Python scrape.VoteEvent类代码示例

def test_fix_bill_id():
    j = create_jurisdiction()
    j.legislative_sessions.create(name='1900', identifier='1900')

    org1 = ScrapeOrganization(name='House', classification='lower')
    bill = ScrapeBill('HB 1', '1900', 'Test Bill ID',
                      classification='bill', chamber='lower')

    oi = OrganizationImporter('jid')

    oi.import_data([org1.as_dict()])

    from pupa.settings import IMPORT_TRANSFORMERS
    IMPORT_TRANSFORMERS['bill'] = {
        'identifier': lambda x: re.sub(r'([A-Z]*)\s*0*([-\d]+)', r'\1 \2', x, 1)
    }

    bi = BillImporter('jid', oi, DumbMockImporter())
    bi.import_data([bill.as_dict()])

    ve = ScrapeVoteEvent(legislative_session='1900', motion_text='passage',
                         start_date='1900-04-02', classification='passage:bill',
                         result='fail', bill_chamber='lower', bill='HB1',
                         identifier='4',
                         bill_action='passage',
                         organization=org1._id)

    VoteEventImporter('jid', DumbMockImporter(), oi, bi).import_data([
        ve.as_dict(),
    ])

    IMPORT_TRANSFORMERS['bill'] = {}

    ve = VoteEvent.objects.get()
    ve.bill.identifier == 'HB 1'

    def add_vote(self, bill, chamber, date, text, url):
        votes = re.findall(r'Ayes,?[\s]?(\d+)[,;]\s+N(?:oes|ays),?[\s]?(\d+)', text)
        yes, no = int(votes[0][0]), int(votes[0][1])

        vtype = 'other'
        for regex, type in motion_classifiers.items():
            if re.match(regex, text):
                vtype = type
                break

        v = VoteEvent(
            chamber=chamber,
            start_date=TIMEZONE.localize(date),
            motion_text=text,
            result='pass' if yes > no else 'fail',
            classification=vtype,
            bill=bill,
        )
        v.pupa_id = url.split('/')[-1]
        v.set_count('yes', yes)
        v.set_count('no', no)

        # fetch the vote itself
        if url:
            v.add_source(url)

            if 'av' in url:
                self.add_house_votes(v, url)
            elif 'sv' in url:
                self.add_senate_votes(v, url)

        return v

 def _get_votes(self, date, actor, action, bill, url):
     vre = r'(?P<leader>.*)(AYES|YEAS):\s+(?P<yeas>\d+)\s+(NOES|NAYS):\s+(?P<nays>\d+).*'
     if 'YEAS' in action.upper() or 'AYES' in action.upper():
         match = re.match(vre, action)
         if match:
             v = match.groupdict()
             yes, no = int(v['yeas']), int(v['nays'])
             vote = VoteEvent(
                 chamber=actor,
                 motion_text=v['leader'],
                 result='pass' if yes > no else 'fail',
                 classification='passage',
                 start_date=TIMEZONE.localize(date),
                 bill=bill,
             )
             vote.add_source(url)
             yield vote

    def get_vote_event(self, bill, act, votes, result):
        '''Make VoteEvent object from given Bill, action, votes and result.'''
        organization = json.loads(act['organization_id'].lstrip('~'))
        vote_event = VoteEvent(legislative_session=bill.legislative_session,
                               motion_text=act['description'],
                               organization=organization,
                               classification=None,
                               start_date=act['date'],
                               result=result,
                               bill=bill)

        legistar_web, legistar_api = [src['url'] for src in bill.sources]

        vote_event.add_source(legistar_web)
        vote_event.add_source(legistar_api + '/histories')

        for vote in votes:
            raw_option = vote['VoteValueName'].lower()

            if raw_option == 'suspended':
                continue

            clean_option = self.VOTE_OPTIONS.get(raw_option, raw_option)
            vote_event.vote(clean_option, vote['VotePersonName'].strip())

        return vote_event

def test_vote_event_identifier_dedupe():
    j = Jurisdiction.objects.create(id='jid', division_id='did')
    j.legislative_sessions.create(name='1900', identifier='1900')

    vote_event = ScrapeVoteEvent(legislative_session='1900', start_date='2013',
                                 classification='anything', result='passed',
                                 motion_text='a vote on something',
                                 identifier='Roll Call No. 1')
    dmi = DumbMockImporter()
    bi = BillImporter('jid', dmi, dmi)

    _, what = VoteEventImporter('jid', dmi, dmi, bi).import_item(vote_event.as_dict())
    assert what == 'insert'
    assert VoteEvent.objects.count() == 1

    # same exact vote event, no changes
    _, what = VoteEventImporter('jid', dmi, dmi, bi).import_item(vote_event.as_dict())
    assert what == 'noop'
    assert VoteEvent.objects.count() == 1

    # new info, update
    vote_event.result = 'failed'
    _, what = VoteEventImporter('jid', dmi, dmi, bi).import_item(vote_event.as_dict())
    assert what == 'update'
    assert VoteEvent.objects.count() == 1

    # new bill, insert
    vote_event.identifier = 'Roll Call 2'
    _, what = VoteEventImporter('jid', dmi, dmi, bi).import_item(vote_event.as_dict())
    assert what == 'insert'
    assert VoteEvent.objects.count() == 2

def test_vote_event_pupa_identifier_dedupe():
    j = create_jurisdiction()
    j.legislative_sessions.create(name='1900', identifier='1900')
    Organization.objects.create(id='org-id', name='Legislature',
                                classification='legislature',
                                jurisdiction=j)

    vote_event = ScrapeVoteEvent(legislative_session='1900', start_date='2013',
                                 classification='anything', result='passed',
                                 motion_text='a vote on something',
                                 identifier='Roll Call No. 1')
    vote_event.pupa_id = 'foo'

    dmi = DumbMockImporter()
    oi = OrganizationImporter('jid')
    bi = BillImporter('jid', dmi, oi)

    _, what = VoteEventImporter('jid', dmi, oi, bi).import_item(vote_event.as_dict())
    assert what == 'insert'
    assert VoteEvent.objects.count() == 1

    # same exact vote event, no changes
    _, what = VoteEventImporter('jid', dmi, oi, bi).import_item(vote_event.as_dict())
    assert what == 'noop'
    assert VoteEvent.objects.count() == 1

    # new info, update
    vote_event.result = 'failed'
    _, what = VoteEventImporter('jid', dmi, oi, bi).import_item(vote_event.as_dict())
    assert what == 'update'
    assert VoteEvent.objects.count() == 1

    # new bill identifier, update
    vote_event.identifier = 'First Roll Call'
    _, what = VoteEventImporter('jid', dmi, oi, bi).import_item(vote_event.as_dict())
    assert what == 'update'
    assert VoteEvent.objects.count() == 1

    # new pupa identifier, insert
    vote_event.pupa_id = 'bar'
    _, what = VoteEventImporter('jid', dmi, oi, bi).import_item(vote_event.as_dict())
    assert what == 'insert'
    assert VoteEvent.objects.count() == 2

def test_full_vote_event():
    j = Jurisdiction.objects.create(id='jid', division_id='did')
    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_vote_event_bill_actions_two_stage():
    # this test is very similar to what we're testing in test_vote_event_bill_actions w/
    # ve3 and ve4, that two bills that reference the same action won't conflict w/ the
    # OneToOneField, but in this case we do it in two stages so that the conflict is found
    # even if the votes weren't in the same scrape
    j = create_jurisdiction()
    j.legislative_sessions.create(name='1900', identifier='1900')
    org1 = ScrapeOrganization(name='House', classification='lower')
    bill = ScrapeBill('HB 1', '1900', 'Axe & Tack Tax Act', from_organization=org1._id)

    bill.add_action(description='passage', date='1900-04-02', chamber='lower')

    ve1 = ScrapeVoteEvent(legislative_session='1900', motion_text='passage',
                          start_date='1900-04-02', classification='passage:bill',
                          result='pass', bill_chamber='lower', bill='HB 1',
                          bill_action='passage',
                          organization=org1._id)
    ve2 = ScrapeVoteEvent(legislative_session='1900', motion_text='passage',
                          start_date='1900-04-02', classification='passage:bill',
                          result='pass', bill_chamber='lower', bill='HB 1',
                          bill_action='passage',
                          organization=org1._id)
    # disambiguate them
    ve1.pupa_id = 'one'
    ve2.pupa_id = 'two'

    oi = OrganizationImporter('jid')
    oi.import_data([org1.as_dict()])

    bi = BillImporter('jid', oi, DumbMockImporter())
    bi.import_data([bill.as_dict()])

    # first imports just fine
    VoteEventImporter('jid', DumbMockImporter(), oi, bi).import_data([
        ve1.as_dict(),
    ])
    votes = list(VoteEvent.objects.all())
    assert len(votes) == 1
    assert votes[0].bill_action is not None

    # when second is imported, ensure that action stays pinned to first just as it would
    # have if they were both in same import
    VoteEventImporter('jid', DumbMockImporter(), oi, bi).import_data([
        ve1.as_dict(),
        ve2.as_dict(),
    ])
    votes = list(VoteEvent.objects.all())
    assert len(votes) == 2
    assert votes[0].bill_action is not None
    assert votes[1].bill_action is None

def viva_voce_votes(root, session, chamber):
    for el in root.xpath(u'//div[starts-with(., "All Members are deemed")]'):
        mv = MaybeViva(el)
        if not mv.is_valid:
            continue

        v = VoteEvent(
            chamber=chamber,
            start_date=None,
            motion_text='passage' if mv.passed else 'other',
            result='pass' if mv.passed else 'fail',
            classification='passage' if mv.passed else 'other',
            legislative_session=session[0:2],
            bill=mv.bill_id,
            bill_chamber=mv.chamber
        )

        v.set_count('yes', 0)
        v.set_count('no', 0)
        v.set_count('absent', 0)
        v.set_count('not voting', 0)

        yield v

def test_vote_event_bill_id_dedupe():
    j = create_jurisdiction()
    session = j.legislative_sessions.create(name='1900', identifier='1900')
    org = Organization.objects.create(id='org-id', name='House', classification='lower',
                                      jurisdiction=j)
    bill = Bill.objects.create(id='bill-1', identifier='HB 1', legislative_session=session,
                               from_organization=org)
    bill2 = Bill.objects.create(id='bill-2', identifier='HB 2', legislative_session=session,
                                from_organization=org)

    vote_event = ScrapeVoteEvent(legislative_session='1900', start_date='2013',
                                 classification='anything', result='passed',
                                 motion_text='a vote on something',
                                 bill=bill.identifier, bill_chamber='lower',
                                 chamber='lower')
    dmi = DumbMockImporter()
    oi = OrganizationImporter('jid')
    bi = BillImporter('jid', dmi, oi)

    _, what = VoteEventImporter('jid', dmi, oi, bi).import_item(vote_event.as_dict())
    assert what == 'insert'
    assert VoteEvent.objects.count() == 1

    # same exact vote event, no changes
    _, what = VoteEventImporter('jid', dmi, oi, bi).import_item(vote_event.as_dict())
    assert what == 'noop'
    assert VoteEvent.objects.count() == 1

    # new info, update
    vote_event.result = 'failed'
    _, what = VoteEventImporter('jid', dmi, oi, bi).import_item(vote_event.as_dict())
    assert what == 'update'
    assert VoteEvent.objects.count() == 1

    # new vote event, insert
    vote_event = ScrapeVoteEvent(legislative_session='1900', start_date='2013',
                                 classification='anything', result='passed',
                                 motion_text='a vote on something',
                                 bill=bill2.identifier, bill_chamber='lower',
                                 chamber='lower')
    _, what = VoteEventImporter('jid', dmi, oi, bi).import_item(vote_event.as_dict())
    assert what == 'insert'
    assert VoteEvent.objects.count() == 2

def test_vote_event_bill_clearing():
    # ensure that we don't wind up with vote events sitting around forever on bills as
    # changes make it look like there are multiple vote events
    j = create_jurisdiction()
    session = j.legislative_sessions.create(name='1900', identifier='1900')
    org = Organization.objects.create(id='org-id', name='House', classification='lower',
                                      jurisdiction=j)
    bill = Bill.objects.create(id='bill-1', identifier='HB 1', legislative_session=session,
                               from_organization=org)
    Bill.objects.create(id='bill-2', identifier='HB 2', legislative_session=session,
                        from_organization=org)
    oi = OrganizationImporter('jid')
    dmi = DumbMockImporter()
    bi = BillImporter('jid', dmi, oi)

    vote_event1 = ScrapeVoteEvent(legislative_session='1900', start_date='2013',
                                  classification='anything', result='passed',
                                  motion_text='a vote on somthing',             # typo intentional
                                  bill=bill.identifier, bill_chamber='lower',
                                  chamber='lower'
                                  )
    vote_event2 = ScrapeVoteEvent(legislative_session='1900', start_date='2013',
                                  classification='anything', result='passed',
                                  motion_text='a vote on something else',
                                  bill=bill.identifier, bill_chamber='lower',
                                  chamber='lower'
                                  )

    # have to use import_data so postimport is called
    VoteEventImporter('jid', dmi, oi, bi).import_data([
        vote_event1.as_dict(),
        vote_event2.as_dict()
    ])
    assert VoteEvent.objects.count() == 2

    # a typo is fixed, we don't want 3 vote events now
    vote_event1.motion_text = 'a vote on something'
    VoteEventImporter('jid', dmi, oi, bi).import_data([
        vote_event1.as_dict(),
        vote_event2.as_dict()
    ])
    assert VoteEvent.objects.count() == 2

    def createVoteEvent(self, motion, agenda_item_version):
        version = agenda_item_version
        date = self.toDate(version['date'])
        v = VoteEvent(
            motion_text=motion['title_text'],
            result=RESULT_MAP[motion['result']],
            classification=motion['action'],
            start_date=date,
            legislative_session=version['session'],
        )

        if motion['mover']:
            v.extras['mover'] = motion['mover']
        if motion['body_text']:
            v.extras['body'] = motion['body_text']

        v.set_bill(version['bill_identifier'])
        v.add_source(version['url'])

        return v

    def process_committee_vote(self, committee_action, bill):
        try:
            date = committee_action["ActionDate"]
            vote_info = committee_action["Vote"]

        except KeyError:
            self.logger.warning("Committee vote has no data. Skipping.")
            return
        date = self.date_format(date)

        other_count = 0
        for v in vote_info:
            vote_count = 0 if v["VoteCount"] == "" else int(v["VoteCount"])

            if v["VoteType"] == "Yes":
                yes_count = vote_count
            elif v["VoteType"] == "No":
                no_count = vote_count
            else:
                other_count += vote_count

        result = 'fail'
        if yes_count > no_count:
            result = 'pass'

        v = VoteEvent(chamber='legislature',
                      start_date=date,
                      motion_text='Committee Vote',
                      result=result,
                      classification='committee',
                      bill=bill
                      )
        v.set_count('yes', yes_count)
        v.set_count('no', no_count)
        v.set_count('other', other_count)

        return v

    def scrape_vote(self, bill, vote_json, session):

        if vote_json['amendmentNumber']:
            motion = '{}: {}'.format(
                vote_json['amendmentNumber'], vote_json['action'])
        else:
            motion = vote_json['action']

        result = 'pass' if vote_json['yesVotesCount'] > vote_json['noVotesCount'] else 'fail'

        v = VoteEvent(
            chamber=self.chamber_abbrev_map[vote_json['chamber']],
            start_date=self.parse_local_date(vote_json['voteDate']),
            motion_text=motion,
            result=result,
            legislative_session=session,
            bill=bill,
            classification='other',
        )

        v.set_count(option='yes', value=vote_json['yesVotesCount'])
        v.set_count('no', vote_json['noVotesCount'])
        v.set_count('absent', vote_json['absentVotesCount'])
        v.set_count('excused', vote_json['excusedVotesCount'])
        v.set_count('other', vote_json['conflictVotesCount'])

        for name in vote_json['yesVotes'].split(','):
            if name.strip():
                v.yes(name.strip())

        for name in vote_json['noVotes'].split(','):
            if name.strip():
                v.no(name.strip())

        # add votes with other classifications
        # option can be 'yes', 'no', 'absent',
        # 'abstain', 'not voting', 'paired', 'excused'
        for name in vote_json['absentVotes'].split(','):
            if name.strip():
                v.vote(option="absent",
                       voter=name)

        for name in vote_json['excusedVotes'].split(','):
            if name.strip():
                v.vote(option="excused",
                       voter=name)

        for name in vote_json['conflictVotes'].split(','):
            if name.strip():
                v.vote(option="other",
                       voter=name)

        source_url = 'http://lso.wyoleg.gov/Legislation/{}/{}'.format(
            session, vote_json['billNumber'])
        v.add_source(source_url)

        yield v

    def scrape_action_page(self, bill, page):
        action_rows = page.xpath('//tbody/tr')
        for row in action_rows:
            action_date = row.xpath('td[1]/text()')[0]
            action_date = datetime.strptime(action_date, '%m/%d/%Y')
            action_year = action_date.year
            action_date = action_date.strftime('%Y-%m-%d')

            if row.xpath('td[2]/text()'):
                action_actor = row.xpath('td[2]/text()')[0]
                action_actor = self.chamber_map_reverse[action_actor.strip()]

            action_name = row.xpath('string(td[3])')

            # House votes
            if "Supplement" in action_name:
                actor = "lower"
                vote_action = action_name.split(' -')[0]
                y = int(action_name.strip().split('-')[1].split('YEAS')[0])
                n = int(action_name.strip().split('YEAS to')[1].split('NAYS')[0])

                # get supplement number
                n_supplement = int(action_name.strip().split('No. ')[1].split(r')')[0])
                cached_vote = VoteEvent(
                    chamber=actor,
                    start_date=action_date,
                    motion_text=vote_action,
                    result='pass' if y > n else 'fail',
                    classification='passage',
                    bill=bill,
                )
                cached_vote.set_count('yes', y)
                cached_vote.set_count('no', n)

                housevote_pdf = 'http://www.mass.gov/legis/journal/combined{}RCs.pdf'.format(
                    action_year
                )
                # note: 2014-2015 different format and no data on website for years prior to 2014
                self.scrape_house_vote(cached_vote, housevote_pdf, n_supplement)
                cached_vote.add_source(housevote_pdf)

                cached_vote.pupa_id = '{}#{}'.format(housevote_pdf, n_supplement)

                yield cached_vote

            # Senate votes
            if "Roll Call" in action_name:
                actor = "upper"
                # placeholder
                vote_action = action_name.split(' -')[0]
                try:
                    y, n = re.search('(\d+) yeas .*? (\d+) nays', action_name.lower()).groups()
                    y = int(y)
                    n = int(n)
                except AttributeError:
                    y = int(re.search(r"yeas\s*(\d*)", action_name.lower()).group(1))
                    n = int(re.search(r"nays\s*(\d*)", action_name.lower()).group(1))

                # TODO: other count isn't included, set later
                cached_vote = VoteEvent(
                    chamber=actor,
                    start_date=action_date,
                    motion_text=vote_action,
                    result='pass' if y > n else 'fail',
                    classification='passage',
                    bill=bill,
                )
                cached_vote.set_count('yes', y)
                cached_vote.set_count('no', n)

                rollcall_pdf = 'http://malegislature.gov' + row.xpath('string(td[3]/a/@href)')
                self.scrape_senate_vote(cached_vote, rollcall_pdf)
                cached_vote.add_source(rollcall_pdf)
                yield cached_vote

            attrs = self.categorizer.categorize(action_name)
            action = bill.add_action(
                action_name.strip(),
                action_date,
                chamber=action_actor,
                classification=attrs['classification'],
            )
            for com in attrs.get('committees', []):
                action.add_related_entity(com, entity_type='organization')