more scrapers
more scrapers


Former-commit-id: 4e165fe3801a38df9f3943299b117f28334cc042

import sys import sys
import os import os
sys.path.insert(0, os.path.join(os.path.dirname(__file__) or '.', '../')) sys.path.insert(0, os.path.join(os.path.dirname(__file__) or '.', '../'))
import scrape import scrape
from bs4 import BeautifulSoup from bs4 import BeautifulSoup
from time import mktime from time import mktime
import feedparser import feedparser
import abc import abc
import unicodedata import unicodedata
import re import re
import dateutil import dateutil
from dateutil.parser import * from dateutil.parser import *
from datetime import * from datetime import *
import codecs import codecs
   
  import difflib
   
from StringIO import StringIO from StringIO import StringIO
   
from pdfminer.pdfparser import PDFDocument, PDFParser from pdfminer.pdfparser import PDFDocument, PDFParser
from pdfminer.pdfinterp import PDFResourceManager, PDFPageInterpreter, process_pdf from pdfminer.pdfinterp import PDFResourceManager, PDFPageInterpreter, process_pdf
from pdfminer.pdfdevice import PDFDevice, TagExtractor from pdfminer.pdfdevice import PDFDevice, TagExtractor
from pdfminer.converter import TextConverter from pdfminer.converter import TextConverter
from pdfminer.cmapdb import CMapDB from pdfminer.cmapdb import CMapDB
from pdfminer.layout import LAParams from pdfminer.layout import LAParams
   
   
class GenericDisclogScraper(object): class GenericDisclogScraper(object):
__metaclass__ = abc.ABCMeta __metaclass__ = abc.ABCMeta
agencyID = None agencyID = None
disclogURL = None disclogURL = None
   
def remove_control_chars(self, input): def remove_control_chars(self, input):
return "".join([i for i in input if ord(i) in range(32, 127)]) return "".join([i for i in input if ord(i) in range(32, 127)])
   
def getAgencyID(self): def getAgencyID(self):
""" disclosr agency id """ """ disclosr agency id """
if self.agencyID is None: if self.agencyID is None:
self.agencyID = os.path.basename(sys.argv[0]).replace(".py", "") self.agencyID = os.path.basename(sys.argv[0]).replace(".py", "")
return self.agencyID return self.agencyID
   
def getURL(self): def getURL(self):
""" disclog URL""" """ disclog URL"""
if self.disclogURL is None: if self.disclogURL is None:
agency = scrape.agencydb.get(self.getAgencyID()) agency = scrape.agencydb.get(self.getAgencyID())
self.disclogURL = agency['FOIDocumentsURL'] self.disclogURL = agency['FOIDocumentsURL']
return self.disclogURL return self.disclogURL
   
@abc.abstractmethod @abc.abstractmethod
def doScrape(self): def doScrape(self):
""" do the scraping """ """ do the scraping """
return return
   
  class GenericHTMLDisclogScraper(GenericDisclogScraper):
   
  def doScrape(self):
  foidocsdb = scrape.couch['disclosr-foidocuments']
  (url, mime_type, rcontent) = scrape.fetchURL(scrape.docsdb,
  self.getURL(), "foidocuments", self.getAgencyID())
  content = rcontent.read()
  dochash = scrape.mkhash(content)
  doc = foidocsdb.get(dochash)
  if doc is None:
  print "saving " + dochash
  description = "This log may have updated but as it was not in a table last time we viewed it, we cannot extract what has changed. Please refer to the agency's website Disclosure Log to see the most recent entries"
  last_attach = scrape.getLastAttachment(scrape.docsdb, self.getURL())
  if last_attach != None:
  html_diff = difflib.HtmlDiff()
  description = description + "\nChanges: "
  description = description + html_diff.make_table(last_attach.read().split('\n'),
  content.split('\n'))
  edate = date.today().strftime("%Y-%m-%d")
  doc = {'_id': dochash, 'agencyID': self.getAgencyID()
  , 'url': self.getURL(), 'docID': dochash,
  "date": edate, "title": "Disclosure Log Updated", "description": description}
  foidocsdb.save(doc)
  else:
  print "already saved"
   
class GenericPDFDisclogScraper(GenericDisclogScraper): class GenericPDFDisclogScraper(GenericDisclogScraper):
   
def doScrape(self): def doScrape(self):
foidocsdb = scrape.couch['disclosr-foidocuments'] foidocsdb = scrape.couch['disclosr-foidocuments']
(url, mime_type, content) = scrape.fetchURL(scrape.docsdb, (url, mime_type, content) = scrape.fetchURL(scrape.docsdb,
self.getURL(), "foidocuments", self.getAgencyID()) self.getURL(), "foidocuments", self.getAgencyID())
laparams = LAParams() laparams = LAParams()
rsrcmgr = PDFResourceManager(caching=True) rsrcmgr = PDFResourceManager(caching=True)
outfp = StringIO() outfp = StringIO()
device = TextConverter(rsrcmgr, outfp, codec='utf-8', device = TextConverter(rsrcmgr, outfp, codec='utf-8',
laparams=laparams) laparams=laparams)
fp = StringIO() fp = StringIO()
fp.write(content.read()) fp.write(content.read())
   
process_pdf(rsrcmgr, device, fp, set(), caching=True, process_pdf(rsrcmgr, device, fp, set(), caching=True,
check_extractable=True) check_extractable=True)
description = outfp.getvalue() description = outfp.getvalue()
fp.close() fp.close()
device.close() device.close()
outfp.close() outfp.close()
dochash = scrape.mkhash(description) dochash = scrape.mkhash(description)
doc = foidocsdb.get(dochash) doc = foidocsdb.get(dochash)
if doc is None: if doc is None:
print "saving " + dochash print "saving " + dochash
edate = date.today().strftime("%Y-%m-%d") edate = date.today().strftime("%Y-%m-%d")
doc = {'_id': dochash, 'agencyID': self.getAgencyID() doc = {'_id': dochash, 'agencyID': self.getAgencyID()
, 'url': self.getURL(), 'docID': dochash, , 'url': self.getURL(), 'docID': dochash,
"date": edate, "title": "Disclosure Log Updated", "description": description} "date": edate, "title": "Disclosure Log Updated", "description": description}
foidocsdb.save(doc) foidocsdb.save(doc)
else: else:
print "already saved" print "already saved"
   
   
class GenericDOCXDisclogScraper(GenericDisclogScraper): class GenericDOCXDisclogScraper(GenericDisclogScraper):
   
def doScrape(self): def doScrape(self):
foidocsdb = scrape.couch['disclosr-foidocuments'] foidocsdb = scrape.couch['disclosr-foidocuments']
(url, mime_type, content) = scrape.fetchURL(scrape.docsdb (url, mime_type, content) = scrape.fetchURL(scrape.docsdb
, self.getURL(), "foidocuments", self.getAgencyID()) , self.getURL(), "foidocuments", self.getAgencyID())
mydoc = zipfile.ZipFile(file) mydoc = zipfile.ZipFile(file)
xmlcontent = mydoc.read('word/document.xml') xmlcontent = mydoc.read('word/document.xml')
document = etree.fromstring(xmlcontent) document = etree.fromstring(xmlcontent)
## Fetch all the text out of the document we just created ## Fetch all the text out of the document we just created
paratextlist = getdocumenttext(document) paratextlist = getdocumenttext(document)
# Make explicit unicode version # Make explicit unicode version
newparatextlist = [] newparatextlist = []
for paratext in paratextlist: for paratext in paratextlist:
newparatextlist.append(paratext.encode("utf-8")) newparatextlist.append(paratext.encode("utf-8"))
## Print our documnts test with two newlines under each paragraph ## Print our documnts test with two newlines under each paragraph
description = '\n\n'.join(newparatextlist).strip(' \t\n\r') description = '\n\n'.join(newparatextlist).strip(' \t\n\r')
dochash = scrape.mkhash(description) dochash = scrape.mkhash(description)
doc = foidocsdb.get(dochash) doc = foidocsdb.get(dochash)
   
if doc is None: if doc is None:
print "saving " + dochash print "saving " + dochash
edate = time().strftime("%Y-%m-%d") edate = time().strftime("%Y-%m-%d")
doc = {'_id': dochash, 'agencyID': self.getAgencyID() doc = {'_id': dochash, 'agencyID': self.getAgencyID()
, 'url': self.getURL(), 'docID': dochash, , 'url': self.getURL(), 'docID': dochash,
"date": edate, "title": "Disclosure Log Updated", "description": description} "date": edate, "title": "Disclosure Log Updated", "description": description}
foidocsdb.save(doc) foidocsdb.save(doc)
else: else:
print "already saved" print "already saved"
   
   
class GenericRSSDisclogScraper(GenericDisclogScraper): class GenericRSSDisclogScraper(GenericDisclogScraper):
   
def doScrape(self): def doScrape(self):
foidocsdb = scrape.couch['disclosr-foidocuments'] foidocsdb = scrape.couch['disclosr-foidocuments']
(url, mime_type, content) = scrape.fetchURL(scrape.docsdb, (url, mime_type, content) = scrape.fetchURL(scrape.docsdb,
self.getURL(), "foidocuments", self.getAgencyID()) self.getURL(), "foidocuments", self.getAgencyID())
feed = feedparser.parse(content) feed = feedparser.parse(content)
for entry in feed.entries: for entry in feed.entries:
#print entry #print entry
print entry.id print entry.id
dochash = scrape.mkhash(entry.id) dochash = scrape.mkhash(entry.id)
doc = foidocsdb.get(dochash) doc = foidocsdb.get(dochash)
#print doc #print doc
if doc is None: if doc is None:
print "saving " + dochash print "saving " + dochash
edate = datetime.fromtimestamp( edate = datetime.fromtimestamp(
mktime(entry.published_parsed)).strftime("%Y-%m-%d") mktime(entry.published_parsed)).strftime("%Y-%m-%d")
doc = {'_id': dochash, 'agencyID': self.getAgencyID(), doc = {'_id': dochash, 'agencyID': self.getAgencyID(),
'url': entry.link, 'docID': entry.id, 'url': entry.link, 'docID': entry.id,
"date": edate, "title": entry.title} "date": edate, "title": entry.title}
self.getDescription(entry, entry, doc) self.getDescription(entry, entry, doc)
foidocsdb.save(doc) foidocsdb.save(doc)
else: else:
print "already saved" print "already saved"
   
def getDescription(self, content, entry, doc): def getDescription(self, content, entry, doc):
""" get description from rss entry""" """ get description from rss entry"""
doc.update({'description': content.summary}) doc.update({'description': content.summary})
return return
   
   
class GenericOAICDisclogScraper(GenericDisclogScraper): class GenericOAICDisclogScraper(GenericDisclogScraper):
__metaclass__ = abc.ABCMeta __metaclass__ = abc.ABCMeta
   
@abc.abstractmethod @abc.abstractmethod
def getColumns(self, columns): def getColumns(self, columns):
""" rearranges columns if required """ """ rearranges columns if required """
return return
   
def getColumnCount(self): def getColumnCount(self):
return 5 return 5
   
def getDescription(self, content, entry, doc): def getDescription(self, content, entry, doc):
""" get description from rss entry""" """ get description from rss entry"""
descriptiontxt = "" descriptiontxt = ""
for string in content.stripped_strings: for string in content.stripped_strings:
descriptiontxt = descriptiontxt + " \n" + string descriptiontxt = descriptiontxt + " \n" + string
doc.update({'description': descriptiontxt}) doc.update({'description': descriptiontxt})
   
def getTitle(self, content, entry, doc): def getTitle(self, content, entry, doc):
doc.update({'title': (''.join(content.stripped_strings))}) doc.update({'title': (''.join(content.stripped_strings))})
   
def getTable(self, soup): def getTable(self, soup):
return soup.table return soup.table
   
def getRows(self, table): def getRows(self, table):
return table.find_all('tr') return table.find_all('tr')
   
def getDate(self, content, entry, doc): def getDate(self, content, entry, doc):
date = ''.join(content.stripped_strings).strip() date = ''.join(content.stripped_strings).strip()
(a, b, c) = date.partition("(") (a, b, c) = date.partition("(")
date = self.remove_control_chars(a.replace("Octber", "October")) date = self.remove_control_chars(a.replace("Octber", "October"))
print date print date
edate = parse(date, dayfirst=True, fuzzy=True).strftime("%Y-%m-%d") edate = parse(date, dayfirst=True, fuzzy=True).strftime("%Y-%m-%d")
print edate print edate
doc.update({'date': edate}) doc.update({'date': edate})
return return
   
def getLinks(self, content, entry, doc): def getLinks(self, content, entry, doc):
links = [] links = []
for atag in entry.find_all("a"): for atag in entry.find_all("a"):
if atag.has_key('href'): if atag.has_key('href'):
links.append(scrape.fullurl(content, atag['href'])) links.append(scrape.fullurl(content, atag['href']))
if links != []: if links != []:
doc.update({'links': links}) doc.update({'links': links})
return return
   
def doScrape(self): def doScrape(self):
foidocsdb = scrape.couch['disclosr-foidocuments'] foidocsdb = scrape.couch['disclosr-foidocuments']
(url, mime_type, content) = scrape.fetchURL(scrape.docsdb, (url, mime_type, content) = scrape.fetchURL(scrape.docsdb,
self.getURL(), "foidocuments", self.getAgencyID()) self.getURL(), "foidocuments", self.getAgencyID())
if content is not None: if content is not None:
if mime_type == "text/html" or mime_type == "application/xhtml+xml" or mime_type == "application/xml": if mime_type == "text/html" or mime_type == "application/xhtml+xml" or mime_type == "application/xml":
# http://www.crummy.com/software/BeautifulSoup/documentation.html # http://www.crummy.com/software/BeautifulSoup/documentation.html
print "parsing" print "parsing"
soup = BeautifulSoup(content) soup = BeautifulSoup(content)
table = self.getTable(soup) table = self.getTable(soup)
for row in self.getRows(table): for row in self.getRows(table):
columns = row.find_all('td') columns = row.find_all('td')
if len(columns) is self.getColumnCount(): if len(columns) is self.getColumnCount():
(id, date, title, (id, date, title,
description, notes) = self.getColumns(columns) description, notes) = self.getColumns(columns)
print self.remove_control_chars( print self.remove_control_chars(
''.join(id.stripped_strings)) ''.join(id.stripped_strings))
if id.string is None: if id.string is None:
dochash = scrape.mkhash( dochash = scrape.mkhash(
self.remove_control_chars( self.remove_control_chars(
url + (''.join(date.stripped_strings)))) url + (''.join(date.stripped_strings))))
else: else:
dochash = scrape.mkhash( dochash = scrape.mkhash(
self.remove_control_chars( self.remove_control_chars(
url + (''.join(id.stripped_strings)))) url + (''.join(id.stripped_strings))))
doc = foidocsdb.get(dochash) doc = foidocsdb.get(dochash)
   
if doc is None: if doc is None:
print "saving " + dochash print "saving " + dochash
doc = {'_id': dochash, doc = {'_id': dochash,
'agencyID': self.getAgencyID(), 'agencyID': self.getAgencyID(),
'url': self.getURL(), 'url': self.getURL(),
'docID': (''.join(id.stripped_strings))} 'docID': (''.join(id.stripped_strings))}
self.getLinks(self.getURL(), row, doc) self.getLinks(self.getURL(), row, doc)
self.getTitle(title, row, doc) self.getTitle(title, row, doc)
self.getDate(date, row, doc) self.getDate(date, row, doc)
self.getDescription(description, row, doc) self.getDescription(description, row, doc)
if notes is not None: if notes is not None:
doc.update({ 'notes': ( doc.update({ 'notes': (
''.join(notes.stripped_strings))}) ''.join(notes.stripped_strings))})
badtitles = ['-','Summary of FOI Request' badtitles = ['-','Summary of FOI Request'
, 'FOI request(in summary form)' , 'FOI request(in summary form)'
, 'Summary of FOI request received by the ASC', , 'Summary of FOI request received by the ASC',
'Summary of FOI request received by agency/minister', 'Summary of FOI request received by agency/minister',
'Description of Documents Requested','FOI request', 'Description of Documents Requested','FOI request',
'Description of FOI Request','Summary of request','Description','Summary', 'Description of FOI Request','Summary of request','Description','Summary',
'Summary of FOIrequest received by agency/minister','Summary of FOI request received','Description of FOI Request',"FOI request",'Results 1 to 67 of 67'] 'Summary of FOIrequest received by agency/minister','Summary of FOI request received','Description of FOI Request',"FOI request",'Results 1 to 67 of 67']
if doc['title'] not in badtitles\ if doc['title'] not in badtitles\
and doc['description'] != '': and doc['description'] != '':
print "saving" print "saving"
foidocsdb.save(doc) foidocsdb.save(doc)
else: else:
print "already saved " + dochash print "already saved " + dochash
   
elif len(row.find_all('th')) is self.getColumnCount(): elif len(row.find_all('th')) is self.getColumnCount():
print "header row" print "header row"