download_sample_data.sh

#!/bin/bash

# Download sample data for Myanmar Ghost project

set -e

echo "📥 Downloading sample data..."

# Create data directories
mkdir -p data/raw/audio
mkdir -p data/raw/transcripts
mkdir -p data/processed

# Download sample audio (placeholder - replace with actual URLs)
# Example: wget -O data/raw/audio/sample.wav "https://example.com/sample.wav"

# Create sample metadata
cat > data/raw/metadata.csv << 'EOF'
id,audio_file,transcript,sentiment
utt_001,session_001/speaker_a.wav,မင်္ဂလာပါ,neutral
utt_002,session_001/speaker_b.wav,ကျေးဇူးပါ,positive
utt_003,session_002/speaker_a.wav,မကျေနပ်ပါဗျ,negative
EOF

# Download sample from HuggingFace datasets (if datasets is installed)
if command -v python &> /dev/null; then
    python -c "
from datasets import load_dataset
# Load a sample dataset
print('Sample data structure created successfully')
"
fi

echo "✅ Sample data downloaded to data/raw/"
echo ""
echo "Next steps:"
echo "  1. Add your actual audio files to data/raw/audio/"
echo "  2. Update transcripts in data/raw/metadata.csv"
echo "  3. Run: bash scripts/run_data_pipeline.sh"