Python で画像のダウンロードを自動化する方法: 包括的なガイド

import os
import requests
from PIL import Image
from io import BytesIO
import random
import time
from datetime import datetime, timedelta

API_KEY = '...'  # Replace with your actual Pexels API key
BASE_URL = 'https://api.pexels.com/v1/search'
OUTPUT_DIR = 'output'
SEARCH_QUERY = 'Subaru Forester'
MIN_WIDTH = 1080
MIN_HEIGHT = 1920
NUM_IMAGES = 16
ORIENTATION = 'vertical'

if not os.path.exists(OUTPUT_DIR):
    os.makedirs(OUTPUT_DIR)

def get_images(query, page=1, per_page=15):
    params = {
        'query': query,
        'page': page,
        'per_page': per_page
    }
    response = requests.get(BASE_URL, headers=headers, params=params)
    response.raise_for_status()
    return response.json()

def download_and_save_image(url, filename):
    response = requests.get(url)
    response.raise_for_status()
    img = Image.open(BytesIO(response.content))
    if img.width >= MIN_WIDTH and img.height >= MIN_HEIGHT:
        if ORIENTATION == 'vertical' and (img.height / img.width > 1.4):
            img.save(filename)
            print(f'Saved {filename}')
            return 1
        elif ORIENTATION == 'horizontal' and (img.width / img.height > 1.4):
            img.save(filename)
            print(f'Saved {filename}')
            return 1
    return 0

def generate_random_date():
    end_date = datetime.now()
    start_date = end_date - timedelta(days=4*365)
    random_date = start_date + timedelta(days=random.randint(0, 4*365))
    return random_date

def main():
    images_downloaded = 0
    page = 1
    seen_images = set()
    target_date = generate_random_date()
    print(f"Target date for filtering: {target_date.strftime('%Y-%m-%d')}")


while images_downloaded < NUM_IMAGES:
        data = get_images(SEARCH_QUERY, page)
        photos = data.get('photos', [])
        if not photos:
            print('No more photos found.')
            break
        random.shuffle(photos)  # Shuffle to get a more random selection
        for photo in photos:
            if images_downloaded >= NUM_IMAGES:
                break
            image_url = photo['src']['original']
            photo_date = datetime.strptime(photo['created_at'], '%Y-%m-%dT%H:%M:%S%z') if 'created_at' in photo else datetime.now()
            if photo_date > target_date and image_url not in seen_images:
                filename = os.path.join(OUTPUT_DIR, f'image_{images_downloaded + 1}.jpg')
                if download_and_save_image(image_url, filename):
                    images_downloaded += 1
                    seen_images.add(image_url)
        page += 1
        time.sleep(random.uniform(1, 2))  # Sleep to avoid hitting API rate limits