import sys,os |
import sys,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 genericScrapers |
import genericScrapers |
import scrape |
import scrape |
from bs4 import BeautifulSoup |
from bs4 import BeautifulSoup |
|
|
#http://www.doughellmann.com/PyMOTW/abc/ |
#http://www.doughellmann.com/PyMOTW/abc/ |
class ScraperImplementation(genericScrapers.GenericOAICDisclogScraper): |
class ScraperImplementation(genericScrapers.GenericOAICDisclogScraper): |
|
def getDocHash(self, id,date, url): |
|
''' url changes on ever request so ignore for hash ''' |
|
return scrape.mkhash( |
|
self.remove_control_chars( |
|
''.join(id.stripped_strings))) |
def getColumnCount(self): |
def getColumnCount(self): |
return 4 |
return 4 |
def getColumns(self,columns): |
def getColumns(self,columns): |
(date, id, title, description) = columns |
(date, id, title, description) = columns |
return (id, date, title, description, None) |
return (id, date, title, description, None) |
|
|
if __name__ == '__main__': |
if __name__ == '__main__': |
print 'Subclass:', issubclass(ScraperImplementation, genericScrapers.GenericOAICDisclogScraper) |
print 'Subclass:', issubclass(ScraperImplementation, genericScrapers.GenericOAICDisclogScraper) |
print 'Instance:', isinstance(ScraperImplementation(), genericScrapers.GenericOAICDisclogScraper) |
print 'Instance:', isinstance(ScraperImplementation(), genericScrapers.GenericOAICDisclogScraper) |
ScraperImplementation().doScrape() |
ScraperImplementation().doScrape() |
|
|