purge_bad_transcripts.py

# It is possible that general influencer data was inspected at a later time.
# In this case, the provided link for subtitles may have already expired.
# This script deletes all failed transcripts.

import os

to_be_deleted: list[str] = []

i = 0

for transcript in os.listdir("./transcripts"):
    with open(os.path.join("transcripts", transcript), encoding="utf-8") as fp:
        # This is a bad hack, but it should work.
        aaa = fp.readlines()[0]
        if "<HTML><HEAD>" in aaa or "<html>" in aaa:
            to_be_deleted.append(transcript)
            i += 1

for bad_transcript in to_be_deleted:
    os.remove(os.path.join("transcripts", bad_transcript))

print(f"{i} transcripts were faulty.")