from playwright.sync_api import Playwright, sync_playwright
from bs4 import BeautifulSoup
import pandas as pd
def run(playwright: Playwright) -> None:
browser = playwright.firefox.launch(headless=True) #slow_mo=50
context = browser.new_context()
# Open new page
page = context.new_page()
# Go to https://portal.librus.pl/rodzina
page.goto("https://portal.librus.pl/rodzina")
# Click text=LIBRUS Synergia >> nth=0
page.locator("text=LIBRUS Synergia").first.click()
# Click nav >> text=Zaloguj >> nth=0
page.locator("nav >> text=Zaloguj").first.click()
page.wait_for_url("https://portal.librus.pl/rodzina/synergia/loguj")
# Click input[name="Login"]
page.frame_locator("#caLoginIframe").locator("input[name=\"Login\"]").click()
# Fill input[name="Login"]
page.frame_locator("#caLoginIframe").locator("input[name=\"Login\"]").fill("XXXXXXXXXX") ### login
# Click input[name="Pass"]
page.frame_locator("#caLoginIframe").locator("input[name=\"Pass\"]").click()
# Fill input[name="Pass"]
page.frame_locator("#caLoginIframe").locator("input[name=\"Pass\"]").fill("XXXXXXXXXXX") ### password
# # Click button:has-text("Zaloguj")
page.frame_locator("#caLoginIframe").locator("button:has-text(\"Zaloguj\")").click()
page.wait_for_url("https://synergia.librus.pl/rodzic/index")
page.is_visible('div.container-background')
# # Click #icon-oceny span
page.locator("#icon-wiadomosci span").click()
page.wait_for_url("https://synergia.librus.pl/wiadomosci")
# Select tbody
html_1 = page.inner_html('xpath=/html/body/div[3]/div[3]/form/div/div/table/tbody/tr/td[2]/table[2]/tbody')
soup = BeautifulSoup(html_1, 'html.parser')
list_temp = []
for tr in soup.find_all('tr'):
slownik = []
for td in tr.find_all('td'):
if '\n' not in td.text:
slownik.append(td.text)
else:
pass
list_temp.append(slownik)
mails = []
for i in list_temp:
slownik = {}
slownik = {
"Autor": i[0],
"Tytuł": i[1],
"Data": i[2]
}
mails.append(slownik)
df = pd.DataFrame(mails)
with open('df.json', 'w', encoding='utf-8') as file:
df.to_json(file, force_ascii=False)
# ---------------------
context.close()
browser.close()
with sync_playwright() as playwright:
run(playwright)