Script to convert unsearchable pdf to a searchable one using tesseract (I should be studying instead of making this)

pdf_to_searchabled_pdf.py

#TESSERACT NEEDS TO BE INSTALLED FOR THIS TO WORK
# %%
#Imports
import os
import fitz
import pytesseract

# %%
#Create dirs if necessary
IMG_OUT_FOLDER = "tmp/imgs/"
PDF_OUT_FOLDER = "tmp/pdfs/"
os.makedirs(IMG_OUT_FOLDER, exist_ok=True)
os.makedirs(PDF_OUT_FOLDER, exist_ok=True)


# %%
#Convert pdf to images
PDF_FILE = "7-Infrastructure d'éxécution"
doc = fitz.open(f"{PDF_FILE}.pdf")

NUMBER_OF_PAGES = len(doc)

for i in range(NUMBER_OF_PAGES):
    page = doc.loadPage(i)  # number of page
    pix = page.getPixmap()
    output = f"{IMG_OUT_FOLDER}img{i}.png"
    pix.writePNG(output)


# %%
#Convert images to searchable pdf
pytesseract.pytesseract.tesseract_cmd = r'C:/Program Files/Tesseract-OCR/tesseract'
for i in range(NUMBER_OF_PAGES):
    pdf = pytesseract.image_to_pdf_or_hocr(f'{IMG_OUT_FOLDER}img{i}.png', extension='pdf')
    with open(f'{PDF_OUT_FOLDER}pdf{i}.pdf', 'w+b') as f:
        f.write(pdf) # pdf type is bytes by default


# %%
#Merges pdfs generated by Tesseract and write final pdf
if NUMBER_OF_PAGES > 1:
    doc = fitz.open(f"{PDF_OUT_FOLDER}pdf0.pdf")

    for i in range(NUMBER_OF_PAGES-1):
        doc.insertPDF(fitz.open(f"{PDF_OUT_FOLDER}pdf{i}.pdf"))
    
    doc.save(f"{PDF_FILE}_converted.pdf")