def createFields(self):
if self.frame["../type"].value in [0xC0, 0xC1]:
# yay, huffman coding!
if not hasattr(self, "huffman_tables"):
self.huffman_tables = {}
for huffman in self.parent.array("huffman"):
for table in huffman["content"].array("huffman_table"):
for _dummy_ in table:
# exhaust table, so the huffman tree is built
pass
self.huffman_tables[table["table_class"].value, table["index"].value] = table.tree
components = [] # sos_comp, samples
max_vert = 0
max_horiz = 0
for component in self.scan.array("component"):
for sof_comp in self.frame.array("component"):
if sof_comp["component_id"].value == component["component_id"].value:
vert = sof_comp["vert_sample"].value
horiz = sof_comp["horiz_sample"].value
components.append((component, vert * horiz))
max_vert = max(max_vert, vert)
max_horiz = max(max_horiz, horiz)
mcu_height = alignValue(self.frame["height"].value, 8 * max_vert) // (8 * max_vert)
mcu_width = alignValue(self.frame["width"].value, 8 * max_horiz) // (8 * max_horiz)
if self.restart_interval and self.restart_offset > 0:
mcu_number = self.restart_interval * self.restart_offset
else:
mcu_number = 0
initial_mcu = mcu_number
while True:
if (self.restart_interval and mcu_number != initial_mcu and mcu_number % self.restart_interval == 0) or\
mcu_number == mcu_height * mcu_width:
padding = paddingSize(self.current_size, 8)
if padding:
yield PaddingBits(self, "padding[]", padding) # all 1s
last_byte = self.stream.readBytes(self.absolute_address + self.current_size - 8, 1)
if last_byte == '\xFF':
next_byte = self.stream.readBytes(self.absolute_address + self.current_size, 1)
if next_byte != '\x00':
raise FieldError("Unexpected byte sequence %r!"%(last_byte + next_byte))
yield NullBytes(self, "stuffed_byte[]", 1)
break
for sos_comp, num_units in components:
for interleave_count in range(num_units):
yield JpegHuffmanImageUnit(self, "block[%i]component[%i][]" % (mcu_number, sos_comp["component_id"].value),
self.huffman_tables[0, sos_comp["dc_coding_table"].value],
self.huffman_tables[1, sos_comp["ac_coding_table"].value])
mcu_number += 1
else:
self.warning("Sorry, only supporting Baseline & Extended Sequential JPEG images so far!")
return
def extend_data(data, length, offset):
"""Extend data using a length and an offset."""
if length >= offset:
new_data = data[-offset:] * (alignValue(length, offset) // offset)
return data + new_data[:length]
else:
return data + data[-offset:-offset+length]
def extend_data(data, length, offset):
"""Extend data using a length and an offset."""
if length >= offset:
new_data = data[-offset:] * (alignValue(length, offset) // offset)
return data + new_data[:length]
else:
return data + data[-offset:-offset + length]
def __init__(self, *args):
FieldSet.__init__(self, *args)
self._size = (8 + alignValue(self["size"].value, 2)) * 8
tag = self["type"].value
if tag in self.TAG_INFO:
self._name, self._description, self._parser = self.TAG_INFO[tag]
else:
self._parser = None
def __init__(self, *args, **kw):
FieldSet.__init__(self, *args, **kw)
try:
self._name, self.handler, self._description = self.TAG_INFO[self["tag"].value]
except KeyError:
self.handler = None
size = self["size"]
self._size = size.address + size.size + alignValue(size.value, 2) * 8
def __init__(self, *args):
FieldSet.__init__(self, *args)
self._size = (8 + alignValue(self["size"].value, 2)) * 8
tag = self["type"].value
if tag in self.TAG_INFO:
self._name, self._description, self._parser = self.TAG_INFO[tag]
else:
self._parser = None
def calc_char_range(start, end):
aligned_start = byte_addr(start)
aligned_end = byte_addr(alignValue(end, 8))
char_start = calc_char_pos(aligned_start)
char_end = calc_char_pos(aligned_end)
return char_start, char_end
def __init__(self, *args, **kw):
FieldSet.__init__(self, *args, **kw)
try:
self._name, self.handler, self._description = self.TAG_INFO[
self["tag"].value]
except KeyError:
self.handler = None
size = self["size"]
self._size = size.address + size.size + alignValue(size.value, 2) * 8
def createFields(self):
if self.has_all_byte:
yield Enum(UInt8(self, "all_defined"), {0: 'False', 1: 'True'})
if self['all_defined'].value:
return
nbytes = alignValue(self.num, 8) // 8
ctr = 0
for i in xrange(nbytes):
for j in reversed(xrange(8)):
yield Bit(self, "bit[%d]" % (ctr + j))
ctr += 8
def createFields(self):
if self.has_all_byte:
yield Enum(UInt8(self, "all_defined"), {0: "False", 1: "True"})
if self["all_defined"].value:
return
nbytes = alignValue(self.num, 8) // 8
ctr = 0
for i in xrange(nbytes):
for j in reversed(xrange(8)):
yield Bit(self, "bit[%d]" % (ctr + j))
ctr += 8
def createFields(self):
group = self["../group_desc/group[%u]" % self.uniq_id]
superblock = self["/superblock"]
block_size = self["/"].block_size
# Read block bitmap
addr = self.absolute_address + 56 * 8
self.superblock_copy = (self.stream.readBytes(addr, 2) == "\x53\xEF")
if self.superblock_copy:
yield SuperBlock(self, "superblock_copy")
# Compute number of block and inodes
block_count = superblock["blocks_per_group"].value
inode_count = superblock["inodes_per_group"].value
block_index = self.uniq_id * block_count
inode_index = self.uniq_id * inode_count
if (block_count % 8) != 0:
raise ParserError("Invalid block count")
if (inode_count % 8) != 0:
raise ParserError("Invalid inode count")
block_count = min(block_count,
superblock["blocks_count"].value - block_index)
inode_count = min(inode_count,
superblock["inodes_count"].value - inode_index)
# Read block bitmap
field = self.seekByte(group["block_bitmap"].value * block_size,
relative=False,
null=True)
if field:
yield field
yield BlockBitmap(self, "block_bitmap", block_index, block_count,
"Block bitmap")
# Read inode bitmap
field = self.seekByte(group["inode_bitmap"].value * block_size,
relative=False)
if field:
yield field
yield InodeBitmap(self, "inode_bitmap", inode_index, inode_count,
"Inode bitmap")
# Read inode table
field = self.seekByte(alignValue(self.current_size // 8, block_size))
if field:
yield field
yield InodeTable(self, "inode_table", inode_index, inode_count)
# Add padding if needed
addr = min(self.parent.size / 8, (self.uniq_id + 1) *
superblock["blocks_per_group"].value * block_size)
yield self.seekByte(addr, "data", relative=False)
def setCheckedSizes(self, size):
# First set size so that end is aligned, if needed
self.real_size = size
size *= 8
if self.ALIGN:
size = alignValue(self.absolute_address+size, 8*self.ALIGN) \
- self.absolute_address
if self._parent._size:
if self._parent.current_size + size > self._parent._size:
size = self._parent._size - self._parent.current_size
self._size = size
def setCheckedSizes(self, size):
# First set size so that end is aligned, if needed
self.real_size = size
size *= 8
if self.ALIGN:
size = alignValue(self.absolute_address+size, 8*self.ALIGN) \
- self.absolute_address
if self._parent._size:
if self._parent.current_size + size > self._parent._size:
size = self._parent._size - self._parent.current_size
self._size = size
def hexadecimal(field):
"""
Convert an integer to hexadecimal in lower case. Returns unicode string.
>>> hexadecimal(type("", (), dict(value=412, size=16)))
u'0x019c'
>>> hexadecimal(type("", (), dict(value=0, size=32)))
u'0x00000000'
"""
assert hasattr(field, "value") and hasattr(field, "size")
size = field.size
padding = alignValue(size, 4) // 4
pattern = u"0x%%0%ux" % padding
return pattern % field.value
def __init__(self, *args, **kw):
FieldSet.__init__(self, *args, **kw)
self._size = (8 + alignValue(self["size"].value, 2)) * 8
tag = self["tag"].value
if tag in self.TAG_INFO:
self.tag_info = self.TAG_INFO[tag]
if tag == "LIST":
subtag = self["subtag"].value
if subtag in self.subtag_info:
info = self.subtag_info[subtag]
self.tag_info = (info[0], None, info[1])
self._name = self.tag_info[0]
self._description = self.tag_info[2]
else:
self.tag_info = ("field[]", None, None)
def __init__(self, *args, **kw):
FieldSet.__init__(self, *args, **kw)
self._size = (8 + alignValue(self["size"].value, 2)) * 8
tag = self["tag"].value
if tag in self.TAG_INFO:
self.tag_info = self.TAG_INFO[tag]
if tag == "LIST":
subtag = self["subtag"].value
if subtag in self.subtag_info:
info = self.subtag_info[subtag]
self.tag_info = (info[0], None, info[1])
self._name = self.tag_info[0]
self._description = self.tag_info[2]
else:
self.tag_info = ("field[]", None, None)
def createFields(self):
group = self["../group_desc/group[%u]" % self.uniq_id]
superblock = self["/superblock"]
block_size = self["/"].block_size
# Read block bitmap
addr = self.absolute_address + 56*8
self.superblock_copy = (self.stream.readBytes(addr, 2) == "\x53\xEF")
if self.superblock_copy:
yield SuperBlock(self, "superblock_copy")
# Compute number of block and inodes
block_count = superblock["blocks_per_group"].value
inode_count = superblock["inodes_per_group"].value
block_index = self.uniq_id * block_count
inode_index = self.uniq_id * inode_count
if (block_count % 8) != 0:
raise ParserError("Invalid block count")
if (inode_count % 8) != 0:
raise ParserError("Invalid inode count")
block_count = min(block_count, superblock["blocks_count"].value - block_index)
inode_count = min(inode_count, superblock["inodes_count"].value - inode_index)
# Read block bitmap
field = self.seekByte(group["block_bitmap"].value * block_size, relative=False, null=True)
if field:
yield field
yield BlockBitmap(self, "block_bitmap", block_index, block_count, "Block bitmap")
# Read inode bitmap
field = self.seekByte(group["inode_bitmap"].value * block_size, relative=False)
if field:
yield field
yield InodeBitmap(self, "inode_bitmap", inode_index, inode_count, "Inode bitmap")
# Read inode table
field = self.seekByte(alignValue(self.current_size//8, block_size))
if field:
yield field
yield InodeTable(self, "inode_table", inode_index, inode_count)
# Add padding if needed
addr = min(self.parent.size / 8,
(self.uniq_id+1) * superblock["blocks_per_group"].value * block_size)
yield self.seekByte(addr, "data", relative=False)
def createFields(self):
yield String(self, "signature", 4, "8BIM signature", charset="ASCII")
if self["signature"].value != "8BIM":
raise ParserError("Stream doesn't look like 8BIM item (wrong signature)!")
yield textHandler(UInt16(self, "tag"), hexadecimal)
if self.stream.readBytes(self.absolute_address + self.current_size, 4) != "\0\0\0\0":
yield PascalString8(self, "name")
size = 2 + (self["name"].size // 8) % 2
yield NullBytes(self, "name_padding", size)
else:
yield String(self, "name", 4, strip="\0")
yield UInt16(self, "size")
size = alignValue(self["size"].value, 2)
if not size:
return
if self.handler:
yield self.handler(self, "content", size=size*8)
else:
yield RawBytes(self, "content", size)
def readBits(self, address, nbits, endian):
assert endian in (BIG_ENDIAN, LITTLE_ENDIAN, MIDDLE_ENDIAN)
if endian is MIDDLE_ENDIAN:
# read an aligned chunk of words
wordaddr, remainder = divmod(address, 16)
wordnbits = alignValue(remainder + nbits, 16)
_, data, missing = self.read(wordaddr * 16, wordnbits)
shift = remainder
else:
shift, data, missing = self.read(address, nbits)
if missing:
raise ReadStreamError(nbits, address)
value = str2long(data, endian)
if endian in (BIG_ENDIAN, MIDDLE_ENDIAN):
value >>= len(data) * 8 - shift - nbits
else:
value >>= shift
return value & (1 << nbits) - 1
def readBits(self, address, nbits, endian):
assert endian in (BIG_ENDIAN, LITTLE_ENDIAN, MIDDLE_ENDIAN)
if endian is MIDDLE_ENDIAN:
# read an aligned chunk of words
wordaddr, remainder = divmod(address, 16)
wordnbits = alignValue(remainder+nbits, 16)
_, data, missing = self.read(wordaddr*16, wordnbits)
shift = remainder
else:
shift, data, missing = self.read(address, nbits)
if missing:
raise ReadStreamError(nbits, address)
value = str2long(data, endian)
if endian in (BIG_ENDIAN, MIDDLE_ENDIAN):
value >>= len(data) * 8 - shift - nbits
else:
value >>= shift
return value & (1 << nbits) - 1
def createFields(self):
yield String(self, "signature", 4, "8BIM signature", charset="ASCII")
if self["signature"].value != "8BIM":
raise ParserError(
"Stream doesn't look like 8BIM item (wrong signature)!")
yield textHandler(UInt16(self, "tag"), hexadecimal)
if self.stream.readBytes(self.absolute_address + self.current_size,
4) != "\0\0\0\0":
yield PascalString8(self, "name")
size = 2 + (self["name"].size // 8) % 2
yield NullBytes(self, "name_padding", size)
else:
yield String(self, "name", 4, strip="\0")
yield UInt16(self, "size")
size = alignValue(self["size"].value, 2)
if not size:
return
if self.handler:
yield self.handler(self, "content", size=size * 8)
else:
yield RawBytes(self, "content", size)
def handle_form():
"""Process submitted data.
See comments for details.
"""
prune_old()
form = cgi.FieldStorage()
if 'file' in form and form['file'].file:
# compute session id
sessid = get_sessid()
if not sessid:
rand = str(time.time())+form['file'].filename+str(random.random())
sessid = hashlib.md5(rand).hexdigest()
# write uploaded file
f = open(tmp_dir+sessid+'.file','wb')
if form['file'].done==-1:
raise ValueError("File upload canceled?")
while f.tell()<2**22: # 4MB limit
chunk = form['file'].file.read(32768) # 32KB chunks
if not chunk:
break
f.write(chunk)
if f.tell() == 0:
f.close()
print_form('Nothing uploaded.')
return
f.close()
# write session variables
try:
fn = unicode(form['file'].filename,'utf-8')
except UnicodeDecodeError:
fn = unicode(form['file'].filename,'iso-8859-1')
# stream "None" represents the original stream
save_data({'filename':fn,'streams':[(None, None, fn)]}, sessid)
# send session id and reset variables
c = SimpleCookie()
c['sess'] = sessid
c['hpath'] = '/' # clear path var.
c['stream'] = '0' # clear stream var
print c # send cookie to client (headers)
print_page() # print AJAX frame page
elif get_sessid(): # or perhaps you already have a file to parse?
if not 'hpath' in form:
print_page()
return
# redirect stderr, so we can catch parser errors
sys.stderr = StringIO()
# load variables
hpath = cgi.escape(form.getfirst('hpath','/'))
stream_id = int(form.getfirst('stream','0'))
path = hpath.split(':')[stream_id]
sessid = get_sessid()
try:
data = cPickle.load(file(tmp_dir+sessid+'.sess','rb'))
except IOError:
print_error('Your file was deleted due to inactivity. '
'Please upload a new one.')
return
stream, parser = get_parser(data, data['streams'][stream_id], sessid)
if parser is None:
return # sorry, couldn't parse file!
if 'save' in form:
# "Download Raw"
f = FileFromInputStream(stream)
fld = parser[path]
f.seek(fld.absolute_address/8)
size = alignValue(fld.size, 8)/8
sys.stdout.write('Content-Type: application/octet-stream\r\n')
sys.stdout.write('Content-Length: %i\r\n'%size)
sys.stdout.write('Content-Disposition: attachment; '
'filename=%s\r\n\r\n'%path.strip('/').split('/')[-1])
sys.stdout.write(f.read(size))
return
elif 'savesub' in form:
# "Download Substream"
stream = parser[path.rstrip('/')].getSubIStream()
filename = path.strip('/').split('/')[-1]
tags = getattr(stream,'tags',[])
for tag in tags:
if tag[0] == 'filename':
filename = tag[1]
sys.stdout.write('Content-Type: application/octet-stream\r\n')
sys.stdout.write('Content-Disposition: attachment; '
'filename=%s\r\n\r\n'%filename)
sys.stdout.write(FileFromInputStream(stream).read())
return
elif 'addStream' in form:
# "Parse Substream"
spath = cgi.escape(form['addStream'].value)
new_stream = parser[spath.rstrip('/')].getSubIStream()
streamdata = FileFromInputStream(new_stream).read()
new_parser = guessParser(new_stream)
if new_parser:
stream = new_stream
parser = new_parser
tags = getattr(stream,'tags',[])
streamname = data['streams'][stream_id][2]+':'
data['streams'].append((tags, streamdata, streamname+spath))
try:
if force_substream_ref:
raise Exception("Use references for all substreams")
save_data(data, sessid)
except Exception:
# many things could go wrong with pickling
data['streams'][-1] = (data['streams'][stream_id],
spath, streamname+spath)
save_data(data, sessid)
path = '/'
hpath += ':/'
stream_id = len(data['streams'])-1
else:
sys.stderr.write("Cannot parse substream %s: "
"No suitable parser\n"%spath)
elif 'delStream' in form:
# "Delete Stream"
n = int(form['delStream'].value)
paths = hpath.split(':')
del paths[n]
del data['streams'][n]
if n >= len(data['streams']):
stream_id = 0
else:
stream_id = n
path = paths[stream_id]
hpath = ':'.join(paths)
save_data(data, sessid)
stream, parser = get_parser(data, data['streams'][stream_id],
sessid)
# update client's variables
c = SimpleCookie()
c['hpath'] = hpath
c['stream'] = str(stream_id)
print c # send cookie to client
# send headers
print 'Content-Type: text/html'
print
# breadcrumb trail path up top
print_path(path, data, stream_id)
# fields
print '''<table id="maintable" border="1">
<tr class="header">
<th class="headertext">Offset</th>
<th class="headertext">Name</th>
<th class="headertext">Type</th>
<th class="headertext">Size</th>
<th class="headertext">Description</th>
<th class="headertext">Data</th>
<th class="headertext">Download Field</th>
</tr>'''
for i in parser[path]:
# determine options
display = i.raw_display if form.getfirst('raw','0') == '1'\
else i.display
disp_off = bits2hex if form.getfirst('hex','0') == '1'\
else bits2dec
addr = i.address if form.getfirst('rel','0') == '1'\
else i.absolute_address
if display == 'None':
display = ''
# clickable name for field sets
if i.is_field_set:
name = '''<span href="#" onClick="goPath('%s%s/')"\
class="fieldlink">%s/</span>'''%(path, i.name, i.name)
else:
name = i.name
print '<tr class="data">'
print '<td class="fldaddress">%s</td>'%disp_off(addr)
print '<td class="fldname">%s</td>'%name
print '<td class="fldtype">%s</td>'%i.__class__.__name__
print '<td class="fldsize">%s</td>'%disp_off(i.size)
print '<td class="flddesc">%s</td>'%i.description
print '<td class="flddisplay">%s</td>'%display
print '<td class="flddownload">'
paths = hpath.split(':')
paths[stream_id] += i.name
url = "%s?hpath=%s&stream=%s"%\
(script_name,':'.join(paths), stream_id)
# hack to determine if a substream is present
# the default getSubIStream() returns InputFieldStream()
# InputFieldStream() then returns an InputSubStream.
# in all the overrides, the return is a different stream type,
# but this is certainly not the safest way to check for
# an overridden method...
# finally, if the field is a SubFile, then it has a custom
# substream, and thus gets the substream features.
if not isinstance(i.getSubIStream(), InputSubStream)\
or isinstance(i, SubFile):
print '<a href="javascript:addStream(\'%s\')"\
class="dllink">Parse Substream</a><br/>'%(path+i.name)
print '<a href="%s&savesub=1"\
class="dllink">Download Substream</a><br/>'%url
print '<a href="%s&save=1"\
class="dllink">Download Raw</a>'%url
else:
print '<a href="%s&save=1"\
class="dllink">Download</a>'%url
print '</td>'
print '</tr>'
print '</table>'
print_path(path, data, stream_id)
if sys.stderr.getvalue():
print_error('Error(s) encountered:', print_headers=False)
print '<pre class="parseerror">%s</pre>'%sys.stderr.getvalue()
else:
print_form('Note: Cookies MUST be enabled!')
def __init__(self, parent, name, format, description=None,
strip=None, charset=None, nbytes=None, truncate=None):
Bytes.__init__(self, parent, name, 1, description)
# Is format valid?
assert format in self.VALID_FORMATS
# Store options
self._format = format
self._strip = strip
self._truncate = truncate
# Check charset and compute character size in bytes
# (or None when it's not possible to guess character size)
if not charset or charset in self.CHARSET_8BIT:
self._character_size = 1 # one byte per character
elif charset in self.UTF_CHARSET:
self._character_size = None
else:
raise FieldError("Invalid charset for %s: \"%s\"" %
(self.path, charset))
self._charset = charset
# It is a fixed string?
if nbytes is not None:
assert self._format == "fixed"
# Arbitrary limits, just to catch some bugs...
if not (1 <= nbytes <= 0xffff):
raise FieldError("Invalid string size for %s: %s" %
(self.path, nbytes))
self._content_size = nbytes # content length in bytes
self._size = nbytes * 8
self._content_offset = 0
else:
# Format with a suffix: Find the end of the string
if self._format in self.SUFFIX_FORMAT:
self._content_offset = 0
# Choose the suffix
suffix = self.suffix_str
# Find the suffix
length = self._parent.stream.searchBytesLength(
suffix, False, self.absolute_address)
if length is None:
raise FieldError("Unable to find end of string %s (format %s)!"
% (self.path, self._format))
if 1 < len(suffix):
# Fix length for little endian bug with UTF-xx charset:
# u"abc" -> "a\0b\0c\0\0\0" (UTF-16-LE)
# search returns length=5, whereas real lenght is 6
length = alignValue(length, len(suffix))
# Compute sizes
self._content_size = length # in bytes
self._size = (length + len(suffix)) * 8
# Format with a prefix: Read prefixed length in bytes
else:
assert self._format in self.PASCAL_FORMATS
# Get the prefix size
prefix_size = self.PASCAL_FORMATS[self._format]
self._content_offset = prefix_size
# Read the prefix and compute sizes
value = self._parent.stream.readBits(
self.absolute_address, prefix_size*8, self._parent.endian)
self._content_size = value # in bytes
self._size = (prefix_size + value) * 8
# For UTF-16 and UTF-32, choose the right charset using BOM
if self._charset in self.UTF_CHARSET:
# Charset requires a BOM?
bomsize, endian = self.UTF_CHARSET[self._charset]
if endian == "BOM":
# Read the BOM value
nbytes = bomsize // 8
bom = self._parent.stream.readBytes(self.absolute_address, nbytes)
# Choose right charset using the BOM
bom_endian = self.UTF_BOM[bomsize]
if bom not in bom_endian:
raise FieldError("String %s has invalid BOM (%s)!"
% (self.path, repr(bom)))
self._charset = bom_endian[bom]
self._content_size -= nbytes
self._content_offset += nbytes
# Compute length in character if possible
if self._character_size:
self._length = self._content_size // self._character_size
else:
self._length = None
def __init__(self, parent, name, format, description=None,
strip=None, charset=None, nbytes=None, truncate=None):
Bytes.__init__(self, parent, name, 1, description)
# Is format valid?
assert format in self.VALID_FORMATS
# Store options
self._format = format
self._strip = strip
self._truncate = truncate
# Check charset and compute character size in bytes
# (or None when it's not possible to guess character size)
if not charset or charset in self.CHARSET_8BIT:
self._character_size = 1 # one byte per character
elif charset in self.UTF_CHARSET:
self._character_size = None
else:
raise FieldError("Invalid charset for %s: \"%s\"" %
(self.path, charset))
self._charset = charset
# It is a fixed string?
if nbytes is not None:
assert self._format == "fixed"
# Arbitrary limits, just to catch some bugs...
if not (1 <= nbytes <= 0xffff):
raise FieldError("Invalid string size for %s: %s" %
(self.path, nbytes))
self._content_size = nbytes # content length in bytes
self._size = nbytes * 8
self._content_offset = 0
else:
# Format with a suffix: Find the end of the string
if self._format in self.SUFFIX_FORMAT:
self._content_offset = 0
# Choose the suffix
suffix = self.suffix_str
# Find the suffix
length = self._parent.stream.searchBytesLength(
suffix, False, self.absolute_address)
if length is None:
raise FieldError("Unable to find end of string %s (format %s)!"
% (self.path, self._format))
if 1 < len(suffix):
# Fix length for little endian bug with UTF-xx charset:
# u"abc" -> "a\0b\0c\0\0\0" (UTF-16-LE)
# search returns length=5, whereas real lenght is 6
length = alignValue(length, len(suffix))
# Compute sizes
self._content_size = length # in bytes
self._size = (length + len(suffix)) * 8
# Format with a prefix: Read prefixed length in bytes
else:
assert self._format in self.PASCAL_FORMATS
# Get the prefix size
prefix_size = self.PASCAL_FORMATS[self._format]
self._content_offset = prefix_size
# Read the prefix and compute sizes
value = self._parent.stream.readBits(
self.absolute_address, prefix_size*8, self._parent.endian)
self._content_size = value # in bytes
self._size = (prefix_size + value) * 8
# For UTF-16 and UTF-32, choose the right charset using BOM
if self._charset in self.UTF_CHARSET:
# Charset requires a BOM?
bomsize, endian = self.UTF_CHARSET[self._charset]
if endian == "BOM":
# Read the BOM value
nbytes = bomsize // 8
bom = self._parent.stream.readBytes(self.absolute_address, nbytes)
# Choose right charset using the BOM
bom_endian = self.UTF_BOM[bomsize]
if bom not in bom_endian:
raise FieldError("String %s has invalid BOM (%s)!"
% (self.path, repr(bom)))
self._charset = bom_endian[bom]
self._content_size -= nbytes
self._content_offset += nbytes
# Compute length in character if possible
if self._character_size:
self._length = self._content_size // self._character_size
else:
self._length = None
def calc_byte_range(start, end):
return byte_addr(start), byte_addr(alignValue(end, 8))
def __init__(self, parent, name):
FieldSet.__init__(self, parent, name)
self._size = alignValue(self["size"].value, 4) * 8
def __init__(self, parent, name, is_32bit=True):
FieldSet.__init__(self, parent, name)
self._size = alignValue(self["size"].value, 4) * 8
self.is_32bit = is_32bit
def __init__(self, parent, name):
FieldSet.__init__(self, parent, name)
self._size = alignValue(self["size"].value, 4) * 8
def handle_form():
"""Process submitted data.
See comments for details.
"""
prune_old()
form = cgi.FieldStorage()
if 'file' in form and form['file'].file:
# compute session id
sessid = get_sessid()
if not sessid:
rand = str(time.time()) + form['file'].filename + str(
random.random())
sessid = hashlib.md5(rand).hexdigest()
# write uploaded file
f = open(tmp_dir + sessid + '.file', 'wb')
if form['file'].done == -1:
raise ValueError("File upload canceled?")
while f.tell() < 2**22: # 4MB limit
chunk = form['file'].file.read(32768) # 32KB chunks
if not chunk:
break
f.write(chunk)
if f.tell() == 0:
f.close()
print_form('Nothing uploaded.')
return
f.close()
# write session variables
try:
fn = unicode(form['file'].filename, 'utf-8')
except UnicodeDecodeError:
fn = unicode(form['file'].filename, 'iso-8859-1')
# stream "None" represents the original stream
save_data({'filename': fn, 'streams': [(None, None, fn)]}, sessid)
# send session id and reset variables
c = SimpleCookie()
c['sess'] = sessid
c['hpath'] = '/' # clear path var.
c['stream'] = '0' # clear stream var
print c # send cookie to client (headers)
print_page() # print AJAX frame page
elif get_sessid(): # or perhaps you already have a file to parse?
if not 'hpath' in form:
print_page()
return
# redirect stderr, so we can catch parser errors
sys.stderr = StringIO()
# load variables
hpath = cgi.escape(form.getfirst('hpath', '/'))
stream_id = int(form.getfirst('stream', '0'))
path = hpath.split(':')[stream_id]
sessid = get_sessid()
try:
data = cPickle.load(file(tmp_dir + sessid + '.sess', 'rb'))
except IOError:
print_error('Your file was deleted due to inactivity. '
'Please upload a new one.')
return
stream, parser = get_parser(data, data['streams'][stream_id], sessid)
if parser is None:
return # sorry, couldn't parse file!
if 'save' in form:
# "Download Raw"
f = FileFromInputStream(stream)
fld = parser[path]
f.seek(fld.absolute_address / 8)
size = alignValue(fld.size, 8) / 8
sys.stdout.write('Content-Type: application/octet-stream\r\n')
sys.stdout.write('Content-Length: %i\r\n' % size)
sys.stdout.write('Content-Disposition: attachment; '
'filename=%s\r\n\r\n' %
path.strip('/').split('/')[-1])
sys.stdout.write(f.read(size))
return
elif 'savesub' in form:
# "Download Substream"
stream = parser[path.rstrip('/')].getSubIStream()
filename = path.strip('/').split('/')[-1]
tags = getattr(stream, 'tags', [])
for tag in tags:
if tag[0] == 'filename':
filename = tag[1]
sys.stdout.write('Content-Type: application/octet-stream\r\n')
sys.stdout.write('Content-Disposition: attachment; '
'filename=%s\r\n\r\n' % filename)
sys.stdout.write(FileFromInputStream(stream).read())
return
elif 'addStream' in form:
# "Parse Substream"
spath = cgi.escape(form['addStream'].value)
new_stream = parser[spath.rstrip('/')].getSubIStream()
streamdata = FileFromInputStream(new_stream).read()
new_parser = guessParser(new_stream)
if new_parser:
stream = new_stream
parser = new_parser
tags = getattr(stream, 'tags', [])
streamname = data['streams'][stream_id][2] + ':'
data['streams'].append((tags, streamdata, streamname + spath))
try:
if force_substream_ref:
raise Exception("Use references for all substreams")
save_data(data, sessid)
except Exception:
# many things could go wrong with pickling
data['streams'][-1] = (data['streams'][stream_id], spath,
streamname + spath)
save_data(data, sessid)
path = '/'
hpath += ':/'
stream_id = len(data['streams']) - 1
else:
sys.stderr.write("Cannot parse substream %s: "
"No suitable parser\n" % spath)
elif 'delStream' in form:
# "Delete Stream"
n = int(form['delStream'].value)
paths = hpath.split(':')
del paths[n]
del data['streams'][n]
if n >= len(data['streams']):
stream_id = 0
else:
stream_id = n
path = paths[stream_id]
hpath = ':'.join(paths)
save_data(data, sessid)
stream, parser = get_parser(data, data['streams'][stream_id],
sessid)
# update client's variables
c = SimpleCookie()
c['hpath'] = hpath
c['stream'] = str(stream_id)
print c # send cookie to client
# send headers
print 'Content-Type: text/html'
print
# breadcrumb trail path up top
print_path(path, data, stream_id)
# fields
print '''<table id="maintable" border="1">
<tr class="header">
<th class="headertext">Offset</th>
<th class="headertext">Name</th>
<th class="headertext">Type</th>
<th class="headertext">Size</th>
<th class="headertext">Description</th>
<th class="headertext">Data</th>
<th class="headertext">Download Field</th>
</tr>'''
for i in parser[path]:
# determine options
display = i.raw_display if form.getfirst('raw','0') == '1'\
else i.display
disp_off = bits2hex if form.getfirst('hex','0') == '1'\
else bits2dec
addr = i.address if form.getfirst('rel','0') == '1'\
else i.absolute_address
if display == 'None':
display = ''
# clickable name for field sets
if i.is_field_set:
name = '''<span href="#" onClick="goPath('%s%s/')"\
class="fieldlink">%s/</span>''' % (path, i.name, i.name)
else:
name = i.name
print '<tr class="data">'
print '<td class="fldaddress">%s</td>' % disp_off(addr)
print '<td class="fldname">%s</td>' % name
print '<td class="fldtype">%s</td>' % i.__class__.__name__
print '<td class="fldsize">%s</td>' % disp_off(i.size)
print '<td class="flddesc">%s</td>' % i.description
print '<td class="flddisplay">%s</td>' % display
print '<td class="flddownload">'
paths = hpath.split(':')
paths[stream_id] += i.name
url = "%s?hpath=%s&stream=%s"%\
(script_name,':'.join(paths), stream_id)
# hack to determine if a substream is present
# the default getSubIStream() returns InputFieldStream()
# InputFieldStream() then returns an InputSubStream.
# in all the overrides, the return is a different stream type,
# but this is certainly not the safest way to check for
# an overridden method...
# finally, if the field is a SubFile, then it has a custom
# substream, and thus gets the substream features.
if not isinstance(i.getSubIStream(), InputSubStream)\
or isinstance(i, SubFile):
print '<a href="javascript:addStream(\'%s\')"\
class="dllink">Parse Substream</a><br/>' % (path + i.name)
print '<a href="%s&savesub=1"\
class="dllink">Download Substream</a><br/>' % url
print '<a href="%s&save=1"\
class="dllink">Download Raw</a>' % url
else:
print '<a href="%s&save=1"\
class="dllink">Download</a>' % url
print '</td>'
print '</tr>'
print '</table>'
print_path(path, data, stream_id)
if sys.stderr.getvalue():
print_error('Error(s) encountered:', print_headers=False)
print '<pre class="parseerror">%s</pre>' % sys.stderr.getvalue()
else:
print_form('Note: Cookies MUST be enabled!')
def createFields(self):
yield String(self, "magic", 4, "Signature (BLP2)")
yield Enum(UInt32(self, "compression", "Compression type"), {
0: "JPEG Compressed",
1: "Uncompressed or DXT/S3TC compressed"
})
yield Enum(UInt8(self, "encoding", "Encoding type"), {
1: "Raw",
2: "DXT/S3TC Texture Compression (a.k.a. DirectX)"
})
yield UInt8(self, "alpha_depth",
"Alpha channel depth, in bits (0 = no alpha)")
yield Enum(
UInt8(self, "alpha_encoding", "Encoding used for alpha channel"), {
0: "DXT1 alpha (0 or 1 bit alpha)",
1: "DXT3 alpha (4 bit alpha)",
7: "DXT5 alpha (8 bit interpolated alpha)"
})
yield Enum(
UInt8(self, "has_mips", "Are mip levels present?"), {
0: "No mip levels",
1:
"Mip levels present; number of levels determined by image size"
})
yield UInt32(self, "width", "Base image width")
yield UInt32(self, "height", "Base image height")
for i in xrange(16):
yield UInt32(self, "mipmap_offset[]")
for i in xrange(16):
yield UInt32(self, "mipmap_size[]")
yield PaletteRGBA(self, "palette", 256)
compression = self["compression"].value
encoding = self["encoding"].value
alpha_depth = self["alpha_depth"].value
alpha_encoding = self["alpha_encoding"].value
width = self["width"].value
height = self["height"].value
if compression == 0: # JPEG Compression
yield UInt32(self, "jpeg_header_len")
yield RawBytes(self, "jpeg_header", self["jpeg_header_len"].value,
"Shared JPEG Header")
offsets = self.array("mipmap_offset")
sizes = self.array("mipmap_size")
for i in xrange(16):
if not offsets[i].value or not sizes[i].value:
continue
padding = self.seekByte(offsets[i].value)
if padding:
yield padding
if compression == 0:
yield RawBytes(
self, "mipmap[%i]" % i, sizes[i].value,
"JPEG data, append to header to recover complete image")
elif compression == 1 and encoding == 1:
yield Generic2DArray(self, "mipmap_indexes[%i]" % i, height,
width, PaletteIndex, "row", "index",
"Indexes into the palette")
if alpha_depth == 1:
yield GenericVector(self, "mipmap_alphas[%i]" % i, height,
width, Bit, "row", "is_opaque",
"Alpha values")
elif alpha_depth == 8:
yield GenericVector(self, "mipmap_alphas[%i]" % i, height,
width, UInt8, "row", "alpha",
"Alpha values")
elif compression == 1 and encoding == 2:
block_height = alignValue(height, 4) // 4
block_width = alignValue(width, 4) // 4
if alpha_depth in [0, 1] and alpha_encoding == 0:
yield Generic2DArray(self, "mipmap[%i]" % i, block_height,
block_width, DXT1, "row", "block",
"DXT1-compressed image blocks")
elif alpha_depth == 8 and alpha_encoding == 1:
yield Generic2DArray(self, "mipmap[%i]" % i, block_height,
block_width, DXT3, "row", "block",
"DXT3-compressed image blocks")
elif alpha_depth == 8 and alpha_encoding == 7:
yield Generic2DArray(self, "mipmap[%i]" % i, block_height,
block_width, DXT5, "row", "block",
"DXT5-compressed image blocks")
width /= 2
height /= 2
def __init__(self, parent, name, width, pixel_class):
FieldSet.__init__(self, parent, name)
self._pixel = pixel_class
self._width = width
self._size = alignValue(self._width * self._pixel.static_size, 32)
def __init__(self, parent, name, is_32bit=True):
FieldSet.__init__(self, parent, name)
self._size = alignValue(self["size"].value, 4) * 8
self.is_32bit = is_32bit
def createFields(self):
yield String(self, "magic", 4, "Signature (BLP2)")
yield Enum(UInt32(self, "compression", "Compression type"), {
0:"JPEG Compressed",
1:"Uncompressed or DXT/S3TC compressed"})
yield Enum(UInt8(self, "encoding", "Encoding type"), {
1:"Raw",
2:"DXT/S3TC Texture Compression (a.k.a. DirectX)"})
yield UInt8(self, "alpha_depth", "Alpha channel depth, in bits (0 = no alpha)")
yield Enum(UInt8(self, "alpha_encoding", "Encoding used for alpha channel"), {
0:"DXT1 alpha (0 or 1 bit alpha)",
1:"DXT3 alpha (4 bit alpha)",
7:"DXT5 alpha (8 bit interpolated alpha)"})
yield Enum(UInt8(self, "has_mips", "Are mip levels present?"), {
0:"No mip levels",
1:"Mip levels present; number of levels determined by image size"})
yield UInt32(self, "width", "Base image width")
yield UInt32(self, "height", "Base image height")
for i in xrange(16):
yield UInt32(self, "mipmap_offset[]")
for i in xrange(16):
yield UInt32(self, "mipmap_size[]")
yield PaletteRGBA(self, "palette", 256)
compression = self["compression"].value
encoding = self["encoding"].value
alpha_depth = self["alpha_depth"].value
alpha_encoding = self["alpha_encoding"].value
width = self["width"].value
height = self["height"].value
if compression == 0: # JPEG Compression
yield UInt32(self, "jpeg_header_len")
yield RawBytes(self, "jpeg_header", self["jpeg_header_len"].value, "Shared JPEG Header")
offsets = self.array("mipmap_offset")
sizes = self.array("mipmap_size")
for i in xrange(16):
if not offsets[i].value or not sizes[i].value:
continue
padding = self.seekByte(offsets[i].value)
if padding:
yield padding
if compression == 0:
yield RawBytes(self, "mipmap[%i]" % i, sizes[i].value, "JPEG data, append to header to recover complete image")
elif compression == 1 and encoding == 1:
yield Generic2DArray(self, "mipmap_indexes[%i]" % i, height, width, PaletteIndex, "row", "index", "Indexes into the palette")
if alpha_depth == 1:
yield GenericVector(self, "mipmap_alphas[%i]" % i, height, width, Bit, "row", "is_opaque", "Alpha values")
elif alpha_depth == 8:
yield GenericVector(self, "mipmap_alphas[%i]" % i, height, width, UInt8, "row", "alpha", "Alpha values")
elif compression == 1 and encoding == 2:
block_height = alignValue(height, 4) // 4
block_width = alignValue(width, 4) // 4
if alpha_depth in [0, 1] and alpha_encoding == 0:
yield Generic2DArray(self, "mipmap[%i]" % i, block_height, block_width, DXT1, "row", "block", "DXT1-compressed image blocks")
elif alpha_depth == 8 and alpha_encoding == 1:
yield Generic2DArray(self, "mipmap[%i]" % i, block_height, block_width, DXT3, "row", "block", "DXT3-compressed image blocks")
elif alpha_depth == 8 and alpha_encoding == 7:
yield Generic2DArray(self, "mipmap[%i]" % i, block_height, block_width, DXT5, "row", "block", "DXT5-compressed image blocks")
width /= 2
height /= 2