def add_inode(self, fd, offset, factories):
""" We think we have a zip file here. """
b = Zip.Buffer(fd=fd)[offset:]
try:
header = Zip.ZipFileHeader(b)
size = int(header['uncompr_size'])
compressed_length = int(header['compr_size'])
## Some zip programs seem to leave this at 0 - because its
## already in the central directory. Unfortunately the
## carver currently does not look at the central directory
## - so we just make it a reasonable value
if compressed_length==0:
compressed_length = 100*1024
name = header['zip_path'].get_value()
if len(name)==0 or invalid_filename.search(name):
pyflaglog.log(pyflaglog.DEBUG, "Thought the name %r is invalid - skipping file" % name[:10])
return 10
header_offset = header['data'].buffer.offset
except:
return 10
new_inode = "%s|Z%s:%s" % (fd.inode, offset, compressed_length)
self._add_inode(new_inode, size, name, fd, factories)
return size
class ZipFile(File):
""" A file like object to read files from within zip files.
We essentially decompress the file on the disk because the file
may be exceptionally large.
"""
specifier = 'Z'
def __init__(self, case, fd, inode):
File.__init__(self, case, fd, inode)
## Make sure our parent is cached:
self.fd.cache()
## Parse out inode - if we got the compressed length provided,
## we use that, otherwise we calculate it from the zipfile
## header
parts = inode.split('|')
ourpart = parts[-1][1:]
try:
offset, size = ourpart.split(":")
self.compressed_length = int(size)
offset = int(offset)
except:
offset = int(ourpart)
self.offset = offset
## Ensure that we can read the file header:
b = Zip.Buffer(fd=fd)[offset:]
self.header = Zip.ZipFileHeader(b)
## This is sometimes invalid and set to zero - should we query
## the db?
self.size = int(self.header['uncompr_size'])
if not self.compressed_length:
self.compressed_length = int(self.header['compr_size'])
self.type = int(self.header['compression_method'])
## Where does the data start?
self.init()