73 lines
3.2 KiB
Python
73 lines
3.2 KiB
Python
from playwright.sync_api import sync_playwright
|
||
import json, time, re
|
||
|
||
def run():
|
||
with sync_playwright() as p:
|
||
try:
|
||
browser = p.chromium.connect_over_cdp('http://127.0.0.1:9222')
|
||
page = browser.contexts[0].pages[0]
|
||
|
||
page.goto('https://my.dom.gosuslugi.ru/citizen-cabinet/#!/appeals/applicant/search')
|
||
page.wait_for_load_state('networkidle')
|
||
time.sleep(10)
|
||
|
||
# Нажимаем 'Найти' для сброса фильтров
|
||
print('Нажимаем Найти...')
|
||
page.locator('button:has-text("Найти")').click()
|
||
time.sleep(10)
|
||
|
||
# Пытаемся выставить 'По 100'
|
||
try:
|
||
page.evaluate('''() => {
|
||
const el = Array.from(document.querySelectorAll('span, div')).find(e => e.innerText === '10');
|
||
if (el) el.click();
|
||
}''')
|
||
time.sleep(2)
|
||
page.locator('div.ui-select-choices-row:has-text("100")').click()
|
||
print('Выбрано по 100.')
|
||
time.sleep(10)
|
||
except: pass
|
||
|
||
full_queue = []
|
||
page_num = 1
|
||
|
||
while True:
|
||
# Очистка оверлеев
|
||
page.evaluate('document.querySelectorAll(".p-dialog-mask, .modal-backdrop").forEach(el => el.remove())')
|
||
|
||
rows = page.locator('tr').all()
|
||
added_on_page = 0
|
||
for row in rows:
|
||
text = row.inner_text()
|
||
if '73-20' in text:
|
||
link = row.locator('a[href*="appeals/view/"]').first
|
||
num_match = re.search(r'73-202\d-\d+', text)
|
||
num = num_match.group(0) if num_match else f"unknown_{time.time()}"
|
||
href = link.get_attribute('href')
|
||
if href and num not in [x['number'] for x in full_queue]:
|
||
full_queue.append({'number': num, 'href': href})
|
||
added_on_page += 1
|
||
|
||
print(f'Страница {page_num}: собрано {added_on_page} новых.')
|
||
|
||
# Ищем стрелку вправо или слово 'следующая'
|
||
next_btn = page.locator('a:has-text("следующая"), .pagination-next a').first
|
||
if next_btn.is_visible() and next_btn.is_enabled():
|
||
page.evaluate('el => el.click()', next_btn.element_handle())
|
||
page.wait_for_load_state('networkidle')
|
||
time.sleep(10)
|
||
page_num += 1
|
||
else:
|
||
break
|
||
|
||
output_path = '/app/services/gis_harvester/data/full_queue.json'
|
||
with open(output_path, 'w', encoding='utf-8') as f:
|
||
json.dump(full_queue, f, ensure_ascii=False, indent=2)
|
||
|
||
print(f'ИТОГО СОБРАНО: {len(full_queue)} целей.')
|
||
browser.close()
|
||
except Exception as e:
|
||
print(f'ERROR: {e}')
|
||
|
||
if __name__ == "__main__":
|
||
run()
|