Skip to content

Instantly share code, notes, and snippets.

@ricardodani
Created September 5, 2024 05:53
Show Gist options
  • Select an option

  • Save ricardodani/8e26d116ba79dced5da27b5e06a40201 to your computer and use it in GitHub Desktop.

Select an option

Save ricardodani/8e26d116ba79dced5da27b5e06a40201 to your computer and use it in GitHub Desktop.
salvar_tabela_brasileirao_excel.py
"""
Required libraries:
- selenium
- pandas
- openpyxl
"""
from selenium import webdriver
from selenium.webdriver.common.by import By
from datetime import date
import pandas
def get_tabela_brasileirao():
URL_TABELA = 'https://ge.globo.com/futebol/brasileirao-serie-a/'
navegador = webdriver.Chrome()
navegador.get(URL_TABELA)
equipes_nome = navegador.find_element(
By.CLASS_NAME, "tabela__equipes"
).find_elements(
By.CLASS_NAME, "classificacao__equipes--nome"
)
equipes_pontos = navegador.find_element(
By.CLASS_NAME, "tabela__pontos"
).find_elements(
By.CLASS_NAME, "classificacao__tabela--linha"
)
tabela_montada = []
for pos in range(len(equipes_nome)):
p, j, v, e, d, gp, gc, sg, per = equipes_pontos[pos].text.split()
tabela_montada.append(
[
pos + 1,
equipes_nome[pos].text,
p, j, v, e, d, gp, gc, sg, per,
]
)
navegador.close()
return (
["#", "Time", "P", "J", "V", "E", "D", "GP", "GC", "SG", "%"],
tabela_montada
)
def salvar_excel():
colunas, tabela = get_tabela_brasileirao()
nome_arquivo_saida = f'brasileirao-{date.today()}.xlsx'
data_frame = pandas.DataFrame(tabela, columns=colunas)
data_frame.to_excel(nome_arquivo_saida, index=False)
print(f"Salvo em {nome_arquivo_saida}")
if __name__ == '__main__':
salvar_excel()
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment