126 lines
3.7 KiB
Python
126 lines
3.7 KiB
Python
#!/usr/bin/env python3
|
|
# -*- coding: utf-8 -*-
|
|
|
|
|
|
|
|
import fitz
|
|
import os
|
|
import re
|
|
import time
|
|
import io
|
|
import pytesseract
|
|
from pathlib import Path
|
|
from PIL import Image
|
|
from FunkceWhereIsDropbox import get_dropbox_path
|
|
|
|
# --- CONFIGURATION ---
|
|
#Get cesta to Dropbox on working computer
|
|
dropbox=get_dropbox_path()
|
|
BASE_DIR=dropbox/"Ordinace"/"Dokumentace_ke_zpracování"/"EKGforProcessing"
|
|
FLAG = "rotated-by-script"
|
|
|
|
|
|
# Point to your Tesseract executable if it's not in your PATH:
|
|
pytesseract.pytesseract.tesseract_cmd = r'C:\Program Files\Tesseract-OCR\tesseract.exe'
|
|
|
|
def ocr_page(page):
|
|
"""Perform OCR using Tesseract with high DPI for better accuracy."""
|
|
# Increasing DPI to 300 is crucial for small text on EKG strips
|
|
pix = page.get_pixmap(dpi=300)
|
|
img = Image.frombytes("RGB", [pix.width, pix.height], pix.samples)
|
|
|
|
# Using 'ces' for Czech language support
|
|
# PSM 6: Assume a single uniform block of text
|
|
text = pytesseract.image_to_string(img, lang='ces', config='--psm 6')
|
|
return text
|
|
|
|
|
|
def extract_rodne_cislo(text):
|
|
"""Extracts RC (6 digits + optional slash/spaces + 3-4 digits)."""
|
|
m = re.search(r"(\d{6})\s*/?\s*(\d{3,4})", text)
|
|
if not m:
|
|
return None
|
|
left = m.group(1)
|
|
right = m.group(2).zfill(4)
|
|
return left + right
|
|
|
|
|
|
def extract_date(text):
|
|
"""Extracts date with high tolerance for OCR spacing errors."""
|
|
# Matches D.M.YYYY, D. M. YYYY, DD.MM.YYYY, etc.
|
|
# It also handles if OCR accidentally puts a comma instead of a dot
|
|
m = re.search(r"(\d{1,2})[\.,]\s*(\d{1,2})[\.,]\s*(\d{4})", text)
|
|
if m:
|
|
return m.groups()
|
|
return None
|
|
|
|
|
|
def convert_date_to_iso(date_tuple):
|
|
"""Converts (D, M, Y) -> YYYY-MM-DD."""
|
|
d, m, y = date_tuple
|
|
return f"{y}-{m.zfill(2)}-{d.zfill(2)}"
|
|
|
|
|
|
# --- MAIN PROCESS ---
|
|
for pdf_path in BASE_DIR.glob("*.pdf"):
|
|
if ".tmp" in pdf_path.name: continue
|
|
|
|
print(f"\n{'=' * 60}")
|
|
print(f"PROCESSING: {pdf_path.name}")
|
|
print(f"{'=' * 60}")
|
|
|
|
doc = fitz.open(pdf_path)
|
|
meta = doc.metadata
|
|
keywords = meta.get("keywords", "") or ""
|
|
|
|
# 1. HANDLE ROTATION
|
|
if FLAG not in keywords:
|
|
print(" [Action] Rotating page...")
|
|
page = doc[0]
|
|
page.set_rotation((page.rotation + 90) % 360)
|
|
if doc.page_count > 1:
|
|
doc.delete_page(1)
|
|
meta["keywords"] = (keywords + " " + FLAG).strip()
|
|
doc.set_metadata(meta)
|
|
tmp = pdf_path.with_suffix(".tmp.pdf")
|
|
doc.save(tmp, deflate=True)
|
|
doc.close()
|
|
os.replace(tmp, pdf_path)
|
|
doc = fitz.open(pdf_path)
|
|
|
|
# 2. PERFORM OCR
|
|
raw_text = ocr_page(doc[0])
|
|
|
|
# --- DEBUG PRINTOUT ---
|
|
print("\n--- START OF EXTRACTED TEXT ---")
|
|
print(raw_text)
|
|
print("--- END OF EXTRACTED TEXT ---\n")
|
|
# ---------------------
|
|
|
|
rc = extract_rodne_cislo(raw_text)
|
|
date_tuple = extract_date(raw_text)
|
|
|
|
print(f"RESULT -> RC: {rc if rc else 'NOT FOUND'}")
|
|
print(f"RESULT -> Date: {date_tuple if date_tuple else 'NOT FOUND'}")
|
|
|
|
doc.close()
|
|
|
|
# 3. RENAME LOGIC
|
|
if rc and date_tuple:
|
|
date_iso = convert_date_to_iso(date_tuple)
|
|
new_name = f"{rc} {date_iso} [EKG] [bez hodnocení].pdf"
|
|
new_path = pdf_path.with_name(new_name)
|
|
|
|
if not new_path.exists():
|
|
try:
|
|
# Close the handle properly before renaming
|
|
pdf_path.rename(new_path)
|
|
print(f"✅ Success: Renamed to {new_name}")
|
|
except Exception as e:
|
|
print(f"❌ Rename error: {e}")
|
|
else:
|
|
print(f"⚠ Skipping: {new_name} already exists.")
|
|
else:
|
|
print("❌ Script could not find all data. Check the extracted text above.")
|
|
|
|
print("\nAll files processed.") |