Files
sisvietnamvn_01/sisvietnamvn_main/parse_pdf.py
T

14 lines
490 B
Python

import pdfplumber
import sys
filename = "/home/x79/sisvietnamvn_01/sisvietnamvn_Trang chính thức hiện tại/DaoTao/2.-CHUONG-TRINH-CAN-THIEP-MACH-MAU-THAN-KINH-NANG-CAO-BO-SUNG-FINAL.pdf"
with pdfplumber.open(filename) as pdf:
text = ""
for i, page in enumerate(pdf.pages):
text += f"--- PAGE {i+1} ---\n"
text += page.extract_text() + "\n"
with open('pdf1_output.txt', 'w', encoding='utf-8') as f:
f.write(text)
print("Saved to pdf1_output.txt")