def _handle_transfer_encoding(headers, body):
if not "Transfer-Encoding" in headers:
return BitStream(body)
if not headers["Transfer-Encoding"].strip().lower() == "chunked":
return BitStream(body)
new_body = BitStream()
# Chunked responses are encoded by separating the response into chunks.
# Each chunk begins with a number followed by \r\n.
# The number is how many bytes the chunk contains.
# After the \r\n\r\n the chunk bytes go. The chunk ends after the
# specified number of byes and another \r\n
start_pos = 0
while start_pos < len(body):
body.pos = start_pos
i = body.find(Bits(bytes=b'\r\n'), start=start_pos, bytealigned=True)
if not i:
break
i = i[0]
logger.debug(i)
body.pos = start_pos # rewind
h = body[start_pos:i].bytes.decode('ascii')
num_bits = int(h, 16) * 8
logger.debug("Chunk with {num_bytes} bytes at binary pos {start_pos}".\
format(num_bytes = num_bits//8, start_pos = start_pos))
if num_bits == 0:
break
chunk_data = body[i + 16:i + 16 + num_bits]
logger.debug("{b} len: {l}".format(b=str(chunk_data),
l=len(chunk_data)))
new_body.insert(chunk_data)
start_pos += i + num_bits + 32
# todo: handle chunk extensions.
body = new_body
return body
# coding: utf-8
from bitstring import BitStream
s = BitStream(filename="t.bmp")
s.pos = 0x436 * 8
while s.pos <= s.len - 7:
s.insert('0b0', s.pos)
s.pos += 7
f = file('gen.bmp', 'wb')
s.tofile(f)
f.close()
print 'Done'
