From: Maxious <maxious@lambdacomplex.org>
Date: Sun, 25 Nov 2012 22:43:31 +0000
Subject: add ausaid again
X-Git-Url: https://maxious.lambdacomplex.org/git/?p=disclosr.git&a=commitdiff&h=7816ac6af5ed23e3f105e802868b58bad4e895e4
---
add ausaid again


Former-commit-id: ccbe1ad50acae0133a4d29c3431230936af84f95
---


--- /dev/null
+++ b/documents/scrapers/ac6cf1ebaa79694e2cc3705bc83c89b3.py
@@ -1,1 +1,31 @@
+import sys,os
+sys.path.insert(0, os.path.join(os.path.dirname(__file__) or '.', '../'))
+import genericScrapers
+import scrape
+from bs4 import BeautifulSoup
+from datetime import datetime
+from datetime import date
 
+#http://www.doughellmann.com/PyMOTW/abc/
+class ScraperImplementation(genericScrapers.GenericOAICDisclogScraper):
+	def getColumnCount(self):
+		return 3
+	def getTable(self,soup):
+		return soup.find(class_ = "internal")
+	def getColumns(self,columns):
+		(id, date, description) = columns
+		return (id, date, description, description, None)
+        def getTitle(self, content, entry, doc):
+                doc.update({'title': content.stripped_strings.next()})
+                return
+        def getDate(self, content, entry, doc):
+                edate = datetime.strptime(content.string.strip(), "%d/%m/%Y").strftime("%Y-%m-%d")
+                print edate
+                doc.update({'date': edate})
+                return
+
+if __name__ == '__main__':
+    print 'Subclass:', issubclass(ScraperImplementation, genericScrapers.GenericOAICDisclogScraper)
+    print 'Instance:', isinstance(ScraperImplementation(), genericScrapers.GenericOAICDisclogScraper)
+    ScraperImplementation().doScrape()
+