lafrite/pdf_auralia
1
0
Fork 0
Code Issues 1 Pull Requests Actions Packages Projects 1 Releases Wiki Activity

Feat: make from_pdf importable and move plumber in it

Browse Source
This commit is contained in:
Bertrand Benjamin 2024-10-16 06:47:25 +02:00
parent 0fc39ed317
commit ab2fdb0541
2 changed files with 4 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
pdf_oralia/__init__.py
View File

@ -0,0 +1 @@
from .extract import from_pdf

6
pdf_oralia/extract.py
View File

@ -38,8 +38,9 @@ def catch_malformed_table(tables):
return tables[0] return tables[0]
def from_pdf(pdf): def from_pdf(pdf_file):
"""Build dataframes one about charges and another on loc""" """Build dataframes one about charges and another on loc"""
pdf = pdfplumber.open(pdf_file)
recapitulatif_tables = [] recapitulatif_tables = []
loc_tables = [] loc_tables = []
charge_tables = [] charge_tables = []
@ -90,8 +91,7 @@ def extract_save(pdf_file, dest):
xls_charge = Path(dest) / f"{pdf_file.stem.replace(' ', '_')}_charge.xlsx" xls_charge = Path(dest) / f"{pdf_file.stem.replace(' ', '_')}_charge.xlsx"
xls_locataire = Path(dest) / f"{pdf_file.stem.replace(' ', '_')}_locataire.xlsx" xls_locataire = Path(dest) / f"{pdf_file.stem.replace(' ', '_')}_locataire.xlsx"
pdf = pdfplumber.open(pdf_file) df_charge, df_loc = from_pdf(pdf_file)
df_charge, df_loc = from_pdf(pdf)
df_charge.to_excel(xls_charge, sheet_name="Charges", index=False) df_charge.to_excel(xls_charge, sheet_name="Charges", index=False)
logging.info(f"{xls_charge} saved") logging.info(f"{xls_charge} saved")