Python3 write gzip file - memoryview: a bytes-like object is required, not 'str'

Question

I want to write a file. Based on the name of the file this may or may not be compressed with the gzip module. Here is my code:

import gzip
filename = 'output.gz'
opener = gzip.open if filename.endswith('.gz') else open
with opener(filename, 'wb') as fd:
    print('blah blah blah'.encode(), file=fd)

I'm opening the writable file in binary mode and encoding my string to be written. However I get the following error:

File "/usr/lib/python3.5/gzip.py", line 258, in write
  data = memoryview(data)
TypeError: memoryview: a bytes-like object is required, not 'str'

Why is my object not a bytes? I get the same error if I open the file with 'w' and skip the encoding step. I also get the same error if I remove the '.gz' from the filename.

I'm using Python3.5 on Ubuntu 16.04

Answer 1

print is a relatively complex function. It writes str to a file but not the str that you pass, it writes the str that is the result of rendering the parameters.

If you have bytes already, you can use fd.write(bytes) directly and take care of adding a newline if you need it.

If you don't have bytes, make sure fd is opened to receive text.

Answer 2

you can convert it to bytes like this.

import gzip 
with gzip.open(filename, 'wb') as fd:
   fd.write('blah blah blah'.encode('utf-8'))

Answer 3

For me, changing the gzip flag to 'wt' did the job. I could write the original string, without "byting" it. (tested on python 3.5, 3.7 on ubuntu 16).

From python 3 gzip doc - quoting: "... The mode argument can be any of 'r', 'rb', 'a', 'ab', 'w', 'wb', 'x' or 'xb' for binary mode, or 'rt', 'at', 'wt', or 'xt' for text mode..."

import gzip

filename = 'output.gz'
opener = gzip.open if filename.endswith('.gz') else open
with opener(filename, 'wt') as fd:
    print('blah blah blah', file=fd)

!zcat output.gz
> blah blah blah

Answer 4

You can serialize it using pickle.

First serializing the object to be written using pickle, then using gzip.

To save the object:

import gzip, pickle
filename = 'non-serialize_object.zip'
# serialize the object    
serialized_obj = pickle.dumps(object)
# writing zip file
with gzip.open(filename, 'wb') as f:
   f.write(serialized_obj)

To load the object:

import gzip, pickle
filename = 'non-serialize_object.zip'

with gzip.open(filename, 'rb') as f:    
   serialized_obj = f.read()
# de-serialize the object
object = pickle.loads(serialized_obj)