From 0ced41ed9afcaa7a529bbc1c283bac01c6a5f3e5 Mon Sep 17 00:00:00 2001
From: Bastien Le Querrec <blq@laquadrature.net>
Date: Fri, 21 Jun 2024 23:28:11 +0200
Subject: [PATCH] pref81: simplification de l'analyse des pages
---
Attrap_pref81.py | 103 +++++++++++++++++++++--------------------------
1 file changed, 45 insertions(+), 58 deletions(-)
diff --git a/Attrap_pref81.py b/Attrap_pref81.py
index 6d6943b..65dab05 100644
--- a/Attrap_pref81.py
+++ b/Attrap_pref81.py
@@ -11,15 +11,7 @@ class Attrap_pref81(Attrap):
# Config
__HOST = 'https://www.tarn.gouv.fr'
- __RAA_PAGE = {
- 'default': f'{__HOST}/Publications/RAA-Recueil-des-Actes-Administratifs/RAA',
- '2024': f'{__HOST}/Publications/RAA-Recueil-des-Actes-Administratifs/RAA/2024',
- '2023': f'{__HOST}/Publications/RAA-Recueil-des-Actes-Administratifs/RAA/2023',
- '2022': f'{__HOST}/Publications/RAA-Recueil-des-Actes-Administratifs/RAA/2022',
- '2021': f'{__HOST}/Publications/RAA-Recueil-des-Actes-Administratifs/RAA/2021',
- '2020': f'{__HOST}/Publications/RAA-Recueil-des-Actes-Administratifs/RAA/2020',
- '2019': f'{__HOST}/Publications/RAA-Recueil-des-Actes-Administratifs/RAA/2019',
- }
+ __RAA_PAGE = f'{__HOST}/Publications/RAA-Recueil-des-Actes-Administratifs/RAA'
__USER_AGENT = 'Mozilla/5.0 (X11; Linux x86_64; rv:109.0) Gecko/20100101 Firefox/115.0'
full_name = 'Préfecture du Tarn'
short_code = 'pref81'
@@ -29,79 +21,74 @@ class Attrap_pref81(Attrap):
self.enable_tor(10)
def get_raa(self, keywords):
- pages_to_parse = []
- if self.not_before.year <= 2024:
- pages_to_parse.append(self.__RAA_PAGE['2024'])
- if self.not_before.year <= 2023:
- pages_to_parse.append(self.__RAA_PAGE['2023'])
- if self.not_before.year <= 2022:
- pages_to_parse.append(self.__RAA_PAGE['2022'])
- if self.not_before.year <= 2021:
- pages_to_parse.append(self.__RAA_PAGE['2021'])
- if self.not_before.year <= 2020:
- pages_to_parse.append(self.__RAA_PAGE['2020'])
- if self.not_before.year <= 2019:
- pages_to_parse.append(self.__RAA_PAGE['2019'])
-
- sub_pages_to_parse = [self.__RAA_PAGE['default']]
-
+ year_pages_to_parse = []
+
+ # On détermine quelles pages d'année parser
+ page_content = self.get_page(self.__RAA_PAGE, 'get').content
+ year_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
+ )
+ for year_page in year_pages:
+ if int(year_page['name'].replace('Année ', '').strip()) >= self.not_before.year:
+ year_pages_to_parse.append(year_page['url'])
+
+ month_pages_to_parse = []
# Pour chaque année, on cherche les sous-pages de mois
- for raa_page in pages_to_parse:
- page_content = self.get_page(raa_page, 'get').content
+ for year_page in year_pages_to_parse:
+ page_content = self.get_page(year_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]
+ for month_page in month_pages:
+ # On filtre les mois ne correspondant pas à la période analysée
+ guessed_date = Attrap.guess_date(month_page['name'], '(.*)')
+ if guessed_date.replace(day=1) >= self.not_before.replace(day=1):
+ month_pages_to_parse.append(month_page['url'])
- # On regarde aussi si sur la page de l'année il n'y aurait pas un
- # RAA mal catégorisé
- for page_to_parse in self.find_raa_card(raa_page):
- sub_pages_to_parse.append(page_to_parse)
+ pages_to_parse = []
+ # Pour chaque page de mois, on cherche les pages de RAA
+ for month_page in month_pages_to_parse:
+ # TODO : il reste à gérer le cas où une page de mois redirige vers un RAA (cela
+ # arrive quand la préfecture n'a publié qu'un seul RAA pendant le mois)
+ pages = self.get_sub_pages_with_pager(
+ month_page,
+ 'div.fr-card.fr-card--horizontal.fr-card--sm.fr-enlarge-link.fr-mb-3w div.fr-card__body div.fr-card__content h2.fr-card__title a.fr-card__link',
+ 'nav.fr-pagination ul.fr-pagination__list li a.fr-pagination__link.fr-pagination__link--next.fr-pagination__link--lg-label',
+ 'div.fr-card.fr-card--horizontal.fr-card--sm.fr-enlarge-link.fr-mb-3w div.fr-card__body div.fr-card__content div.fr-card__end p.fr-card__detail',
+ self.__HOST
+ )[::-1]
+ for page in pages:
+ guessed_date = datetime.datetime.strptime(page['details'].replace('Publié le ', '').strip(), '%d/%m/%Y')
+ if guessed_date.replace(day=1) >= self.not_before.replace(day=1):
+ pages_to_parse.append(page['url'])
- # Pour chaque mois, on cherche les pages des RAA
- for month_page in month_pages:
- year = Attrap.guess_date(month_page['name'], '(.*)').year
- for page_to_parse in self.find_raa_card(month_page['url'], year):
- sub_pages_to_parse.append(page_to_parse)
- # On ajoute aussi la page des mois à parser au cas où il y ait
- # eu une redirection vers un RAA
- sub_pages_to_parse.append(month_page['url'])
+ # On ajoute également la page racine, qui peut contenir des RAA mal catégorisés
+ pages_to_parse.append(self.__RAA_PAGE)
- # On parse les pages contenant des RAA
elements = []
- for page in sub_pages_to_parse:
+ # On parse les pages contenant des RAA
+ for page in pages_to_parse:
page_content = self.get_page(page, 'get').content
for element in self.get_raa_elements(page_content):
elements.append(element)
+ # On parse les RAA
self.parse_raa(elements, keywords)
self.mailer()
- def find_raa_card(self, page, year=None):
- pages = []
- card_pages = self.get_sub_pages_with_pager(
- page,
- 'div.fr-card__body div.fr-card__content h2.fr-card__title a.fr-card__link',
- 'ul.fr-pagination__list li a.fr-pagination__link.fr-pagination__link--next',
- 'div.fr-card__body div.fr-card__content div.fr-card__end p.fr-card__detail',
- self.__HOST
- )[::-1]
- for card_page in card_pages:
- # On filtre les pages de RAA ne correspondant pas à la période analysée
- guessed_date = datetime.datetime.strptime(card_page['details'].replace('Publié le ', '').strip(), '%d/%m/%Y')
- if guessed_date >= self.not_before:
- pages.append(card_page['url'])
- return pages
-
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-downloads-group.fr-downloads-group--bordered ul li a'):
+ for a in soup.select('div.fr-grid-row div.fr-downloads-group.fr-downloads-group--bordered ul li a'):
if a.get('href') and a['href'].endswith('.pdf'):
if a['href'].startswith('/'):
url = f"{self.__HOST}{a['href']}"
--
GitLab