shazeghi/Report-Parser

1import os2import PyPDF23from pathlib import Path4from pdf2image import convert_from_path5 6def get_pg_x(infile, outfile, x=0):7    reader = PyPDF2.PdfReader(infile)8    first_pg = reader.pages[x]9    writer = PyPDF2.PdfWriter(outfile)10    writer.add_page(first_pg)11    writer.write(outfile)12 13 14 15def get_mult_pgs(infile, out):16    reader = PyPDF2.PdfReader(infile)17    pg_ind = list(range(99, 123,2)) # Basically get every other page (a very naive way of splitting it up)18    pages = [reader.pages[i] for i in pg_ind]19 20    if not os.path.exists(out):21        os.makedirs(out)22    for i,p in enumerate(pages):23        writer = PyPDF2.PdfWriter()24        filepath = os.path.join(out, f'case_{i+5}.pdf')25        writer.add_page(p)26        writer.write(filepath)27 28 29def splitall(infile, out):30    path = Path(out)31    if not path.exists():32        path.mkdir()33    reader = PyPDF2.PdfReader(infile)34    35    for i in range(len(reader.pages)):36        writer = PyPDF2.PdfWriter()37        if i%2 == 1:38            dest = (path/'other')39            dest.mkdir(exist_ok=True)40            fpath = dest.joinpath(f'other{i//2}.pdf')41            writer.add_page(reader.pages[i])42            writer.write(fpath)43        else:44            dest = (path/'cases')45            dest.mkdir(exist_ok=True)46            fpath = dest.joinpath(f'case{i//2}.pdf')47            writer.add_page(reader.pages[i])48            writer.write(fpath)49 50def folder_to_img(folder):51    for root, _, filelist in os.walk(folder):52        for file in filelist:53            if file.lower().endswith('.pdf'):54                pdf_path = os.path.join(root, file)55                outfolder = root56                try:57                    im = convert_from_path(pdf_path)58                    outfile = os.path.join(outfolder,f'{os.path.splitext(os.path.basename(pdf_path))[0]}.jpg')59                    im[0].save(outfile, 'JPEG')60                except Exception as e:61                    print(f"Error converting {pdf_path}: {e}")62                Path(pdf_path).unlink()63 64 65 66if __name__=='__main__':67    # splitall('report_batch.pdf', 'data')68    folder_to_img('data/cases')69    folder_to_img('data/other')70 71 72 73 74    75