Grammalecte  Check-in [37ac74a3bb]

Overview
Comment:[graphspell][py] remove deprecated code
Downloads: Tarball | ZIP archive | SQL archive
Timelines: family | ancestors | descendants | both | trunk | graphspell
Files: files | file ages | folders
SHA3-256: 37ac74a3bb8e7f3c9984f0b8801c0e6344380bc5c88d925dd49458b525ff19eb
User & Date: olr on 2020-05-28 17:28:06
Other Links: manifest | tags
Context
2020-05-30
17:00
[fr] merge dicfac branch: dictionary update and new affixation rules check-in: 33eba8eb3d user: olr tags: trunk, fr
2020-05-28
18:21
merge trunk check-in: 3ebde2f15f user: olr tags: dicfac
17:28
[graphspell][py] remove deprecated code check-in: 37ac74a3bb user: olr tags: trunk, graphspell
15:20
[graphspell] suggestions: end of experiment -> aborted, too slow check-in: 4cd155b03e user: olr tags: trunk, graphspell
Changes

Modified graphspell/ibdawg.py from [b622d0fc42] to [db890a0452].

385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
        "returns True if sChars in known 2grams"
        if len(sChars) != 2:
            return True
        if not self.a2grams:
            return True
        return sChars in self.a2grams

    #@timethis
    def suggest2 (self, sWord, nSuggLimit=10):
        "returns a set of suggestions for <sWord>"
        sWord = cp.spellingNormalization(sWord)
        sPfx, sWord, sSfx = cp.cut(sWord)
        oSuggResult = SuggResult(sWord)
        self._suggest2(oSuggResult)
        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 _suggest2 (self, oSuggResult, nDeep=0, iAddr=0, sNewWord=""):
        # recursive function
        #logging.info((nDeep * "  ") + sNewWord)
        if nDeep >= oSuggResult.nDistLimit:
            sCleanNewWord = cp.simplifyWord(sNewWord)
            if st.distanceSift4(oSuggResult.sCleanWord[:len(sCleanNewWord)], sCleanNewWord) > oSuggResult.nDistLimit:
                return
        if int.from_bytes(self.byDic[iAddr:iAddr+self.nBytesArc], byteorder='big') & self._finalNodeMask:
            oSuggResult.addSugg(sNewWord, nDeep)
        for cChar, jAddr in self._getCharArcs(iAddr, oSuggResult.sWord[nDeep:nDeep+1]):
            self._suggest2(oSuggResult, nDeep+1, jAddr, sNewWord+cChar)
        return

    def _getCharArcs (self, iAddr):
        "generator: yield all chars and addresses from node at address <iAddr>"
        for nVal, jAddr in self._getArcs(iAddr):
            if nVal <= self.nChar:
                yield (self.dCharVal[nVal], jAddr)

    def _getTails (self, iAddr, sTail="", n=2):







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







385
386
387
388
389
390
391


























392
393
394
395
396
397
398
        "returns True if sChars in known 2grams"
        if len(sChars) != 2:
            return True
        if not self.a2grams:
            return True
        return sChars in self.a2grams



























    def _getCharArcs (self, iAddr):
        "generator: yield all chars and addresses from node at address <iAddr>"
        for nVal, jAddr in self._getArcs(iAddr):
            if nVal <= self.nChar:
                yield (self.dCharVal[nVal], jAddr)

    def _getTails (self, iAddr, sTail="", n=2):