def table_row_to_legislator_and_profile_url(table_row_element, chamber):
"""Derive a Legislator from an HTML table row lxml Element, and a link to their profile"""
td_elements = table_row_element.xpath('td')
(role_element, name_element, district_element, party_element,
phone_element, email_element) = td_elements
# Name comes in the form Last, First
# last_name_first_name = name_element.text_content().strip()
# full_name = last_name_first_name_to_full_name(last_name_first_name)
full_name = name_element.text_content().strip()
district = district_element.text_content().strip()
party = party_element.text_content().strip()
if party == 'Democrat':
party = 'Democratic'
role = role_element.text_content().strip()
address = co_address_from_role(role)
phone = phone_element.text_content().strip()
email = email_element.text_content().strip()
(profile_url, ) = name_element.xpath('a/@href')
print(chamber, district, party)
legislator = Person(primary_org=chamber,
name=full_name,
district=district,
party=party)
legislator.add_contact_detail(type='address', value=address, note='Capitol Office')
legislator.add_contact_detail(type='voice', value=phone, note='Capitol Office')
legislator.add_contact_detail(type='email', value=email, note='Capitol Office')
return legislator, profile_url
def test_deduplication_no_jurisdiction_overlap():
create_person()
# make sure we get a new person if we're in a different org
person = ScrapePerson('Dwayne Johnson')
pd = person.as_dict()
PersonImporter('new-jurisdiction-id').import_data([pd])
assert Person.objects.all().count() == 2
def test_deduplication_no_name_overlap():
create_person()
# make sure we're not just being ridiculous and avoiding importing anything in the same org
person = ScrapePerson('CM Punk')
pd = person.as_dict()
PersonImporter('jurisdiction-id').import_data([pd])
assert Person.objects.all().count() == 2
def scrape_legislator(self, chamber, name, url):
html = self.get(url).text
page = lxml.html.fromstring(html)
page.make_links_absolute(url)
district = page.xpath('//h1[contains(., "DISTRICT")]/text()').pop() \
.split()[1].strip().lstrip('0')
party = page.xpath('//h2').pop().text_content()
party = re.search(r'\((R|D|I)[ \-\]]', party).group(1)
if party == 'D':
party = 'Democratic'
elif party == 'R':
party = 'Republican'
elif party == 'I':
party = 'Independent'
photo_url = page.xpath(
"//img[contains(@src, 'images/members/')]")[0].attrib['src']
leg = Person(name, district=district, party=party, image=photo_url, primary_org=chamber)
leg.add_link(url)
leg.add_source(url)
self.scrape_offices(leg, page)
yield leg
def get_council(self):
council_doc = self.lxmlize(self.COUNCIL_URL)
member_urls = council_doc.xpath(
'//table[@summary="City Directory"]/tr//'
'a[contains(@href, "/directory.aspx?EID=")]/@href')
for member_url in member_urls:
member_doc = self.lxmlize(member_url)
(name, ) = member_doc.xpath('//h1[@class="BioName"]/text()')
(name, ) = re.findall(r'^(?:Mr\.|Mrs\.|Hon\.)?\s*(.*?)\s*$', name)
# Returning everything into a list because the number of values returned varies
# depending on if the person has an email or not
text_list = member_doc.xpath(
'//a[@class="BioLink"]/parent::div/text()')
title = text_list[1].strip()
(title, ) = re.findall(
r'^Title: (Council Member,?(?: Ward \d)|Mayor)\s*$', title)
try:
(image_url, ) = member_doc.xpath(
'//span[@class="BioText"]//img/@src')
except ValueError:
image_url = ''
member = Person(name=name,
image=image_url,
primary_org='legislature',
role=title)
member.add_source(member_url)
yield member
def get_council(self):
council_doc = self.lxmlize(self.COUNCIL_URL)
member_urls = council_doc.xpath(
'//table[@summary="City Directory"]/tr//' 'a[contains(@href, "/directory.aspx?EID=")]/@href'
)
for member_url in member_urls:
member_doc = self.lxmlize(member_url)
(name,) = member_doc.xpath('//span[@class="BioName"]/span/text()')
(name,) = re.findall(r"^(?:Mr\.|Mrs\.|Hon\.)?\s*(.*?)\s*$", name)
(title,) = member_doc.xpath('//a[@class="BioLink"]/following-sibling::text()')
(title,) = re.findall(r"^Title: (Council Member(?: Ward \d)|Mayor)\s*$", title)
try:
(image_url,) = member_doc.xpath('//span[@class="BioText"]//img/@src')
except ValueError:
image_url = ""
member = Person(name=name, image=image_url, primary_org="legislature", role=title)
member.add_source(member_url)
yield member
def scrape_csv(self, reader):
for row in reader:
contributor = Person(
name="{Contact First Name} {Contact Last Name}".format(**row)
)
contributor.add_source(SEARCH_URL)
yield contributor
def test_bill_sponsor_by_identifier():
create_jurisdiction()
org = create_org()
bill = ScrapeBill('HB 1', '1900', 'Axe & Tack Tax Act',
classification='tax bill', chamber='lower')
bill.add_sponsorship_by_identifier(name="SNODGRASS",
classification='sponsor',
entity_type='person',
primary=True,
identifier="TOTALLY_REAL_ID",
scheme="TOTALLY_REAL_SCHEME")
oi = OrganizationImporter('jid')
pi = PersonImporter('jid')
zs = ScrapePerson(name='Zadock Snodgrass')
zs.add_identifier(identifier='TOTALLY_REAL_ID',
scheme='TOTALLY_REAL_SCHEME')
pi.import_data([zs.as_dict()])
za_db = Person.objects.get()
Membership.objects.create(person_id=za_db.id,
organization_id=org.id)
BillImporter('jid', oi, pi).import_data([bill.as_dict()])
obj = Bill.objects.get()
(entry,) = obj.sponsorships.all()
assert entry.person.name == "Zadock Snodgrass"
def test_deduplication_other_name_exists():
create_person()
# Rocky is already saved in other_names
person = ScrapePerson('Rocky')
pd = person.as_dict()
PersonImporter('jurisdiction-id').import_data([pd])
assert Person.objects.all().count() == 1
def scrape_member(self, chamber, link):
name = link.text.strip()
leg_url = link.get('href')
district = link.xpath("string(../../td[3])")
party = link.xpath("string(../../td[4])")
# we get email on the next page now
# email = link.xpath("string(../../td[5])")
if party == 'Democrat':
party = 'Democratic'
elif party == 'No Party Specified':
party = 'Independent'
pid = re.search(r"personID=(\d+)", link.attrib['href']).group(1)
photo_url = ("https://www.legis.iowa.gov/photo"
"?action=getPhoto&ga=%s&pid=%s" % (self.latest_session(), pid))
leg = Person(
name=name,
primary_org=chamber,
district=district,
party=party,
image=photo_url)
leg.add_link(leg_url)
leg.add_source(leg_url)
leg_page = lxml.html.fromstring(self.get(link.attrib['href']).text)
self.scrape_member_page(leg, leg_page)
yield leg
def test_deduplication_same_name():
create_person()
# simplest case- just the same name
person = ScrapePerson('Dwayne Johnson')
pd = person.as_dict()
PersonImporter('jurisdiction-id').import_data([pd])
assert Person.objects.all().count() == 1
def test_multiple_orgs_of_same_class():
"""
We should be able to set memberships on organizations with the
same classification within the same jurisdictions
"""
Organization.objects.create(id="fnd", name="Foundation", classification="foundation",
jurisdiction_id="fnd-jid")
Organization.objects.create(id="fdr", name="Federation", classification="foundation",
jurisdiction_id="fnd-jid")
hari = ScrapePerson('Hari Seldon',
primary_org='foundation',
role='founder',
primary_org_name='Foundation')
picard = ScrapePerson('Jean Luc Picard',
primary_org='foundation',
role='founder',
primary_org_name='Federation')
person_imp = PersonImporter('fnd-jid')
person_imp.import_data([hari.as_dict()])
person_imp.import_data([picard.as_dict()])
# try to import a membership
org_imp = OrganizationImporter('fnd-jid')
dumb_imp = DumbMockImporter()
memimp = MembershipImporter('fnd-jid', person_imp, org_imp, dumb_imp)
memimp.import_data([hari._related[0].as_dict(),
picard._related[0].as_dict()])
assert Person.objects.get(name='Hari Seldon').memberships.get().organization.name == 'Foundation'
assert Person.objects.get(name='Jean Luc Picard').memberships.get().organization.name == 'Federation'
def scrape_counciler(self, url):
page = self.lxmlize(url)
who, = page.xpath("//h3[@class='subtitle']/text()")
district, = page.xpath("//div[@class='right-bar']//h2/text()")
image, = page.xpath(
"//div[@class='left-bar']//a[@class='image lightbox']//img"
)
member = Person(
primary_org='legislature',
name=who, district=district,
image=image.attrib['src']
)
member.add_source(url)
details = page.xpath("//table[@align='center']//td")
for detail in details:
detail = detail.text_content().strip()
if detail is None or detail == "":
continue
type_, value = detail.split(":", 1)
cdtype = {
"Home Phone": "voice",
"Address": "address",
"Email": "email",
"Cell Phone": "voice",
}[type_]
member.add_contact_detail(type=cdtype,
note=type_,
value=value)
yield member
def scrape_upper_chamber(self, term):
url = "http://oksenate.gov/Senators/Default.aspx"
html = self.get(url).text
doc = lxml.html.fromstring(html)
doc.make_links_absolute(url)
for a in doc.xpath('//table[@summary]')[0]. \
xpath('.//td//a[contains(@href, "biographies")]'):
tail = a.xpath('..')[0].tail
if tail:
district = tail.split()[1]
else:
district = a.xpath('../../span')[1].text.split()[1]
if a.text is None or a.text.strip() == 'Vacant':
self.warning("District {} appears to be empty".format(district))
continue
else:
match = re.match(r'(.+) \(([A-Z])\)', a.text.strip())
name, party = match.group(1), self._parties[match.group(2)]
url = a.get('href')
person = Person(primary_org='upper',
district=district,
name=name.strip(),
party=party,
)
person.add_link(url)
person.add_source(url)
self.scrape_upper_offices(person, url)
yield person
def scrape_chamber(self, chamber):
self._party_map = {
'Democrat': 'Democratic',
'Republican': 'Republican',
'Non Affiliated': 'Independent',
'Not Affiliated': 'Independent',
}
if chamber == 'upper':
url = 'http://senate.legis.state.ak.us/'
else:
url = 'http://house.legis.state.ak.us/'
page = self.lxmlize(url)
items = page.xpath('//ul[@class="item"]')[1].getchildren()
for item in items:
photo_url = item.xpath('.//img/@src')[0]
name = item.xpath('.//strong/text()')[0]
leg_url = item.xpath('.//a/@href')[0]
email = item.xpath('.//a[text()="Email Me"]/@href')
if email:
email = email[0].replace('mailto:', '')
else:
self.warning('no email for ' + name)
party = district = None
skip = False
for dt in item.xpath('.//dt'):
dd = dt.xpath('following-sibling::dd')[0].text_content()
label = dt.text.strip()
if label == 'Party:':
party = dd
elif label == 'District:':
district = dd
elif label.startswith('Deceased'):
skip = True
self.warning('skipping deceased ' + name)
break
if skip:
continue
person = Person(
primary_org=chamber,
district=district,
name=name,
party=self._party_map[party],
image=photo_url,
)
person.add_source(leg_url)
person.add_link(leg_url)
# scrape offices
self._scrape_offices(person, leg_url, email)
yield person
def test_deduplication_other_name_overlaps():
create_person()
# Person has other_name that overlaps w/ existing name
person = ScrapePerson('The Rock')
person.add_name('Dwayne Johnson')
pd = person.as_dict()
PersonImporter('jurisdiction-id').import_data([pd])
assert Person.objects.all().count() == 1
def test_invalid_fields_related_item():
p1 = ScrapePerson('Dwayne')
p1.add_link('http://example.com')
p1 = p1.as_dict()
p1['links'][0]['test'] = 3
with pytest.raises(DataImportError):
PersonImporter('jid').import_data([p1])
def test_legislator_related_chamber_district():
leg = Person('John Adams', district='1', primary_org='upper')
leg.pre_save('jurisdiction-id')
assert len(leg._related) == 1
assert leg._related[0].person_id == leg._id
assert get_pseudo_id(leg._related[0].organization_id) == {'classification': 'upper'}
assert get_pseudo_id(leg._related[0].post_id) == {"organization__classification": "upper",
"label": "1"}
def test_person_add_term():
p = Person('Eternal')
p.add_term('eternal', 'council', start_date='0001', end_date='9999')
p._related[0].validate()
assert get_pseudo_id(p._related[0].organization_id) == {
'classification': 'council',
}
assert p._related[0].start_date == '0001'
assert p._related[0].end_date == '9999'
def test_person_add_membership():
p = Person('Bob B. Bear')
p.add_source('http://example.com')
o = Organization('test org', classification='unknown')
p.add_membership(o, role='member', start_date='2007')
assert len(p._related) == 1
p._related[0].validate()
assert p._related[0].person_id == p._id
assert p._related[0].organization_id == o._id
assert p._related[0].start_date == '2007'
def test_person_add_membership_name():
p = Person('Leonardo DiCaprio')
p.add_membership('Academy of Motion Picture Arts and Sciences',
role='winner', start_date='2016')
p._related[0].validate()
assert get_pseudo_id(p._related[0].organization_id) == {
'name': 'Academy of Motion Picture Arts and Sciences'}
assert p._related[0].person_id == p._id
assert p._related[0].role == 'winner'
assert p._related[0].start_date == '2016'
def test_legislator_related_district():
l = Person('John Adams', district='1', primary_org='legislature')
l.pre_save('jurisdiction-id')
assert len(l._related) == 1
assert l._related[0].person_id == l._id
assert get_pseudo_id(l._related[0].organization_id) == {'classification': 'legislature'}
assert get_pseudo_id(l._related[0].post_id) == {"organization__classification": "legislature",
"label": "1", "role": "member"}
assert l._related[0].role == 'member'
def test_legislator_related_party():
l = Person('John Adams', party='Democratic-Republican')
l.pre_save('jurisdiction-id')
# a party membership
assert len(l._related) == 1
assert l._related[0].person_id == l._id
assert get_pseudo_id(l._related[0].organization_id) == {'classification': 'party',
'name': 'Democratic-Republican'}
assert l._related[0].role == 'member'
def handle_list_item(self, item):
photo_url = item.xpath('./img/@src')[0]
url = item.xpath('.//h5/a/@href')[0]
name_text = item.xpath('.//h5/a/b/text()')[0]
name_match = re.match(r'^(.+)\(([0-9]{2}[AB]), ([A-Z]+)\)$', name_text)
name = name_match.group(1).strip()
district = name_match.group(2).lstrip('0').upper()
party_text = name_match.group(3)
party = PARTIES[party_text]
info_texts = [x.strip() for x in item.xpath(
'./div/text()[normalize-space()]'
) if x.strip()]
address = '\n'.join((info_texts[0], info_texts[1]))
phone_text = info_texts[2]
if validate_phone_number(phone_text):
phone = phone_text
email_text = item.xpath('.//a/@href')[1].replace('mailto:', '').strip()
if validate_email_address(email_text):
email = email_text
rep = Person(name=name, district=district, party=party,
primary_org='lower', role='Representative',
image=photo_url)
rep.add_link(url)
rep.add_contact_detail(type='address', value=address, note='capitol')
rep.add_contact_detail(type='voice', value=phone, note='capitol')
rep.add_contact_detail(type='email', value=email, note='capitol')
rep.add_source(self.url)
yield rep
def scrape_lower(self, chamber):
url = 'http://www.house.mi.gov/mhrpublic/frmRepList.aspx'
table = [
"website",
"district",
"name",
"party",
"location",
"phone",
"email"
]
data = self.get(url).text
doc = lxml.html.fromstring(data)
# skip two rows at top
for row in doc.xpath('//table[@id="grvRepInfo"]/*'):
tds = row.xpath('.//td')
if len(tds) == 0:
continue
metainf = {}
for i in range(0, len(table)):
metainf[table[i]] = tds[i]
district = str(int(metainf['district'].text_content().strip()))
party = metainf['party'].text_content().strip()
phone = metainf['phone'].text_content().strip()
email = metainf['email'].text_content().strip()
leg_url = metainf['website'].xpath("./a")[0].attrib['href']
name = metainf['name'].text_content().strip()
if name == 'Vacant' or re.match(r'^District \d{1,3}$', name):
self.warning('District {} appears vacant, and will be skipped'.format(district))
continue
office = metainf['location'].text_content().strip()
office = re.sub(
' HOB',
' Anderson House Office Building\n124 North Capitol Avenue\nLansing, MI 48933',
office
)
office = re.sub(
' CB',
' State Capitol Building\nLansing, MI 48909',
office
)
photo_url = self.get_photo_url(leg_url)
person = Person(name=name, district=district, party=abbr[party],
primary_org='lower', image=photo_url[0] if photo_url else None)
person.add_link(leg_url)
person.add_source(leg_url)
person.add_contact_detail(type='address', value=office, note='Capitol Office')
person.add_contact_detail(type='voice', value=phone, note='Capitol Office')
person.add_contact_detail(type='email', value=email, note='Capitol Office')
yield person
def test_save_related():
s = Scraper('jurisdiction', '/tmp/')
p = Person('Michael Jordan')
p.add_source('http://example.com')
o = Organization('Chicago Bulls')
o.add_source('http://example.com')
p._related.append(o)
with mock.patch('json.dump') as json_dump:
s.save_object(p)
assert json_dump.mock_calls == [mock.call(p.as_dict(), mock.ANY, cls=mock.ANY),
mock.call(o.as_dict(), mock.ANY, cls=mock.ANY)]
def test_save_object_basics():
# ensure that save object dumps a file
s = Scraper('jurisdiction', '/tmp/')
p = Person('Michael Jordan')
p.add_source('http://example.com')
with mock.patch('json.dump') as json_dump:
s.save_object(p)
# ensure object is saved in right place
filename = 'person_' + p._id + '.json'
assert filename in s.output_names['person']
json_dump.assert_called_once_with(p.as_dict(), mock.ANY, cls=mock.ANY)
def scrape_member_page(self, chamber, url):
page = self.get(url).text
page = lxml.html.fromstring(page)
page.make_links_absolute(url)
for legislator in page.xpath(
"//div[contains(concat(' ', normalize-space(@class), ' '), "
"' memberModule ')]"
):
img = legislator.xpath(
".//div[@class='thumbnail']//img")[0].attrib['src']
data = legislator.xpath(".//div[@class='data']")[0]
homepage = data.xpath(".//a[@class='black']")[0]
full_name = homepage.text_content()
if "Vacant" in full_name:
continue
homepage = homepage.attrib['href']
party = data.xpath(
".//span[@class='partyLetter']")[0].text_content()
party = {"R": "Republican", "D": "Democratic"}[party]
office_lines = data.xpath("child::text()")
phone = office_lines.pop(-1)
office = "\n".join(office_lines)
h3 = data.xpath("./h3")
if len(h3):
h3 = h3[0]
district = h3.xpath("./br")[0].tail.replace("District", ""
).strip()
else:
district = re.findall(
r"\d+\.png",
legislator.attrib['style']
)[-1].split(".", 1)[0]
full_name = re.sub(r"\s+", " ", full_name).strip()
email = (
'rep{0:0{width}}@ohiohouse.gov'
if chamber == 'lower' else
'sd{0:0{width}}@ohiosenate.gov'
).format(int(district), width=2)
leg = Person(name=full_name, district=district,
party=party, primary_org=chamber,
image=img)
leg.add_contact_detail(type='address', value=office, note='Capitol Office')
leg.add_contact_detail(type='voice', value=phone, note='Capitol Office')
leg.add_contact_detail(type='email', value=email, note='Capitol Office')
self.scrape_homepage(leg, chamber, homepage)
leg.add_source(url)
leg.add_link(homepage)
yield leg
def scrape_alderman(self, ward_num):
ward_url = "{}/ward-{}".format(Utils.ALDERMEN_HOME, ward_num)
alderman_url = self.alderman_url(ward_url)
alderman_page = self.lxmlize(alderman_url)
# person's name is the only <h1> tag on the page
name = alderman_page.xpath("//h1/text()")[0]
# initialize person object with appropriate data so that pupa can
# automatically create a membership object linking this person to
# a post in the jurisdiction's "Board of Aldermen" organization
district = "Ward {} Alderman".format(ward_num)
person = Person(name=name, district=district, role="Alderman",
primary_org="legislature")
# set additional fields
person.image = alderman_page.xpath("//div/img/@src")[0]
phone_number = alderman_page.xpath("//strong[text()='Phone:']/../text()")[1].strip()
person.add_contact_detail(type="voice", value=phone_number)
# add sources
person.add_source(alderman_url, note="profile")
person.add_source(ward_url, note="ward")
return person
def test_no_membership_for_person():
org = Organization.objects.create(id="fnd", name="Foundation", classification="foundation",
jurisdiction_id="fnd-jid")
# import a person with no memberships
p = ScrapePerson('a man without a country')
person_imp = PersonImporter('fnd-jid')
person_imp.import_data([p.as_dict()])
# try to import a membership
dumb_imp = DumbMockImporter()
memimp = MembershipImporter('fnd-jid', person_imp, dumb_imp, dumb_imp)
with pytest.raises(NoMembershipsError):
memimp.import_data([])
def test_save_object_invalid():
s = Scraper('jurisdiction', '/tmp/')
p = Person('Michael Jordan')
# no source, won't validate
with pytest.raises(ValueError):
s.save_object(p)
def test_committee_add_member_person():
c = Organization('Defense', classification='committee')
p = Person('John Adams')
c.add_member(p, role='chairman')
assert c._related[0].person_id == p._id
assert c._related[0].organization_id == c._id
assert c._related[0].role == 'chairman'
def scrape_senator_page(self, chamber, url):
page = self.get(url).text
page = lxml.html.fromstring(page)
page.make_links_absolute(url)
for legislator in page.xpath(
"//div[@id='senators']//div[contains(concat(' ', normalize-space(@class), ' '), "
"' portraitContainer ')]"):
img = legislator.xpath(".//div[@class='profileThumbnailBoundingBox']/@style")[0]
img = img[img.find('(')+1:img.find(')')]
full_name = legislator.xpath(".//div[@class='profileName']/a/text()")[0]
homepage_url = legislator.xpath(".//a[@class='profileImageLink']")[0].attrib['href']
district = legislator.xpath(".//div[@class='profileDistrict']"
"/a/text()")[0].split("#")[1]
if "Vacant" in full_name:
continue
homepage = self.get(homepage_url).text
page = lxml.html.fromstring(homepage)
phone = page.xpath("//div[@class='phone']/span/text()")[0]
address_lines = page.xpath("//div[@class='address']/span/text()")
address = "\n".join(address_lines)
party_image = page.xpath('//div[@class="senatorParty"]/img/@src')[0]
if 'Republican' in party_image:
party = 'Republican'
elif 'Democrat' in party_image:
party = 'Democratic'
email = (
'rep{0:0{width}}@ohiohouse.gov'
if chamber == 'lower' else
'sd{0:0{width}}@ohiosenate.gov'
).format(int(district), width=2)
leg = Person(name=full_name, district=district,
primary_org=chamber, image=img, party=party)
leg.add_contact_detail(type='address', value=address, note='Capitol Office')
leg.add_contact_detail(type='voice', value=phone, note='Capitol Office')
leg.add_contact_detail(type='email', value=email, note='Capitol Office')
leg.add_source(url)
leg.add_link(homepage_url)
yield leg
def test_full_vote_event():
j = create_jurisdiction()
j.legislative_sessions.create(name='1900', identifier='1900')
sp1 = ScrapePerson('John Smith', primary_org='lower')
sp2 = ScrapePerson('Adam Smith', primary_org='lower')
org = ScrapeOrganization(name='House', classification='lower')
bill = ScrapeBill('HB 1', '1900', 'Axe & Tack Tax Act', from_organization=org._id)
vote_event = ScrapeVoteEvent(legislative_session='1900', motion_text='passage',
start_date='1900-04-01', classification='passage:bill',
result='pass', bill_chamber='lower', bill='HB 1',
organization=org._id)
vote_event.set_count('yes', 20)
vote_event.yes('John Smith')
vote_event.no('Adam Smith')
oi = OrganizationImporter('jid')
oi.import_data([org.as_dict()])
pi = PersonImporter('jid')
pi.import_data([sp1.as_dict(), sp2.as_dict()])
mi = MembershipImporter('jid', pi, oi, DumbMockImporter())
mi.import_data([sp1._related[0].as_dict(), sp2._related[0].as_dict()])
bi = BillImporter('jid', oi, pi)
bi.import_data([bill.as_dict()])
VoteEventImporter('jid', pi, oi, bi).import_data([vote_event.as_dict()])
assert VoteEvent.objects.count() == 1
ve = VoteEvent.objects.get()
assert ve.legislative_session == LegislativeSession.objects.get()
assert ve.motion_classification == ['passage:bill']
assert ve.bill == Bill.objects.get()
count = ve.counts.get()
assert count.option == 'yes'
assert count.value == 20
votes = list(ve.votes.all())
assert len(votes) == 2
for v in ve.votes.all():
if v.voter_name == 'John Smith':
assert v.option == 'yes'
assert v.voter == Person.objects.get(name='John Smith')
else:
assert v.option == 'no'
assert v.voter == Person.objects.get(name='Adam Smith')
def test_multiple_memberships():
# there was a bug where two or more memberships to the same jurisdiction
# would cause an ORM error, this test ensures that it is fixed
p = Person.objects.create(name='Dwayne Johnson')
o = Organization.objects.create(name='WWE',
jurisdiction_id='jurisdiction-id')
Membership.objects.create(person=p, organization=o)
o = Organization.objects.create(name='WWF',
jurisdiction_id='jurisdiction-id')
Membership.objects.create(person=p, organization=o)
person = ScrapePerson('Dwayne Johnson')
pd = person.as_dict()
PersonImporter('jurisdiction-id').import_data([pd])
# deduplication should still work
assert Person.objects.all().count() == 1
def test_no_membership_for_person():
Organization.objects.create(id="fnd",
name="Foundation",
classification="foundation",
jurisdiction_id="fnd-jid")
# import a person with no memberships
p = ScrapePerson('a man without a country')
person_imp = PersonImporter('fnd-jid')
person_imp.import_data([p.as_dict()])
# try to import a membership
dumb_imp = DumbMockImporter()
memimp = MembershipImporter('fnd-jid', person_imp, dumb_imp, dumb_imp)
with pytest.raises(NoMembershipsError):
memimp.import_data([])
def scrape(self):
urls = Urls(dict(list=legislators_url), self)
council = Organization('Temecula City Council',
classification='legislature')
council.add_source(urls.list.url)
yield council
for tr in urls.list.xpath('//table[2]//tr')[1:]:
# Parse some attributes.
name, role = tr.xpath('td/p[1]//font/text()')
image = tr.xpath('td/img/@src').pop()
# Create legislator.
person = Person(name, image=image)
# Add membership on council.
memb = person.add_membership(council, role=role)
# Add email address.
email, detail_url = tr.xpath('td//a/@href')
email = email[7:]
memb.contact_details.append(
dict(type='email', value=email, note='work'))
# Add sources.
person.add_source(urls.list.url)
person.add_source(detail_url)
yield person
def scrape_upper_chamber(self, term):
url = 'https://senado.pr.gov/Pages/Senadores.aspx'
doc = self.lxmlize(url)
links = self.get_nodes(doc, '//ul[@class="senadores-list"]/li/a/@href')
for link in links:
senator_page = self.lxmlize(link)
profile_links = self.get_nodes(senator_page, '//ul[@class="profiles-links"]/li')
name_text = self.get_node(senator_page, '//span[@class="name"]').text_content().strip()
name = re.sub(r'^Hon\.', '', name_text, flags=re.IGNORECASE).strip()
party = profile_links[0].text_content().strip()
photo_url = self.get_node(senator_page, '//div[@class="avatar"]//img/@src')
if profile_links[1].text_content().strip() == "Senador por Distrito":
district_text = self.get_node(
senator_page,
'//div[@class="module-distrito"]//span[@class="headline"]').text_content()
district = district_text.replace('DISTRITO', '', 1).replace('\u200b', '').strip()
elif profile_links[1].text_content().strip() == "Senador por Acumulación":
district = "At-Large"
phone_node = self.get_node(senator_page, '//a[@class="contact-data tel"]')
phone = phone_node.text_content().strip()
email_node = self.get_node(senator_page, '//a[@class="contact-data email"]')
email = email_node.text_content().replace('\u200b', '').strip()
person = Person(primary_org='upper',
district=district,
name=name,
party=party,
image=photo_url)
person.add_contact_detail(type='email',
value=email,
note='Capitol Office')
person.add_contact_detail(type='voice',
value=phone,
note='Capitol Office')
person.add_link(link)
person.add_source(link)
yield person
def scrape_rep(self, url):
page = self.get(url).text
page = lxml.html.fromstring(page)
page.make_links_absolute(url)
main = page.xpath('//div[@id="main-info"]')[0]
if "Resigned" in main.text_content():
print("Member resigned {}".format(url))
raise StopIteration # don't yield anything
name = page.xpath('//div[@class="member-name"]/text()')[0].strip()
name = re.sub(r"\s+", " ", name)
district_number = page.xpath(
'//span[contains(text(), "House District:")]'
"/following-sibling::span/text()")[0].strip()
# remove anything after first whitespace
district_number = re.sub(r"\s.*", "", district_number.strip())
email = None
email_content = page.xpath(
'//a[./i[contains(@class,"fa-envelope")]]/text()')
if email_content and email_content[0].strip():
email = email_content[0].strip()
photo_url = page.xpath('//header[@id="home"]/img/@src')[0]
party = self.get_rep_table_by_header(page,
"Party Affiliation").text.strip()
party = _party_map[party[0]] # standardize
main_p_text = page.xpath('//div[@id="main-info"]/p/text()')
address = [t.strip() for t in main_p_text if t.strip()][0]
person = Person(
name=name,
district=district_number,
primary_org="lower",
party=party,
image=photo_url,
)
person.add_contact_detail(type="address",
value=address,
note="District Office")
person.add_contact_detail(type="email",
value=email,
note="District Office")
person.add_source(url)
yield person
def handle_list_item(self, item):
photo_url = item.xpath('./td[1]/a/img/@src')[0]
info_nodes = item.xpath('./td[2]/p/a')
name_text = info_nodes[0].xpath('./b/text()')[0]
url = info_nodes[0].get('href')
name_match = re.match(r'^(.+)\(([0-9]{2}[AB]), ([A-Z]+)\)$', name_text)
name = name_match.group(1).strip()
district = name_match.group(2).lstrip('0').upper()
party_text = name_match.group(3)
party = PARTIES[party_text]
info_texts = [
x.strip() for x in item.xpath(
'./td[2]/p/text()[normalize-space() and preceding-sibling::br]'
) if x.strip()
]
address = '\n'.join((info_texts[0], info_texts[1]))
phone_text = info_texts[2]
if validate_phone_number(phone_text):
phone = phone_text
email_node = info_nodes[1]
email_text = email_node.text
email_text = email_text.replace('Email: ', '').strip()
if validate_email_address(email_text):
email = email_text
rep = Person(name=name,
district=district,
party=party,
primary_org='lower',
role='Representative',
image=photo_url)
rep.add_link(url)
rep.add_contact_detail(type='address', value=address, note='capitol')
rep.add_contact_detail(type='voice', value=phone, note='capitol')
rep.add_contact_detail(type='email', value=email, note='capitol')
rep.add_source(self.url)
yield rep
def handle_list_item(self, item):
photo_url = item.xpath("./img/@src")[0]
url = item.xpath(".//h5/a/@href")[0]
name_text = item.xpath(".//h5/a/b/text()")[0]
name_match = re.match(r"^(.+)\(([0-9]{2}[AB]), ([A-Z]+)\)$", name_text)
name = name_match.group(1).strip()
district = name_match.group(2).lstrip("0").upper()
party_text = name_match.group(3)
party = PARTIES[party_text]
info_texts = [
x.strip()
for x in item.xpath("./div/text()[normalize-space()]")
if x.strip()
]
address = "\n".join((info_texts[0], info_texts[1]))
phone_text = info_texts[2]
if validate_phone_number(phone_text):
phone = phone_text
email_text = item.xpath(".//a/@href")[1].replace("mailto:", "").strip()
if validate_email_address(email_text):
email = email_text
rep = Person(
name=name,
district=district,
party=party,
primary_org="lower",
role="Representative",
image=photo_url,
)
rep.add_link(url)
rep.add_contact_detail(type="address", value=address, note="capitol")
rep.add_contact_detail(type="voice", value=phone, note="capitol")
rep.add_contact_detail(type="email", value=email, note="capitol")
rep.add_source(self.url)
yield rep
def test_multiple_orgs_of_same_class():
"""
We should be able to set memberships on organizations with the
same classification within the same jurisdictions
"""
create_jurisdiction()
Organization.objects.create(id="fnd",
name="Foundation",
classification="foundation",
jurisdiction_id="fnd-jid")
Organization.objects.create(id="fdr",
name="Federation",
classification="foundation",
jurisdiction_id="fnd-jid")
hari = ScrapePerson('Hari Seldon',
primary_org='foundation',
role='founder',
primary_org_name='Foundation')
picard = ScrapePerson('Jean Luc Picard',
primary_org='foundation',
role='founder',
primary_org_name='Federation')
person_imp = PersonImporter('fnd-jid')
person_imp.import_data([hari.as_dict()])
person_imp.import_data([picard.as_dict()])
# try to import a membership
org_imp = OrganizationImporter('fnd-jid')
dumb_imp = DumbMockImporter()
memimp = MembershipImporter('fnd-jid', person_imp, org_imp, dumb_imp)
memimp.import_data(
[hari._related[0].as_dict(), picard._related[0].as_dict()])
assert Person.objects.get(
name='Hari Seldon').memberships.get().organization.name == 'Foundation'
assert Person.objects.get(name='Jean Luc Picard').memberships.get(
).organization.name == 'Federation'
def parse_row(self, row, chamber):
print(row)
display = '{} {}'.format(row['First Name'], row['Surname'])
# TODO: map state to ocd
# TODO: https://www.aph.gov.au/Senators_and_Members/Senators/Senators_by_service_expiry_date
person = Person(
name=display,
district=row['State'],
role='member',
primary_org=chamber,
gender=row['Gender'].lower(),
party=row['Political Party'],
)
person.extras['given_name'] = row['First Name']
person.extras['family_name'] = row['Surname']
return person
def scrape_member(self, chamber, link):
name = link.text.strip()
leg_url = link.get("href")
district = link.xpath("string(../../td[3])")
party = link.xpath("string(../../td[4])")
# we get email on the next page now
# email = link.xpath("string(../../td[5])")
if party == "Democrat":
party = "Democratic"
elif party == "No Party Specified":
party = "Independent"
pid = re.search(r"personID=(\d+)", link.attrib["href"]).group(1)
photo_url = ("https://www.legis.iowa.gov/photo"
"?action=getPhoto&ga=%s&pid=%s" %
(self.latest_session(), pid))
leg = Person(
name=name,
primary_org=chamber,
district=district,
party=party,
image=photo_url,
)
leg.add_link(leg_url)
leg.add_source(leg_url)
leg_page = lxml.html.fromstring(self.get(link.attrib["href"]).text)
self.scrape_member_page(leg, leg_page)
yield leg
def scrape_upper_chamber(self, term):
url = "http://oksenate.gov/Senators/Default.aspx"
html = self.get(url).text
doc = lxml.html.fromstring(html)
doc.make_links_absolute(url)
for a in doc.xpath('//table[@summary]')[0]. \
xpath('.//td//a[contains(@href, "biographies")]'):
tail = a.xpath('..')[0].tail
if tail:
district = tail.split()[1]
else:
district = a.xpath('../../span')[1].text.split()[1]
if a.text is None or a.text.strip() == 'Vacant':
self.warning(
"District {} appears to be empty".format(district))
continue
else:
match = re.match(r'(.+) \(([A-Z])\)', a.text.strip())
name, party = match.group(1), self._parties[match.group(2)]
url = a.get('href')
person = Person(
primary_org='upper',
district=district,
name=name.strip(),
party=party,
)
person.add_link(url)
person.add_source(url)
self.scrape_upper_offices(person, url)
yield person
def scrape_legislator(self, chamber, name, url):
html = self.get(url).text
page = lxml.html.fromstring(html)
page.make_links_absolute(url)
district = page.xpath('//h1[contains(., "DISTRICT")]/text()').pop() \
.split()[1].strip().lstrip('0')
party = page.xpath('//h2').pop().text_content()
party = re.search(r'\((R|D|I)[ \-\]]', party).group(1)
if party == 'D':
party = 'Democratic'
elif party == 'R':
party = 'Republican'
elif party == 'I':
party = 'Independent'
photo_url = page.xpath(
"//img[contains(@src, 'images/members/')]")[0].attrib['src']
leg = Person(name,
district=district,
party=party,
image=photo_url,
primary_org=chamber)
leg.add_link(url)
leg.add_source(url)
self.scrape_offices(leg, page)
yield leg
def scrape_lower_legislator(self, url, leg_info):
page = self.lxmlize(url)
name = page.xpath(
'//span[@id="body_FormView5_FULLNAMELabel"]/text()')[0].strip()
if name.startswith("District ") or name.startswith("Vacant "):
self.warning("Seat is vacant: {}".format(name))
return
photo = page.xpath(
'//img[contains(@src, "/h_reps/RepPics")]')[0].attrib["src"]
party_flags = {
"Democrat": "Democratic",
"Republican": "Republican",
"Independent": "Independent",
}
party_info = page.xpath(
'//span[@id="body_FormView5_PARTYAFFILIATIONLabel"]/text()'
)[0].strip()
party = party_flags[party_info]
try:
email = page.xpath(
'//span[@id="body_FormView6_EMAILADDRESSPUBLICLabel"]/text()'
)[0].strip()
except IndexError:
email = None
district = leg_info["dist"].replace("Dist", "").strip()
person = Person(name=name,
party=party,
district=district,
primary_org="lower",
image=photo)
contacts = [
(leg_info["office"], "address"),
(leg_info["phone"], "voice"),
(email, "email"),
]
for value, key in contacts:
if value:
person.add_contact_detail(type=key,
value=value,
note="District Office")
person.add_source(url)
person.add_link(url)
yield person
def test_no_membership_for_person_including_party():
"""
even though party is specified we should still get a no memberships error because it doesn't
bind the person to a jurisdiction, thus causing duplication
"""
create_jurisdiction()
Organization.objects.create(id="fnd", name="Foundation", classification="foundation",
jurisdiction_id="fnd-jid")
Organization.objects.create(id="dem", name="Democratic", classification="party")
# import a person with no memberships
p = ScrapePerson('a man without a country', party='Democratic')
person_imp = PersonImporter('fnd-jid')
org_imp = OrganizationImporter('fnd-jid')
person_imp.import_data([p.as_dict()])
# try to import a membership
dumb_imp = DumbMockImporter()
memimp = MembershipImporter('fnd-jid', person_imp, org_imp, dumb_imp)
with pytest.raises(NoMembershipsError):
memimp.import_data([p._related[0].as_dict()])
def scrape_chamber(self, chamber):
self._party_map = {
'Democrat': 'Democratic',
'Republican': 'Republican',
'Non Affiliated': 'Independent',
'Not Affiliated': 'Independent',
}
if chamber == 'upper':
url = 'http://senate.legis.state.ak.us/'
else:
url = 'http://house.legis.state.ak.us/'
page = self.lxmlize(url)
items = page.xpath('//ul[@class="item"]')[1].getchildren()
for item in items:
photo_url = item.xpath('.//img/@src')[0]
name = item.xpath('.//strong/text()')[0]
leg_url = item.xpath('.//a/@href')[0]
email = item.xpath('.//a[text()="Email Me"]/@href')
if email:
email = email[0].replace('mailto:', '')
else:
self.warning('no email for ' + name)
party = district = None
skip = False
for dt in item.xpath('.//dt'):
dd = dt.xpath('following-sibling::dd')[0].text_content()
label = dt.text.strip()
if label == 'Party:':
party = dd
elif label == 'District:':
district = dd
elif label.startswith('Deceased'):
skip = True
self.warning('skipping deceased ' + name)
break
if skip:
continue
person = Person(
primary_org=chamber,
district=district,
name=name,
party=self._party_map[party],
image=photo_url,
)
person.add_source(leg_url)
person.add_link(leg_url)
# scrape offices
self._scrape_offices(person, leg_url, email)
yield person
def scrape(self):
# lower
url = 'http://164.100.47.194/Loksabha/Members/AlphabeticalList.aspx'
entry = self.get(url).content
page = lxml.html.fromstring(entry)
page.make_links_absolute(url)
for tr in page.xpath(
'//table[contains(@class,"member_list_table")]/tr'):
name = tr.xpath('td[2]/a[1]/@title')[0]
bio = tr.xpath('td[2]/a[1]/@href')[0]
photo_url = tr.xpath('td[2]/a[1]/img/@src')[0]
party = tr.xpath('td[3]/text()')[0].strip()
state = tr.xpath('td[4]/text()')[0].strip()
member = Person(name=name,
role="member",
primary_org="lower",
party=party,
image=photo_url,
district=state)
member.add_source('http://164.100.47.5/Newmembers/memberlist.aspx')
yield member
# upper
url = 'http://164.100.47.5/Newmembers/memberlist.aspx'
entry = self.get(url).content
page = lxml.html.fromstring(entry)
page.make_links_absolute(url)
for tr in page.xpath(
'//table[@id="ContentPlaceHolder1_GridView2"]/tr')[1:]:
name = tr.xpath('td[2]/font/a/text()')[0]
party_abbr = tr.xpath('td[3]/font/text()')[0].strip()
state = tr.xpath('td[4]/font/text()')[0].strip()
member = Person(name=name,
role="member",
primary_org="upper",
party=party_abbr,
district=state)
member.add_source('http://164.100.47.5/Newmembers/memberlist.aspx')
yield member
def test_person_add_membership():
p = Person('Bob B. Bear')
p.add_source('http://example.com')
o = Organization('test org')
p.add_membership(o, 'member', start_date='2007')
assert len(p._related) == 1
p._related[0].validate()
assert p._related[0].person_id == p._id
assert p._related[0].organization_id == o._id
assert p._related[0].start_date == '2007'
def test_person_add_membership_org():
p = Person('Bob B. Bear')
p.add_source('http://example.com')
o = Organization('test org', classification='unknown')
p.add_membership(o, role='member', start_date='2007', end_date=datetime.date(2015, 5, 8))
assert len(p._related) == 1
p._related[0].validate()
assert p._related[0].person_id == p._id
assert p._related[0].organization_id == o._id
assert p._related[0].start_date == '2007'
assert p._related[0].end_date == datetime.date(2015, 5, 8)
def test_same_name_people_other_name():
create_jurisdiction()
# ensure we're taking other_names into account for the name collision code
Organization.objects.create(name='WWE', jurisdiction_id='jid')
p1 = ScrapePerson('Dwayne Johnson', image='http://example.com/1')
p2 = ScrapePerson('Rock', image='http://example.com/2')
p2.add_name('Dwayne Johnson')
# the people have the same name but are apparently different
with pytest.raises(SameNameError):
PersonImporter('jid').import_data([p1.as_dict(), p2.as_dict()])
def table_row_to_legislator_and_profile_url(table_row_element, chamber):
"""Derive a Legislator from an HTML table row lxml Element, and a link to their profile"""
td_elements = table_row_element.xpath("td")
(
role_element,
name_element,
district_element,
party_element,
phone_element,
email_element,
) = td_elements
# Name comes in the form Last, First
# last_name_first_name = name_element.text_content().strip()
# full_name = last_name_first_name_to_full_name(last_name_first_name)
full_name = name_element.text_content().strip()
if full_name.count(", ") == 1:
full_name = " ".join(full_name.split(", ")[::-1]).strip()
district = district_element.text_content().strip()
party = party_element.text_content().strip()
if party == "Democrat":
party = "Democratic"
elif party == "Unaffiliated":
party = "Independent"
role = role_element.text_content().strip()
address = co_address_from_role(role)
phone = phone_element.text_content().strip()
email = email_element.text_content().strip()
(profile_url,) = name_element.xpath("a/@href")
print(chamber, district, party)
legislator = Person(
primary_org=chamber, name=full_name, district=district, party=party
)
legislator.add_contact_detail(type="address", value=address, note="Capitol Office")
if phone:
legislator.add_contact_detail(type="voice", value=phone, note="Capitol Office")
if email:
legislator.add_contact_detail(type="email", value=email, note="Capitol Office")
return legislator, profile_url
def test_same_name_second_import():
create_jurisdiction()
# ensure two people with the same name don't import without birthdays
o = Organization.objects.create(name='WWE', jurisdiction_id='jid')
p1 = ScrapePerson('Dwayne Johnson', image='http://example.com/1')
p2 = ScrapePerson('Dwayne Johnson', image='http://example.com/2')
p1.birth_date = '1970'
p2.birth_date = '1930'
# when we give them birth dates all is well though
PersonImporter('jid').import_data([p1.as_dict(), p2.as_dict()])
# fake some memberships so future lookups work on these people
for p in Person.objects.all():
Membership.objects.create(person=p, organization=o)
p3 = ScrapePerson('Dwayne Johnson', image='http://example.com/3')
with pytest.raises(SameNameError):
PersonImporter('jid').import_data([p3.as_dict()])
def test_bill_sponsor_limit_lookup():
create_jurisdiction()
org = create_org()
bill = ScrapeBill('HB 1',
'1900',
'Axe & Tack Tax Act',
classification='tax bill',
chamber='lower')
bill.add_sponsorship_by_identifier(name="SNODGRASS",
classification='sponsor',
entity_type='person',
primary=True,
identifier="TOTALLY_REAL_ID",
scheme="TOTALLY_REAL_SCHEME")
oi = OrganizationImporter('jid')
pi = PersonImporter('jid')
zs = ScrapePerson(name='Zadock Snodgrass', birth_date="1800-01-01")
zs.add_identifier(identifier='TOTALLY_REAL_ID',
scheme='TOTALLY_REAL_SCHEME')
pi.import_data([zs.as_dict()])
za_db = Person.objects.get()
Membership.objects.create(person_id=za_db.id, organization_id=org.id)
zs2 = ScrapePerson(name='Zadock Snodgrass', birth_date="1900-01-01")
zs2.add_identifier(identifier='TOTALLY_REAL_ID',
scheme='TOTALLY_REAL_SCHEME')
# This is contrived and perhaps broken, but we're going to check this.
# We *really* don't want to *ever* cross jurisdiction bounds.
PersonImporter('another-jurisdiction').import_data([zs.as_dict()])
BillImporter('jid', oi, pi).import_data([bill.as_dict()])
obj = Bill.objects.get()
(entry, ) = obj.sponsorships.all()
assert entry.person.name == "Zadock Snodgrass"
assert entry.person.birth_date == "1800-01-01"
def handle_list_item(self, item):
name = item.text
if 'resigned' in name.lower() or 'vacated' in name.lower():
return
if (name in CHAMBER_MOVES and (self.chamber != CHAMBER_MOVES[name])):
return
name, action, date = clean_name(name)
leg = Person(name=name)
leg.add_source(self.url)
leg.add_source(item.get('href'))
leg.add_link(item.get('href'))
yield from self.scrape_page(
self.detail_page,
item.get('href'),
session=self.kwargs['session'],
obj=leg,
)
yield leg
def table_row_to_legislator_and_profile_url(table_row_element, chamber):
"""Derive a Legislator from an HTML table row lxml Element, and a link to their profile"""
td_elements = table_row_element.xpath('td')
(role_element, name_element, district_element, party_element,
phone_element, email_element) = td_elements
# Name comes in the form Last, First
# last_name_first_name = name_element.text_content().strip()
# full_name = last_name_first_name_to_full_name(last_name_first_name)
full_name = name_element.text_content().strip()
district = district_element.text_content().strip()
party = party_element.text_content().strip()
if party == 'Democrat':
party = 'Democratic'
role = role_element.text_content().strip()
address = co_address_from_role(role)
phone = phone_element.text_content().strip()
email = email_element.text_content().strip()
(profile_url, ) = name_element.xpath('a/@href')
print(chamber, district, party)
legislator = Person(primary_org=chamber,
name=full_name,
district=district,
party=party)
legislator.add_contact_detail(type='address',
value=address,
note='Capitol Office')
if phone:
legislator.add_contact_detail(type='voice',
value=phone,
note='Capitol Office')
if email:
legislator.add_contact_detail(type='email',
value=email,
note='Capitol Office')
return legislator, profile_url
def handle_list_item(self, item):
name = item.text
lname = name.lower()
if "resigned" in lname or "vacated" in lname or "retired" in lname:
return
if name in CHAMBER_MOVES and (self.chamber != CHAMBER_MOVES[name]):
return
name, action, date = clean_name(name)
leg = Person(name=name)
leg.add_source(self.url)
leg.add_source(item.get("href"))
leg.add_link(item.get("href"))
yield from self.scrape_page(
self.detail_page,
item.get("href"),
session=self.kwargs["session"],
committees=self.kwargs["committees"],
obj=leg,
)
yield leg
def handle_list_item(self, item):
name = " ".join(item.xpath('.//text()'))
name = re.sub(r'\s+', " ", name).replace(" ,", ",").strip()
if 'Vacant' in name:
return
district = item.xpath("string(../../td[1])")
party = item.xpath("string(../../td[2])")
if party == 'Democrat':
party = 'Democratic'
leg_url = item.get('href')
name = fix_name(name)
leg = Person(name=name, district=district, party=party,
primary_org='upper', role='Senator')
leg.add_link(leg_url)
leg.add_source(self.url)
leg.add_source(leg_url)
self.scrape_page(SenDetail, leg_url, obj=leg)
return leg
