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

Feat: nothing

Browse Source
This commit is contained in:
Bertrand Benjamin 2025-03-01 13:55:45 +01:00
parent d0fc473134
commit 8df9be825f
2 changed files with 10 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
pdf_oralia/extract.py
View File

@ -13,6 +13,10 @@ extract_table_settings = {
} }
class ExtractError(Exception):
pass
def extract_date(page_text): def extract_date(page_text):
"""Extract date from a page """Extract date from a page
@ -123,12 +127,10 @@ def extract_save(pdf_file, dest, save=[]):
pdf_file = Path(pdf_file) pdf_file = Path(pdf_file)
xlss = extract_plan(pdf_file, dest) xlss = extract_plan(pdf_file, dest)
if save != []: dfs = from_pdf(pdf_file)
dfs = from_pdf(pdf_file)
for s in save: for s in save:
dfs[s].to_excel(xlss[s], sheet_name=s, index=False) dfs[s].to_excel(xlss[s], sheet_name=s, index=False)
logging.info(f"{xlss[s]} saved") logging.info(f"{xlss[s]} saved")
return {k: v for k, v in xlss.items() if k in save}
return xlss return {k: v for k, v in xlss.items() if k in save}

1
pdf_oralia/scripts.py
View File

@ -62,6 +62,7 @@ def on(pdf_file, dest, force, only_plan):
else: else:
dest = Path(dest) dest = Path(dest)
assert pdf_file.exists()
logging.info(f"Found {pdf_file}") logging.info(f"Found {pdf_file}")
plan_dest = extract_plan(pdf_file, dest) plan_dest = extract_plan(pdf_file, dest)