8

my question is simple:

Is there any way to code in a pythonic way that bash command?

hexdump -e '2/1 "%02x"' file.dat

Obviously, without using os, popen, or any shortcut ;)

EDIT: although I've not explicitly specified, it would be great if the code was functional in Python3.x

Thanks!

peluzza
  • 329
  • 2
  • 4
  • 11
  • What does `2/1 "%02x"` mean? – anatoly techtonik Dec 04 '14 at 14:15
  • "%02x" prints each byte as a 2 character, 0 prefixed capital hex number. – peluzza Jan 27 '15 at 15:19
  • And what about `2/1`? The question would be much clearer for those who know Python, but now familiar with hexfump cli. – anatoly techtonik Jan 27 '15 at 16:43
  • Take a look at that : << http://256.com/gray/docs/misc/hexdump_manual_how_to.html >> "An interation count which defaults to 1 if not supplied but has to be supplied if you want a byte count. This tells how many times to do the conversion before we print the end string. So if you were decoding 4 things, each of 1 byte, you'd say 4/1. " – peluzza Jan 27 '15 at 21:15

3 Answers3

16

The standard library is your friend. Try binascii.hexlify().

i0x539
  • 4,763
  • 2
  • 20
  • 28
Raymond Hettinger
  • 216,523
  • 63
  • 388
  • 485
14

If you only care about Python 2.x, line.encode('hex') will encode a chunk of binary data into hex. So:

with open('file.dat', 'rb') as f:
    for chunk in iter(lambda: f.read(32), b''):
        print chunk.encode('hex')

(IIRC, hexdump by default prints 32 pairs of hex per line; if not, just change that 32 to 16 or whatever it is…)

If the two-argument iter looks baffling, click the help link; it's not too complicated once you get the idea.

If you care about Python 3.x, encode only works for codecs that convert Unicode strings to bytes; any codecs that convert the other way around (or any other combination), you have to use codecs.encode to do it explicitly:

with open('file.dat', 'rb') as f:
    for chunk in iter(lambda: f.read(32), b''):
        print(codecs.encode(chunk, 'hex'))

Or it may be better to use hexlify:

with open('file.dat', 'rb') as f:
    for chunk in iter(lambda: f.read(32), b''):
        print(binascii.hexlify(chunk))

If you want to do something besides print them out, rather than read the whole file into memory, you probably want to make an iterator. You could just put this in a function and change that print to a yield, and that function returns exactly the iterator you want. Or use a genexpr or map call:

with open('file.dat', 'rb') as f:
    chunks = iter(lambda: f.read(32), b'')
    hexlines = map(binascii.hexlify, chunks)
abarnert
  • 354,177
  • 51
  • 601
  • 671
5

Simply read() the whole file and encode('hex'). What could be more pythonic?

with open('file.dat', 'rb') as f:
    hex_content = f.read().encode('hex')
Kijewski
  • 25,517
  • 12
  • 101
  • 143
  • Except you almost certainly want to open it in `rb` mode so it doesn't translate newlines. Also, this is Python 2-specific; in Python 3, you can't `encode` bytes. Still +1. – abarnert Jul 28 '14 at 23:18
  • 2
    Great approach, it works, but only in python2. This is the output in Py3.4: UnicodeDecodeError: 'utf-8' codec can't decode byte 0xac in position 0: invalid start byte – peluzza Jul 28 '14 at 23:20
  • 1
    @peluzza: Do you need Python 3? – abarnert Jul 28 '14 at 23:20
  • Well, I'm doing my best to code only for 3.X, but the gaps are so deep, not only working with hex dumps ;) – peluzza Jul 28 '14 at 23:23
  • @peluzza: See Raymond Hettinger's answer, or mine if you need more details. – abarnert Jul 28 '14 at 23:25