Grammalecte  Diff

Differences From Artifact [059d031769]:

To Artifact [f1e3328ca8]:


305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
                sEntry = sWord + "\t" + self.funcStemming(sWord, self.lArcVal[nVal])
                for nMorphVal, _ in oNextNode.arcs.items():
                    if not zPattern or zPattern.search(self.lArcVal[nMorphVal]):
                        yield sEntry + "\t" + self.lArcVal[nMorphVal]


    # BINARY CONVERSION
    def createBinary (self, sPathFile, nCompressionMethod, bDebug=False):
        print(" > Write DAWG as an indexable binary dictionary [method: %d]" % nCompressionMethod)
        if nCompressionMethod == 1:
            self.nBytesArc = ( (self.nArcVal.bit_length() + 2) // 8 ) + 1   # We add 2 bits. See DawgNode.convToBytes1()
            self.nBytesOffset = 0
            self._calcNumBytesNodeAddress()
            self._calcNodesAddress1()
        elif nCompressionMethod == 2:







|







305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
                sEntry = sWord + "\t" + self.funcStemming(sWord, self.lArcVal[nVal])
                for nMorphVal, _ in oNextNode.arcs.items():
                    if not zPattern or zPattern.search(self.lArcVal[nMorphVal]):
                        yield sEntry + "\t" + self.lArcVal[nMorphVal]


    # BINARY CONVERSION
    def _calculateBinary (self, nCompressionMethod):
        print(" > Write DAWG as an indexable binary dictionary [method: %d]" % nCompressionMethod)
        if nCompressionMethod == 1:
            self.nBytesArc = ( (self.nArcVal.bit_length() + 2) // 8 ) + 1   # We add 2 bits. See DawgNode.convToBytes1()
            self.nBytesOffset = 0
            self._calcNumBytesNodeAddress()
            self._calcNodesAddress1()
        elif nCompressionMethod == 2:
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
            self._calcNodesAddress3()
        else:
            print(" # Error: unknown compression method")
        print("   Arc values (chars, affixes and tags): {}  ->  {} bytes".format( self.nArcVal, len("\t".join(self.lArcVal).encode("utf-8")) ))
        print("   Arc size: {} bytes, Address size: {} bytes   ->   {} * {} = {} bytes".format( self.nBytesArc, self.nBytesNodeAddress, \
                                                                                                self.nBytesArc+self.nBytesNodeAddress, self.nArc, \
                                                                                                (self.nBytesArc+self.nBytesNodeAddress)*self.nArc ))
        self._writeBinary(sPathFile, nCompressionMethod)
        self._writeAsJSObject(sPathFile, nCompressionMethod)
        if bDebug:
            self._writeNodes(sPathFile, nCompressionMethod)

    def _calcNumBytesNodeAddress (self):
        "how many bytes needed to store all nodes/arcs in the binary dictionary"
        self.nBytesNodeAddress = 1
        while ((self.nBytesArc + self.nBytesNodeAddress) * self.nArc) > (2 ** (self.nBytesNodeAddress * 8)):
            self.nBytesNodeAddress += 1








<
<
<
<







329
330
331
332
333
334
335




336
337
338
339
340
341
342
            self._calcNodesAddress3()
        else:
            print(" # Error: unknown compression method")
        print("   Arc values (chars, affixes and tags): {}  ->  {} bytes".format( self.nArcVal, len("\t".join(self.lArcVal).encode("utf-8")) ))
        print("   Arc size: {} bytes, Address size: {} bytes   ->   {} * {} = {} bytes".format( self.nBytesArc, self.nBytesNodeAddress, \
                                                                                                self.nBytesArc+self.nBytesNodeAddress, self.nArc, \
                                                                                                (self.nBytesArc+self.nBytesNodeAddress)*self.nArc ))





    def _calcNumBytesNodeAddress (self):
        "how many bytes needed to store all nodes/arcs in the binary dictionary"
        self.nBytesNodeAddress = 1
        while ((self.nBytesArc + self.nBytesNodeAddress) * self.nArc) > (2 ** (self.nBytesNodeAddress * 8)):
            self.nBytesNodeAddress += 1

384
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
425
426
427
428
429
430
431
432
433
434
435
436









437
438
439
440
441
442
443
444
445
446
447
                for oNextNode in self.lSortedNodes[i].arcs.values():
                    if 1 < (oNextNode.addr - self.lSortedNodes[i].addr) < self.nMaxOffset:
                        nSize -= nDiff
                if self.lSortedNodes[i].size != nSize:
                    self.lSortedNodes[i].size = nSize
                    bEnd = False

    def _writeAsJSObject (self, spfDst, nCompressionMethod, bInJSModule=False, bBinaryDictAsHexString=True):
        if not spfDst.endswith(".json"):
            spfDst += "."+str(nCompressionMethod)+".json"
        byDic = b""
        if nCompressionMethod == 1:
            byDic = self.oRoot.convToBytes1(self.nBytesArc, self.nBytesNodeAddress)
            for oNode in self.lMinimizedNodes:
                byDic += oNode.convToBytes1(self.nBytesArc, self.nBytesNodeAddress)
        elif nCompressionMethod == 2:
            byDic = self.oRoot.convToBytes2(self.nBytesArc, self.nBytesNodeAddress)
            for oNode in self.lSortedNodes:
                byDic += oNode.convToBytes2(self.nBytesArc, self.nBytesNodeAddress)
        elif nCompressionMethod == 3:
            byDic = self.oRoot.convToBytes3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset)
            for oNode in self.lSortedNodes:
                byDic += oNode.convToBytes3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset)

        with open(spfDst, "w", encoding="utf-8", newline="\n") as hDst:
            if bInJSModule:
                hDst.write('// JavaScript\n// Generated data (do not edit)\n\n"use strict";\n\nconst dictionary = ')
            hDst.write(json.dumps({
                            "sHeader": "/pyfsa/",
                            "sLangCode": self.sLangCode,
                            "sLangName": self.sLangName,
                            "sDicName": self.sDicName,
                            "sFileName": self.sFileName,
                            "sDate": self._getDate(),
                            "nEntry": self.nEntry,
                            "nChar": self.nChar,
                            "nAff": self.nAff,
                            "nTag": self.nTag,
                            "cStemming": self.cStemming,
                            "dChar": self.dChar,
                            "nNode": self.nNode,
                            "nArc": self.nArc,
                            "nArcVal": self.nArcVal,
                            "lArcVal": self.lArcVal,
                            "nCompressionMethod": 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": byDic.hex()  if bBinaryDictAsHexString  else [ e  for e in byDic ]
                        }, ensure_ascii=False))









            if bInJSModule:
                hDst.write(";\n\nexports.dictionary = dictionary;\n")

    def _writeBinary (self, sPathFile, nCompressionMethod):
        """
        Format of the binary indexable dictionary:
        Each section is separated with 4 bytes of \0
        
        - Section Header:
            /pyfsa/[compression method]
                * compression method is an ASCII string







|
<
|













<
<
<
<
<
|



















|



|
>
>
>
>
>
>
>
>
>



|







380
381
382
383
384
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
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
                for oNextNode in self.lSortedNodes[i].arcs.values():
                    if 1 < (oNextNode.addr - self.lSortedNodes[i].addr) < self.nMaxOffset:
                        nSize -= nDiff
                if self.lSortedNodes[i].size != nSize:
                    self.lSortedNodes[i].size = nSize
                    bEnd = False

    def getBinaryAsJSON (self, nCompressionMethod=1, bBinaryDictAsHexString=True):

        self._calculateBinary(nCompressionMethod)
        byDic = b""
        if nCompressionMethod == 1:
            byDic = self.oRoot.convToBytes1(self.nBytesArc, self.nBytesNodeAddress)
            for oNode in self.lMinimizedNodes:
                byDic += oNode.convToBytes1(self.nBytesArc, self.nBytesNodeAddress)
        elif nCompressionMethod == 2:
            byDic = self.oRoot.convToBytes2(self.nBytesArc, self.nBytesNodeAddress)
            for oNode in self.lSortedNodes:
                byDic += oNode.convToBytes2(self.nBytesArc, self.nBytesNodeAddress)
        elif nCompressionMethod == 3:
            byDic = self.oRoot.convToBytes3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset)
            for oNode in self.lSortedNodes:
                byDic += oNode.convToBytes3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset)





        return json.dumps({ "sHeader": "/pyfsa/",
                            "sLangCode": self.sLangCode,
                            "sLangName": self.sLangName,
                            "sDicName": self.sDicName,
                            "sFileName": self.sFileName,
                            "sDate": self._getDate(),
                            "nEntry": self.nEntry,
                            "nChar": self.nChar,
                            "nAff": self.nAff,
                            "nTag": self.nTag,
                            "cStemming": self.cStemming,
                            "dChar": self.dChar,
                            "nNode": self.nNode,
                            "nArc": self.nArc,
                            "nArcVal": self.nArcVal,
                            "lArcVal": self.lArcVal,
                            "nCompressionMethod": nCompressionMethod,
                            "nBytesArc": self.nBytesArc,
                            "nBytesNodeAddress": self.nBytesNodeAddress,
                            "nBytesOffset": self.nBytesOffset,
                            # 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": byDic.hex()  if bBinaryDictAsHexString  else [ e  for e in byDic ]
                        }, ensure_ascii=False)


    def writeAsJSObject (self, spfDst, nCompressionMethod, bInJSModule=False, bBinaryDictAsHexString=True):
        if not spfDst.endswith(".json"):
            spfDst += "."+str(nCompressionMethod)+".json"
        with open(spfDst, "w", encoding="utf-8", newline="\n") as hDst:
            if bInJSModule:
                hDst.write('// JavaScript\n// Generated data (do not edit)\n\n"use strict";\n\nconst dictionary = ')
            hDst.write( self.getBinaryAsJSON(nCompressionMethod, bBinaryDictAsHexString) )
            if bInJSModule:
                hDst.write(";\n\nexports.dictionary = dictionary;\n")

    def writeBinary (self, sPathFile, nCompressionMethod, bDebug=False):
        """
        Format of the binary indexable dictionary:
        Each section is separated with 4 bytes of \0
        
        - Section Header:
            /pyfsa/[compression method]
                * compression method is an ASCII string
468
469
470
471
472
473
474

475
476
477
478
479
480
481
        - Section Values:
                * a list of strings encoded in binary from utf-8, each value separated with a tabulation
        
        - Section Word Graph (nodes / arcs)
                * A list of nodes which are a list of arcs with an address of the next node.
                  See DawgNode.convToBytes() for details.
        """

        if not sPathFile.endswith(".bdic"):
            sPathFile += "."+str(nCompressionMethod)+".bdic"
        with open(sPathFile, 'wb') as hDst:
            # header
            hDst.write("/pyfsa/{}/".format(nCompressionMethod).encode("utf-8"))
            hDst.write(b"\0\0\0\0")
            # infos







>







467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
        - Section Values:
                * a list of strings encoded in binary from utf-8, each value separated with a tabulation
        
        - Section Word Graph (nodes / arcs)
                * A list of nodes which are a list of arcs with an address of the next node.
                  See DawgNode.convToBytes() for details.
        """
        self._calculateBinary(nCompressionMethod)
        if not sPathFile.endswith(".bdic"):
            sPathFile += "."+str(nCompressionMethod)+".bdic"
        with open(sPathFile, 'wb') as hDst:
            # header
            hDst.write("/pyfsa/{}/".format(nCompressionMethod).encode("utf-8"))
            hDst.write(b"\0\0\0\0")
            # infos
496
497
498
499
500
501
502
503

504
505
506
507
508
509
510
                hDst.write(self.oRoot.convToBytes2(self.nBytesArc, self.nBytesNodeAddress))
                for oNode in self.lSortedNodes:
                    hDst.write(oNode.convToBytes2(self.nBytesArc, self.nBytesNodeAddress))
            elif nCompressionMethod == 3:
                hDst.write(self.oRoot.convToBytes3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset))
                for oNode in self.lSortedNodes:
                    hDst.write(oNode.convToBytes3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset))
            hDst.close()


    def _getDate (self):
        return time.strftime("%Y.%m.%d, %H:%M")

    def _writeNodes (self, sPathFile, nCompressionMethod):
        "for debugging only"
        print(" > Write nodes")







|
>







496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
                hDst.write(self.oRoot.convToBytes2(self.nBytesArc, self.nBytesNodeAddress))
                for oNode in self.lSortedNodes:
                    hDst.write(oNode.convToBytes2(self.nBytesArc, self.nBytesNodeAddress))
            elif nCompressionMethod == 3:
                hDst.write(self.oRoot.convToBytes3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset))
                for oNode in self.lSortedNodes:
                    hDst.write(oNode.convToBytes3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset))
        if bDebug:
            self._writeNodes(sPathFile, nCompressionMethod)

    def _getDate (self):
        return time.strftime("%Y.%m.%d, %H:%M")

    def _writeNodes (self, sPathFile, nCompressionMethod):
        "for debugging only"
        print(" > Write nodes")
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
                for oNode in self.lSortedNodes:
                    hDst.write(oNode.getTxtRepr2(self.nBytesArc, self.nBytesNodeAddress, self.lArcVal)+"\n")
            if nCompressionMethod == 3:
                hDst.write(self.oRoot.getTxtRepr3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset, self.lArcVal)+"\n")
                #hDst.write( ''.join( [ "%02X " %  z  for z in self.oRoot.convToBytes3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset) ] ).strip() )
                for oNode in self.lSortedNodes:
                    hDst.write(oNode.getTxtRepr3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset, self.lArcVal)+"\n")
            hDst.close()
    
    def writeResults (self, sPathFile):
        bFileExits = os.path.isfile("_lexicons.res.txt")
        with open("_lexicons.res.txt", "a", encoding='utf-8', newline="\n") as hDst:
            sFormat1 = "{:<12} {:>12} {:>5} {:>8} {:>8} {:>6} {:>8} {:>9} {:>9} {:>15} {:>12} {:>12}\n"
            sFormat2 = "{:<12} {:>12,} {:>5,} {:>8,} {:>8} {:>6,} {:>8,} {:>9,} {:>9,} {:>15,} {:>12,} {:>12,}\n"
            if not bFileExits:
                hDst.write(sFormat1.format("Lexicon", "Entries", "Chars", "Affixes", "Stemming", "Tags", "Values", "Nodes", "Arcs", "Lexicon (Kb)", "Dict (Kb)", "LT Dict (Kb)"))
            hDst.write(sFormat2.format(self.sLangName, self.nEntry, self.nChar, self.nAff, self.cStemming + "FX", self.nTag, self.nArcVal, \
                                       self.nNode, self.nArc, os.path.getsize(self.sFileName), os.path.getsize(sPathFile), \
                                       os.path.getsize("cfsa/dict/{}.dict".format(self.sLangName)) if os.path.isfile("cfsa/dict/{}.dict".format(self.sLangName)) else 0))
            hDst.close()



class DawgNode:
    NextId = 0
    NextPos = 1 # (version 2)
    







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







520
521
522
523
524
525
526













527
528
529
530
531
532
533
                for oNode in self.lSortedNodes:
                    hDst.write(oNode.getTxtRepr2(self.nBytesArc, self.nBytesNodeAddress, self.lArcVal)+"\n")
            if nCompressionMethod == 3:
                hDst.write(self.oRoot.getTxtRepr3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset, self.lArcVal)+"\n")
                #hDst.write( ''.join( [ "%02X " %  z  for z in self.oRoot.convToBytes3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset) ] ).strip() )
                for oNode in self.lSortedNodes:
                    hDst.write(oNode.getTxtRepr3(self.nBytesArc, self.nBytesNodeAddress, self.nBytesOffset, self.lArcVal)+"\n")
















class DawgNode:
    NextId = 0
    NextPos = 1 # (version 2)