83 lines
2.0 KiB
Python
83 lines
2.0 KiB
Python
import logging
|
|
from logging.config import dictConfig
|
|
from pathlib import Path
|
|
|
|
import click
|
|
import pandas as pd
|
|
|
|
from .extract import extract_save
|
|
|
|
logging_config = dict(
|
|
version=1,
|
|
formatters={"f": {"format": "%(levelname)-8s %(name)-12s %(message)s"}},
|
|
handlers={
|
|
"h": {
|
|
"class": "logging.StreamHandler",
|
|
"formatter": "f",
|
|
"level": logging.DEBUG,
|
|
}
|
|
},
|
|
root={
|
|
"handlers": ["h"],
|
|
"level": logging.DEBUG,
|
|
},
|
|
)
|
|
|
|
dictConfig(logging_config)
|
|
|
|
|
|
@click.group()
|
|
def main():
|
|
pass
|
|
|
|
|
|
@main.group()
|
|
def extract():
|
|
pass
|
|
|
|
|
|
@extract.command()
|
|
@click.argument("pdf_file", required=1)
|
|
@click.option("--dest", help="Où mettre les fichiers produits", default="")
|
|
def on(pdf_file, dest):
|
|
if not dest:
|
|
pdf_path = Path(pdf_file)
|
|
dest = pdf_path.parent
|
|
|
|
extract_save(pdf_file, dest)
|
|
|
|
|
|
@extract.command()
|
|
@click.option("--folder", help="Tous les fichiers dans folder", default="./")
|
|
@click.option("--dest", help="Où mettre les fichiers produits", default="./")
|
|
def all(folder, dest):
|
|
p = Path(folder)
|
|
|
|
d = Path(dest)
|
|
d.mkdir(exist_ok=True)
|
|
|
|
pdf_files = [x for x in p.iterdir() if ".pdf" in str(x)]
|
|
for pdf_file in pdf_files:
|
|
logging.info(f"Found {pdf_file}")
|
|
extract_save(pdf_file, d)
|
|
|
|
|
|
@main.command()
|
|
@click.option("--folder", help="Tous les fichiers dans folder", default="./")
|
|
@click.option("--dest", help="Où mettre les fichiers produits", default="")
|
|
def join(folder, dest):
|
|
p = Path(folder)
|
|
dfs = {
|
|
"charge": [],
|
|
"locataire": [],
|
|
}
|
|
for file in p.glob("*.xlsx"):
|
|
year, month, immeuble, table = file.stem.split("_")
|
|
df = pd.read_excel(file).assign(annee=year, mois=month, immeuble=immeuble)
|
|
dfs[table].append(df)
|
|
for tablename, datas in dfs.items():
|
|
df = pd.concat(datas)
|
|
destination = Path(dest) / f"{tablename}.xlsx"
|
|
df.to_excel(destination, index=False)
|
|
logging.info(f"{destination} written")
|