ffmpeg-slide-gen/youtube-transcript.py at main · haridsv/ffmpeg-slide-gen · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
import sys

import click
from youtube_transcript_api import YouTubeTranscriptApi, FetchedTranscriptSnippet


# Return aggregated lines at the specified interval duration
def lines_by_dur(fetched_transcript, int_dur):
    cur_st = 0
    cur_text = ""
    next_st = int_dur
    line_iter = iter(fetched_transcript)
    while True:
        try:
            line_map = next(line_iter)
        except StopIteration:
            line_map = None
        if line_map and line_map.start < next_st:
            cur_st = cur_st or line_map.start
            cur_text = cur_text + " " + line_map.text
            continue
        elif cur_text:
            res_map = FetchedTranscriptSnippet(start=cur_st, text=cur_text, duration=0)
            yield res_map
        if not line_map:
            break
        else:
            cur_st = line_map.start
            cur_text = line_map.text
            next_st = cur_st + int_dur


def lines_by_words(fetched_transcript, word_count):
    cur_st = 0
    cur_text = ""
    cur_wc = 0
    line_iter = iter(fetched_transcript)
    while True:
        try:
            line_map = next(line_iter)
        except StopIteration:
            line_map = None
        new_wc = line_map and len(line_map.text.split()) or 0
        if line_map and (cur_wc + new_wc) < word_count:
            cur_st = cur_st or line_map.start
            cur_text = cur_text + " " + line_map.text
            cur_wc += new_wc
            continue
        if line_map and (cur_wc + new_wc) >= word_count:
            res_map = FetchedTranscriptSnippet(start=cur_st, text=cur_text, duration=0)
            yield res_map
        if not line_map:
            break
        else:
            cur_st = line_map.start
            cur_text = line_map.text
            cur_wc = len(cur_text.split())


@click.command()
@click.option("--interval-duration", "--id", "-d", default=60, help="Interval duration for generating timestamps, use 0 to disable and produce raw output")
@click.option("--word-count", "--wc", "-w", default=0, help="Number of words for generating timestamps, applicable when --interval-duration=0")
@click.option("--timestamp-in-seconds", "--tis", "-s", is_flag=True, default=False, help="Specify to see duration in seconds instead of min:sec")
@click.argument("video-id")
def main(interval_duration, timestamp_in_seconds, word_count, video_id):
    # Retrieve the available transcripts
    ytt_api = YouTubeTranscriptApi()
    fetched_transcript = ytt_api.fetch(video_id)
    print("<html><body><dl>")
    lines_iter = (interval_duration <= 0 and word_count <= 0) and fetched_transcript or (
        interval_duration > 0 and lines_by_dur(fetched_transcript, interval_duration)) or lines_by_words(fetched_transcript, word_count)

    for line_map in lines_iter:
        link_to_tstmp = f"https://youtu.be/{video_id}?t={int(line_map.start)}"
        if timestamp_in_seconds:
            tstmp_str = str(int(line_map.start))
        else:
            st_min = int(line_map.start / 60)
            st_sec = int(line_map.start - st_min * 60)
            tstmp_str = ("%2d:%-2d" % (st_min, st_sec)).replace(" ", "&nbsp;")
        print("""<dt><a href="%s">%s</a></dt><dd>%s</dd>""" % (link_to_tstmp, tstmp_str, line_map.text))
    print("</dl></html></body>")


if __name__ == '__main__':
    main()