--- a/documents/genericScrapers.py +++ b/documents/genericScrapers.py @@ -2,12 +2,14 @@ sys.path.insert(0, os.path.join(os.path.dirname(__file__) or '.', '../')) import scrape from bs4 import BeautifulSoup -import parsedatetime as pdt from time import mktime -from datetime import datetime import feedparser import abc import unicodedata, re +import dateutil +from dateutil.parser import * +from datetime import * +import codecs class GenericDisclogScraper(object): __metaclass__ = abc.ABCMeta @@ -54,7 +56,7 @@ doc = foidocsdb.get(hash) #print doc if doc == None: - print "saving" + print "saving "+ hash edate = datetime.fromtimestamp(mktime( entry.published_parsed)).strftime("%Y-%m-%d") doc = {'_id': hash, 'agencyID': self.getAgencyID(), 'url': entry.link, 'docID': entry.id, "date": edate,"title": entry.title} @@ -82,11 +84,32 @@ descriptiontxt = descriptiontxt + " \n" + string doc.update({'description': descriptiontxt}) return + def getTitle(self, content, entry, doc): + doc.update({'title': (''.join(content.stripped_strings))}) + return def getTable(self, soup): return soup.table + def getRows(self, table): + return table.find_all('tr') + def getDate(self, content, entry, doc): + date = ''.join(content.stripped_strings).strip() + (a,b,c) = date.partition("(") + date = self.remove_control_chars(a.replace("Octber","October")) + print date + edate = parse(date, dayfirst=True, fuzzy=True).strftime("%Y-%m-%d") + print edate + doc.update({'date': edate}) + return + def getLinks(self, content, entry, doc): + links = [] + for atag in entry.find_all("a"): + if atag.has_key('href'): + links.append(scrape.fullurl(content,atag['href'])) + if links != []: + doc.update({'links': links}) + return def doScrape(self): - cal = pdt.Calendar() foidocsdb = scrape.couch['disclosr-foidocuments'] (url,mime_type,content) = scrape.fetchURL(scrape.docsdb, self.getURL(), "foidocuments", self.getAgencyID()) if content != None: @@ -94,37 +117,26 @@ # http://www.crummy.com/software/BeautifulSoup/documentation.html soup = BeautifulSoup(content) table = self.getTable(soup) - for row in table.find_all('tr'): + for row in self.getRows(table): columns = row.find_all('td') if len(columns) == self.getColumnCount(): - (id, date, description, title, notes) = self.getColumns(columns) - print id.string + (id, date, title, description, notes) = self.getColumns(columns) + print self.remove_control_chars(''.join(id.stripped_strings)) if id.string == None: - hash = scrape.mkhash(self.remove_control_chars(url+date.string)) + hash = scrape.mkhash(self.remove_control_chars(url+(''.join(date.stripped_strings)))) else: - hash = scrape.mkhash(self.remove_control_chars(url+id.string)) - links = [] - for atag in row.find_all("a"): - if atag.has_key('href'): - links.append(scrape.fullurl(url,atag['href'])) + hash = scrape.mkhash(self.remove_control_chars(url+(''.join(id.stripped_strings)))) doc = foidocsdb.get(hash) if doc == None: - print "saving" - dtresult = cal.parseDateText(date.string) - if len(dtresult) == 2: - (dtdate,dtr) = dtresult - print dtdate - edate = ""+str(dtdate[0])+'-'+str(dtdate[1])+'-'+str(dtdate[2]) - else: - edate = datetime.strptime(date.string.strip(), "%d %B %Y").strftime("%Y-%m-%d") - doc = {'_id': hash, 'agencyID': self.getAgencyID(), 'url': self.getURL(), 'docID': id.string, - "date": edate,"title": title.string} - if links != []: - doc.update({'links': links}) + print "saving " +hash + doc = {'_id': hash, 'agencyID': self.getAgencyID(), 'url': self.getURL(), 'docID': (''.join(id.stripped_strings))} + self.getLinks(self.getURL(),row,doc) + self.getTitle(title,row, doc) + self.getDate(date,row, doc) self.getDescription(description,row, doc) - if notes != None: - doc.update({ 'notes': notes.string}) + if notes != None: + doc.update({ 'notes': (''.join(notes.stripped_strings))}) foidocsdb.save(doc) else: print "already saved "+hash