feat: ✨ Add multi-threading to improve the process speed

main.py

@@ -1,19 +1,45 @@
 import string
-from urllib.request import urlopen
+from urllib.request import urlopen, urlretrieve
-
 from PIL import Image, ImageDraw, ImageFont
 from sys import argv
 import random
 import requests
 import os
+from multiprocessing import Pool
+
+
+CACHE_FOLDER = "cache/fonts"
+
+# Créer le dossier de cache s'il n'existe pas
+if not os.path.exists(CACHE_FOLDER):
+    os.makedirs(CACHE_FOLDER, exist_ok=True)
+
+
+def get_font_path(font_url):
+    """
+    Check if the font is already cached locally. If not, download it.
+    :param font_url: URL of the TTF font file.
+    :return: Path to the locally cached TTF file.
+    """
+    font_name = font_url.split("/")[-1]  # Extract the font filename from the URL
+    local_path = os.path.join(CACHE_FOLDER, font_name)
+
+    if not os.path.exists(local_path):
+        print(f"Downloading font: {font_name}")
+        urlretrieve(font_url, local_path)  # Download and save the file locally
+    else:
+        print(f"Using cached font: {font_name}")
+
+    return local_path
+
 
 def getRandomFont(apiKey: str, usedFont: list = None) -> str:
     """
     Function to fetch a random font from a list of available fonts
     provided by Google Font
     :param apiKey: API Key used to connect to Google Fonts.
-    :param usedFont: List of already used fonts. Set to none by default
+    :param usedFont: List of already used fonts. Set to None by default
-    :return: Link to the ttf file
+    :return: Local path to the cached TTF file.
     """
     if usedFont is None:
         usedFont = []
@@ -26,7 +52,7 @@ def getRandomFont(apiKey: str, usedFont: list=None) -> str:
     if "items" not in fonts_data:
         raise Exception("No fonts found in the API response.")
     available_fonts = fonts_data["items"]
-    unused_fonts = [font for font in available_fonts if usedFont is None or font["family"] not in usedFont]
+    unused_fonts = [font for font in available_fonts if font["family"] not in usedFont]
 
     if not unused_fonts:
         raise Exception("No unused fonts available to select from.")
@@ -41,15 +67,15 @@ def getRandomFont(apiKey: str, usedFont: list=None) -> str:
         raise Exception(f"No TTF link found for the font: {chosen_font['family']}")
 
     print(f"Selected font: {chosen_font['family']}")
-    return ttf_link
+    return get_font_path(ttf_link)
 
 
-def create_letter_image(letter, output_path, font):
+def create_letter_image(letter, output_path, font_path):
     """
     Function to create an image for a letter passed as an argument
     :param letter: the letter
     :param output_path: the path to save the image
-    :param font: font to be used (should be an online Font URL)
+    :param font_path: local path to the TTF font file
     :return: print a message and save the image
     """
     width, height = 13, 13
@@ -59,9 +85,9 @@ def create_letter_image(letter, output_path, font):
     draw = ImageDraw.Draw(image)
 
     try:
-        font = ImageFont.truetype(urlopen(font), font_size)
+        font = ImageFont.truetype(font_path, font_size)
     except IOError:
-        raise Exception("Cannot open font URL")
+        raise Exception(f"Cannot load font file: {font_path}")
 
     bbox = draw.textbbox((0, 0), letter, font=font)
     text_width, text_height = bbox[2] - bbox[0], bbox[3] - bbox[1]
@@ -74,29 +100,41 @@ def create_letter_image(letter, output_path, font):
     image.save(output_path)
     print(f"Image saved at {output_path}")
 
-def createImageForEachLetter(output_folder, font, index=0):
+
+def createImageForEachLetter(args):
     """
     Function to create an image for each of the letters passed as an argument
-    :param output_folder: the folder where to save the image
+    :param args: Tuple containing output folder, font path, and index
-    :param font: Google Font URL
-    :param index: the current index of the iteration (default to 0 if there is no iteration)
-    :return: Save the image into the correct subfolder
     """
-    alphabet = string.ascii_letters
+    output_folder, font_path, index = args
-    if not os.path.exists(output_folder):
+    alphabet = string.ascii_letters  # Inclut à la fois les lettres minuscules et majuscules
-        os.makedirs(output_folder)
-    for element in alphabet:
-        if not os.path.exists(os.path.join(output_folder, element)):
-            os.makedirs(os.path.join(output_folder, element))
-        create_letter_image(element, f"{output_folder}/{element}/{element}-{index}.png", font)
 
+    os.makedirs(output_folder, exist_ok=True)  # Ensure base output folder exists
+
+    for element in alphabet:
+        letter_folder = os.path.join(output_folder, element)
+        os.makedirs(letter_folder, exist_ok=True)
+
+        create_letter_image(element, f"{letter_folder}/{element}-{index}.png", font_path)
 
 if __name__ == "__main__":
     if len(argv) < 3:
-        raise Exception("Usage: " + argv[0] + " <number-of-letter> <api-key>")
+        raise Exception("Usage: " + argv[0] + " <number-of-iterations> <api-key>")
     if input(f"This script will generate {argv[1]} images for each letter in the alphabet. Continue? Y/n ") != "Y":
         print("Exiting...")
         exit(0)
+
+    # Configuration
+    num_iterations = int(argv[1])
+    api_key = argv[2]
     usedFont = []
-    for i in range(int(argv[1])):
+
-        createImageForEachLetter("dataset", getRandomFont(argv[2], usedFont), i)
+    # Fetch fonts and cache them
+    fonts = [getRandomFont(api_key, usedFont) for _ in range(num_iterations)]
+
+    # Prepare arguments for multiprocessing
+    args = [("dataset", fonts[i], i) for i in range(num_iterations)]
+
+    # Use multiprocessing pool
+    with Pool() as pool:
+        pool.map(createImageForEachLetter, args)