fblanke/tvstud-download-digital
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Add options for dry run and grouping

Browse Source
This commit is contained in:
Felix Blanke 2023-08-17 22:21:31 +02:00
parent f34e5faecc
commit 0d9774ac00
1 changed files with 13 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
download_digital.py
View File

@ -74,7 +74,10 @@ bundesland_dict = {
def main( def main(
url: str = "https://beschaeftigtenbefragung.verdi.de/", tag: str = "bez_data_2" url: str = "https://beschaeftigtenbefragung.verdi.de/",
tag: str = "bez_data_2",
dry_run: bool = False,
grouped: bool = False,
) -> pd.DataFrame: ) -> pd.DataFrame:
options = Options() options = Options()
options.add_argument("--headless") options.add_argument("--headless")
@ -107,10 +110,16 @@ def main(
data["Digitale Befragung"] = pd.Series(tot_col_data, index=tot_col_index) data["Digitale Befragung"] = pd.Series(tot_col_data, index=tot_col_index)
df = pd.DataFrame(data=data) df = pd.DataFrame(data=data)
filename = f"data/{datetime.today().strftime('%Y-%m-%d')}_data.ods" filename = f"data/{datetime.today().strftime('%Y-%m-%d')}_data.ods"
if Path(filename).exists(): if grouped:
print("File already exists!") df = df.groupby("Bundesland")[["Digitale Befragung"]].sum()
if dry_run:
df.loc["Total"] = df.sum(numeric_only=True)
print(df)
else: else:
df.to_excel(filename, sheet_name="digital") if Path(filename).exists():
print("File already exists!")
else:
df.to_excel(filename, sheet_name="digital")
if __name__ == "__main__": if __name__ == "__main__":