1

我正在尝试解压缩文件进行一些更改，然后再次压缩它。解压缩似乎工作正常。我只需要一种方法来压缩它，以便它像以前一样被压缩。

这是我的代码：

import zlib

path = './input'
pathOut = './output'

def getInt(intArray):
    summe = 0
    for i in range(len(list(intArray))):
        summe += intArray[i]*256**i
        return(summe)


print(path)
inputfile = open(path, 'rb')
header = {}

header.update({"intro":inputfile.read(28)})
print("intro",(header["intro"]))

for key in ["header_size", "c_size", "header_v", "u_size", "blocks"]:
    header.update({key:inputfile.read(4)})
    print(key,getInt(header[key]))


inputfile.seek(getInt(header['header_size']))

blocks_count = getInt(header['blocks'])
data = []
for i in range(blocks_count):
    block_header = {}
    block_header.update({"c_size":inputfile.read(2)})
    print("c_size",getInt(block_header["c_size"]))
    block_header.update({"u_size":inputfile.read(2)})
    print("u_size",getInt(block_header["u_size"]))
    block_header.update({"checksum":inputfile.read(4)})
    print("checksum",getInt(block_header["checksum"]))
    temp = inputfile.read(getInt(block_header['c_size']))[2:-4]
    data.append(zlib.decompressobj().decompress(b'x\x9c' + temp))
output = b''
inputfile.seek(0)
output = inputfile.read(getInt(header["header_size"]))
inputfile.close()

compressor = zlib.compressobj(1)

for block in data:
    compressor.compress(block)
    output += compressor.flush(zlib.Z_SYNC_FLUSH)

print("output length",len(output))
print("c_size",getInt(header["c_size"]))
outputFile = open(pathOut, 'wb')
outputFile.write(output)
outputFile.close()

当我尝试解压缩我的输出时，它说：

Traceback (most recent call last):
    File "C:\Users\LSDesktop\Desktop\bla - Copy.py", line 45, in <module>
        data.append(zlib.decompressobj().decompress(b'x\x9c' + temp))
zlib.error: Error -3 while decompressing data: invalid stored block lengths

这是前 50 个字节，包括第一个块的标头。它们应该是怎样的（在我压缩之前它们是怎样的）：

b'B\x0c\x00 rWfyx\x01\xacY\tX\x14G\x16\xae\x9e)p\x00\r\x18\x15\xb9Lk\xc7\x03\xaf\x88\x1a\x8c\x0c\xa8\x084(\x82\x17\x87\x17\x88\xa3\x8c\x1c;r\x0b'

以及压缩后的前 50 个字节，包括第一个块的标头：

b'\xacY\tX\x14G\x16\xae\x9e)p\x00\r\x18\x15\xb9Lk\xc7\x03\xaf\x88\x1a\x8c\x0c\xa8\x084(\x82\x17\x87\x17\x88\xa3\x8c\x1c;r\x0b\x82&\nk4\x1c\x89fQ\x89'
4

1 回答 1

0

Try Z_FINISH instead of Z_SYNC_FLUSH. You are decompressing separate zlib streams, but you then try to make a new single zlib stream that is never terminated. To make separate zlib streams, you need to use Z_FINISH each time. You must use Z_FINISH at least once at the end, or you will never generate a proper zlib stream.

As for the wrappers around that, if you are attempting to write out the same format you read in, you're not doing that. You need to recreate the header information, i.e. the lengths and checksums. Where is this format documented? Did you copy the code to read the input from somewhere else?

This is not the problem, but you should not be stripping the zlib header and trailer with [2:-4], and then adding a header back with b'x\x9c' +. In order to decompress. Just decompress the zlib stream directly.

于 2013-08-01T04:53:03.590 回答