Grammalecte  Diff

Annotate Patch Side-by-side Diff

Differences From Artifact [e811d4895f]:

To Artifact [a255097656]:

︙
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

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








-



+
-
-
-
-
+
+
+
+
+

-
-
-
-
+
+
+
+








                self.aSugg.add(sSugg)
                if nDist < self.nMinDist:
                    self.nMinDist = nDist
                self.nDistLimit = min(self.nDistLimit, self.nMinDist+2)

    def getSuggestions (self, nSuggLimit=10, nDistLimit=-1):
        "return a list of suggestions"
        lRes = []
        if self.dSugg[0]:
            # we sort the better results with the original word
            self.dSugg[0].sort(key=lambda sSugg: st.distanceDamerauLevenshtein(self.sWord, sSugg))
        lRes = self.dSugg.pop(0)
        for lSugg in self.dSugg.values():
            lRes.extend(lSugg)
            if len(lRes) > nSuggLimit:
                break
        for nDist, lSugg in self.dSugg.items():
            if nDist <= self.nDistLimit:
                lRes.extend(lSugg)
                if len(lRes) > nSuggLimit:
                    break
        lRes = list(cp.filterSugg(lRes))
        if self.sWord.istitle():
            lRes = list(map(lambda sSugg: sSugg.title(), lRes))
        elif self.sWord.isupper():
            lRes = list(map(lambda sSugg: sSugg.upper(), lRes))
        if self.sWord.isupper():
            lRes = list(map(lambda sSugg: sSugg.upper(), lRes))
        elif self.sWord[0:1].isupper():
            lRes = list(map(lambda sSugg: sSugg[0:1].upper()+sSugg[1:], lRes))  # dont’ use <.istitle>
        return lRes[:nSuggLimit]

    def reset (self):
        self.aSugg.clear()
        self.dSugg.clear()
︙
131
132
133
134
135
136
137
138
139


140
141
142
143
144
145
146

132
133
134
135
136
137
138


139
140
141
142
143
144
145
146
147








-
-
+
+








            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.bOptNumSigle = False
        self.bOptNumAtLast = False
        self.bAcronymValid = True
        self.bNumAtLastValid = False

    def _initBinary (self):
        "initialize with binary structure file"
        if self.by[0:17] != b"/grammalecte-fsa/":
            raise TypeError("# Error. Not a grammalecte-fsa binary dictionary. Header: {}".format(self.by[0:9]))
        if not(self.by[17:18] == b"1" or self.by[17:18] == b"2" or self.by[17:18] == b"3"):
            raise ValueError("# Error. Unknown dictionary version: {}".format(self.by[17:18]))
︙
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
260
261
262
263
264

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
260
261
262
263
264
265
266
267








+
+










-
-





-
+





+
+








        sToken = cp.spellingNormalization(sToken)
        if self.isValid(sToken):
            return True
        if "-" in sToken:
            if sToken.count("-") > 4:
                return True
            return all(self.isValid(sWord)  for sWord in sToken.split("-"))
        if "." in sToken or "·" in sToken:
            return True
        return False

    def isValid (self, sWord):
        "checks if <sWord> is valid (different casing tested if the first letter is a capital)"
        if not sWord:
            return None
        if "’" in sWord: # ugly hack
            sWord = sWord.replace("’", "'")
        if self.lookup(sWord):
            return True
        if sWord.isdigit():
            return True
        if sWord[0:1].isupper():
            if len(sWord) > 1:
                if sWord.istitle():
                    return self.lookup(sWord.lower())
                if sWord.isupper():
                    if self.bOptNumSigle:
                    if self.bAcronymValid:
                        return True
                    return self.lookup(sWord.lower()) or self.lookup(sWord.capitalize())
                return self.lookup(sWord[:1].lower() + sWord[1:])
            else:
                return self.lookup(sWord.lower())
        if sWord[0:1].isdigit():
            return True
        return False

    def lookup (self, sWord):
        "returns True if <sWord> in dictionary (strict verification)"
        iAddr = 0
        for c in sWord:
            if c not in self.dChar:
︙
277
278
279
280
281
282
283

284
285
286
287
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
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

280
281
282
283
284
285
286
287
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
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








+





+

-
+
-
-
-
-






-
+








+
+



-
-
-
+
+
+
+
+
+




-
+



-
+


-
+


-
+

-
+



-
+

-
+

-
+








            if sWord.isupper() and len(sWord) > 1:
                l.extend(self.morph(sWord.capitalize()))
        return l

    #@timethis
    def suggest (self, sWord, nSuggLimit=10):
        "returns a set of suggestions for <sWord>"
        sWord = sWord.rstrip(".")   # useful for LibreOffice
        sWord = cp.spellingNormalization(sWord)
        sPfx, sWord, sSfx = cp.cut(sWord)
        nMaxSwitch = max(len(sWord) // 3, 1)
        nMaxDel = len(sWord) // 5
        nMaxHardRepl = max((len(sWord) - 5) // 4, 1)
        nMaxJump = max(len(sWord) // 4, 1)
        oSuggResult = SuggResult(sWord)
        self._suggest(oSuggResult, sWord, nMaxSwitch=nMaxSwitch, nMaxDel=nMaxDel, nMaxHardRepl=nMaxHardRepl)
        self._suggest(oSuggResult, sWord, nMaxSwitch, nMaxDel, nMaxHardRepl, nMaxJump)
        if sWord.istitle():
            self._suggest(oSuggResult, sWord.lower(), nMaxSwitch=nMaxSwitch, nMaxDel=nMaxDel, nMaxHardRepl=nMaxHardRepl)
        elif sWord.islower():
            self._suggest(oSuggResult, sWord.title(), nMaxSwitch=nMaxSwitch, nMaxDel=nMaxDel, nMaxHardRepl=nMaxHardRepl)
        aSugg = oSuggResult.getSuggestions(nSuggLimit)
        if sSfx or sPfx:
            # we add what we removed
            return list(map(lambda sSug: sPfx + sSug + sSfx, aSugg))
        return aSugg

    def _suggest (self, oSuggResult, sRemain, nMaxSwitch=0, nMaxDel=0, nMaxHardRepl=0, nDeep=0, iAddr=0, sNewWord="", bAvoidLoop=False):
    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 not sRemain:
            if int.from_bytes(self.byDic[iAddr:iAddr+self.nBytesArc], byteorder='big') & self._finalNodeMask:
                oSuggResult.addSugg(sNewWord, nDeep)
            for sTail in self._getTails(iAddr):
                oSuggResult.addSugg(sNewWord+sTail, nDeep)
            return
        if nDist > oSuggResult.nDistLimit:
            return
        cCurrent = sRemain[0:1]
        for cChar, jAddr in self._getCharArcs(iAddr):
            if cChar in cp.d1to1.get(cCurrent, cCurrent):
                self._suggest(oSuggResult, sRemain[1:], nMaxSwitch, nMaxDel, nMaxHardRepl, nDeep+1, jAddr, sNewWord+cChar)
            elif not bAvoidLoop and nMaxHardRepl:
                self._suggest(oSuggResult, sRemain[1:], nMaxSwitch, nMaxDel, nMaxHardRepl-1, nDeep+1, jAddr, sNewWord+cChar, True)
                self._suggest(oSuggResult, sRemain[1:], nMaxSwitch, nMaxDel, nMaxHardRepl, nMaxJump, nDist, nDeep+1, jAddr, sNewWord+cChar)
            elif not bAvoidLoop:
                if nMaxHardRepl:
                    self._suggest(oSuggResult, sRemain[1:], nMaxSwitch, nMaxDel, nMaxHardRepl-1, nMaxJump, nDist+1, nDeep+1, jAddr, sNewWord+cChar, True)
                if nMaxJump:
                    self._suggest(oSuggResult, sRemain, nMaxSwitch, nMaxDel, nMaxHardRepl, nMaxJump-1, nDist+1, nDeep+1, jAddr, sNewWord+cChar, True)
        if not bAvoidLoop: # avoid infinite loop
            if len(sRemain) > 1:
                if cCurrent == sRemain[1:2]:
                    # same char, we remove 1 char without adding 1 to <sNewWord>
                    self._suggest(oSuggResult, sRemain[1:], nMaxSwitch, nMaxDel, nMaxHardRepl, nDeep+1, iAddr, sNewWord)
                    self._suggest(oSuggResult, sRemain[1:], nMaxSwitch, nMaxDel, nMaxHardRepl, nMaxJump, nDist, nDeep+1, iAddr, sNewWord)
                else:
                    # switching chars
                    if nMaxSwitch:
                        self._suggest(oSuggResult, sRemain[1:2]+sRemain[0:1]+sRemain[2:], nMaxSwitch-1, nMaxDel, nMaxHardRepl, nDeep+1, iAddr, sNewWord, True)
                        self._suggest(oSuggResult, sRemain[1:2]+sRemain[0:1]+sRemain[2:], nMaxSwitch-1, nMaxDel, nMaxHardRepl, nMaxJump, nDist+1, nDeep+1, iAddr, sNewWord, True)
                    # delete char
                    if nMaxDel:
                        self._suggest(oSuggResult, sRemain[1:], nMaxSwitch, nMaxDel-1, nMaxHardRepl, nDeep+1, iAddr, sNewWord, True)
                        self._suggest(oSuggResult, sRemain[1:], nMaxSwitch, nMaxDel-1, nMaxHardRepl, nMaxJump, nDist+1, nDeep+1, iAddr, sNewWord, True)
                # Phonetic replacements
                for sRepl in cp.get1toXReplacement(sNewWord[-1:], cCurrent, sRemain[1:2]):
                    self._suggest(oSuggResult, sRepl + sRemain[1:], nMaxSwitch, nMaxDel, nMaxHardRepl, nDeep+1, iAddr, sNewWord, True)
                    self._suggest(oSuggResult, sRepl + sRemain[1:], nMaxSwitch, nMaxDel, nMaxHardRepl, nMaxJump, nDist, nDeep+1, iAddr, sNewWord, True)
                for sRepl in cp.d2toX.get(sRemain[0:2], ()):
                    self._suggest(oSuggResult, sRepl + sRemain[2:], nMaxSwitch, nMaxDel, nMaxHardRepl, nDeep+1, iAddr, sNewWord, True)
                    self._suggest(oSuggResult, sRepl + sRemain[2:], nMaxSwitch, nMaxDel, nMaxHardRepl, nMaxJump, nDist, nDeep+1, iAddr, sNewWord, True)
            # end of word
            if len(sRemain) == 2:
                for sRepl in cp.dFinal2.get(sRemain, ()):
                    self._suggest(oSuggResult, sRepl, nMaxSwitch, nMaxDel, nMaxHardRepl, nDeep+1, iAddr, sNewWord, True)
                    self._suggest(oSuggResult, sRepl, nMaxSwitch, nMaxDel, nMaxHardRepl, nMaxJump, nDist, nDeep+1, iAddr, sNewWord, True)
            elif len(sRemain) == 1:
                self._suggest(oSuggResult, "", nMaxSwitch, nMaxDel, nMaxHardRepl, nDeep+1, iAddr, sNewWord, True) # remove last char and go on
                self._suggest(oSuggResult, "", nMaxSwitch, nMaxDel, nMaxHardRepl, nMaxJump, nDist, nDeep+1, iAddr, sNewWord, True) # remove last char and go on
                for sRepl in cp.dFinal1.get(sRemain, ()):
                    self._suggest(oSuggResult, sRepl, nMaxSwitch, nMaxDel, nMaxHardRepl, nDeep+1, iAddr, sNewWord, True)
                    self._suggest(oSuggResult, sRepl, nMaxSwitch, nMaxDel, nMaxHardRepl, nMaxJump, nDist, nDeep+1, iAddr, sNewWord, True)

    #@timethis
    def suggest2 (self, sWord, nMaxSugg=10):
        "returns a set of suggestions for <sWord>"
        sWord = cp.spellingNormalization(sWord)
        sPfx, sWord, sSfx = cp.cut(sWord)
        oSuggResult = SuggResult(sWord)
︙
Fossil 2.24 [8be0372c10] 2024-04-23 13:25:26
This page was generated in about 0.006s