import_representatives.py 11.8 KB
Newer Older
Nicolas Joyard's avatar
Nicolas Joyard committed
1
2
3
# coding: utf-8

import logging
4
import re
Nicolas Joyard's avatar
Nicolas Joyard committed
5
6
7
8
9
10
11
12
13
14
import sys
from datetime import datetime

import django.dispatch
import ijson
import django
from django.apps import apps
from django.db import transaction
from django.utils import timezone

15
from representatives.models import (Country, Mandate, Email, Address, WebSite,
16
17
                                    Representative, Constituency, Phone, Group,
                                    Chamber)
18
19
from variants import FranceDataVariants

Nicolas Joyard's avatar
Nicolas Joyard committed
20
21
22
23
24
25
26
27
28

logger = logging.getLogger(__name__)

representative_pre_import = django.dispatch.Signal(
    providing_args=['representative_data'])


def _get_mdef_item(mdef, item, json, default=None):
    if item in mdef:
29
30
31
32
        try:
            return mdef[item] % json
        except:
            return default
Nicolas Joyard's avatar
Nicolas Joyard committed
33
34
35
36
37
38
39
40
41
42
43
44
45
46

    if '%s_path' % item in mdef:
        return _get_path(json, mdef['%s_path' % item])

    if '%s_fn' % item in mdef:
        return mdef['%s_fn' % item](json)

    return default


def _parse_date(date):
    return datetime.strptime(date, "%Y-%m-%d").date()


47
48
def _create_mandate(representative, group, constituency, role='',
                    begin_date=None, end_date=None):
Nicolas Joyard's avatar
Nicolas Joyard committed
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
    mandate, _ = Mandate.objects.get_or_create(
        representative=representative,
        group=group,
        constituency=constituency,
        role=role,
        begin_date=begin_date,
        end_date=end_date
    )

    if _:
        logger.debug('Created mandate %s', mandate.pk)


def _get_path(dict_, path):
    '''
    Get value at specific path in dictionary. Path is specified by slash-
    separated string, eg _get_path(foo, 'bar/baz') returns foo['bar']['baz']
    '''
    cur = dict_
    for part in path.split('/'):
        cur = cur[part]
    return cur


class GenericImporter(object):

    def pre_import(self):
        self.import_start_datetime = timezone.now()

    def touch_model(self, model, **data):
        '''
        This method create or look up a model with the given data
        it saves the given model if it exists, updating its
        updated field
        '''
84

Nicolas Joyard's avatar
Nicolas Joyard committed
85
86
87
88
89
90
91
92
93
        instance, created = model.objects.get_or_create(**data)

        if not created:
            if instance.updated < self.import_start_datetime:
                instance.save()     # Updates updated field

        return (instance, created)


Nicolas Joyard's avatar
Nicolas Joyard committed
94
95
96
97
98
99
100
101
102
103
104
105
def ensure_chambers():
    """
    Ensures chambers are created
    """
    france = Country.objects.get(name="France")
    for key in ('AN', 'SEN'):
        variant = FranceDataVariants[key]
        Chamber.objects.get_or_create(name=variant['chamber'],
                                      abbreviation=variant['abbreviation'],
                                      country=france)


Nicolas Joyard's avatar
Nicolas Joyard committed
106
class FranceDataImporter(GenericImporter):
107
    url = 'http://francedata.future/data/parlementaires.json'
Nicolas Joyard's avatar
Nicolas Joyard committed
108
109
110
111
112

    def parse_date(self, date):
        return _parse_date(date)

    def __init__(self, variant):
113
        self.france = Country.objects.get(name="France")
Nicolas Joyard's avatar
Nicolas Joyard committed
114
        self.variant = FranceDataVariants[variant]
Nicolas Joyard's avatar
Nicolas Joyard committed
115
        self.chamber = Chamber.objects.get(name=self.variant['chamber'])
116
117
        self.ch_constituency, _ = Constituency.objects.get_or_create(
            name=self.variant['chamber'], country=self.france)
Nicolas Joyard's avatar
Nicolas Joyard committed
118
119
120
121
122
123
124
125
126

    @transaction.atomic
    def manage_rep(self, rep_json):
        '''
        Import a rep as a representative from the json dict fetched from
        FranceData (which comes from nosdeputes.fr)
        '''
        remote_id = rep_json[self.variant['remote_id_field']]

127
128
129
        if rep_json['num_circo'] == 'non disponible':
            rep_json['num_circo'] = 'nd'

Nicolas Joyard's avatar
Nicolas Joyard committed
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
        if not remote_id:
            logger.warning('Skipping MEP without UID %s %s',
                           rep_json['nom'],
                           rep_json[self.variant['remote_id_field']])
            return

        # Some versions of memopol will connect to this and skip inactive reps.
        responses = representative_pre_import.send(sender=self,
                representative_data=rep_json)

        for receiver, response in responses:
            if response is False:
                logger.debug(
                    'Skipping MEP %s', rep_json['nom'])
                return

        try:
            representative = Representative.objects.get(remote_id=remote_id)
        except Representative.DoesNotExist:
            representative = Representative(remote_id=remote_id)

        # Save representative attributes
        self.import_representative_details(representative, rep_json)

        representative.save()

        self.add_mandates(representative, rep_json)

        self.add_contacts(representative, rep_json)

        logger.debug('Imported MEP %s', unicode(representative))

        return representative

    def import_representative_details(self, representative, rep_json):
        representative.active = True
166
167
168
169
        if rep_json.get("ancien_depute", 0) == 1:
            representative.active = False
        if rep_json.get("ancien_senateur", 0) == 1:
            representative.active = False
Nicolas Joyard's avatar
Nicolas Joyard committed
170
171
172
173
174
175
176

        if rep_json.get("date_naissance"):
            representative.birth_date = _parse_date(rep_json["date_naissance"])
        if rep_json.get("lieu_naissance"):
            representative.birth_place = rep_json["lieu_naissance"]

        representative.photo = rep_json['photo_url']
177
178
        representative.first_name = rep_json['prenom']
        representative.last_name = rep_json['nom_de_famille']
Nicolas Joyard's avatar
Nicolas Joyard committed
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
        representative.full_name = rep_json["nom"]

        gender_convertion_dict = {
            u"F": 1,
            u"H": 2
        }
        if 'sexe' in rep_json:
            representative.gender = gender_convertion_dict.get(rep_json[
                                                               'sexe'], 0)
        else:
            representative.gender = 0

        representative.slug = rep_json['slug']

    def add_mandates(self, representative, rep_json):
        '''
        Create mandates from rep data based on variant configuration
        '''

198
199
200
201
202
203
204
205
206
207
208
209
210
        # Mandate in country group for party constituency
        if rep_json.get('parti_ratt_financier'):
            constituency, _ = Constituency.objects.get_or_create(
                name=rep_json.get('parti_ratt_financier'), country=self.france)

            group, _ = self.touch_model(model=Group,
                                        abbreviation=self.france.code,
                                        kind='country',
                                        name=self.france.name)

            _create_mandate(representative, group, constituency, 'membre')

        # Configurable mandates
Nicolas Joyard's avatar
Nicolas Joyard committed
211
        for mdef in self.variant['mandates']:
212
213
214
215
216
            if mdef.get('chamber', False):
                chamber = self.chamber
            else:
                chamber = None

Nicolas Joyard's avatar
Nicolas Joyard committed
217
218
219
220
221
222
223
224
225
226
227
228
            if 'from' in mdef:
                elems = mdef['from'](rep_json)
            else:
                elems = [rep_json]

            for elem in elems:
                name = _get_mdef_item(mdef, 'name', elem, '')
                abbr = _get_mdef_item(mdef, 'abbr', elem, '')

                group, _ = self.touch_model(model=Group,
                                            abbreviation=abbr,
                                            kind=mdef['kind'],
229
                                            chamber=chamber,
Nicolas Joyard's avatar
Nicolas Joyard committed
230
231
232
233
234
235
236
237
238
239
                                            name=name)

                role = _get_mdef_item(mdef, 'role', elem, 'membre')
                start = _get_mdef_item(mdef, 'start', elem, None)
                if start is not None:
                    start = _parse_date(start)
                end = _get_mdef_item(mdef, 'end', elem, None)
                if end is not None:
                    end = _parse_date(end)

240
241
                _create_mandate(representative, group, self.ch_constituency,
                                role, start, end)
Nicolas Joyard's avatar
Nicolas Joyard committed
242
243
244
245
246
247
248
249
250

                logger.debug(
                    '%s => %s: %s of "%s" (%s) %s-%s' % (rep_json['slug'],
                    mdef['kind'], role, name, abbr, start, end))

    def add_contacts(self, representative, rep_json):
        # Websites
        websites = rep_json.get('sites_web', [])
        for site in websites:
251
252
253
254
255
256
257
            if re.search(r'facebook\.com', site['site']):
                self.touch_model(model=WebSite,
                                 url=site['site'],
                                 kind='facebook',
                                 representative=representative
                                 )
            elif not re.search(r'twitter\.com', site['site']):
Nicolas Joyard's avatar
Nicolas Joyard committed
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
                self.touch_model(model=WebSite,
                                 url=site['site'],
                                 representative=representative
                                 )

        # Twitter
        if rep_json.get('twitter'):
            tid = rep_json.get('twitter')
            self.touch_model(model=WebSite,
                             representative=representative,
                             kind='twitter',
                             url='http://twitter.com/%s' % tid
                             )

        # E-mails
        emails = rep_json.get('emails', [])
        for email in emails:
            mail = email['email']
            self.touch_model(
                model=Email,
                representative=representative,
                kind=('official' if mail.endswith(self.variant['mail_domain'])
                    else 'other'),
                email=mail)

283
284
285
286
287
288
289
290
291
292
293
294
295
        # Official address
        off_name = self.variant['off_name']
        official_addr, _ = self.touch_model(model=Address,
                                            representative=representative,
                                            country=self.france,
                                            city=self.variant['off_city'],
                                            street=self.variant['off_street'],
                                            number=self.variant['off_number'],
                                            postcode=self.variant['off_code'],
                                            kind='official',
                                            name=off_name
                                            )

Nicolas Joyard's avatar
Nicolas Joyard committed
296
297
        # Addresses & phone numbers
        addresses = rep_json.get('adresses', [])
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
        for item in addresses:
            addr = None
            if 'geo' in item:
                props = item['geo'].get('properties', {})
                name = ''

                if item['adresse'].lower().startswith('permanence'):
                    name = 'Permanence'

                addr, _ = self.touch_model(model=Address,
                                           representative=representative,
                                           country=self.france,
                                           city=props.get('city', ''),
                                           street=props.get('street', ''),
                                           number=props.get('housenumber', ''),
                                           postcode=props.get('postcode', ''),
                                           kind='',
                                           name=name
                                           )
            elif item['adresse'].lower().startswith(off_name.lower()):
                addr = official_addr

            if 'tel' in item:
                self.touch_model(model=Phone, address=addr,
Nicolas Joyard's avatar
Nicolas Joyard committed
322
                                 representative=representative,
323
                                 kind='', number=item['tel']
Nicolas Joyard's avatar
Nicolas Joyard committed
324
325
326
327
328
329
330
                                 )


def main(stream=None):
    if not apps.ready:
        django.setup()

Nicolas Joyard's avatar
Nicolas Joyard committed
331
332
    ensure_chambers()

333
334
335
336
337
    an_importer = FranceDataImporter('AN')
    GenericImporter.pre_import(an_importer)

    sen_importer = FranceDataImporter('SEN')
    GenericImporter.pre_import(sen_importer)
Nicolas Joyard's avatar
Nicolas Joyard committed
338
339
340

    for data in ijson.items(stream or sys.stdin, ''):
        for rep in data:
341
342
343
344
            if rep['chambre'] == 'AN':
                an_importer.manage_rep(rep)
            elif rep['chambre'] == 'SEN':
                sen_importer.manage_rep(rep)