Tumblr audio posts crawler using Tumblr API v1 & Python 3 - API docs at https://www.tumblr.com/docs/en/api/v1#api_read

crawler.py

#!/usr/bin/env python3

import requests, json, sqlite3

from xml.etree import ElementTree as ET
from urllib.parse import urlparse, parse_qs

blog_domain = 'tracks.tumblr.com' # change this
post_type = 'audio' # leave empty for all post types
posts_per_request = 50 # max is 50

session = requests.session()

db_filename = '%s.db' % blog_domain
db_connection = sqlite3.connect(db_filename)
db_cursor = db_connection.cursor()

db_cursor.execute('''CREATE TABLE IF NOT EXISTS posts
             (id integer primary key, title text, artist text, album text, avatar text, file_url text)''')

start = 0

while True:
	print('Starting at %d, requesting %d posts' % (start, posts_per_request))

	request_url = 'https://%s/api/read/json?num=%d&type=%s&start=%d' % \
		(blog_domain, posts_per_request, post_type, start)

	request = session.get(request_url)
	response = request.text.replace('var tumblr_api_read = ', '')[:-2]
	json_response = json.loads(response)

	posts = json_response['posts']

	if len(posts) == 0:
		break

	print('Found %d posts' % len(posts))

	for post in posts:
		
		embed_code = post['audio-embed'] if 'audio-embed' in post else ''

		iframe = ET.fromstring(embed_code)
		iframe_src = iframe.attrib.get('src')

		iframe_src_url = urlparse(iframe_src)
		iframe_src_query = parse_qs(iframe_src_url.query)

		file_url = iframe_src_query['audio_file'][0] if 'audio_file' in iframe_src_query else iframe_src

		post_data = {
			'id': int(post['id']),
			'title': post['id3-title'] if 'id3-title' in post else '',
			'artist': post['id3-artist'] if 'id3-artist' in post else '',
			'album': post['id3-album'] if 'id3-album' in post else '',
			'avatar': post['tumblelog']['avatar_url_512'] if 'avatar_url_512' in post['tumblelog'] else '',
			'file_url': file_url
		}

		print(post_data['id'])

		db_cursor.execute("INSERT OR REPLACE INTO posts VALUES (:id,:title,:artist,:album,:avatar,:file_url)", post_data)

	start += posts_per_request
	db_connection.commit()

print('All done.')
print('Saved to %s' % db_filename)
db_connection.close()