def pulverizeData(self):
"""
Pulverizes the original raw data building up
a map of links to be visited as well as cleaned up data
"""
self.anchors = set()
if not self.rawData:
return
text_content = None
soup = BeautifulSoup(self.rawData)
# Get rid of all script style and link rules
for elem in soup.findAll(["script", "link", "style"]):
elem.extract()
text_content = str(soup.html.body)
# Clean all unworthy characters in the text document
cleaner = HtmlCleaner()
self.content = cleaner.clean_text(text_content)
# Ensure that the current uri is not crawled again by putting it
# into the list of crawled uri
PhrasetankSink.addSeenURI(self.currentUri)
# Extract all links on the document
links = soup.findAll("a", dict(href=True))
for link in links:
uri = PhrasetankRule.isCrawlable(self.baseUri, link["href"])
if uri is not None:
# Check if we had seen this link before
if not PhrasetankSink.hasSeenURI(uri):
# Add this for later crawling
self.anchors.add(uri)
def fetchURIData(self,uri):
"""
Visit the given uri and fetch its page content
"""
contentsource = None
print "Visiting page at "+uri
try:
client = HTTPClient()
response = client.fetch(uri)
contentsource = response.body or None
except:
pass
if contentsource:
# Send the content source to the pulverizer for processing
pulverize = PhrasetankPulverizer()
pulverize.setBaseURI(self.uri)
pulverize.setCurrentURI(uri)
pulverize.setRawData(contentsource)
#process the data
pulverize.pulverizeData()
links = pulverize.getLinks() or []
content = pulverize.getTextContent() or ''
if links and len(links):
#self.uriTank.extend(links)
for l in links:
self.uriTank.add(l)
if content and len(content):
# call the consumer for further processing of the data content
consumer = PhrasetankConsumer()
consumer.setDataContent(content)
consumer.setProducerName(self.name)
consumer.start()
else:
# Perhaps the link has an issue we need to not visit it again
PhrasetankSink.addSeenURI(uri)
logMessage("ERROR",'Failed to read content source for '+uri)
def run(self):
"""
Start running the producer
"""
logMessage('START','PhrasetankProducer '+self.name+' is starting to crawl '+self.uri+' ...')
while True:
uri = None
try:
uri = self.uriTank.pop()
except:
pass
if not uri:
break
# make sure that we have not visited this link before
if PhrasetankSink.hasSeenURI(uri):
#Log the seen status
logMessage('SEEN',uri+' has been visited before now...')
continue
#Ok we are ready to visit the link
self.fetchURIData(uri)
