Python Batch Client - Revision history

Padro at 11:01, 21 January 2016

2016-01-21T11:01:51Z

Padro at 12:08, 20 January 2016

2016-01-20T12:08:54Z

Padro: Created page with "#! /usr/bin/python ######################################################################## # # Example client to submit a batch request to TextServer # SERVICENAME service..."

2016-01-20T12:08:15Z

Created page with "#! /usr/bin/python ######################################################################## # # Example client to submit a batch request to TextServer # SERVICENAME service..."

New page

#! /usr/bin/python

########################################################################
#
# Example client to submit a batch request to TextServer
# SERVICENAME service, and wait for the job to be finished.
#
# Input must be a ZIP file containing one or more text files to analyze
# Output will be a ZIP file containing results for each input file in
# the requested format (XML, json, conll)
#
########################################################################

# import required libraries
import urllib2
# You may need to install "poster" python module for these two:
from poster.encode import multipart_encode
from poster.streaminghttp import register_openers
import xml.etree.ElementTree
import time

# Register the streaming http handlers with urllib2
register_openers()

# set query elements
fname = raw_input('Input ZIP file: ')
outfname = raw_input('Output ZIP file: ')
lang = raw_input('Language: ')
out = raw_input('Output format (xml,json,conll,naf): ')
user = raw_input('TextServer Username: ')
pwd = raw_input('TextServer Password: ')

# Encode query in a form-data.
# 'headers' contains the necessary Content-Type and Content-Length.
# 'datagen' is a generator object that yields the encoded parameters.
datagen, headers = multipart_encode({'username' : user,
'password' : pwd,
'file' : open(fname,"rb"),
'language' : lang,
'output' : out,
'interactive':'0'
} )
# service URL
TextServer_URL = "http://frodo.lsi.upc.edu:8080/TextWS/textservlet/ws"
service = "SERVICENAME"

# Create the Request object
request = urllib2.Request(TextServer_URL + "/processQuery/" + service,
datagen,
headers)

try:
# Actually do the request, and get the response
resp = urllib2.urlopen(request).read()

except urllib2.HTTPError, e:
# handle connection errors
print e, "-", e.read()
exit()

# Server response should include a job tokenID, retrieve it
dom = xml.etree.ElementTree.XML(resp)
tkid = dom.text
print "Job sumbitted. Token id=",tkid

# prepare request to poll for completion and retrieve results
datagen, headers = multipart_encode({'username' : user,
'tokenID' : tkid
})
request = urllib2.Request(TextServer_URL + "/resultRetrieve",
datagen,
headers)

# Periodically poll server until batch job is ended.
nseconds=20
while (1) :

print "Sleeping for",nseconds,"seconds"
time.sleep(nseconds)
try:
# send request to check for results
print "Polling server for job completion..."
resp = urllib2.urlopen(request).read()

except urllib2.HTTPError, e:
if (e.code == 503 and e.read()[0:8] == "[TS-125]" ) :
# http status 503, textserver code TS-125 means the job is not finished yet. keep waiting
print "Job not finished yet"
continue
else :
# some actual error happened. abort
print e, "-", e.read()
exit()

# if status=200, we a got ZIP response, the job is done, exit loop.
break;

print "Job finished. Saving results to",outfname
# print results to output file
outf=open(outfname, 'w+')
print >>outf, resp
outf.close()

@@ Line 26: / Line 26: @@
 # set query elements
 fname = raw_input('Input ZIP file: ')
 lang = raw_input('Language: ')
 out = raw_input('Output format (xml,json,conll,naf): ')
@@ Line 59: / Line 58: @@
    print e, "-", e.read()
    exit()
 # Server response should include a job tokenID, retrieve it
@@ Line 65: / Line 63: @@
 tkid = dom.text
 print "Job sumbitted. Token id=",tkid
 </syntaxhighlight>

@@ Line 1: / Line 1: @@
 #! /usr/bin/python
@@ Line 102: / Line 103: @@
 print >>outf, resp
 outf.close()