def decodeWordListToBits(words,
maxDigits,
markovChain,
previousWord=config.startSymbol,
wordsPerState=1):
bitsRange = ["0", "1"]
bitsField = bigBitField.BigBitField()
wordsUsed = 0
markovChainDict = utils.markovChainToDictionary(markovChain)
lastTime = time.time()
secondsForStatusPrint = 20
for word in words:
bitsRange = decodeWordToBitsRange(
word, previousWord, markovChainDict,
maxDigits - bitsField.totalFieldLen(), bitsRange)
wordsUsed = wordsUsed + 1
# compute previous word (or bigram) for next iteration
if wordsPerState == 1:
previousWord = word
elif wordsPerState == 2:
if word == config.startSymbol:
previousWord = (config.startSymbol, config.startSymbol)
else:
previousWord = (previousWord[1], word)
# simplify range, remove bits and add them to the field
bitsRange2 = utils.removeCommonBitsInRange(bitsRange)
bitsRemovedLen = len(bitsRange[0]) - len(bitsRange2[0])
if bitsRemovedLen + bitsField.totalFieldLen() > maxDigits:
bitsRemovedLen = maxDigits - bitsField.totalFieldLen()
bitsField.pushQueueNBits(bitsRange[0][0:bitsRemovedLen])
bitsRange = bitsRange2
if time.time() - lastTime > secondsForStatusPrint:
print(" - decoded bits so far: " + repr(bitsField.totalFieldLen()))
lastTime = time.time()
# we exit when our range describes only one number
if bitsField.totalFieldLen() == maxDigits:
break
if bitsField.totalFieldLen(
) == maxDigits - 1 and bitsRange[0][0] == bitsRange[1][0]:
bitsField.pushQueueNBits(bitsRange[0][0])
break
return (bitsField, wordsUsed)
def encodeBitsToWordList(bitsField,
markovChain,
startWord=config.startSymbol,
wordsPerState=1):
bitsField = bitsField.copy()
lastTime = time.time()
secondsForStatusPrint = 20
words = []
nextRange = ["0", "1"]
markovChainDict = utils.markovChainToDictionary(markovChain)
while True:
# encode to one word
(word, nextRange) = encodeBitsToWord(bitsField, nextRange, startWord,
markovChainDict)
words.append(word)
# compute previous word (or bigram) for next iteration
if wordsPerState == 1:
startWord = word
elif wordsPerState == 2:
if word == config.startSymbol:
startWord = (config.startSymbol, config.startSymbol)
else:
startWord = (startWord[1], word)
# optimization, remove start of range when it is identical in both sides
nextRange2 = utils.removeCommonBitsInRange(nextRange)
bitsField.popFirstNBits(len(nextRange[0]) - len(nextRange2[0]))
nextRange = nextRange2
if time.time() - lastTime > secondsForStatusPrint:
print " - remaining bits: " + repr(bitsField.totalFieldLen())
lastTime = time.time()
# we exit when our range describes only to our number
if bitsField.totalFieldLen() == 0 or (bitsField.totalFieldLen() == 1
and nextRange[0][0]
== nextRange[1][0]):
break
return words
def decodeWordListToBits(words, maxDigits, markovChain, previousWord = config.startSymbol, wordsPerState = 1): bitsRange = ["0", "1"] bitsField = bigBitField.BigBitField() wordsUsed = 0 markovChainDict = utils.markovChainToDictionary(markovChain) lastTime = time.time() secondsForStatusPrint = 20 for word in words: bitsRange = decodeWordToBitsRange(word, previousWord, markovChainDict, maxDigits - bitsField.totalFieldLen(), bitsRange) wordsUsed = wordsUsed + 1 # compute previous word (or bigram) for next iteration if wordsPerState == 1: previousWord = word elif wordsPerState == 2: if word == config.startSymbol: previousWord = (config.startSymbol, config.startSymbol) else: previousWord = (previousWord[1], word) # simplify range, remove bits and add them to the field bitsRange2 = utils.removeCommonBitsInRange(bitsRange) bitsRemovedLen = len(bitsRange[0]) - len(bitsRange2[0]) if bitsRemovedLen + bitsField.totalFieldLen() > maxDigits: bitsRemovedLen = maxDigits - bitsField.totalFieldLen() bitsField.pushQueueNBits(bitsRange[0][0:bitsRemovedLen]) bitsRange = bitsRange2 if time.time()-lastTime > secondsForStatusPrint: print " - decoded bits so far: " + repr(bitsField.totalFieldLen()) lastTime = time.time() # we exit when our range describes only one number if bitsField.totalFieldLen() == maxDigits: break if bitsField.totalFieldLen() == maxDigits - 1 and bitsRange[0][0] == bitsRange[1][0]: bitsField.pushQueueNBits(bitsRange[0][0]) break return (bitsField, wordsUsed)
def encodeBitsToWordList(bitsField, markovChain, startWord = config.startSymbol, wordsPerState = 1): bitsField = bitsField.copy() lastTime = time.time() secondsForStatusPrint = 20 words = [] nextRange = ["0", "1"] markovChainDict = utils.markovChainToDictionary(markovChain) while True: # encode to one word (word, nextRange) = encodeBitsToWord(bitsField, nextRange, startWord, markovChainDict) words.append(word) # compute previous word (or bigram) for next iteration if wordsPerState == 1: startWord = word elif wordsPerState == 2: if word == config.startSymbol: startWord = (config.startSymbol, config.startSymbol) else: startWord = (startWord[1], word) # optimization, remove start of range when it is identical in both sides nextRange2 = utils.removeCommonBitsInRange(nextRange) bitsField.popFirstNBits(len(nextRange[0])-len(nextRange2[0])) nextRange = nextRange2 if time.time()-lastTime > secondsForStatusPrint: print " - remaining bits: " + repr(bitsField.totalFieldLen()) lastTime = time.time() # we exit when our range describes only to our number if bitsField.totalFieldLen() == 0 or (bitsField.totalFieldLen() == 1 and nextRange[0][0] == nextRange[1][0]): break return words
lastTime = time.time()
# we exit when our range describes only to our number
if bitsField.totalFieldLen() == 0 or (bitsField.totalFieldLen() == 1
and nextRange[0][0]
== nextRange[1][0]):
break
return words
if __name__ == '__main__':
print "testing fixedSizeCode.py"
testMarkov = config.testMarkov
testMarkovDict = utils.markovChainToDictionary(testMarkov)
testMarkov2 = config.testMarkov2
testMarkovDict2 = utils.markovChainToDictionary(testMarkov2)
# this is "01000110 01011010 11111111"
testBitField = bigBitField.BigBitField([70, 90, 255])
print "A:"
print encodeBitsToWord(testBitField, ["0", "1"], config.startSymbol,
testMarkovDict) == ('A', ('00', '01'))
print "A2:"
print encodeBitsToWord(testBitField, ["0", "1"],
(config.startSymbol, config.startSymbol),
testMarkovDict2) == ('A', ('00', '01'))
print "B:"
if time.time()-lastTime > secondsForStatusPrint:
print " - remaining bits: " + repr(bitsField.totalFieldLen())
lastTime = time.time()
# we exit when our range describes only to our number
if bitsField.totalFieldLen() == 0 or (bitsField.totalFieldLen() == 1 and nextRange[0][0] == nextRange[1][0]):
break
return words
if __name__ == '__main__':
print "testing fixedSizeCode.py"
testMarkov = config.testMarkov
testMarkovDict = utils.markovChainToDictionary(testMarkov)
testMarkov2 = config.testMarkov2
testMarkovDict2 = utils.markovChainToDictionary(testMarkov2)
# this is "01000110 01011010 11111111"
testBitField = bigBitField.BigBitField([70, 90, 255])
print "A:"
print encodeBitsToWord(testBitField, ["0", "1"], config.startSymbol, testMarkovDict) == ('A', ('00', '01'))
print "A2:"
print encodeBitsToWord(testBitField, ["0", "1"], (config.startSymbol, config.startSymbol), testMarkovDict2) == ('A', ('00', '01'))
print "B:"
print encodeBitsToWord(testBitField, ["0", "1"], "A", testMarkovDict) == ('B', ('01', '01'))
print "B2:"
print encodeBitsToWord(testBitField, ["0", "1"], (config.startSymbol, "A"), testMarkovDict2) == ('B', ('01', '01'))