Finds YouTube videos you're interested in. PyCon US talks finder example.

youtube_find_relevant.py

import re
import requests
import html

API_KEY = ''
'''Google API (YouTube Data API v3) key from https://console.developers.google.com/apis/.'''

# Put titles you're interested into RELEVANT string:
# one title per line. These are regular expressions.
# NOTE: escape (), [], etc.
RELEVANT = '''

'''

CHANNEL = 'UCxs2IIVXaEHHA4BtTiWZ2mQ'
'''YouTube channel ID here.'''

TITLE_POSTFIX = 'PyCon 2019'
'''Postfix to strip from titles.'''

#############################################

_SPACES = re.compile('\s+')
_BASE_URL = 'https://www.googleapis.com/youtube/v3/search?order=date&part=snippet&channelId=%(channel)s&maxResults=50&key=%(key)s%(page)s'


def traverse(page=0):
    params = {
        'channel': CHANNEL,
        'key': API_KEY,
        'page': '',
    }

    if page:
        params['page'] = '&pageToken=%s' % page

    url = _BASE_URL % params

    response = requests.get(url)
    json = response.json()

    error = json.get('error')
    if error:
        raise Exception(error['message'])

    next_page = json.get('nextPageToken')

    for item in json['items']:

        if item['id']['kind'] != 'youtube#video':
            continue

        video_id = item['id']['videoId']
        title = html.unescape(item['snippet']['title'])

        title = title.replace(TITLE_POSTFIX, '').strip(' -')
        split = title.split(' - ', 1)

        prefix = split[0]
        prefix = prefix.replace('/', ',')

        if len(prefix.split(' ')) in {2, 3} or (',' in prefix):
            # strip person name
            try:
                title = split[1]
            except IndexError:
                pass

        title = _SPACES.sub(' ', title)
        title = title.strip(' -')

        yield video_id, title

    if next_page:
        yield from traverse(page=next_page)


def find_relevant():
    print('Channel: https://www.youtube.com/channel/%s/videos?view=0&sort=dd&flow=list\n' % CHANNEL)

    relevant_lines = []

    for line in RELEVANT.splitlines():
        line = line.strip()
        if line:
            line = _SPACES.sub(' ', line)
            relevant_lines.append(line)

    total_relevant = len(relevant_lines)
    total_traversed = 0
    seen = []

    traversed = [item for item in traverse()][::-1]  # eldest first
    len_traversed = len(str(len(traversed)))

    for video_id, title in traversed:
        total_traversed += 1

        matched = False
        for line in relevant_lines:
            matched = re.match(line, title)

            if matched:
                relevant_lines.remove(line)

        url = 'https://youtu.be/%s' % video_id
        seen.append((title, url, bool(matched)))

    for idx, seen_item in enumerate(sorted(seen), 1):
        title, url, matched = seen_item
        print('%s. %s%s' % (str(idx).zfill(len_traversed), '[!] ' if matched else '', title))

        if url:
            print('%s%s' % (' ' * (len_traversed + 2), url))

    total_missing = len(relevant_lines)

    print(
        '\nSummary: among %s found %s of %s, missing %s\n' % (
            total_traversed,
            total_relevant - total_missing,
            total_relevant,
            total_missing
        ))

    print('====' * 20)
    print('Missing:\n')

    for idx, line in enumerate(sorted(relevant_lines), 1):
        print('%s. %s' % (idx, line))


find_relevant()