class CrawlerDB(Thread):
def __init__(self,
delegate,
initialLink=None,
max_links=0,
no_workers=10,
id=str(uuid.uuid4())):
Thread.__init__(self)
self.noOfWorkers = no_workers
self.workers = []
self.running = True
self.paused = False
self.condition = RLock()
self.delegate = Delegate()
self.listeners = [
] # A list of listeners that want to listen to messages (ex: progress) from Crawler
self.id = id
self.initialLink = initialLink
if initialLink is not None:
self.add_initial_url(initialLink)
self.max_links = max_links
try:
self.domain_regex = re.compile(get_domain(initialLink))
except Exception as ex:
logging.error("Exception {}".format(ex))
# def __del__(self):
# self.delegate.get_session().close()
def add_initial_url(self, address):
logger.info("Add initial URL")
with self.condition:
url = Url(url=address,
absolute_url=address,
type=Url.TYPE_INTERNAL,
crawl_id=self.id,
job_status=Url.JOB_STATUS_NOT_VISITED)
self.delegate.url_create(url)
def no_unvisited_urls(self):
with self.condition:
return self.delegate.url_count_unvisited(self.id)
def no_pending_urls(self):
with self.condition:
return self.delegate.url_count_pending(self.id)
def all_unvisited_urls(self):
with self.condition:
return self.delegate.url_get_all_unvisited(self.id)
def no_visited_urls(self):
with self.condition:
return self.delegate.url_count_visited(self.id)
def no_visited_resources(self):
with self.condition:
return self.delegate.resource_count_visited(self.id)
def no_external_urls(self):
with self.condition:
return self.delegate.url_count_external(self.id)
def next_unvisited_link_id(self):
link_id = -1
with self.condition:
url = self.delegate.url_get_first_unvisited(self.id)
if url is not None:
url.job_status = Url.JOB_STATUS_IN_PROGRESS # Set Url as in progress
self.delegate.url_update(url)
# self.increaseNoOfJobs()
link_id = url.id
return link_id
def mark_url_as_visited(self, url_id):
with self.condition:
url = self.delegate.url_get_by_id(url_id)
url.job_status = Url.JOB_STATUS_VISITED
self.delegate.url_update(url)
def _type_links(self, links):
for link in links:
if is_internal(get_domain(self.initialLink),
link['absolute']): # internal link
link['type'] = 'internal'
else: # external link
link['type'] = 'external'
def _get_links(self, link_id):
with self.condition:
link = self.delegate.url_get_by_id(link_id)
(page, links) = get_links(link.absolute_url)
self._type_links(links)
return page, links
def link2url(self, link):
url = Url(crawl_id=self.id)
# url=link['href'], absolute_url=link['absolute'], type=link['type'],
if 'href' in link:
url.url = link['href']
if 'absolute' in link:
url.absolute_url = link['absolute']
if 'type' in link:
url.type = link['type']
if 'content' in link:
url.raw_content = str(link['content'])
url.text = str(
link['content']
) # TODO: Parse the raw_content and used only the text without HTML tags or other stuff
return url
def page2resource(self, page):
resource = Resource(crawl_id=self.id)
if 'url' in page:
resource.absolute_url = page['url']
if 'content' in page:
resource.content = page['content']
if 'elapsed' in page:
resource.elapsed = page['elapsed']
return resource
def add_links(self, links, src_resource_id=None, status_code=200):
"""Add a bunch of URLs using the resource id as source (page where found it)"""
with self.condition:
for link in links:
url = self.link2url(link)
if src_resource_id is not None:
url.src_resource_id = src_resource_id
# Check if destination resource exists, and if does mark it as visited
try:
src_resource = self.delegate.resource_get_by_id(
src_resource_id)
dest_resource = self.delegate.resource_get_by_absolute_url_and_crawl_id(
url.absolute_url, src_resource.crawl_id)
if dest_resource is not None:
url.job_status = Url.JOB_STATUS_VISITED
url.dst_resource_id = dest_resource.id
url.status_code = status_code
except Exception as e:
logger.warning("Exception {}".format(e))
self.delegate.url_create(url)
def add_resource(self, page):
with self.condition:
if not self.delegate.resource_is_present(crawlId=self.id):
resource = self.page2resource(page)
self.delegate.resource_create(resource)
def connect_url_to_destination(self, url_id, resource_id):
with self.condition:
url = self.delegate.url_get_by_id(url_id)
url.dst_resource_id = resource_id
self.delegate.url_update(url)
def resource_get_by_absolute_url_and_crawl_id(self, address, crawler_id):
with self.condition:
resource = self.delegate.resource_get_by_absolute_url_and_crawl_id(
address, crawler_id)
return resource
def resource_create(self, page):
with self.condition:
try:
resource = self.page2resource(page)
self.delegate.resource_create(resource)
except Exception as e:
logger.warn("{} Exception {}}.".format(
currentThread().getName(), e))
return resource
def run(self):
# Initialize workers
for i in range(self.noOfWorkers):
self.workers.append(
Thread(target=self.workerJob,
kwargs={"crawlId": self.id},
name="Thread-{}".format(i)))
# Start workers
self._start_all_workers()
while self.running:
logger.debug("[%s] Crawler thread cycle started." %
(currentThread().getName()))
if self.paused:
logger.debug("[%s] Crawler paused." %
(currentThread().getName()))
continue
logger.debug("[%s] Crawler check if jobs are done." %
(currentThread().getName()))
if self._are_jobs_done():
logger.debug("Crawler is shutting down")
self.setRunning(False)
break
else:
logger.debug("[%s] Crawler's jos are NOT done." %
(currentThread().getName()))
logger.debug("[%s] Crawler sleep." % (currentThread().getName()))
time.sleep(1)
# Join them
self._join_all_workers()
# self.delegate.get_session().close()
msg = {
"status": "done",
"visited": self.no_visited_urls(),
"to_visit": self.no_unvisited_urls(),
"max_links": self.max_links,
"crawlId": self.id
}
self.notify(msg)
def workerJob(self, crawlId):
while self.running:
logger.debug("[%s] Worker thread cycle started." %
(currentThread().getName()))
if self.paused:
continue
# If max pages specified see if we already reached it
if self.max_links > 0:
no_pages_visited = self.no_visited_resources()
if no_pages_visited >= self.max_links:
continue
# Grab next job
link_id = self.next_unvisited_link_id()
logger.debug("[%s] Next link [%d]." %
(currentThread().getName(), link_id))
if 'link_id' in locals() and link_id != -1:
logger.debug("[%s] Current link : %d" %
(currentThread().getName(), link_id))
page, links = self._get_links(link_id)
logger.debug("[%s] Discovered [%d] links." %
(currentThread().getName(), len(links)))
try:
with self.condition:
# Update links status code
url = self.delegate.url_get_by_id(link_id)
url.status_code = page['status-code']
self.delegate.url_update(url)
if page['status-code'] == 200:
# 1.Add Resource 2.Link URLs to (new | existing) Resources
resource = self.resource_get_by_absolute_url_and_crawl_id(
page['url'], self.id)
if resource is None:
#Add it only if max links not reached
maximum_reached = False
if self.max_links > 0: # We have a max_link specified
no_pages_visited = self.no_visited_resources(
)
if no_pages_visited >= self.max_links:
maximum_reached = True
if not maximum_reached:
resource = self.resource_create(page)
self.connect_url_to_destination(
link_id, resource.id)
logger.debug(
"[%s] Adding links to DB linked to resource [%d]"
% (currentThread().getName(),
resource.id))
self.add_links(links, resource.id,
page['status-code'])
else:
# Resource already added only make the end connection
self.connect_url_to_destination(
link_id, resource.id)
else:
pass
self.mark_url_as_visited(link_id)
msg = {
"status": "in_progress",
"visited": self.no_visited_urls(),
"to_visit": self.no_unvisited_urls(),
"max_links": self.max_links,
"crawlId": crawlId,
"currentWorker": currentThread().getName()
}
self.notify(msg)
except Exception as e:
print("Error {}".format(e))
logger.debug("[%s] cycle ended." % (currentThread().getName()))
else:
logger.debug("[%s] is shutting down." %
(currentThread().getName()))
# self.delegate.get_session().close()
def stop(self):
self.setRunning(False)
def pause(self):
self.paused = True
def resume(self):
if self.paused:
self.paused = False
def _start_all_workers(self):
for w in self.workers:
w.start()
def _are_jobs_done(self):
# Test if noOfJobs == 0 and to_visit == 0
# no_of_jobs = self.getNoOfJobs()
# FIXME: If a thread grabs the initial link, while here, no_unvisited_urls() will
# return zero (on next line) , also the no_of_jobs are zero so the Crawler
# will initiate shutdown
no_pending_urls = self.no_pending_urls()
logger.debug("Crawler: _are_jobs_done(...) : no_pendind_urls = %d " %
(no_pending_urls, ))
if no_pending_urls == 0:
return True
# Test if we have reached the max no of pages
if self.max_links > 0:
no_pages_visited = self.no_visited_resources()
if no_pages_visited >= self.max_links:
return True
return False
def _join_all_workers(self):
for w in self.workers:
w.join()
def setRunning(self, status):
self.running = status
def addListener(self, callback):
self.listeners.append(callback)
def removeListener(self, callback):
self.listeners.remove(callback)
def notify(self, msg):
for callback in self.listeners:
callback(msg)
def test_link(self):
delegate = XDelegate()
print("test_page started")
# Site 1
site1 = Site()
site1.name = "Site1"
site1.url = 'http://foo.com'
delegate.site_create(site1)
# Crawl
crawl = Crawl(site_id=site1.id)
delegate.crawl_create(crawl)
assert crawl.id > 0
# Page
page = Resource()
page.crawl_id = crawl.id
page.content = "Ala bala portocala"
page.absolute_url = "https://scriptoid.com/index.php"
delegate.resource_create(page)
# Link
# Test url_is_present()
p1 = delegate.url_is_present('https://scriptoid.com/index.php',
crawl.id)
assert not p1
# Test url_count_unvisited()
n1 = delegate.url_count_unvisited(crawl_id=crawl.id)
assert n1 == 0, 'n1 is {}'.format(n1)
# Test url_get_all_by_crawl_id()
crawl_urls = delegate.url_get_all_by_crawl_id(crawl.id)
assert len(crawl_urls) == 0
# Test url_count_incoming_for_resource()
uc1 = delegate.url_count_incoming_for_resource(page.id)
assert uc1 == 0
# Test url_count_internal_full()
cif = delegate.url_count_internal_full(crawl.id)
assert cif == 0
url1 = Url()
url1.src_resource_id = page.id
url1.url = '/team'
url1.absolute_url = 'https://scriptoid.com/team'
url1.type = Url.TYPE_INTERNAL
url1.crawl_id = crawl.id
url1.job_status = Url.JOB_STATUS_IN_PROGRESS
lid1 = delegate.url_create(url1)
assert url1.id > 0
assert lid1 == url1.id
url2 = Url()
url2.src_resource_id = page.id
url2.dst_resource_id = page.id
url2.url = '/contact'
url2.absolute_url = 'https://scriptoid.com/index.php'
url2.type = Url.TYPE_INTERNAL
url2.crawl_id = crawl.id
delegate.url_create(url2)
assert url2.id > 0
url3 = Url()
url3.dst_resource_id = page.id
url3.url = '/jobs'
url3.absolute_url = 'https://scriptoid.com/jobs.php'
url3.type = Url.TYPE_INTERNAL
url3.crawl_id = crawl.id
delegate.url_create(url3)
assert url3.id > 0
# Test url_count_incoming_for_resource()
uc1 = delegate.url_count_incoming_for_resource(page.id)
assert uc1 == 1
# Test url_get_by_id()
u1 = delegate.url_get_by_id(url1.id)
assert u1.id == url1.id
# Test url_is_present()
p1 = delegate.url_is_present('https://scriptoid.com/index.php',
crawl.id)
assert p1
# Test url_get_all_by_crawl_id()
crawl_urls = delegate.url_get_all_by_crawl_id(crawl.id)
assert len(crawl_urls) == 3
# Test first unvisited link
l1 = delegate.url_get_first_unvisited(crawl_id=crawl.id)
assert l1.id == url2.id, 'l1.id = {} and url.id = {}'.format(
l1.id, url2.id)
# Test url_get_all_unvisited()
unvisited1 = delegate.url_get_all_unvisited(crawl.id)
assert len(unvisited1) == 2
# Test url_count_unvisited()
n1 = delegate.url_count_unvisited(crawl_id=crawl.id)
assert n1 == 2, 'n1 is {}'.format(n1)
n2 = delegate.url_count_visited(crawl_id=crawl.id)
assert n2 == 0, 'Actually n2 is {}'.format(n2)
url1.job_status = Url.JOB_STATUS_VISITED
delegate.url_update(url1)
l1 = delegate.url_get_first_unvisited(crawl_id=crawl.id)
assert l1.id == url2.id
n1 = delegate.url_count_unvisited(crawl_id=crawl.id)
assert n1 == 2, 'n1 is {}'.format(n1)
n2 = delegate.url_count_visited(crawl_id=crawl.id)
assert n2 == 1, 'n2 is {}'.format(n2)
# Test url_count_internal_full()
cif = delegate.url_count_internal_full(crawl.id)
assert cif == 1
# Test url_count_pending()
ucp = delegate.url_count_pending(crawl.id)
assert ucp == 2
# Test url_delete_all()
delegate.url_delete_all()
links = delegate.url_get_all()
assert len(links) == 0, "When actually there are {}".format(len(links))
# Test url_count_external()
uce = delegate.url_count_external(crawl.id)
assert uce == 0
url4 = Url()
url4.dst_resource_id = page.id
url4.url = '/jobs'
url4.absolute_url = 'https://scriptoid.com/jobs.php'
url4.type = Url.TYPE_EXTERNAL
url4.crawl_id = crawl.id
delegate.url_create(url4)
assert url4.id > 0
uce = delegate.url_count_external(crawl.id)
assert uce == 1
assert delegate.url_delete_by_id(url4.id)
# Test a cascade delete from parent Page resource_delete_all() to Link
url = Url()
url.src_resource_id = page.id
url.url = '/contact'
url.absolute_url = 'https://scriptoid.com/index.php'
url.type = Url.TYPE_INTERNAL
url.crawl_id = crawl.id
delegate.url_create(url)
assert url.id > 0
delegate.resource_delete_all()
links = delegate.url_get_all()
assert len(links) == 0, "When actually there are {}".format(len(links))
# Clean up
# delegate.link_delete_all()
delegate.resource_delete_all()
delegate.crawl_delete_all()
delegate.site_delete_all()
print("test_page done")