def parse(self, response):
"""Scrape a spider's HttpRequest.Response for links"""
# sanity check
if self._link_extractor is None:
self._link_extractor = SgmlLinkExtractor()
# use scrapy SgmlLinkExtractor to extract links
try:
links = self._link_extractor.extract_links(response)
except SGMLParseError as e:
# Page was poorly formatted, oh well
_linkscraper_logger.error('Exception encountered when link extracting page')
return []
# add these links to our Url item
urls = list()
for link in links:
url = ScrapedUrl()
url['url'] = link.url
url['domain'] = UrlUtility.get_domain(link.url)
url['last_visited'] = datetime(1, 1, 1)
if url not in urls:
urls.append(url)
return urls
def parse(self, response):
"""Scrape a spider's HttpRequest.Response for links"""
# sanity check
if self._link_extractor is None:
self._link_extractor = SgmlLinkExtractor()
# use scrapy SgmlLinkExtractor to extract links
try:
links = self._link_extractor.extract_links(response)
except SGMLParseError as e:
# Page was poorly formatted, oh well
_linkscraper_logger.error(
'Exception encountered when link extracting page')
return []
# add these links to our Url item
urls = list()
for link in links:
url = ScrapedUrl()
url['url'] = link.url
url['domain'] = UrlUtility.get_domain(link.url)
url['last_visited'] = datetime(1, 1, 1)
if url not in urls:
urls.append(url)
return urls
def clean_organization_url(self):
url = self.cleaned_data['organization_url']
if url:
try:
return UrlUtility.get_domain(url)
except:
raise ValidationError("Oops! We couldn't recognize that URL's domain.")
else:
return None
def clean_organization_url(self):
url = self.cleaned_data['organization_url']
if url:
try:
return UrlUtility.get_domain(url)
except:
raise ValidationError(
"Oops! We couldn't recognize that URL's domain.")
else:
return None
def __init__(self, org_model, id):
"""
Constructs a new SmallOrganization instance.
Arguments:
org_model (Organization): The organization from which to create a SmallOrganization.
id (ObjectId): The database ID of the organization.
"""
self.id = id
self.page_rank_info = org_model.page_rank_info
try:
self.org_domain = UrlUtility.get_domain(org_model.organization_url, no_exception=False)
except Exception:
self.org_domain = None
self.page_rank = None
self.page_rank_weight = None
def __init__(self, org_model, id):
"""
Constructs a new SmallOrganization instance.
Arguments:
org_model (Organization): The organization from which to create a SmallOrganization.
id (ObjectId): The database ID of the organization.
"""
self.id = id
self.page_rank_info = org_model.page_rank_info
try:
self.org_domain = UrlUtility.get_domain(org_model.organization_url,
no_exception=False)
except Exception:
self.org_domain = None
self.page_rank = None
self.page_rank_weight = None
def _merge_page_rank_info(self, new_references, existing_references, organization_url):
if existing_references is None:
return new_references
org_domain = UrlUtility().get_domain(organization_url)
for ref in new_references.references:
ref_exists = False
# Search for existing references from one organization to another
for exist_ref in existing_references.references:
if ref.org_domain == exist_ref.org_domain:
# We found existing data for references from Org A to Org B
ref_exists = True
for page in ref.pages:
page_exists = False
# Search if we have data from this specific URL to this specific organization
for exist_page in exist_ref.pages:
if page.url == exist_page.url:
# We found existing data for references from URL A to Org B
page_exists = True
count_diff = page.count - exist_page.count
if count_diff != 0:
# This page must have changed b/c the number of references is different
# update everything
exist_page.count = page.count
exist_ref.count += count_diff
existing_references.total_with_self += count_diff
if exist_ref.org_domain != org_domain:
# This value only updated if Organization A and B are different
existing_references.total += count_diff
break
if not page_exists:
# We have recorded other references to this organization, but none from this url
exist_ref.pages.append(page)
exist_ref.count += page.count
existing_references.total_with_self += page.count
if exist_ref.org_domain != org_domain:
existing_references.total += page.count
break
# If this organization has not yet referenced the specified outside org, add it
if not ref_exists:
existing_references.references.append(ref)
existing_references.total_with_self += ref.count
if ref.org_domain != org_domain:
existing_references.total += ref.count
return existing_references
def check_valid_org(self, response):
"""
Checks if the current page is a valid page for an organization's homepage.
Arguments:
reponse (Response): Scrapy Response object of the page that is to be scraped.
Returns:
True if it's a valid organization page or already in the database.
False if it's not the homepage.
"""
# If already in database, then valid
url = OrgUrlScraper().parse(response)
org_dto = self.org_dao().find(organization_url=url)
if org_dto:
return True
# If not homepage, then return false and make sure homepage is added to scrape:
home_url_obj = urlparse(response.request.url)
if home_url_obj.path and home_url_obj.path is not '/':
home_url = home_url_obj.scheme + '://' + home_url_obj.netloc + '/'
home_domain = UrlUtility.get_domain(home_url)
meta = URLMetadata(url=home_url,
domain=home_domain,
last_visited=datetime(1, 1, 1))
self.url_frontier.put_url(meta)
return False
else:
# this is homepage, scrape for keywords
hxs = HtmlXPathSelector(response)
site_text = hxs.select('//html//text()').extract()
site_text = [
element.strip() for element in site_text
if element.strip() != ''
]
for word in self._required_words:
for sentence in site_text:
sentence = self._punctuation.sub(' ', sentence)
if word in sentence.lower():
return True
# no keyword found, check if we already added organization
return False
def request_organization(request):
"""
Sends a request to the Request Organization page if the user is logged in.
Returns:
A rendered page containing the Request Organization form.
"""
if 'user_id' not in request.session:
logger.error('Bad request made for organization seed without login')
return unauthorized(request)
else:
user_id = request.session['user_id']
form = RequestOrgForm(request.POST or None)
error = ''
success = ''
if request.method == 'POST':
if form.is_valid():
url = form.cleaned_data['url']
dao = ctx.get_object('URLMetadataDAO')
try:
metadata = URLMetadata(url=url,
domain=UrlUtility.get_domain(url))
except ValueError:
error = "Oops! We don't recognize that domain. Please try another."
if not error:
try:
dto = DTOConverter.to_dto(URLMetadataDTO, metadata)
dao.create_update(dto)
logger.info(
'Org seed with url={0} requested by user={1}'.format(
url, user_id))
success = 'Your request has been sent successfully!'
except:
error = 'Something went wrong with your request. Please try again later.'
return render(request, 'organization/request_organization.html', {
'form': form,
'success': success,
'error': error
})
def clean_url(self):
url = self.cleaned_data['url']
ctx = ApplicationContext(DAOContext())
org_dao = ctx.get_object('OrganizationDAO')
url_metadata_dao = ctx.get_object('URLMetadataDAO')
try:
domain = UrlUtility().get_domain(url)
except:
raise ValidationError(
"Oops! We couldn't find information on that domain.")
if org_dao.find(organization_url=domain) or url_metadata_dao.find(
domain=domain):
raise ValidationError(
"Oops! Looks like we already have information on that organization."
)
return url
def check_valid_org(self, response):
"""
Checks if the current page is a valid page for an organization's homepage.
Arguments:
reponse (Response): Scrapy Response object of the page that is to be scraped.
Returns:
True if it's a valid organization page or already in the database.
False if it's not the homepage.
"""
# If already in database, then valid
url = OrgUrlScraper().parse(response)
org_dto = self.org_dao().find(organization_url=url)
if org_dto:
return True
# If not homepage, then return false and make sure homepage is added to scrape:
home_url_obj = urlparse(response.request.url)
if home_url_obj.path and home_url_obj.path is not '/':
home_url = home_url_obj.scheme + '://' + home_url_obj.netloc + '/'
home_domain = UrlUtility.get_domain(home_url)
meta = URLMetadata(url=home_url, domain=home_domain, last_visited=datetime(1, 1, 1))
self.url_frontier.put_url(meta)
return False
else:
# this is homepage, scrape for keywords
hxs = HtmlXPathSelector(response)
site_text = hxs.select('//html//text()').extract()
site_text = [element.strip() for element in site_text if element.strip() != '']
for word in self._required_words:
for sentence in site_text:
sentence = self._punctuation.sub(' ', sentence)
if word in sentence.lower():
return True
# no keyword found, check if we already added organization
return False
def request_organization(request):
"""
Sends a request to the Request Organization page if the user is logged in.
Returns:
A rendered page containing the Request Organization form.
"""
if 'user_id' not in request.session:
logger.error('Bad request made for organization seed without login')
return unauthorized(request)
else:
user_id = request.session['user_id']
form = RequestOrgForm(request.POST or None)
error = ''
success = ''
if request.method == 'POST':
if form.is_valid():
url = form.cleaned_data['url']
dao = ctx.get_object('URLMetadataDAO')
try:
metadata = URLMetadata(url=url, domain=UrlUtility.get_domain(url))
except ValueError:
error = "Oops! We don't recognize that domain. Please try another."
if not error:
try:
dto = DTOConverter.to_dto(URLMetadataDTO, metadata)
dao.create_update(dto)
logger.info('Org seed with url={0} requested by user={1}'.format(url, user_id))
success = 'Your request has been sent successfully!'
except:
error = 'Something went wrong with your request. Please try again later.'
return render(request, 'organization/request_organization.html', {'form': form, 'success': success, 'error': error})
def parse(self, response):
"""Scrape a spider's HttpRequest.Response for links"""
# get domain
try:
org_domain = UrlUtility.get_domain(response.request.url, False)
except Exception as e:
_linkscraper_logger.error('Exception encountered when trying to find the domain of ' + response.request.url)
# sanity check
if self._link_extractor is None:
self._link_extractor = SgmlLinkExtractor()
# use scrapy SgmlLinkExtractor to extract links
try:
links = self._link_extractor.extract_links(response)
except SGMLParseError as e:
# Page was poorly formatted, oh well
_linkscraper_logger.error('Exception encountered when PageRankInfo scraping page')
return None
# add these links to our Page Rank Info
page_rank_info = {
"total": 0,
"total_with_self": 0,
"references": []
}
for link in links:
url = link.url
try:
domain = UrlUtility.get_domain(url, False)
except Exception as e:
_linkscraper_logger.error('Exception encountered when trying to find the domain of ' + url)
continue
ref_found = False
for ref in page_rank_info["references"]:
if ref["org_domain"] == domain:
ref_found = True
ref["count"] += 1
ref["pages"][0]["count"] += 1
page_rank_info["total_with_self"] += 1
if domain != org_domain:
page_rank_info["total"] += 1
break;
if not ref_found:
page_rank_info["references"].append(
{
"org_domain": domain,
"count": 1,
"pages": [
{
"url": response.url,
"count": 1
}
]
}
)
page_rank_info["total_with_self"] += 1
if domain != org_domain:
page_rank_info["total"] += 1
return page_rank_info
def parse(self, response):
"""Scrape a spider's HttpRequest.Response for links"""
# get domain
try:
org_domain = UrlUtility.get_domain(response.request.url, False)
except Exception as e:
_linkscraper_logger.error(
'Exception encountered when trying to find the domain of ' +
response.request.url)
# sanity check
if self._link_extractor is None:
self._link_extractor = SgmlLinkExtractor()
# use scrapy SgmlLinkExtractor to extract links
try:
links = self._link_extractor.extract_links(response)
except SGMLParseError as e:
# Page was poorly formatted, oh well
_linkscraper_logger.error(
'Exception encountered when PageRankInfo scraping page')
return None
# add these links to our Page Rank Info
page_rank_info = {"total": 0, "total_with_self": 0, "references": []}
for link in links:
url = link.url
try:
domain = UrlUtility.get_domain(url, False)
except Exception as e:
_linkscraper_logger.error(
'Exception encountered when trying to find the domain of '
+ url)
continue
ref_found = False
for ref in page_rank_info["references"]:
if ref["org_domain"] == domain:
ref_found = True
ref["count"] += 1
ref["pages"][0]["count"] += 1
page_rank_info["total_with_self"] += 1
if domain != org_domain:
page_rank_info["total"] += 1
break
if not ref_found:
page_rank_info["references"].append({
"org_domain":
domain,
"count":
1,
"pages": [{
"url": response.url,
"count": 1
}]
})
page_rank_info["total_with_self"] += 1
if domain != org_domain:
page_rank_info["total"] += 1
return page_rank_info
