%%capture
%mkdir yearbook
%cd yearbook
!pip install --upgrade --no-cache-dir gdown
!gdown --id "1NHT8NN8ClBEnUC5VqkP3wr2KhyiIQzyU"
!unzip PHfiles.zip
%mkdir images
!pip install PyMuPDF
!pip install dlib
!pip install DeepFace
import os, shutil, fitz, cv2, numpy as np, pandas as pd, dlib, tensorflow as tf
from os.path import dirname, join
from deepface import DeepFace

path = r'./'
pdfs = [f for f in os.listdir(path) if f.endswith('.pdf')]
for pdf in pdfs:
    os.chdir(os.path.join('./images'))
    os.mkdir((pdf.split(".")[0]))
    newdir = (os.path.join('./images/' + os.path.join(pdf.split(".")[0])))
    os.chdir("..")
    print ("Now copying images into " + (newdir))
    shutil.copy(pdf, newdir)
    os.chdir(newdir)
    doc = fitz.open(pdf)
    for page in doc:
      pix = page.get_pixmap()
      pix.save("page-%i.png" % page.number)
    os.chdir(os.path.dirname(os.getcwd()))
    os.chdir("..")



path = r'./'

os.chdir(os.path.join(path + 'images'))
dirs = os.listdir(path)
for dir in dirs:
    os.chdir(os.path.join(path + dir))
    pngs = [f for f in os.listdir(path) if f.endswith('.png')]

    if not os.path.exists((dir) + ' faces'):
        print("New 'faces' directory created in " + (dir) + " folder")
        os.makedirs((dir) + ' faces')

        count = 0
        for png in pngs:
            image = cv2.imread(png)

            greyscale_image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)

            face_cascade = cv2.CascadeClassifier(cv2.data.haarcascades + "haarcascade_frontalface_default.xml")

            detected_faces = face_cascade.detectMultiScale(image=greyscale_image, scaleFactor=1.9, minNeighbors=4)

            count = 0
            for (x,y,w,h) in detected_faces:
                try:
                    xpadding = 20
                    ypadding = 40
                    crop_face = image[y-ypadding: y + h+ypadding, x-xpadding: x + w+xpadding]
                    count+=1
                    face = cv2.rectangle(crop_face,(x,y),(x+w,y+h),(255,0,0),2)
                    cv2.imwrite(path + (dir) + ' faces/' + str(count) + '_' + png, face)
                except (Exception):
                    print("An error happened")
                    continue
            os.remove(os.path.join(path, png))
    os.chdir("..")


%cd ..
number_smiles = 0
smile_counts = []
number_nonsmiles = 0
nonsmile_counts = []
num_errors = 0
error_counts = []
pngs = []

file_count = 0
file_count_list = []

years = ['1911', '1921', '1931', '1941', '1951', '1961']

for year in years:
    path = r'./images' + '/' + year
    for root, dirs, files in os.walk(path):
        for dir in dirs:
            path = path + '/' + (year + ' faces')
            if(file_count != 0):
                file_count_list.append(file_count)
            file_count = 0
            for f in os.listdir(path):
                if f.endswith('.png'):
                    pngs.append(path + '/' + f)
                    file_count = file_count + 1

file_count_list.append(file_count)

total_loops = 0
count = 0
iterator = 0
for png in pngs:
    try:
        total_loops = total_loops + 1
        count = count + 1
        if(count != (file_count_list[iterator] + 1)):
            demography = DeepFace.analyze(png, actions = ['emotion'])
            print(demography)
            if(demography[0]['dominant_emotion'] == 'happy'):
                number_smiles = number_smiles + 1
            else:
                number_nonsmiles = number_nonsmiles + 1

        else:
            count = count - 1
            smile_counts.append(number_smiles / count)
            nonsmile_counts.append(number_nonsmiles / count)
            error_counts.append(num_errors / count)
            number_smiles = 0
            number_nonsmiles = 0
            num_errors = 0
            iterator = iterator + 1
            count = 0

    except (Exception):
        num_errors = num_errors + 1
        print("An error happened")
        continue

smile_counts.append(number_smiles / count)
nonsmile_counts.append(number_nonsmiles / count)
error_counts.append(num_errors / count)

dict = {'Years': years, 'Smiles': smile_counts, 'Non-Smiles': nonsmile_counts, "Error Weight": error_counts}
data = pd.DataFrame(dict)
data.to_csv('YearbookOutput.csv', index=False)
print(count)


from google.colab import files
files.download('YearbookOutput.csv')