Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > comp.lang.python > #101758 > unrolled thread

Writing a stream of bytes

Started byjmp <jeanmichel@sequans.com>
First post2016-01-15 16:55 +0100
Last post2016-01-15 16:04 +0000
Articles 2 — 2 participants

Back to article view | Back to comp.lang.python


Contents

  Writing a stream of bytes jmp <jeanmichel@sequans.com> - 2016-01-15 16:55 +0100
    Re: Writing a stream of bytes BartC <bc@freeuk.com> - 2016-01-15 16:04 +0000

#101758 — Writing a stream of bytes

Fromjmp <jeanmichel@sequans.com>
Date2016-01-15 16:55 +0100
SubjectWriting a stream of bytes
Message-ID<mailman.16.1452873321.15297.python-list@python.org>
Hi pyple !


I'd like to write a stream of bytes into a file. I'd like to use the 
struct (instead of bytearray) module because I will have to write more 
than bytes.

let's say I want a file with 4 bytes in that order:

01 02 03 04

None of these work:

import struct

with open('toto', 'wb') as f: f.write(struct.pack('4B', *[1,2,3,4]))
with open('toto', 'wb') as f: f.write(struct.pack('<4B', *[1,2,3,4]))
with open('toto', 'wb') as f: f.write(struct.pack('>4B', *[1,2,3,4]))

I always end up with the following bytes on file:
!hexdump toto
0000000 0201 0403

Note: the '<' and '>' tells struct to pack for a litle/big endian 
architecture.

The only solution I came up with is

with open('toto', 'wb') as f: f.write(struct.pack('>4B', *[2,1,4,3]))

But I'd rather not manipulate the stream, as it seems to me that this 
would be the job of struct. Or maybe I completely overlooked the actual 
issue ?

Cheers,

jm

[toc] | [next] | [standalone]


#101759

FromBartC <bc@freeuk.com>
Date2016-01-15 16:04 +0000
Message-ID<n7b54j$nn2$1@dont-email.me>
In reply to#101758
On 15/01/2016 15:55, jmp wrote:
> Hi pyple !
>
>
> I'd like to write a stream of bytes into a file. I'd like to use the
> struct (instead of bytearray) module because I will have to write more
> than bytes.
>
> let's say I want a file with 4 bytes in that order:
>
> 01 02 03 04
>
> None of these work:
>
> import struct
>
> with open('toto', 'wb') as f: f.write(struct.pack('4B', *[1,2,3,4]))
> with open('toto', 'wb') as f: f.write(struct.pack('<4B', *[1,2,3,4]))
> with open('toto', 'wb') as f: f.write(struct.pack('>4B', *[1,2,3,4]))
>
> I always end up with the following bytes on file:
> !hexdump toto
> 0000000 0201 0403

Why is the hex dump doing 16-bits at a time? Get a byte dump, then it 
might actually be the right output.

Anyway, this seems to work:

data=bytearray([1,2,3,4])

f = open("output", "wb")
f.write(data)
f.close()

Doubtless there are plenty of other ways.

-- 
Bartc

[toc] | [prev] | [standalone]


Back to top | Article view | comp.lang.python


csiph-web