lxml not working with django, scraperwiki - django

Question

I'm working on a django app that goes through Illinois' General Assembly website to scrape some pdfs. While deployed on my desktop it works fine until urllib2 times out. When I try to deploy on my Bluehost server, the lxml part of the code throws up an error. Any help would be appreciated.
import scraperwiki
from bs4 import BeautifulSoup
import urllib2
import lxml.etree
import re
from django.core.management.base import BaseCommand
from legi.models import Votes
class Command(BaseCommand):
def handle(self, *args, **options):
chmbrs =['http://www.ilga.gov/house/', 'http://www.ilga.gov/senate/']
for chmbr in chmbrs:
site = chmbr
url = urllib2.urlopen(site)
content = url.read()
soup = BeautifulSoup(content)
links = []
linkStats = []
x=0
y=0
table = soup.find('table', cellpadding=3)
for a in soup.findAll('a',href=True):
if re.findall('Bills', a['href']):
l = (site + a['href']+'&Primary=True')
links.append(str(l))
x+=1
print x
for link in links:
url = urllib2.urlopen(link)
content = url.read()
soup = BeautifulSoup(content)
table = soup.find('table', cellpadding=3)
for a in table.findAll('a',href=True):
if re.findall('BillStatus', a['href']):
linkStats.append(str('http://ilga.gov'+a['href']))
for linkStat in linkStats:
url = urllib2.urlopen(linkStat)
content = url.read()
soup = BeautifulSoup(content)
for a in soup.findAll('a',href=True):
if re.findall('votehistory', a['href']):
vl = 'http://ilga.gov/legislation/'+a['href']
url = urllib2.urlopen(vl)
content = url.read()
soup = BeautifulSoup(content)
for b in soup.findAll('a',href=True):
if re.findall('votehistory', b['href']):
llink = 'http://ilga.gov'+b['href']
try:
u = urllib2.urlopen(llink)
x = scraperwiki.pdftoxml(u.read())
root = lxml.etree.fromstring(x)
pages = list(root)
chamber = str()
for page in pages:
print "working_1"
for el in page:
print "working_2"
if el.tag == 'text':
if int(el.attrib['top']) == 168:
chamber = el.text
if re.findall("Senate Vote", chamber):
if int(el.attrib['top']) >= 203 and int(el.attrib['top']) < 231:
title = el.text
if (re.findall('House', title)):
title = (re.findall('[0-9]+', title))
title = "HB"+title[0]
elif (re.findall('Senate', title)):
title = (re.findall('[0-9]+', title))
title = "SB"+title[0]
if int(el.attrib['top']) >350 and int(el.attrib['top']) <650:
r = el.text
names = re.findall(r'[A-z-\u00F1]{3,}',r)
vs = re.findall(r'[A-Z]{1,2}\s',r)
for name in names:
legi = name
for vote in vs:
v = vote
if Votes.objects.filter(legislation=title).exists() == False:
c = Votes(legislation=title, legislator=legi, vote=v)
c.save()
print 'saved'
else:
print 'not saved'
elif int(el.attrib['top']) == 189:
chamber = el.text
if re.findall("HOUSE ROLL CALL", chamber):
if int(el.attrib['top']) > 200 and int(el.attrib['top']) <215:
title = el.text
if (re.findall('HOUSE', title)):
title = (re.findall('[0-9]+', title))
title = "HB"+title[0]
elif (re.findall('SENATE', title)):
title = (re.findall('[0-9]+', title))
title = "SB"+title[0]
if int(el.attrib['top']) >385 and int(el.attrib['top']) <1000:
r = el.text
names = re.findall(r'[A-z-\u00F1]{3,}',r)
votes = re.findall(r'[A-Z]{1,2}\s',r)
for name in names:
legi = name
for vote in votes:
v = vote
if Votes.objects.filter(legislation=title).exists() == False:
c = Votes(legislation=title, legislator=legi, vote=v)
c.save()
print 'saved'
else:
print 'not saved'
except:
pass
EDIT 1
Here's the error trace
Traceback (most recent call last):
File "manage.py", line 10, in <module>
execute_from_command_line(sys.argv)
File "/home7/maythirt/python27/lib/python2.7/site-packages/django/core/management/__init__.py", line 399, in execute_from_command_line
utility.execute()
File "/home7/maythirt/python27/lib/python2.7/site-packages/django/core/management/__init__.py", line 392, in execute
self.fetch_command(subcommand).run_from_argv(self.argv)
File "/home7/maythirt/python27/lib/python2.7/site-packages/django/core/management/base.py", line 242, in run_from_argv
self.execute(*args, **options.__dict__)
File "/home7/maythirt/python27/lib/python2.7/site-packages/django/core/management/base.py", line 285, in execute
output = self.handle(*args, **options)
File "/home7/maythirt/GAB/legi/management/commands/vote.py", line 51, in handle
root = lxml.etree.fromstring(x)
File "lxml.etree.pyx", line 3032, in lxml.etree.fromstring (src/lxml/lxml.etree.c:68121)
File "parser.pxi", line 1786, in lxml.etree._parseMemoryDocument (src/lxml/lxml.etree.c:102470)
File "parser.pxi", line 1674, in lxml.etree._parseDoc (src/lxml/lxml.etree.c:101299)
File "parser.pxi", line 1074, in lxml.etree._BaseParser._parseDoc (src/lxml/lxml.etree.c:96481)
File "parser.pxi", line 582, in lxml.etree._ParserContext._handleParseResultDoc (src/lxml/lxml.etree.c:91290)
File "parser.pxi", line 683, in lxml.etree._handleParseResult (src/lxml/lxml.etree.c:92476)
File "parser.pxi", line 633, in lxml.etree._raiseParseError (src/lxml/lxml.etree.c:91939)
lxml.etree.XMLSyntaxError: None

Answer 1

As Jonathan mentioned, it may be the output of scraperwiki.pdftoxml() that's causing a problem. You could display or log the value of x to confirm it.
Specifically, pdftoxml() runs an external program pdftohtml and uses temporary files to store the PDF and XML.
What I'd also check for is:
Is pdftohtml correctly set up on your server?
If so, does the conversion to XML work if you directly run it in a shell on the server with the PDF that the code's failing on? The command it's executing is pdftohtml -xml -nodrm -zoom 1.5 -enc UTF-8 -noframes "input.pdf" "output.xml"
If there's an issue when you directly run the command, then that's there your issue lies. With the way pdftohtml runs in the scraperwiki code, there's no easy way you'd be able to tell if the command fails.

Answer 2

They way I would go about this is add a try: except: clause and when you get the error you simply save the xml file as well as the link down to your harddrive. That way you can inspect the xml file separately.
It might be that scraperwiki.pdftoxml makes an illegal xml file for some reason. I've had that happen to me when using another pdftoxml tool.
And please refactor your code into more functions it will become a lot easier to read and maintain :).
Another way would of course to download all of the pdfs first, and then parse them all. That way you can avoid hitting the website several times whenever you fail for some reason.