Python Batch Client

1. ! /usr/bin/python

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

1. import required libraries

import urllib2

1. 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

1. Register the streaming http handlers with urllib2

register_openers()

1. 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: ')

1. Encode query in a form-data.
2. 'headers' contains the necessary Content-Type and Content-Length.
3. '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'
                                    } )

1. service URL

TextServer_URL = "http://frodo.lsi.upc.edu:8080/TextWS/textservlet/ws" service = "SERVICENAME"

1. 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()

1. Server response should include a job tokenID, retrieve it

dom = xml.etree.ElementTree.XML(resp) tkid = dom.text print "Job sumbitted. Token id=",tkid

1. 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)

1. 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

1. print results to output file

outf=open(outfname, 'w+') print >>outf, resp outf.close()