main.py

import os
import argparse
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import seaborn as sns
from tabulate import tabulate
from PIL import Image, ExifTags
from pillow_heif import register_heif_opener
from progressbar import ProgressBar, Percentage, Bar, ETA


# set dimensiouns of visualizations
WIDTH = 10
HEIGHT = 5

TIMEZONE = 'Europe/Berlin'

# file extentions to ignore
ignore_ext = ('.md', '.mov', '.mp4', '.cr2', '.raf')

# returns specified EXIF tag value for image
def get_field (exif, field):
    for (k,v) in exif.items():
        if ExifTags.TAGS.get(k) == field:
            return v


def create_df(folder):

    # creates pandas DatFrame with relevant columns
    data_frame = pd.DataFrame(columns=('photo', 'capture_date', 'camera'))
    
    # total number of photos to scan
    total_photos = 0
    for root, dirs, files in os.walk(folder):
        total_photos += len(files)

    # loop through all directories and subdirectories
    i = 0
    pbar = ProgressBar(widgets=["Reading photos", Percentage(), Bar(), ETA()], maxval=total_photos).start()
    for subdir, dirs, files in os.walk(folder):
        for filename in files:
            # ignores hidden files and files in ignore list
            if not filename.startswith('.') and not filename.lower().endswith(ignore_ext):
                f = os.path.join(subdir, filename)
                if os.path.isfile(f):
                    img = Image.open(f)
                    img.verify()
                    # gets raw EXIF data for image
                    img_exif = img.getexif()

                    # if image has EXIF data it saves specified tag values in dataframe
                    if img_exif is None:
                        print("no EXIF data for " + filename)
                    else:
                        capture_date = pd.to_datetime(get_field(img_exif, 'DateTime'), format='%Y:%m:%d %H:%M:%S')
                        camera = get_field(img_exif, 'Model')
                        data_frame.loc[i] = [filename, capture_date, camera]
                        i +=1
                        pbar.update(i)
    pbar.finish()
    return data_frame



def show_cam(data_frame):
    # adds new column with number of pictures taken for each camera
    data_frame = data_frame.groupby(['camera']).size().sort_values(ascending=False)

    fig_cam, axs_cam = plt.subplots(figsize=[WIDTH,HEIGHT])
    cam = data_frame.plot.bar(ax=axs_cam, title="Photos by Camera", rot=0)
    cam.bar_label(cam.containers[0])

def show_date(data_frame):

    data_frame['date'] = data_frame['capture_date'].dt.date

    data_frame = data_frame.groupby(['date', 'camera']).size().unstack()

    # fill empty dates with 0
    min_date = pd.to_datetime(data_frame.index).min()
    max_date = pd.to_datetime(data_frame.index).max()
    date_range = pd.date_range(min_date, max_date)
    data_frame = data_frame.reindex(date_range).fillna(0)
    
    plt_date, axs_date = plt.subplots(figsize=[WIDTH, HEIGHT])
    data_frame.plot(kind='line', ax=axs_date)

# instantiates argument parser
parser = argparse.ArgumentParser()
parser.add_argument("folder_path", help="path to photos directory which should be scanned")
parser.add_argument("--cam", help="output number of photos by camera", action="store_true")
parser.add_argument("--date", help="output photos taken by camera per day", action="store_true")
args = parser.parse_args()

# function to open HEIF files
register_heif_opener()

# creates pandas DataFrame
df = create_df(args.folder_path)

if args.cam:
    show_cam(df)

if args.date:
    show_date(df)


print(df.info(memory_usage='deep'))

plt.show()