import pandas as pd
from PIL import Image
from pytesseract import pytesseract


def convert_image_to_df(path_to_image: str) -> pd.DataFrame:
    """
    Function takes images and converts text/numbers into usable dataframe
    params:
        - path_to_image (str) - location of the image file to process
    returns:
        -  dataframe (pandas.DataFrame) - Dataframe containing converted text
    """

    # Define path to tessaract executable
    pytesseract.tesseract_cmd = r"/opt/homebrew/Cellar/tesseract/5.2.0/bin/tesseract"

    # Open image with PIL
    img = Image.open(path_to_image)
    # Extract text from image
    text = pytesseract.image_to_string(img)

    # we could also convert image directly to dataframe but this didn't seem to work as expected
    # dataframe = pytesseract.image_to_data(img, lang='eng', output_type='data.frame')
    # print(dataframe)

    id_list = []
    text_list = []

    chunks = text.split("\n")
    for c in chunks[1:]:
        if len(c)>0:
            id_list.append(c.split(" ", 1)[0])
            text_list.append(c.split(" ", 1)[1])

    dataframe = pd.DataFrame(
        list(zip(id_list, text_list)),
        columns=["id", "text"],
    )
    dataframe["id"] = dataframe["id"].astype(str).astype(int)

    return dataframe


if __name__ == "__main__":
    df = convert_image_to_df('image/image.png')