[SOLVED] standard output changes in python3 versus python2

sharky · 10-16-2018, 03:20 PM

This is a pyqt5 issue with readAllStandardOutput.

I am using that function to get the return from a bash script.

Code:

features = str(self.runner.readAllStandardOutput())
print(features)

With pyqt5 and python2.7 the result of the print statement is;

Quote:

sometext 234 0 0
somemoretext 234 0 0
textagain 234 45 0.192308
moretext 234 0 0

With pyqt5 and python3.7 the result of the print statement is;

Quote:

b'sometext 234 0 0\nsomemoretext 234 0 0\ntextagain 234 45 0.192308\nmoretext 234 0 0\n'

Why is b'' wrapped around the output? Subsequent statements slit 'features' using either features.split('\n') or features.splitlines() (I've tried both) but python3.7 will not split the string. The output of features.splitlines or features.split('\n') is;

Quote:

["b'sometext 234 0 0\nsomemoretext 234 0 0\ntextagain 234 45 0.192308\nmoretext 234 0 0\n'"]

It doesn't split anything at all. Just puts features into a list as a single element. Not sure how to fix this. Is there a new method with python3.x to grab standard out?

Hope I made the issue clear.

Cheers,

individual · 10-16-2018, 04:13 PM

b'String' means it is a series of bytes, rather than a string. If you want to get it into string form, add .decode() to the end of your string.

Code:

self.runner.readAllStandardOutput().decode()

Here is how you can tell the type:

Code:

s = b'Hello, World!'
type(s)
# <class 'bytes'>
type(s.decode())
# <class 'str'>

sharky · 10-16-2018, 04:19 PM

Quote:

Originally Posted by individual

b'String' means it is a series of bytes, rather than a string. If you want to get it into string form, add .decode() to the end of your string.

Code:

self.runner.readAllStandardOutput().decode()

Here is how you can tell the type:

Code:

s = b'Hello, World!'
type(s)
# <class 'bytes'>
type(s.decode())
# <class 'str'>

This is what I get.

Quote:

Traceback (most recent call last):
File "./myScript.py", line 581, in get_featuresStd
features = str(self.runner.readAllStandardOutput().decode())
AttributeError: 'QByteArray' object has no attribute 'decode'
Abort (core dumped)

individual · 10-16-2018, 04:27 PM

Quote:

Originally Posted by sharky

This is what I get.

The fact that the return value is QByteArray changes things. I'm not familiar with pyqt5, but their documentation doesn't provide available methods for that class. It refers the reader to the QT5 C++ documentation, which doesn't do any good. Try

Code:

print(dir(self.runner.readAllStandardOutput()))

That will show you available methods for that type.

sharky · 10-16-2018, 04:28 PM

I found something that works for me.

Code:

features = str(self.runner.readAllStandardOutput(), encoding = 'utf-8')

individual · 10-16-2018, 04:30 PM

Quote:

Originally Posted by sharky

I found something that works for me.

Code:

features = str(self.runner.readAllStandardOutput(), encoding = 'utf-8')

That's interesting that setting the encoding would turn it into a string, but just calling str wouldn't.