


                }
                this.dSugg.get(nDist).push(sSugg);
                this.aSugg.add(sSugg);

                if (nDist < this.nMinDist) {
                    this.nMinDist = nDist;
                }
                this.nDistLimit = Math.min(this.nDistLimit, this.nMinDist+1);
            }





        }
    }

    getSuggestions (nSuggLimit=10) {
        // return a list of suggestions
        let lRes = [];

        let bFirstListSorted = false;
        for (let [nDist, lSugg] of this.dSugg.entries()) {
            if (nDist > this.nDistLimit) {
                break;


            }

            if (!bFirstListSorted && lSugg.length > 1) {



                lRes.sort((a, b) => { return str_transform.distanceDamerauLevenshtein(this.sWord, a) - str_transform.distanceDamerauLevenshtein(this.sWord, b); });
                bFirstListSorted = true;





            }



            lRes.push(...lSugg);
            if (lRes.length > nSuggLimit) {
                break;
            }
        }

        if (this.sWord.gl_isUpperCase()) {
            lRes = lRes.map((sSugg) => { return sSugg.toUpperCase(); });
            lRes = [...new Set(lRes)];
        }
        else if (this.sWord.slice(0,1).gl_isUpperCase()) {
            lRes = lRes.map((sSugg) => { return sSugg.slice(0,1).toUpperCase() + sSugg.slice(1); });
            lRes = [...new Set(lRes)];
        }
        return lRes.slice(0, nSuggLimit);
    }

    reset () {
        this.aSugg.clear();
        this.dSugg.clear();

    }
}


class IBDAWG {
    // INDEXABLE BINARY DIRECT ACYCLIC WORD GRAPH









|




>
|
|
|
>
>
>
>
>


|





<
>
|
|
|
|
>
>
|
|
|
>
>
|
<
<
>
|
|
|
|
|
>
>
>
>
>



|


>
|
|
|
<
>
>

>
|
>
>
>
|
<
>
>
>
>
>
|
>
>
>
|
<
<


>








|



|
|
>








18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46

47
48
49
50
51
52
53
54
55
56
57
58
59


60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80

81
82
83
84
85
86
87
88
89

90
91
92
93
94
95
96
97
98
99


100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124

    var char_player = require("./char_player.js");
}


class SuggResult {
    // Structure for storing, classifying and filtering suggestions

    constructor (sWord, nSuggLimit=10, nDistLimit=-1) {
        this.sWord = sWord;
        this.sSimplifiedWord = str_transform.simplifyWord(sWord);
        this.nDistLimit = (nDistLimit >= 0) ? nDistLimit :  Math.floor(sWord.length / 3) + 1;
        this.nMinDist = 1000;
        // Temporary sets
        this.aAllSugg = new Set();  // All suggestions, even the one rejected
        this.dGoodSugg = new Map(); // Acceptable suggestions
        this.dBestSugg = new Map(); // Best suggestions
        // Parameters
        this.nSuggLimit = nSuggLimit;
        this.nSuggLimitExt = nSuggLimit + 2;                // we add few entries in case suggestions merge after casing modifications
        this.nBestSuggLimit = Math.floor(nSuggLimit * 1.5); // n times the requested limit
        this.nGoodSuggLimit = nSuggLimit * 15;              // n times the requested limit
    }

    addSugg (sSugg) {
        // add a suggestion
        if (this.aAllSugg.has(sSugg)) {
            return;
        }
        this.aAllSugg.add(sSugg);

        // jaro 0->1 1 les chaines sont égale
        let nDistJaro = 1 - str_transform.distanceJaroWinkler(this.sSimplifiedWord, str_transform.simplifyWord(sSugg));
        let nDist = Math.floor(nDistJaro * 10);
        if (nDistJaro < .11) {        // Best suggestions
            this.dBestSugg.set(sSugg, Math.round(nDistJaro*1000));
            if (this.dBestSugg.size > this.nBestSuggLimit) {
                this.nDistLimit = -1; // make suggest() to end search
            }
        } else if (nDistJaro < .33) { // Good suggestions
            this.dGoodSugg.set(sSugg, Math.round(nDistJaro*1000));
            if (this.dGoodSugg.size > this.nGoodSuggLimit) {
                this.nDistLimit = -1; // make suggest() to end search
            }


        } else {
            if (nDist < this.nMinDist) {
                this.nMinDist = nDist;
            }
            this.nDistLimit = Math.min(this.nDistLimit, this.nMinDist);
        }
        if (nDist <= this.nDistLimit) {
            if (nDist < this.nMinDist) {
                this.nMinDist = nDist;
            }
            this.nDistLimit = Math.min(this.nDistLimit, this.nMinDist+1);
        }
    }

    getSuggestions () {
        // return a list of suggestions
        let lRes = [];
        if (this.dBestSugg.size > 0) {
            // sort only with simplified words
            let lResTmp = [...this.dBestSugg.entries()].sort((a, b) => { return a[1] - b[1]; });
            let nSize = Math.min(this.nSuggLimitExt, lResTmp.length);

            for (let i=0;  i < nSize;  i++){
                lRes.push(lResTmp[i][0]);
            }
        }
        if (lRes.length < this.nSuggLimitExt) {
            // sort with simplified words and original word
            let lResTmp = [...this.dGoodSugg.entries()].sort((a, b) => {
                // Low precision to rely more on simplified words
                let nJaroA = Math.round(str_transform.distanceJaroWinkler(this.sWord, a[0]) * 10);

                let nJaroB = Math.round(str_transform.distanceJaroWinkler(this.sWord, b[0]) * 10);
                if (nJaroA == nJaroB) {
                    return a[1] - b[1];     // warning: both lists are NOT sorted the same way (key: a-b)
                } else {
                    return nJaroB - nJaroA; // warning: both lists are NOT sorted the same way (key: b-a)
                }
            }).slice(0, this.nSuggLimitExt);
            let nSize = Math.min(this.nSuggLimitExt, lResTmp.length);
            for (let i=0;  i < nSize;  i++){
                lRes.push(lResTmp[i][0]);


            }
        }
        // casing
        if (this.sWord.gl_isUpperCase()) {
            lRes = lRes.map((sSugg) => { return sSugg.toUpperCase(); });
            lRes = [...new Set(lRes)];
        }
        else if (this.sWord.slice(0,1).gl_isUpperCase()) {
            lRes = lRes.map((sSugg) => { return sSugg.slice(0,1).toUpperCase() + sSugg.slice(1); });
            lRes = [...new Set(lRes)];
        }
        return lRes.slice(0, this.nSuggLimit);
    }

    reset () {
        this.dSugg.clear();
        this.dGoodSugg.clear();
        this.dBestSugg.clear();
    }
}


class IBDAWG {
    // INDEXABLE BINARY DIRECT ACYCLIC WORD GRAPH
︙︙
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159












160

161


162


163




164
165
166
167
168

169

170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208

        }
        /*
            Properties:
            sName, nCompressionMethod, sHeader, lArcVal, nArcVal, sByDic, sLang, nChar, nBytesArc, nBytesNodeAddress,
            nEntry, nNode, nArc, nAff, cStemming, nTag, dChar, nBytesOffset,
        */

        /*
            Bug workaround.
            Mozilla’s JS parser sucks. Can’t read file bigger than 4 Mb!
            So we convert huge hexadecimal string to list of numbers…
            https://github.com/mozilla/addons-linter/issues/1361
        */
        let lTemp = [];
        for (let i = 0;  i < this.sByDic.length;  i+=2) {
            lTemp.push(parseInt(this.sByDic.slice(i, i+2), 16));
        }
        this.byDic = lTemp;
        //this.byDic = new Uint8Array(lTemp);  // not quicker, even slower
        /* end of bug workaround */

        if (!(this.sHeader.startsWith("/grammalecte-fsa/") || this.sHeader.startsWith("/pyfsa/"))) {
            throw TypeError("# Error. Not a grammalecte-fsa binary dictionary. Header: " + this.sHeader);
        }
        if (!(this.nCompressionMethod == 1 || this.nCompressionMethod == 2 || this.nCompressionMethod == 3)) {
            throw RangeError("# Error. Unknown dictionary compression method: " + this.nCompressionMethod);
        }
        // <dChar> to get the value of an arc, <dCharVal> to get the char of an arc with its value
        this.dChar = helpers.objectToMap(this.dChar);
        this.dCharVal = this.dChar.gl_reverse();
        this.a2grams = (this.l2grams) ? new Set(this.l2grams) : null;

        if (this.cStemming == "S") {
            this.funcStemming = str_transform.changeWordWithSuffixCode;
        } else if (this.cStemming == "A") {
            this.funcStemming = str_transform.changeWordWithAffixCode;
        } else {
            this.funcStemming = str_transform.noStemming;
        }













        this._arcMask = (2 ** ((this.nBytesArc * 8) - 3)) - 1;

        this._finalNodeMask = 1 << ((this.nBytesArc * 8) - 1);


        this._lastArcMask = 1 << ((this.nBytesArc * 8) - 2);








        // Configuring DAWG functions according to nCompressionMethod
        switch (this.nCompressionMethod) {
            case 1:
                this.morph = this._morph1;

                this.stem = this._stem1;

                this._lookupArcNode = this._lookupArcNode1;
                this._getArcs = this._getArcs1;
                this._writeNodes = this._writeNodes1;
                break;
            case 2:
                this.morph = this._morph2;
                this.stem = this._stem2;
                this._lookupArcNode = this._lookupArcNode2;
                this._getArcs = this._getArcs2;
                this._writeNodes = this._writeNodes2;
                break;
            case 3:
                this.morph = this._morph3;
                this.stem = this._stem3;
                this._lookupArcNode = this._lookupArcNode3;
                this._getArcs = this._getArcs3;
                this._writeNodes = this._writeNodes3;
                break;
            default:
                throw ValueError("# Error: unknown code: " + this.nCompressionMethod);
        }
        //console.log(this.getInfo());
        this.bAcronymValid = true;
        this.bNumAtLastValid = false;

        // lexicographer module ?
        this.lexicographer = null;
        // JS still sucks: we’ll try importation when importation will be available in Workers. Still waiting...
        if (self && self.hasOwnProperty("lexgraph_"+this.sLangCode)) { // self is the Worker
            this.lexicographer = self["lexgraph_"+this.sLangCode];
        }

    }

    getInfo () {
        return  `  Language: ${this.sLangName}   Lang code: ${this.sLangCode}   Dictionary name: ${this.sDicName}\n` +
                `  Compression method: ${this.nCompressionMethod}   Date: ${this.sDate}   Stemming: ${this.cStemming}FX\n` +
                `  Arcs values:  ${this.nArcVal} = ${this.nChar} characters,  ${this.nAff} affixes,  ${this.nTag} tags\n` +
                `  Dictionary: ${this.nEntry} entries,    ${this.nNode} nodes,   ${this.nArc} arcs\n` +








<
<
<
<
<
<
<
<
<
<
<
<
<
<



















>
>
>
>
>
>
>
>
>
>
>
>
|
>
|
>
>
|
>
>
|
>
>
>
>
|
<
<
|
<
>
|
>
|
<
<
<
<
|
|
|
<
<
<
<
<
<
<
<
<
<
<
<
|










<








147
148
149
150
151
152
153














154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198


199

200
201
202
203




204
205
206












207
208
209
210
211
212
213
214
215
216
217

218
219
220
221
222
223
224

        }
        /*
            Properties:
            sName, nCompressionMethod, sHeader, lArcVal, nArcVal, sByDic, sLang, nChar, nBytesArc, nBytesNodeAddress,
            nEntry, nNode, nArc, nAff, cStemming, nTag, dChar, nBytesOffset,
        */















        if (!(this.sHeader.startsWith("/grammalecte-fsa/") || this.sHeader.startsWith("/pyfsa/"))) {
            throw TypeError("# Error. Not a grammalecte-fsa binary dictionary. Header: " + this.sHeader);
        }
        if (!(this.nCompressionMethod == 1 || this.nCompressionMethod == 2 || this.nCompressionMethod == 3)) {
            throw RangeError("# Error. Unknown dictionary compression method: " + this.nCompressionMethod);
        }
        // <dChar> to get the value of an arc, <dCharVal> to get the char of an arc with its value
        this.dChar = helpers.objectToMap(this.dChar);
        this.dCharVal = this.dChar.gl_reverse();
        this.a2grams = (this.l2grams) ? new Set(this.l2grams) : null;

        if (this.cStemming == "S") {
            this.funcStemming = str_transform.changeWordWithSuffixCode;
        } else if (this.cStemming == "A") {
            this.funcStemming = str_transform.changeWordWithAffixCode;
        } else {
            this.funcStemming = str_transform.noStemming;
        }

        /*
            Bug workaround.
            Mozilla’s JS parser sucks. Can’t read file bigger than 4 Mb!
            So we convert huge hexadecimal string to list of numbers…
            https://github.com/mozilla/addons-linter/issues/1361
        */
        /*
            Performance trick:
            Instead of converting bytes to integers each times we parse the binary dictionary,
            we do it once, then parse the array
        */
        let nAcc = 0;
        let lBytesBuffer = [];
        let lTemp = [];
        let nDivisor = (this.nBytesArc + this.nBytesNodeAddress) / 2;
        for (let i = 0;  i < this.sByDic.length;  i+=2) {
            lBytesBuffer.push(parseInt(this.sByDic.slice(i, i+2), 16));
            if (nAcc == (this.nBytesArc - 1)) {
                lTemp.push(this._convBytesToInteger(lBytesBuffer));
                lBytesBuffer = [];
            }
            else if (nAcc == (this.nBytesArc + this.nBytesNodeAddress - 1)) {
                lTemp.push(Math.round(this._convBytesToInteger(lBytesBuffer) / nDivisor));  // Math.round should be useless, BUT with JS who knowns what can happen…
                lBytesBuffer = [];
                nAcc = -1;
            }


            nAcc = nAcc + 1;

        }
        this.byDic = lTemp;
        /* end of bug workaround */





        this._arcMask = (2 ** ((this.nBytesArc * 8) - 3)) - 1;
        this._finalNodeMask = 1 << ((this.nBytesArc * 8) - 1);
        this._lastArcMask = 1 << ((this.nBytesArc * 8) - 2);













        //console.log(this.getInfo());
        this.bAcronymValid = true;
        this.bNumAtLastValid = false;

        // lexicographer module ?
        this.lexicographer = null;
        // JS still sucks: we’ll try importation when importation will be available in Workers. Still waiting...
        if (self && self.hasOwnProperty("lexgraph_"+this.sLangCode)) { // self is the Worker
            this.lexicographer = self["lexgraph_"+this.sLangCode];
        }

    }

    getInfo () {
        return  `  Language: ${this.sLangName}   Lang code: ${this.sLangCode}   Dictionary name: ${this.sDicName}\n` +
                `  Compression method: ${this.nCompressionMethod}   Date: ${this.sDate}   Stemming: ${this.cStemming}FX\n` +
                `  Arcs values:  ${this.nArcVal} = ${this.nChar} characters,  ${this.nAff} affixes,  ${this.nTag} tags\n` +
                `  Dictionary: ${this.nEntry} entries,    ${this.nNode} nodes,   ${this.nArc} arcs\n` +
︙︙
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343

344
345
346
347
348
349
350
351
352
353
354
355
356

                return false;
            }
            iAddr = this._lookupArcNode(this.dChar.get(c), iAddr);
            if (iAddr === null) {
                return false;
            }
        }
        return Boolean(this._convBytesToInteger(this.byDic.slice(iAddr, iAddr+this.nBytesArc)) & this._finalNodeMask);
    }

    getMorph (sWord) {
        // retrieves morphologies list, different casing allowed
        if (!sWord) {
            return [];
        }
        sWord = str_transform.spellingNormalization(sWord);
        let l = this.morph(sWord);
        if (sWord[0].gl_isUpperCase()) {
            l.push(...this.morph(sWord.toLowerCase()));
            if (sWord.gl_isUpperCase() && sWord.length > 1) {
                l.push(...this.morph(sWord.gl_toCapitalize()));
            }
        }
        return l;
    }

    suggest (sWord, nSuggLimit=10, bSplitTrailingNumbers=false) {
        // returns a array of suggestions for <sWord>
        //console.time("Suggestions for " + sWord);
        sWord = str_transform.spellingNormalization(sWord);
        let sPfx = "";
        let sSfx = "";
        if (this.lexicographer) {
            [sPfx, sWord, sSfx] = this.lexicographer.split(sWord);
        }
        let nMaxSwitch = Math.max(Math.floor(sWord.length / 3), 1);
        let nMaxDel = Math.floor(sWord.length / 5);
        let nMaxHardRepl = Math.max(Math.floor((sWord.length - 5) / 4), 1);
        let nMaxJump = Math.max(Math.floor(sWord.length / 4), 1);
        let oSuggResult = new SuggResult(sWord);

        if (bSplitTrailingNumbers) {
            this._splitTrailingNumbers(oSuggResult, sWord);
        }
        this._splitSuggest(oSuggResult, sWord);
        this._suggest(oSuggResult, sWord, nMaxSwitch, nMaxDel, nMaxHardRepl, nMaxJump);
        let aSugg = oSuggResult.getSuggestions(nSuggLimit);
        if (this.lexicographer) {
            aSugg = this.lexicographer.filterSugg(aSugg);
        }
        if (sSfx || sPfx) {
            // we add what we removed
            return aSugg.map( (sSugg) => { return sPfx + sSugg + sSfx; } );
        }








|








|

|

|


















|
>





|








320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373

                return false;
            }
            iAddr = this._lookupArcNode(this.dChar.get(c), iAddr);
            if (iAddr === null) {
                return false;
            }
        }
        return Boolean(this.byDic[iAddr] & this._finalNodeMask);
    }

    getMorph (sWord) {
        // retrieves morphologies list, different casing allowed
        if (!sWord) {
            return [];
        }
        sWord = str_transform.spellingNormalization(sWord);
        let l = this._morph(sWord);
        if (sWord[0].gl_isUpperCase()) {
            l.push(...this._morph(sWord.toLowerCase()));
            if (sWord.gl_isUpperCase() && sWord.length > 1) {
                l.push(...this._morph(sWord.gl_toCapitalize()));
            }
        }
        return l;
    }

    suggest (sWord, nSuggLimit=10, bSplitTrailingNumbers=false) {
        // returns a array of suggestions for <sWord>
        //console.time("Suggestions for " + sWord);
        sWord = str_transform.spellingNormalization(sWord);
        let sPfx = "";
        let sSfx = "";
        if (this.lexicographer) {
            [sPfx, sWord, sSfx] = this.lexicographer.split(sWord);
        }
        let nMaxSwitch = Math.max(Math.floor(sWord.length / 3), 1);
        let nMaxDel = Math.floor(sWord.length / 5);
        let nMaxHardRepl = Math.max(Math.floor((sWord.length - 5) / 4), 1);
        let nMaxJump = Math.max(Math.floor(sWord.length / 4), 1);
        let oSuggResult = new SuggResult(sWord, nSuggLimit);
        sWord = str_transform.cleanWord(sWord);
        if (bSplitTrailingNumbers) {
            this._splitTrailingNumbers(oSuggResult, sWord);
        }
        this._splitSuggest(oSuggResult, sWord);
        this._suggest(oSuggResult, sWord, nMaxSwitch, nMaxDel, nMaxHardRepl, nMaxJump);
        let aSugg = oSuggResult.getSuggestions();
        if (this.lexicographer) {
            aSugg = this.lexicographer.filterSugg(aSugg);
        }
        if (sSfx || sPfx) {
            // we add what we removed
            return aSugg.map( (sSugg) => { return sPfx + sSugg + sSfx; } );
        }
︙︙
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390

            }
        }
    }

    _suggest (oSuggResult, sRemain, nMaxSwitch=0, nMaxDel=0, nMaxHardRepl=0, nMaxJump=0, nDist=0, nDeep=0, iAddr=0, sNewWord="", bAvoidLoop=false) {
        // returns a set of suggestions
        // recursive function
        if (this._convBytesToInteger(this.byDic.slice(iAddr, iAddr+this.nBytesArc)) & this._finalNodeMask) {
            if (sRemain == "") {
                oSuggResult.addSugg(sNewWord);
                for (let sTail of this._getTails(iAddr)) {
                    oSuggResult.addSugg(sNewWord+sTail);
                }
                return;
            }








|








393
394
395
396
397
398
399
400
401
402
403
404
405
406
407

            }
        }
    }

    _suggest (oSuggResult, sRemain, nMaxSwitch=0, nMaxDel=0, nMaxHardRepl=0, nMaxJump=0, nDist=0, nDeep=0, iAddr=0, sNewWord="", bAvoidLoop=false) {
        // returns a set of suggestions
        // recursive function
        if (this.byDic[iAddr] & this._finalNodeMask) {
            if (sRemain == "") {
                oSuggResult.addSugg(sNewWord);
                for (let sTail of this._getTails(iAddr)) {
                    oSuggResult.addSugg(sNewWord+sTail);
                }
                return;
            }
︙︙
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514

    }

    _getTails (iAddr, sTail="", n=2) {
        // return a list of suffixes ending at a distance of <n> from <iAddr>
        let aTails = new Set();
        for (let [nVal, jAddr] of this._getArcs(iAddr)) {
            if (nVal <= this.nChar) {
                if (this._convBytesToInteger(this.byDic.slice(jAddr, jAddr+this.nBytesArc)) & this._finalNodeMask) {
                    aTails.add(sTail + this.dCharVal.get(nVal));
                }
                if (n && aTails.size == 0) {
                    aTails.gl_update(this._getTails(jAddr, sTail+this.dCharVal.get(nVal), n-1));
                }
            }
        }
        return aTails;
    }

    // morph (sWord) {
    //     is defined in constructor
    // }

    getSimilarEntries (sWord, nSuggLimit=10) {
        // return a list of tuples (similar word, stem, morphology)
        if (sWord == "") {
            return [];
        }
        let lResult = [];
        for (let sSimilar of this.suggest(sWord, nSuggLimit)) {








|










<
<
<
<








503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520




521
522
523
524
525
526
527

    }

    _getTails (iAddr, sTail="", n=2) {
        // return a list of suffixes ending at a distance of <n> from <iAddr>
        let aTails = new Set();
        for (let [nVal, jAddr] of this._getArcs(iAddr)) {
            if (nVal <= this.nChar) {
                if (this.byDic[jAddr] & this._finalNodeMask) {
                    aTails.add(sTail + this.dCharVal.get(nVal));
                }
                if (n && aTails.size == 0) {
                    aTails.gl_update(this._getTails(jAddr, sTail+this.dCharVal.get(nVal), n-1));
                }
            }
        }
        return aTails;
    }





    getSimilarEntries (sWord, nSuggLimit=10) {
        // return a list of tuples (similar word, stem, morphology)
        if (sWord == "") {
            return [];
        }
        let lResult = [];
        for (let sSimilar of this.suggest(sWord, nSuggLimit)) {
︙︙
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692

            zFlexPattern = (sFlexPattern !== "") ? new RegExp(sFlexPattern) : null;
            zTagsPattern = (sTagsPattern !== "") ? new RegExp(sTagsPattern) : null;
        }
        catch (e) {
            console.log("Error in regex pattern");
            console.log(e.message);
        }
        yield* this._select1(zFlexPattern, zTagsPattern, 0, "");
    }

    // VERSION 1

    * _select1 (zFlexPattern, zTagsPattern, iAddr, sWord) {
        // recursive generator
        for (let [nVal, jAddr] of this._getArcs1(iAddr)) {
            if (nVal <= this.nChar) {
                // simple character
                yield* this._select1(zFlexPattern, zTagsPattern, jAddr, sWord + this.lArcVal[nVal]);
            } else {
                if (!zFlexPattern || zFlexPattern.test(sWord)) {
                    let sStem = this.funcStemming(sWord, this.lArcVal[nVal]);
                    for (let [nMorphVal, _] of this._getArcs1(jAddr)) {
                        if (!zTagsPattern || zTagsPattern.test(this.lArcVal[nMorphVal])) {
                            yield [sWord, sStem, this.lArcVal[nMorphVal]];
                        }
                    }
                }
            }
        }
    }

    _morph1 (sWord) {
        // returns morphologies of sWord
        let iAddr = 0;
        for (let c of sWord) {
            if (!this.dChar.has(c)) {
                return [];
            }
            iAddr = this._lookupArcNode(this.dChar.get(c), iAddr);
            if (iAddr === null) {
                return [];
            }
        }
        if (this._convBytesToInteger(this.byDic.slice(iAddr, iAddr+this.nBytesArc)) & this._finalNodeMask) {
            let l = [];
            let nRawArc = 0;
            while (!(nRawArc & this._lastArcMask)) {
                let iEndArcAddr = iAddr + this.nBytesArc;
                nRawArc = this._convBytesToInteger(this.byDic.slice(iAddr, iEndArcAddr));
                let nArc = nRawArc & this._arcMask;
                if (nArc > this.nChar) {
                    // This value is not a char, this is a stemming code
                    let sStem = ">" + this.funcStemming(sWord, this.lArcVal[nArc]);
                    // Now , we go to the next node and retrieve all following arcs values, all of them are tags
                    let iAddr2 = this._convBytesToInteger(this.byDic.slice(iEndArcAddr, iEndArcAddr+this.nBytesNodeAddress));
                    let nRawArc2 = 0;
                    while (!(nRawArc2 & this._lastArcMask)) {
                        let iEndArcAddr2 = iAddr2 + this.nBytesArc;
                        nRawArc2 = this._convBytesToInteger(this.byDic.slice(iAddr2, iEndArcAddr2));
                        l.push(sStem + "/" + this.lArcVal[nRawArc2 & this._arcMask]);
                        iAddr2 = iEndArcAddr2+this.nBytesNodeAddress;
                    }
                }
                iAddr = iEndArcAddr + this.nBytesNodeAddress;
            }
            return l;
        }
        return [];
    }

    _stem1 (sWord) {
        // returns stems list of sWord
        let iAddr = 0;
        for (let c of sWord) {
            if (!this.dChar.has(c)) {
                return [];
            }
            iAddr = this._lookupArcNode(this.dChar.get(c), iAddr);
            if (iAddr === null) {
                return [];
            }
        }
        if (this._convBytesToInteger(this.byDic.slice(iAddr, iAddr+this.nBytesArc)) & this._finalNodeMask) {
            let l = [];
            let nRawArc = 0;
            while (!(nRawArc & this._lastArcMask)) {
                let iEndArcAddr = iAddr + this.nBytesArc;
                nRawArc = this._convBytesToInteger(this.byDic.slice(iAddr, iEndArcAddr));
                let nArc = nRawArc & this._arcMask;
                if (nArc > this.nChar) {
                    // This value is not a char, this is a stemming code
                    l.push(this.funcStemming(sWord, this.lArcVal[nArc]));
                }
                iAddr = iEndArcAddr + this.nBytesNodeAddress;
            }
            return l;
        }
        return [];
    }

    _lookupArcNode1 (nVal, iAddr) {
        // looks if nVal is an arc at the node at iAddr, if yes, returns address of next node else None
        while (true) {
            let iEndArcAddr = iAddr+this.nBytesArc;
            let nRawArc = this._convBytesToInteger(this.byDic.slice(iAddr, iEndArcAddr));
            if (nVal == (nRawArc & this._arcMask)) {
                // the value we are looking for
                // we return the address of the next node
                return this._convBytesToInteger(this.byDic.slice(iEndArcAddr, iEndArcAddr+this.nBytesNodeAddress));
            }
            else {
                // value not found
                if (nRawArc & this._lastArcMask) {
                    return null;
                }
                iAddr = iEndArcAddr + this.nBytesNodeAddress;
            }
        }
    }

    * _getArcs1 (iAddr) {
        // generator: return all arcs at <iAddr> as tuples of (nVal, iAddr)
        while (true) {
            let iEndArcAddr = iAddr+this.nBytesArc;
            let nRawArc = this._convBytesToInteger(this.byDic.slice(iAddr, iEndArcAddr));
            yield [nRawArc & this._arcMask, this._convBytesToInteger(this.byDic.slice(iEndArcAddr, iEndArcAddr+this.nBytesNodeAddress))];
            if (nRawArc & this._lastArcMask) {
                break;
            }
            iAddr = iEndArcAddr+this.nBytesNodeAddress;
        }
    }

    // VERSION 2
    _morph2 (sWord) {
        // to do
    }

    _stem2 (sWord) {
        // to do
    }

    _lookupArcNode2 (nVal, iAddr) {
        // to do
    }


    // VERSION 3
    _morph3 (sWord) {
        // to do
    }

    _stem3 (sWord) {
        // to do
    }

    _lookupArcNode3 (nVal, iAddr) {
        // to do
    }
}


if (typeof(exports) !== 'undefined') {
    exports.IBDAWG = IBDAWG;
}








|


<
<
|

|


|



|









|











|



|
|





|


|
|

|


|






|











|



|
|





|






|


|
|



|






|




|


|
|
|



|

<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<








541
542
543
544
545
546
547
548
549
550


551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669



























670
671
672
673
674
675
676

            zFlexPattern = (sFlexPattern !== "") ? new RegExp(sFlexPattern) : null;
            zTagsPattern = (sTagsPattern !== "") ? new RegExp(sTagsPattern) : null;
        }
        catch (e) {
            console.log("Error in regex pattern");
            console.log(e.message);
        }
        yield* this._select(zFlexPattern, zTagsPattern, 0, "");
    }



    * _select (zFlexPattern, zTagsPattern, iAddr, sWord) {
        // recursive generator
        for (let [nVal, jAddr] of this._getArcs(iAddr)) {
            if (nVal <= this.nChar) {
                // simple character
                yield* this._select(zFlexPattern, zTagsPattern, jAddr, sWord + this.lArcVal[nVal]);
            } else {
                if (!zFlexPattern || zFlexPattern.test(sWord)) {
                    let sStem = this.funcStemming(sWord, this.lArcVal[nVal]);
                    for (let [nMorphVal, _] of this._getArcs(jAddr)) {
                        if (!zTagsPattern || zTagsPattern.test(this.lArcVal[nMorphVal])) {
                            yield [sWord, sStem, this.lArcVal[nMorphVal]];
                        }
                    }
                }
            }
        }
    }

    _morph (sWord) {
        // returns morphologies of sWord
        let iAddr = 0;
        for (let c of sWord) {
            if (!this.dChar.has(c)) {
                return [];
            }
            iAddr = this._lookupArcNode(this.dChar.get(c), iAddr);
            if (iAddr === null) {
                return [];
            }
        }
        if (this.byDic[iAddr] & this._finalNodeMask) {
            let l = [];
            let nRawArc = 0;
            while (!(nRawArc & this._lastArcMask)) {
                let iEndArcAddr = iAddr + 1;
                nRawArc = this.byDic[iAddr];
                let nArc = nRawArc & this._arcMask;
                if (nArc > this.nChar) {
                    // This value is not a char, this is a stemming code
                    let sStem = ">" + this.funcStemming(sWord, this.lArcVal[nArc]);
                    // Now , we go to the next node and retrieve all following arcs values, all of them are tags
                    let iAddr2 = this.byDic[iEndArcAddr];
                    let nRawArc2 = 0;
                    while (!(nRawArc2 & this._lastArcMask)) {
                        let iEndArcAddr2 = iAddr2 + 1;
                        nRawArc2 = this.byDic[iAddr2];
                        l.push(sStem + "/" + this.lArcVal[nRawArc2 & this._arcMask]);
                        iAddr2 = iEndArcAddr2 + 1;
                    }
                }
                iAddr = iEndArcAddr + 1;
            }
            return l;
        }
        return [];
    }

    _stem (sWord) {
        // returns stems list of sWord
        let iAddr = 0;
        for (let c of sWord) {
            if (!this.dChar.has(c)) {
                return [];
            }
            iAddr = this._lookupArcNode(this.dChar.get(c), iAddr);
            if (iAddr === null) {
                return [];
            }
        }
        if (this.byDic[iAddr] & this._finalNodeMask) {
            let l = [];
            let nRawArc = 0;
            while (!(nRawArc & this._lastArcMask)) {
                let iEndArcAddr = iAddr + 1;
                nRawArc = this.byDic[iAddr];
                let nArc = nRawArc & this._arcMask;
                if (nArc > this.nChar) {
                    // This value is not a char, this is a stemming code
                    l.push(this.funcStemming(sWord, this.lArcVal[nArc]));
                }
                iAddr = iEndArcAddr + 1;
            }
            return l;
        }
        return [];
    }

    _lookupArcNode (nVal, iAddr) {
        // looks if nVal is an arc at the node at iAddr, if yes, returns address of next node else None
        while (true) {
            let iEndArcAddr = iAddr+1;
            let nRawArc = this.byDic[iAddr];
            if (nVal == (nRawArc & this._arcMask)) {
                // the value we are looking for
                // we return the address of the next node
                return this.byDic[iEndArcAddr];
            }
            else {
                // value not found
                if (nRawArc & this._lastArcMask) {
                    return null;
                }
                iAddr = iEndArcAddr + 1;
            }
        }
    }

    * _getArcs (iAddr) {
        // generator: return all arcs at <iAddr> as tuples of (nVal, iAddr)
        while (true) {
            let iEndArcAddr = iAddr+1;
            let nRawArc = this.byDic[iAddr];
            yield [nRawArc & this._arcMask, this.byDic[iEndArcAddr]];
            if (nRawArc & this._lastArcMask) {
                break;
            }
            iAddr = iEndArcAddr+1;
        }



























    }
}


if (typeof(exports) !== 'undefined') {
    exports.IBDAWG = IBDAWG;
}

Modified graphspell-js/str_transform.js from [4c4ee6009e] to [3bf8e3b480]. 

1
2
3
4
5
6
7
8
9
10
11
12
13

14
15
16
17
18
19
20

// STRING TRANSFORMATION

/* jshint esversion:6, -W097 */
/* jslint esversion:6 */
/* global exports, console */

"use strict";


if (typeof(process) !== 'undefined') {
    var char_player = require("./char_player.js");
}



// Note: 48 is the ASCII code for "0"

var str_transform = {

    getNgrams: function (sWord, n=2) {
        let lNgrams = [];














>








1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

// STRING TRANSFORMATION

/* jshint esversion:6, -W097 */
/* jslint esversion:6 */
/* global exports, console */

"use strict";


if (typeof(process) !== 'undefined') {
    var char_player = require("./char_player.js");
}



// Note: 48 is the ASCII code for "0"

var str_transform = {

    getNgrams: function (sWord, n=2) {
        let lNgrams = [];
︙︙
61
62
63
64
65
66
67






68
69
70
71
72
73
74

            if (c != sWord.slice(i, i+1) || (c == 'e' && sWord.slice(i, i+2) != "ee")) {  // exception for <e> to avoid confusion between crée / créai
                sNewWord += c;
            }
            i++;
        }
        return sNewWord.replace(/eau/g, "o").replace(/au/g, "o").replace(/ai/g, "éi").replace(/ei/g, "é").replace(/ph/g, "f");
    },







    _xTransNumbersToExponent: new Map([
        ["0", "⁰"], ["1", "¹"], ["2", "²"], ["3", "³"], ["4", "⁴"], ["5", "⁵"], ["6", "⁶"], ["7", "⁷"], ["8", "⁸"], ["9", "⁹"]
    ]),

    numbersToExponent: function (sWord) {
        let sNewWord = "";








>
>
>
>
>
>








62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81

            if (c != sWord.slice(i, i+1) || (c == 'e' && sWord.slice(i, i+2) != "ee")) {  // exception for <e> to avoid confusion between crée / créai
                sNewWord += c;
            }
            i++;
        }
        return sNewWord.replace(/eau/g, "o").replace(/au/g, "o").replace(/ai/g, "éi").replace(/ei/g, "é").replace(/ph/g, "f");
    },

    cleanWord: function (sWord) {
        // word clean for the user who make commun and preditive error help suggest
        // remove letters repeated more than 2 times
        return sWord.replace(/(.)\1{2,}/ig,'$1$1');
    },

    _xTransNumbersToExponent: new Map([
        ["0", "⁰"], ["1", "¹"], ["2", "²"], ["3", "³"], ["4", "⁴"], ["5", "⁵"], ["6", "⁶"], ["7", "⁷"], ["8", "⁸"], ["9", "⁹"]
    ]),

    numbersToExponent: function (sWord) {
        let sNewWord = "";
︙︙
148
149
150
151
152
153
154



























































































155
156
157
158
159
160
161

            }
            return matrix[nLen1][nLen2];
        }
        catch (e) {
            console.error(e);
        }
    },




























































































    showDistance (s1, s2) {
        console.log(`Distance: ${s1} / ${s2} = ${this.distanceDamerauLevenshtein(s1, s2)})`);
    },

    // Suffix only
    defineSuffixCode: function (sFlex, sStem) {








>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>








155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259

            }
            return matrix[nLen1][nLen2];
        }
        catch (e) {
            console.error(e);
        }
    },

    distanceJaroWinkler: function(a, b, boost = .666) {
        // https://github.com/thsig/jaro-winkler-JS
        //if (a == b) { return 1.0; }
        let a_len = a.length;
        let b_len = b.length;
        let a_flag = [];
        let b_flag = [];
        let search_range = Math.floor(Math.max(a_len, b_len) / 2) - 1;
        let minv = Math.min(a_len, b_len);

        // Looking only within the search range, count and flag the matched pairs.
        let Num_com = 0;
        let yl1 = b_len - 1;
        for (let i = 0; i < a_len; i++) {
          let lowlim = (i >= search_range) ? i - search_range : 0;
          let hilim  = ((i + search_range) <= yl1) ? (i + search_range) : yl1;
          for (let j = lowlim; j <= hilim; j++) {
            if (b_flag[j] !== 1 && a[j] === b[i]) {
              a_flag[j] = 1;
              b_flag[i] = 1;
              Num_com++;
              break;
            }
          }
        }

        // Return if no characters in common
        if (Num_com === 0) { return 0.0; }

        // Count the number of transpositions
        let k = 0;
        let N_trans = 0;
        for (let i = 0; i < a_len; i++) {
          if (a_flag[i] === 1) {
            let j;
            for (j = k; j < b_len; j++) {
              if (b_flag[j] === 1) {
                k = j + 1;
                break;
              }
            }
            if (a[i] !== b[j]) { N_trans++; }
          }
        }
        N_trans = Math.floor(N_trans / 2);

        // Adjust for similarities in nonmatched characters
        let N_simi = 0;
        let adjwt = char_player.oDistanceBetweenChars;
        if (minv > Num_com) {
          for (let i = 0; i < a_len; i++) {
            if (!a_flag[i]) {
              for (let j = 0; j < b_len; j++) {
                if (!b_flag[j]) {
                  if (adjwt[a[i]] && adjwt[a[i]][b[j]]) {
                    N_simi += adjwt[a[i]][b[j]];
                    b_flag[j] = 2;
                    break;
                  }
                }
              }
            }
          }
        }

        let Num_sim = (N_simi / 10.0) + Num_com;

        // Main weight computation
        let weight = Num_sim / a_len + Num_sim / b_len + (Num_com - N_trans) / Num_com;
        weight = weight / 3;

        // Continue to boost the weight if the strings are similar
        if (weight > boost) {
          // Adjust for having up to the first 4 characters in common
          let j = (minv >= 4) ? 4 : minv;
          let i;
          for (i = 0; (i < j) && a[i] === b[i]; i++) { }
          if (i) { weight += i * 0.1 * (1.0 - weight) };

          // Adjust for long strings.
          // After agreeing beginning chars, at least two more must agree
          // and the agreeing characters must be more than half of the
          // remaining characters.
          if (minv > 4 && Num_com > i + 1 && 2 * Num_com >= minv + i) {
            weight += (1 - weight) * ((Num_com - i - 1) / (a_len * b_len - i*2 + 2));
          }
        }

        return weight;
    },

    showDistance (s1, s2) {
        console.log(`Distance: ${s1} / ${s2} = ${this.distanceDamerauLevenshtein(s1, s2)})`);
    },

    // Suffix only
    defineSuffixCode: function (sFlex, sStem) {
︙︙

Modified graphspell/ibdawg.py from [d16ed0d683] to [bda5a789eb].

︙︙
116
117
118
119
120
121
122




















123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164

            else:
                raise OSError("# Error. Unknown file type: "+source)
        else:
            self._initJSON(source)

        self.sFileName = source  if isinstance(source, str)  else "[None]"





















        self._arcMask = (2 ** ((self.nBytesArc * 8) - 3)) - 1
        self._finalNodeMask = 1 << ((self.nBytesArc * 8) - 1)
        self._lastArcMask = 1 << ((self.nBytesArc * 8) - 2)
        self._addrBitMask = 1 << ((self.nBytesArc * 8) - 3)  # version 2

        # function to decode the affix/suffix code
        if self.cStemming == "S":
            self.funcStemming = st.changeWordWithSuffixCode
        elif self.cStemming == "A":
            self.funcStemming = st.changeWordWithAffixCode
        else:
            self.funcStemming = st.noStemming

        # Configuring DAWG functions according to nCompressionMethod
        if self.nCompressionMethod == 1:
            self.morph = self._morph1
            self.stem = self._stem1
            self._lookupArcNode = self._lookupArcNode1
            self._getArcs = self._getArcs1
            self._writeNodes = self._writeNodes1
        elif self.nCompressionMethod == 2:
            self.morph = self._morph2
            self.stem = self._stem2
            self._lookupArcNode = self._lookupArcNode2
            self._getArcs = self._getArcs2
            self._writeNodes = self._writeNodes2
        elif self.nCompressionMethod == 3:
            self.morph = self._morph3
            self.stem = self._stem3
            self._lookupArcNode = self._lookupArcNode3
            self._getArcs = self._getArcs3
            self._writeNodes = self._writeNodes3
        else:
            raise ValueError("  # Error: unknown code: {}".format(self.nCompressionMethod))

        self.bAcronymValid = False
        self.bNumAtLastValid = False

        # lexicographer module ?
        self.lexicographer = None
        try:
            self.lexicographer = importlib.import_module(".lexgraph_"+self.sLangCode, "grammalecte.graphspell")








>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>



<









<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<








116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145

146
147
148
149
150
151
152
153
154






















155
156
157
158
159
160
161

            else:
                raise OSError("# Error. Unknown file type: "+source)
        else:
            self._initJSON(source)

        self.sFileName = source  if isinstance(source, str)  else "[None]"

        # Performance trick:
        #     Instead of converting bytes to integers each times we parse the binary dictionary,
        #     we do it once, then parse the array
        nAcc = 0
        byBuffer = b""
        lTemp = []
        nDivisor = (self.nBytesArc + self.nBytesNodeAddress) / 2
        for i in range(0, len(self.byDic)):
            byBuffer += self.byDic[i:i+1]
            if nAcc == (self.nBytesArc - 1):
                lTemp.append(int.from_bytes(byBuffer, byteorder="big"))
                byBuffer = b""
            elif nAcc == (self.nBytesArc + self.nBytesNodeAddress - 1):
                lTemp.append(round(int.from_bytes(byBuffer, byteorder="big") / nDivisor))
                byBuffer = b""
                nAcc = -1
            nAcc = nAcc + 1
        self.byDic = lTemp;

        # masks
        self._arcMask = (2 ** ((self.nBytesArc * 8) - 3)) - 1
        self._finalNodeMask = 1 << ((self.nBytesArc * 8) - 1)
        self._lastArcMask = 1 << ((self.nBytesArc * 8) - 2)


        # function to decode the affix/suffix code
        if self.cStemming == "S":
            self.funcStemming = st.changeWordWithSuffixCode
        elif self.cStemming == "A":
            self.funcStemming = st.changeWordWithAffixCode
        else:
            self.funcStemming = st.noStemming























        self.bAcronymValid = False
        self.bNumAtLastValid = False

        # lexicographer module ?
        self.lexicographer = None
        try:
            self.lexicographer = importlib.import_module(".lexgraph_"+self.sLangCode, "grammalecte.graphspell")
︙︙
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214

        self.cStemming = l.pop(0)
        self.nTag = self.nArcVal - self.nChar - self.nAff
        # <dChar> to get the value of an arc, <dCharVal> to get the char of an arc with its value
        self.dChar = {}
        for i in range(1, self.nChar+1):
            self.dChar[self.lArcVal[i]] = i
        self.dCharVal = { v: k  for k, v in self.dChar.items() }
        self.nBytesOffset = 1 # version 3

    def _initJSON (self, oJSON):
        "initialize with a JSON text file"
        self.sByDic = ""  # init to prevent pylint whining
        self.__dict__.update(oJSON)
        self.byDic = binascii.unhexlify(self.sByDic)
        self.dCharVal = { v: k  for k, v in self.dChar.items() }








<








197
198
199
200
201
202
203

204
205
206
207
208
209
210

        self.cStemming = l.pop(0)
        self.nTag = self.nArcVal - self.nChar - self.nAff
        # <dChar> to get the value of an arc, <dCharVal> to get the char of an arc with its value
        self.dChar = {}
        for i in range(1, self.nChar+1):
            self.dChar[self.lArcVal[i]] = i
        self.dCharVal = { v: k  for k, v in self.dChar.items() }


    def _initJSON (self, oJSON):
        "initialize with a JSON text file"
        self.sByDic = ""  # init to prevent pylint whining
        self.__dict__.update(oJSON)
        self.byDic = binascii.unhexlify(self.sByDic)
        self.dCharVal = { v: k  for k, v in self.dChar.items() }
︙︙
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258

                "nNode": self.nNode,
                "nArc": self.nArc,
                "nArcVal": self.nArcVal,
                "lArcVal": self.lArcVal,
                "nCompressionMethod": self.nCompressionMethod,
                "nBytesArc": self.nBytesArc,
                "nBytesNodeAddress": self.nBytesNodeAddress,
                "nBytesOffset": self.nBytesOffset,
                # JavaScript is a pile of shit, so Mozilla’s JS parser don’t like file bigger than 4 Mb!
                # So, if necessary, we use an hexadecimal string, that we will convert later in Firefox’s extension.
                # https://github.com/mozilla/addons-linter/issues/1361
                "sByDic": self.byDic.hex()  if bBinaryDictAsHexString  else [ e  for e in self.byDic ],
                "l2grams": list(self.a2grams)
            }, ensure_ascii=False))
            if bInJSModule:








<








240
241
242
243
244
245
246

247
248
249
250
251
252
253

                "nNode": self.nNode,
                "nArc": self.nArc,
                "nArcVal": self.nArcVal,
                "lArcVal": self.lArcVal,
                "nCompressionMethod": self.nCompressionMethod,
                "nBytesArc": self.nBytesArc,
                "nBytesNodeAddress": self.nBytesNodeAddress,

                # JavaScript is a pile of shit, so Mozilla’s JS parser don’t like file bigger than 4 Mb!
                # So, if necessary, we use an hexadecimal string, that we will convert later in Firefox’s extension.
                # https://github.com/mozilla/addons-linter/issues/1361
                "sByDic": self.byDic.hex()  if bBinaryDictAsHexString  else [ e  for e in self.byDic ],
                "l2grams": list(self.a2grams)
            }, ensure_ascii=False))
            if bInJSModule:
︙︙
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321

        iAddr = 0
        for c in sWord:
            if c not in self.dChar:
                return False
            iAddr = self._lookupArcNode(self.dChar[c], iAddr)
            if iAddr is None:
                return False
        return bool(int.from_bytes(self.byDic[iAddr:iAddr+self.nBytesArc], byteorder='big') & self._finalNodeMask)

    def getMorph (self, sWord):
        "retrieves morphologies list, different casing allowed"
        if not sWord:
            return []
        sWord = st.spellingNormalization(sWord)
        l = self.morph(sWord)
        if sWord[0:1].isupper():
            l.extend(self.morph(sWord.lower()))
            if sWord.isupper() and len(sWord) > 1:
                l.extend(self.morph(sWord.capitalize()))
        return l

    #@timethis
    def suggest (self, sWord, nSuggLimit=10, bSplitTrailingNumbers=False):
        "returns a set of suggestions for <sWord>"
        sWord = sWord.rstrip(".")   # useful for LibreOffice
        sWord = st.spellingNormalization(sWord)








|






|

|

|








291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316

        iAddr = 0
        for c in sWord:
            if c not in self.dChar:
                return False
            iAddr = self._lookupArcNode(self.dChar[c], iAddr)
            if iAddr is None:
                return False
        return bool(self.byDic[iAddr] & self._finalNodeMask)

    def getMorph (self, sWord):
        "retrieves morphologies list, different casing allowed"
        if not sWord:
            return []
        sWord = st.spellingNormalization(sWord)
        l = self._morph(sWord)
        if sWord[0:1].isupper():
            l.extend(self._morph(sWord.lower()))
            if sWord.isupper() and len(sWord) > 1:
                l.extend(self._morph(sWord.capitalize()))
        return l

    #@timethis
    def suggest (self, sWord, nSuggLimit=10, bSplitTrailingNumbers=False):
        "returns a set of suggestions for <sWord>"
        sWord = sWord.rstrip(".")   # useful for LibreOffice
        sWord = st.spellingNormalization(sWord)
︙︙
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366

                sWord1, sWord2 = sWord.split(cSplitter, 1)
                if self.isValid(sWord1) and self.isValid(sWord2):
                    oSuggResult.addSugg(sWord1+" "+sWord2)

    def _suggest (self, oSuggResult, sRemain, nMaxSwitch=0, nMaxDel=0, nMaxHardRepl=0, nMaxJump=0, nDist=0, nDeep=0, iAddr=0, sNewWord="", bAvoidLoop=False):
        # recursive function
        #logging.info((nDeep * "  ") + sNewWord + ":" + sRemain)
        if int.from_bytes(self.byDic[iAddr:iAddr+self.nBytesArc], byteorder='big') & self._finalNodeMask:
            if not sRemain:
                oSuggResult.addSugg(sNewWord, nDeep)
                for sTail in self._getTails(iAddr):
                    oSuggResult.addSugg(sNewWord+sTail, nDeep)
                return
            if (len(sNewWord) + len(sRemain) == len(oSuggResult.sWord)) and oSuggResult.sWord.lower().startswith(sNewWord.lower()) and self.isValid(sRemain):
                if self.sLangCode == "fr" and sNewWord.lower() in ("l", "d", "n", "m", "t", "s", "c", "j", "qu", "lorsqu", "puisqu", "quoiqu", "jusqu", "quelqu") and sRemain[0:1] in cp.aVowel:








|








347
348
349
350
351
352
353
354
355
356
357
358
359
360
361

                sWord1, sWord2 = sWord.split(cSplitter, 1)
                if self.isValid(sWord1) and self.isValid(sWord2):
                    oSuggResult.addSugg(sWord1+" "+sWord2)

    def _suggest (self, oSuggResult, sRemain, nMaxSwitch=0, nMaxDel=0, nMaxHardRepl=0, nMaxJump=0, nDist=0, nDeep=0, iAddr=0, sNewWord="", bAvoidLoop=False):
        # recursive function
        #logging.info((nDeep * "  ") + sNewWord + ":" + sRemain)
        if self.byDic[iAddr] & self._finalNodeMask:
            if not sRemain:
                oSuggResult.addSugg(sNewWord, nDeep)
                for sTail in self._getTails(iAddr):
                    oSuggResult.addSugg(sNewWord+sTail, nDeep)
                return
            if (len(sNewWord) + len(sRemain) == len(oSuggResult.sWord)) and oSuggResult.sWord.lower().startswith(sNewWord.lower()) and self.isValid(sRemain):
                if self.sLangCode == "fr" and sNewWord.lower() in ("l", "d", "n", "m", "t", "s", "c", "j", "qu", "lorsqu", "puisqu", "quoiqu", "jusqu", "quelqu") and sRemain[0:1] in cp.aVowel:
︙︙
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433

                yield (self.dCharVal[nVal], jAddr)

    def _getTails (self, iAddr, sTail="", n=2):
        "return a list of suffixes ending at a distance of <n> from <iAddr>"
        aTails = set()
        for nVal, jAddr in self._getArcs(iAddr):
            if nVal <= self.nChar:
                if int.from_bytes(self.byDic[jAddr:jAddr+self.nBytesArc], byteorder='big') & self._finalNodeMask:
                    aTails.add(sTail + self.dCharVal[nVal])
                if n and not aTails:
                    aTails.update(self._getTails(jAddr, sTail+self.dCharVal[nVal], n-1))
        return aTails

    def drawPath (self, sWord, iAddr=0):
        "show the path taken by <sWord> in the graph"








|








414
415
416
417
418
419
420
421
422
423
424
425
426
427
428

                yield (self.dCharVal[nVal], jAddr)

    def _getTails (self, iAddr, sTail="", n=2):
        "return a list of suffixes ending at a distance of <n> from <iAddr>"
        aTails = set()
        for nVal, jAddr in self._getArcs(iAddr):
            if nVal <= self.nChar:
                if self.byDic[jAddr] & self._finalNodeMask:
                    aTails.add(sTail + self.dCharVal[nVal])
                if n and not aTails:
                    aTails.update(self._getTails(jAddr, sTail+self.dCharVal[nVal], n-1))
        return aTails

    def drawPath (self, sWord, iAddr=0):
        "show the path taken by <sWord> in the graph"
︙︙
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798

            if sFlexPattern:
                zFlexPattern = re.compile(sFlexPattern)
            if sTagsPattern:
                zTagsPattern = re.compile(sTagsPattern)
        except re.error:
            print("# Error in regex pattern")
            traceback.print_exc()
        yield from self._select1(zFlexPattern, zTagsPattern, 0, "")

    # def morph (self, sWord):
    #     is defined in __init__

    # VERSION 1
    def _select1 (self, zFlexPattern, zTagsPattern, iAddr, sWord):
        # recursive generator
        for nVal, jAddr in self._getArcs1(iAddr):
            if nVal <= self.nChar:
                # simple character
                yield from self._select1(zFlexPattern, zTagsPattern, jAddr, sWord + self.lArcVal[nVal])
            else:
                if not zFlexPattern or zFlexPattern.search(sWord):
                    sStem = self.funcStemming(sWord, self.lArcVal[nVal])
                    for nMorphVal, _ in self._getArcs1(jAddr):
                        if not zTagsPattern or zTagsPattern.search(self.lArcVal[nMorphVal]):
                            yield [sWord, sStem, self.lArcVal[nMorphVal]]

    def _morph1 (self, sWord):
        "returns morphologies of <sWord>"
        iAddr = 0
        for c in sWord:
            if c not in self.dChar:
                return []
            iAddr = self._lookupArcNode(self.dChar[c], iAddr)
            if iAddr is None:
                return []
        if int.from_bytes(self.byDic[iAddr:iAddr+self.nBytesArc], byteorder='big') & self._finalNodeMask:
            l = []
            nRawArc = 0
            while not nRawArc & self._lastArcMask:
                iEndArcAddr = iAddr + self.nBytesArc
                nRawArc = int.from_bytes(self.byDic[iAddr:iEndArcAddr], byteorder='big')
                nArc = nRawArc & self._arcMask
                if nArc > self.nChar:
                    # This value is not a char, this is a stemming code
                    sStem = ">" + self.funcStemming(sWord, self.lArcVal[nArc])
                    # Now , we go to the next node and retrieve all following arcs values, all of them are tags
                    iAddr2 = int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesNodeAddress], byteorder='big')
                    nRawArc2 = 0
                    while not nRawArc2 & self._lastArcMask:
                        iEndArcAddr2 = iAddr2 + self.nBytesArc
                        nRawArc2 = int.from_bytes(self.byDic[iAddr2:iEndArcAddr2], byteorder='big')
                        l.append(sStem + "/" + self.lArcVal[nRawArc2 & self._arcMask])
                        iAddr2 = iEndArcAddr2+self.nBytesNodeAddress
                iAddr = iEndArcAddr+self.nBytesNodeAddress
            return l
        return []

    def _stem1 (self, sWord):
        "returns stems list of <sWord>"
        iAddr = 0
        for c in sWord:
            if c not in self.dChar:
                return []
            iAddr = self._lookupArcNode(self.dChar[c], iAddr)
            if iAddr is None:
                return []
        if int.from_bytes(self.byDic[iAddr:iAddr+self.nBytesArc], byteorder='big') & self._finalNodeMask:
            l = []
            nRawArc = 0
            while not nRawArc & self._lastArcMask:
                iEndArcAddr = iAddr + self.nBytesArc
                nRawArc = int.from_bytes(self.byDic[iAddr:iEndArcAddr], byteorder='big')
                nArc = nRawArc & self._arcMask
                if nArc > self.nChar:
                    # This value is not a char, this is a stemming code
                    l.append(self.funcStemming(sWord, self.lArcVal[nArc]))
                iAddr = iEndArcAddr+self.nBytesNodeAddress
            return l
        return []

    def _lookupArcNode1 (self, nVal, iAddr):
        "looks if <nVal> is an arc at the node at <iAddr>, if yes, returns address of next node else None"
        while True:
            iEndArcAddr = iAddr+self.nBytesArc
            nRawArc = int.from_bytes(self.byDic[iAddr:iEndArcAddr], byteorder='big')
            if nVal == (nRawArc & self._arcMask):
                # the value we are looking for
                # we return the address of the next node
                return int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesNodeAddress], byteorder='big')
            # value not found
            if nRawArc & self._lastArcMask:
                return None
            iAddr = iEndArcAddr+self.nBytesNodeAddress

    def _getArcs1 (self, iAddr):
        "generator: return all arcs at <iAddr> as tuples of (nVal, iAddr)"
        while True:
            iEndArcAddr = iAddr+self.nBytesArc
            nRawArc = int.from_bytes(self.byDic[iAddr:iEndArcAddr], byteorder='big')
            yield nRawArc & self._arcMask, int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesNodeAddress], byteorder='big')
            if nRawArc & self._lastArcMask:
                break
            iAddr = iEndArcAddr+self.nBytesNodeAddress

    def _writeNodes1 (self, spfDest):
        "for debugging only"
        print(" > Write binary nodes")
        with open(spfDest, 'w', 'utf-8', newline="\n") as hDst:
            iAddr = 0
            hDst.write("i{:_>10} -- #{:_>10}\n".format("0", iAddr))
            while iAddr < len(self.byDic):
                iEndArcAddr = iAddr+self.nBytesArc
                nRawArc = int.from_bytes(self.byDic[iAddr:iEndArcAddr], byteorder='big')
                nArc = nRawArc & self._arcMask
                hDst.write("  {:<20}  {:0>16}  i{:>10}   #{:_>10}\n".format(self.lArcVal[nArc], bin(nRawArc)[2:], "?", \
                                                                            int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesNodeAddress], \
                                                                                           byteorder='big')))
                iAddr = iEndArcAddr+self.nBytesNodeAddress
                if (nRawArc & self._lastArcMask) and iAddr < len(self.byDic):
                    hDst.write("\ni{:_>10} -- #{:_>10}\n".format("?", iAddr))
            hDst.close()

    # VERSION 2
    def _morph2 (self, sWord):
        "returns morphologies of <sWord>"
        iAddr = 0
        for c in sWord:
            if c not in self.dChar:
                return []
            iAddr = self._lookupArcNode(self.dChar[c], iAddr)
            if iAddr is None:
                return []
        if int.from_bytes(self.byDic[iAddr:iAddr+self.nBytesArc], byteorder='big') & self._finalNodeMask:
            l = []
            nRawArc = 0
            while not nRawArc & self._lastArcMask:
                iEndArcAddr = iAddr + self.nBytesArc
                nRawArc = int.from_bytes(self.byDic[iAddr:iEndArcAddr], byteorder='big')
                nArc = nRawArc & self._arcMask
                if nArc > self.nChar:
                    # This value is not a char, this is a stemming code
                    sStem = ">" + self.funcStemming(sWord, self.lArcVal[nArc])
                    # Now , we go to the next node and retrieve all following arcs values, all of them are tags
                    if not nRawArc & self._addrBitMask:
                        iAddr2 = int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesNodeAddress], byteorder='big')
                    else:
                        # we go to the end of the node
                        iAddr2 = iEndArcAddr
                        while not nRawArc & self._lastArcMask:
                            nRawArc = int.from_bytes(self.byDic[iAddr2:iAddr2+self.nBytesArc], byteorder='big')
                            iAddr2 += self.nBytesArc + self.nBytesNodeAddress
                    nRawArc2 = 0
                    while not nRawArc2 & self._lastArcMask:
                        iEndArcAddr2 = iAddr2 + self.nBytesArc
                        nRawArc2 = int.from_bytes(self.byDic[iAddr2:iEndArcAddr2], byteorder='big')
                        l.append(sStem + "/" + self.lArcVal[nRawArc2 & self._arcMask])
                        iAddr2 = iEndArcAddr2+self.nBytesNodeAddress  if not nRawArc2 & self._addrBitMask else iEndArcAddr2
                iAddr = iEndArcAddr+self.nBytesNodeAddress  if not nRawArc & self._addrBitMask  else iEndArcAddr
            return l
        return []

    def _stem2 (self, sWord):
        "returns stems list of <sWord>"
        iAddr = 0
        for c in sWord:
            if c not in self.dChar:
                return []
            iAddr = self._lookupArcNode(self.dChar[c], iAddr)
            if iAddr is None:
                return []
        if int.from_bytes(self.byDic[iAddr:iAddr+self.nBytesArc], byteorder='big') & self._finalNodeMask:
            l = []
            nRawArc = 0
            while not nRawArc & self._lastArcMask:
                iEndArcAddr = iAddr + self.nBytesArc
                nRawArc = int.from_bytes(self.byDic[iAddr:iEndArcAddr], byteorder='big')
                nArc = nRawArc & self._arcMask
                if nArc > self.nChar:
                    # This value is not a char, this is a stemming code
                    l.append(self.funcStemming(sWord, self.lArcVal[nArc]))
                    # Now , we go to the next node
                    if not nRawArc & self._addrBitMask:
                        iAddr2 = int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesNodeAddress], byteorder='big')
                    else:
                        # we go to the end of the node
                        iAddr2 = iEndArcAddr
                        while not nRawArc & self._lastArcMask:
                            nRawArc = int.from_bytes(self.byDic[iAddr2:iAddr2+self.nBytesArc], byteorder='big')
                            iAddr2 += self.nBytesArc + self.nBytesNodeAddress
                iAddr = iEndArcAddr+self.nBytesNodeAddress  if not nRawArc & self._addrBitMask  else iEndArcAddr
            return l
        return []

    def _lookupArcNode2 (self, nVal, iAddr):
        "looks if <nVal> is an arc at the node at <iAddr>, if yes, returns address of next node else None"
        while True:
            iEndArcAddr = iAddr+self.nBytesArc
            nRawArc = int.from_bytes(self.byDic[iAddr:iEndArcAddr], byteorder='big')
            if nVal == (nRawArc & self._arcMask):
                # the value we are looking for
                if not nRawArc & self._addrBitMask:
                    # we return the address of the next node
                    return int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesNodeAddress], byteorder='big')
                # we go to the end of the node
                iAddr = iEndArcAddr
                while not nRawArc & self._lastArcMask:
                    nRawArc = int.from_bytes(self.byDic[iAddr:iAddr+self.nBytesArc], byteorder='big')
                    iAddr += self.nBytesArc + self.nBytesNodeAddress  if not nRawArc & self._addrBitMask  else self.nBytesArc
                return iAddr
            # value not found
            if nRawArc & self._lastArcMask:
                return None
            iAddr = iEndArcAddr+self.nBytesNodeAddress  if not nRawArc & self._addrBitMask  else iEndArcAddr

    def _writeNodes2 (self, spfDest):
        "for debugging only"
        print(" > Write binary nodes")
        with open(spfDest, 'w', 'utf-8', newline="\n") as hDst:
            iAddr = 0
            hDst.write("i{:_>10} -- #{:_>10}\n".format("0", iAddr))
            while iAddr < len(self.byDic):
                iEndArcAddr = iAddr+self.nBytesArc
                nRawArc = int.from_bytes(self.byDic[iAddr:iEndArcAddr], byteorder='big')
                nArc = nRawArc & self._arcMask
                if not nRawArc & self._addrBitMask:
                    iNextNodeAddr = int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesNodeAddress], byteorder='big')
                    hDst.write("  {:<20}  {:0>16}  i{:>10}   #{:_>10}\n".format(self.lArcVal[nArc], bin(nRawArc)[2:], "?", iNextNodeAddr))
                    iAddr = iEndArcAddr+self.nBytesNodeAddress
                else:
                    hDst.write("  {:<20}  {:0>16}\n".format(self.lArcVal[nArc], bin(nRawArc)[2:]))
                    iAddr = iEndArcAddr
                if nRawArc & self._lastArcMask:
                    hDst.write("\ni{:_>10} -- #{:_>10}\n".format("?", iAddr))
            hDst.close()

    # VERSION 3
    def _morph3 (self, sWord):
        "returns morphologies of <sWord>"
        iAddr = 0
        for c in sWord:
            if c not in self.dChar:
                return []
            iAddr = self._lookupArcNode(self.dChar[c], iAddr)
            if iAddr is None:
                return []
        if int.from_bytes(self.byDic[iAddr:iAddr+self.nBytesArc], byteorder='big') & self._finalNodeMask:
            l = []
            nRawArc = 0
            iAddrNode = iAddr
            while not nRawArc & self._lastArcMask:
                iEndArcAddr = iAddr + self.nBytesArc
                nRawArc = int.from_bytes(self.byDic[iAddr:iEndArcAddr], byteorder='big')
                nArc = nRawArc & self._arcMask
                if nArc > self.nChar:
                    # This value is not a char, this is a stemming code
                    sStem = ">" + self.funcStemming(sWord, self.lArcVal[nArc])
                    # Now , we go to the next node and retrieve all following arcs values, all of them are tags
                    if not nRawArc & self._addrBitMask:
                        iAddr2 = int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesNodeAddress], byteorder='big')
                    else:
                        iAddr2 = iAddrNode + int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesOffset], byteorder='big')
                    nRawArc2 = 0
                    while not nRawArc2 & self._lastArcMask:
                        iEndArcAddr2 = iAddr2 + self.nBytesArc
                        nRawArc2 = int.from_bytes(self.byDic[iAddr2:iEndArcAddr2], byteorder='big')
                        l.append(sStem + "/" + self.lArcVal[nRawArc2 & self._arcMask])
                        iAddr2 = iEndArcAddr2+self.nBytesNodeAddress  if not nRawArc2 & self._addrBitMask  else iEndArcAddr2+self.nBytesOffset
                iAddr = iEndArcAddr+self.nBytesNodeAddress  if not nRawArc & self._addrBitMask  else iEndArcAddr+self.nBytesOffset
            return l
        return []

    def _stem3 (self, sWord):
        "returns stems list of <sWord>"
        iAddr = 0
        for c in sWord:
            if c not in self.dChar:
                return []
            iAddr = self._lookupArcNode(self.dChar[c], iAddr)
            if iAddr is None:
                return []
        if int.from_bytes(self.byDic[iAddr:iAddr+self.nBytesArc], byteorder='big') & self._finalNodeMask:
            l = []
            nRawArc = 0
            #iAddrNode = iAddr
            while not nRawArc & self._lastArcMask:
                iEndArcAddr = iAddr + self.nBytesArc
                nRawArc = int.from_bytes(self.byDic[iAddr:iEndArcAddr], byteorder='big')
                nArc = nRawArc & self._arcMask
                if nArc > self.nChar:
                    # This value is not a char, this is a stemming code
                    l.append(self.funcStemming(sWord, self.lArcVal[nArc]))
                iAddr = iEndArcAddr+self.nBytesNodeAddress  if not nRawArc & self._addrBitMask  else iEndArcAddr+self.nBytesOffset
            return l
        return []

    def _lookupArcNode3 (self, nVal, iAddr):
        "looks if <nVal> is an arc at the node at <iAddr>, if yes, returns address of next node else None"
        iAddrNode = iAddr
        while True:
            iEndArcAddr = iAddr+self.nBytesArc
            nRawArc = int.from_bytes(self.byDic[iAddr:iEndArcAddr], byteorder='big')
            if nVal == (nRawArc & self._arcMask):
                # the value we are looking for
                if not nRawArc & self._addrBitMask:
                    return int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesNodeAddress], byteorder='big')
                return iAddrNode + int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesOffset], byteorder='big')
            # value not found
            if nRawArc & self._lastArcMask:
                return None
            iAddr = iEndArcAddr+self.nBytesNodeAddress  if not nRawArc & self._addrBitMask  else iEndArcAddr+self.nBytesOffset

    def _writeNodes3 (self, spfDest):
        "for debugging only"
        print(" > Write binary nodes")
        with open(spfDest, 'w', 'utf-8', newline="\n") as hDst:
            iAddr = 0
            hDst.write("i{:_>10} -- #{:_>10}\n".format("0", iAddr))
            while iAddr < len(self.byDic):
                iEndArcAddr = iAddr+self.nBytesArc
                nRawArc = int.from_bytes(self.byDic[iAddr:iEndArcAddr], byteorder='big')
                nArc = nRawArc & self._arcMask
                if not nRawArc & self._addrBitMask:
                    iNextNodeAddr = int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesNodeAddress], byteorder='big')
                    hDst.write("  {:<20}  {:0>16}  i{:>10}   #{:_>10}\n".format(self.lArcVal[nArc], bin(nRawArc)[2:], "?", iNextNodeAddr))
                    iAddr = iEndArcAddr+self.nBytesNodeAddress
                else:
                    iNextNodeAddr = int.from_bytes(self.byDic[iEndArcAddr:iEndArcAddr+self.nBytesOffset], byteorder='big')
                    hDst.write("  {:<20}  {:0>16}  i{:>10}   +{:_>10}\n".format(self.lArcVal[nArc], bin(nRawArc)[2:], "?", iNextNodeAddr))
                    iAddr = iEndArcAddr+self.nBytesOffset
                if nRawArc & self._lastArcMask:
                    hDst.write("\ni{:_>10} -- #{:_>10}\n".format("?", iAddr))
            hDst.close()








|

<
<
<
<
|

|


|



|



|








|



|
|





|


|
|

|
|



|








|



|
|




|



|


|
|



|



|

|


|
|
|


|

|






|
|

|
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
|
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
|
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<
<



462
463
464
465
466
467
468
469
470




471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572

























































































573


























574


































































































575
576

            if sFlexPattern:
                zFlexPattern = re.compile(sFlexPattern)
            if sTagsPattern:
                zTagsPattern = re.compile(sTagsPattern)
        except re.error:
            print("# Error in regex pattern")
            traceback.print_exc()
        yield from self._select(zFlexPattern, zTagsPattern, 0, "")





    def _select (self, zFlexPattern, zTagsPattern, iAddr, sWord):
        # recursive generator
        for nVal, jAddr in self._getArcs(iAddr):
            if nVal <= self.nChar:
                # simple character
                yield from self._select(zFlexPattern, zTagsPattern, jAddr, sWord + self.lArcVal[nVal])
            else:
                if not zFlexPattern or zFlexPattern.search(sWord):
                    sStem = self.funcStemming(sWord, self.lArcVal[nVal])
                    for nMorphVal, _ in self._getArcs(jAddr):
                        if not zTagsPattern or zTagsPattern.search(self.lArcVal[nMorphVal]):
                            yield [sWord, sStem, self.lArcVal[nMorphVal]]

    def _morph (self, sWord):
        "returns morphologies of <sWord>"
        iAddr = 0
        for c in sWord:
            if c not in self.dChar:
                return []
            iAddr = self._lookupArcNode(self.dChar[c], iAddr)
            if iAddr is None:
                return []
        if self.byDic[iAddr] & self._finalNodeMask:
            l = []
            nRawArc = 0
            while not nRawArc & self._lastArcMask:
                iEndArcAddr = iAddr + 1
                nRawArc = self.byDic[iAddr]
                nArc = nRawArc & self._arcMask
                if nArc > self.nChar:
                    # This value is not a char, this is a stemming code
                    sStem = ">" + self.funcStemming(sWord, self.lArcVal[nArc])
                    # Now , we go to the next node and retrieve all following arcs values, all of them are tags
                    iAddr2 = self.byDic[iEndArcAddr]
                    nRawArc2 = 0
                    while not nRawArc2 & self._lastArcMask:
                        iEndArcAddr2 = iAddr2 + 1
                        nRawArc2 = self.byDic[iAddr2]
                        l.append(sStem + "/" + self.lArcVal[nRawArc2 & self._arcMask])
                        iAddr2 = iEndArcAddr2 + 1
                iAddr = iEndArcAddr + 1
            return l
        return []

    def _stem (self, sWord):
        "returns stems list of <sWord>"
        iAddr = 0
        for c in sWord:
            if c not in self.dChar:
                return []
            iAddr = self._lookupArcNode(self.dChar[c], iAddr)
            if iAddr is None:
                return []
        if self.byDic[iAddr] & self._finalNodeMask:
            l = []
            nRawArc = 0
            while not nRawArc & self._lastArcMask:
                iEndArcAddr = iAddr + 1
                nRawArc = self.byDic[iAddr]
                nArc = nRawArc & self._arcMask
                if nArc > self.nChar:
                    # This value is not a char, this is a stemming code
                    l.append(self.funcStemming(sWord, self.lArcVal[nArc]))
                iAddr = iEndArcAddr + 1
            return l
        return []

    def _lookupArcNode (self, nVal, iAddr):
        "looks if <nVal> is an arc at the node at <iAddr>, if yes, returns address of next node else None"
        while True:
            iEndArcAddr = iAddr + 1
            nRawArc = self.byDic[iAddr]
            if nVal == (nRawArc & self._arcMask):
                # the value we are looking for
                # we return the address of the next node
                return self.byDic[iEndArcAddr]
            # value not found
            if nRawArc & self._lastArcMask:
                return None
            iAddr = iEndArcAddr + 1

    def _getArcs (self, iAddr):
        "generator: return all arcs at <iAddr> as tuples of (nVal, iAddr)"
        while True:
            iEndArcAddr = iAddr + 1
            nRawArc = self.byDic[iAddr]
            yield nRawArc & self._arcMask, self.byDic[iEndArcAddr]
            if nRawArc & self._lastArcMask:
                break
            iAddr = iEndArcAddr + 1

    def _writeNodes (self, spfDest):
        "for debugging only"
        print(" > Write binary nodes")
        with open(spfDest, 'w', 'utf-8', newline="\n") as hDst:
            iAddr = 0
            hDst.write("i{:_>10} -- #{:_>10}\n".format("0", iAddr))
            while iAddr < len(self.byDic):
                iEndArcAddr = iAddr + 1
                nRawArc = self.byDic[iAddr]
                nArc = nRawArc & self._arcMask
                hDst.write("  {:<20}  {:0>16}  i{:>10}   #{:_>10}\n".format(self.lArcVal[nArc], bin(nRawArc)[2:], "?", self.byDic[iEndArcAddr]))

























































































                iAddr = iEndArcAddr + 1


























                if (nRawArc & self._lastArcMask) and iAddr < len(self.byDic):


































































































                    hDst.write("\ni{:_>10} -- #{:_>10}\n".format("?", iAddr))
            hDst.close()

Fossil 2.24 [8be0372c10] 2024-04-23 13:25:26
This page was generated in about 0.016s