Download all images on danbooru by tag

danbooru_download.py

import sys, os, requests, json, time, math

if len(sys.argv) < 2:
    print("Tags not specified")
    sys.exit(1)
elif sys.argv[1].count(" ") > 1:
    print("Maximum of 2 arguments")
    sys.exit(1)

tags = sys.argv[1]
tags_url = tags.replace(" ", "+")

output_path = "./"

if len(sys.argv) < 3:
    output_path += "".join([c for c in sys.argv[1] if c.isalpha() or c.isdigit() or c == " " or c == "(" or c == ")" or c == "_"]).rstrip().replace(" ", "_")
else:
    output_path = sys.argv[2]

try:
    os.mkdir(output_path)
except FileExistsError:
    pass

print("Tags:\t\t{0}".format(tags))
print("Output:\t\t{0}".format(output_path))

post_count = json.loads(requests.get("https://danbooru.donmai.us/counts/posts.json?tags={0}".format(tags_url)).text)["counts"]["posts"]

print("# of posts:\t{0}".format(post_count))

page_count = math.ceil(post_count / 200)

count = 0
start = time.time()
for p in range(page_count):
    posts = json.loads(requests.get("https://danbooru.donmai.us/posts.json?tags={0}&page={1}&limit=200".format(tags_url, p + 1)).text)

    for post in posts:
        count += 1
        try:
            print("Downloading {0}, {1} of {2}, elapsed: {3} seconds".format(post["id"], count, post_count, round(time.time() - start, 2)))

            image_fname = post["file_url"][post["file_url"].rfind("/") + 1:]
            image = requests.get(post["file_url"]).content
        except KeyError:
            print("Skipping {0}, {1} of {2}, elapsed: {3} seconds".format(post["id"], count, post_count, round(time.time() - start, 2)))

        f = open("{0}/{1}".format(output_path, image_fname), "wb")
        f.write(image)
        f.close()
        
print("Downloaded {0} posts in {1} seconds".format(count, round(time.time() - start, 2)))

howto.md

python danbooru_download.py <tags> [output_path]

Tags: maximum of 2 tags Output path: defaults to a path based on the tags if not specified