rerun_diffs

#!/usr/bin/env python3

"""
This program will rerun the diff generation using the transcription output that
is already on disk. You need to tell it where the output is, for example:

    ./rerun_diffs docs/output-2024-04-11/
"""

import json
import sys
from pathlib import Path

import pandas

from transcribe.utils import compare_transcripts

output_dir = Path(sys.argv[1])

# get the list of files
data = pandas.read_csv("data.csv", index_col="druid")

for transcript_file in output_dir.glob("*.json"):
    druid, transcript_type, run_count = transcript_file.name.split("-")
    run_count = int(run_count.replace(".json", ""))
    transcript = json.load(open(transcript_file))

    # update the file_metadata with things needed for the diff generation
    file_metadata = dict(data.loc[druid])
    file_metadata["druid"] = druid
    file_metadata["run_count"] = run_count

    print(f"rerunning compare_transcript for {transcript_file.name}")
    compare_transcripts(file_metadata, transcript, transcript_type, output_dir)