|
import csv |
|
import shutil |
|
from pathlib import Path |
|
|
|
folder_path = Path("darija10") |
|
folder_path.mkdir(exist_ok=True) |
|
audio_folders_path: Path = folder_path / "data" |
|
audio_folders_path.mkdir(exist_ok=True) |
|
all_rows = [] |
|
for file in Path("dvoice-v1.0/transcriptions/csv").iterdir(): |
|
with file.open("r") as f: |
|
reader = csv.DictReader(f, delimiter=",") |
|
split = file.stem |
|
split_audios_path = audio_folders_path / split |
|
split_audios_path.mkdir(exist_ok=True) |
|
columns = reader.fieldnames |
|
for row in reader: |
|
audio_path = Path(row["path"]) |
|
row["path"] = Path("data") / split / audio_path.name |
|
shutil.copy(Path("dvoice-v1.0/audios_fixed") / audio_path, split_audios_path) |
|
all_rows.append(row) |
|
|
|
columns = ["file_name", "sample_rate", "transcript", "duration", "audio_id"] |
|
with (Path("darija10") / "metadata.csv").open("w", newline='', encoding="utf-8") as f: |
|
writer = csv.DictWriter(f, fieldnames=columns) |
|
writer.writeheader() |
|
for row in all_rows: |
|
row["file_name"] = row["path"] |
|
row["audio_id"] = Path(row["file_name"]).stem |
|
del row["path"] |
|
writer.writerow(row) |
|
|
|
|
|
|
