Newer
Older
import os
import datetime
from bs4 import BeautifulSoup
from urllib.parse import unquote
from Attrap import Attrap
class Attrap_pref31(Attrap):
# Config
__HOST = 'https://www.haute-garonne.gouv.fr'
__RAA_PAGE = f'{__HOST}/Publications/Recueil-des-Actes-Administratifs/Recueil-des-Actes-Administratifs-Haute-Garonne'
__USER_AGENT = 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36'
full_name = 'Préfecture de la Haute-Garonne'
short_code = 'pref31'
def __init__(self, data_dir):
super().__init__(data_dir, self.__USER_AGENT)
self.set_sleep_time(30)
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
def get_raa(self, keywords):
# On cherche les pages de chaque mois
page_content = self.get_page(self.__RAA_PAGE, 'get').content
month_pages = self.get_sub_pages(
page_content,
'.fr-card.fr-card--sm.fr-card--grey.fr-enlarge-link div.fr-card__body div.fr-card__content h2.fr-card__title a',
self.__HOST,
False
)[::-1]
pages_to_parse = []
# On filtre les pages de mois pour limiter le nombre de requêtes
for month_page in month_pages:
guessed_date = Attrap.guess_date(month_page['name'], '([a-zéû]* [0-9]{4})')
if guessed_date >= self.not_before.replace(day=1):
pages_to_parse.append(month_page['url'])
elements = []
# On parse les pages des mois qu'on veut analyser
for element in self.get_raa_with_pager(
pages_to_parse,
".fr-pagination__link.fr-pagination__link--next",
self.__HOST
):
elements.append(element)
self.parse_raa(elements, keywords)
self.mailer()
def get_raa_elements(self, page_content):
elements = []
# On charge le parser
soup = BeautifulSoup(page_content, 'html.parser')
# On récupère chaque balise a
for a in soup.select('div.fr-card__body div.fr-card__content h2.fr-card__title a.fr-card__link.menu-item-link'):
if a.get('href') and a['href'].endswith('.pdf'):
if a['href'].startswith('/'):
url = f"{self.__HOST}{a['href']}"
else:
url = a['href']
url = unquote(url)
name = a.get_text().strip().capitalize()
date = datetime.datetime.strptime(a['title'].split(' - ')[-1].strip(), '%d/%m/%Y')
raa = Attrap.RAA(url, date, name)
elements.append(raa)
return elements