habernal / feng-hirst-rst-parser-acl-2014 Goto Github PK

My code just hangs at this line when it attempts to read from the stdout. It seems that it is a possible deadlock due with popen. As per https://docs.python.org/2/library/subprocess.html

Note Do not use stdout=PIPE or stderr=PIPE with this function as that can deadlock based on the child process output volume. Use Popen with the communicate() method when you need pipes.

If I use communicate, it works for the first sentence but then terminates with an error
subprocess I/O operation on closed file
This is because communicate also terminates the process.

To fix it, I tried to modify the code as

def classify(self, vectors):
self.classifier = subprocess.Popen( self.classifier_cmd,shell = True, stdin = subprocess.PIPE,stdout=subprocess.PIPE, stderr = subprocess.PIPE)
if self.classifier.poll():
raise OSError('crf_classifier subprocess died')
stdoutdata, stderrdata=self.classifier.communicate('\n'.join(vectors) + "\n\n") #amita add
lines = stdoutdata.splitlines()

This works but I am not sure if this could break the code elsewhere.

Thank you for the great tool!

Can one use a fresher version of NLTK? Or it would just mean that the original behaviour is not guaranteed?

Best regards.