Python: Writing int to a binary file

Question

I have a program which calculates the offset(difference) and then stores them in an 16 bit unsigned int using numPy and I want to store this int into a binary file as it is in binary form. i.e. if the value of offset is 05, I want the file to show "01010000 00000000", but not as a string. The code I have written is:

target = open(file_cp, 'wb')
target.write('Entries')
target.write('\n')
Start = f.tell()
while(!EOF):
    f.read(lines)
    Current = f.tell()
    offset = np.uint16(Current-Start)
    target.write(offset)

there is some processing after f.read(lines) but thats sort of the idea. The code works fine as long as the offset is less than 127. As soon as the offset goes above 127, a 0xC2 appears in the file along with the binary data.

data in the file appears as follows (hex view, little indian): 00 00 05 00 0e 00 17 00 20 00 3c 00 4e 00 7b 00 c2 8d 00 c2 92 00 c2 9f 00

Could someone suggest a solution to the problem?

alxg · Accepted Answer · 2016-05-10 06:48:20Z

2

Try this.

import numpy as np
a=int(4)
binwrite=open('testint.in','wb')
np.array([a]).tofile(binwrite)
binwrite.close()

b=np.fromfile('testint.in',dtype=np.int16)
print b[0], type(b[0])

output: 4 type 'numpy.int16'

I Hope this is wha you are looking for. Works for n>127 But read and writes numpy arrays... binwrite=open('testint.in','ab') will let you append more ints to the file.

answered May 10, 2016 at 6:48

alxg

3681 gold badge4 silver badges15 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

phihag Over a year ago

This answer could be improved by using the with statement (so that the file always gets closed, the code becomes more idiomatic/readable/shorter) and print as a function - as it stands, your program contains a syntax error on current Python versions.

omkar joshi Over a year ago

Thank you for the suggestion. It still does the same thing. instead of 4, i tried 148 and the file read (hex view) : c2 94 00

John Zwinck · Accepted Answer · 2016-05-10 04:31:37Z

0

You should use the built-in struct module. Instead of this:

np.uint16(Current-Start)

Try this:

struct.pack('H', Current-Start)

answered May 10, 2016 at 4:31

John Zwinck

252k44 gold badges346 silver badges459 bronze badges

2 Comments

mgilson Over a year ago

I think that struct.pack('H', np_uint16_instance) is the same as np_uint16_instance.tostring()

omkar joshi Over a year ago

This will write a string to the file which is not what I'm looking for. I have update the question, sorry if it wasn't clear before.

Collectives™ on Stack Overflow

Python: Writing int to a binary file

2 Answers 2

2 Comments

2 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

2 Comments

2 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related