deploy.sh

#!/usr/bin/env bash
#
# Deploy script for the RIR-ASR Evaluation Leaderboard.
#
# The HF Space runs evaluation directly — no Lambda needed.
# AWS S3 is used only for private dataset storage and results.
#
# Prerequisites:
#   - AWS CLI configured (aws configure)
#   - HuggingFace CLI installed (pip install huggingface-hub)
#
# Usage:
#   1. Edit the variables below
#   2. ./deploy.sh setup-s3      # One-time: create S3 bucket + IAM user
#   3. ./deploy.sh upload-data   # Upload private dataset to S3
#   4. ./deploy.sh deploy-space  # Create/update HuggingFace Space

set -euo pipefail

# ============================================================================
# CONFIGURATION — edit these
# ============================================================================
AWS_REGION="us-east-1"

# S3
S3_BUCKET="FFASR_Leaderboard-storage"
S3_PREFIX="generated_with_treble10"

# IAM user for HF Space to access S3
IAM_USER_NAME="whojavumusic"

# HuggingFace Space (full repo id: org/SpaceName)
HF_SPACE_NAME="treble-technologies/FFASR_Leaderboard"
# Required for deploy-space: export HF_TOKEN before running (HF Settings → Access Tokens)
HF_TOKEN="${HF_TOKEN:-}"

# Local paths
DATASET_DIR="../data/packed/"
SPACE_DIR="."

# ============================================================================
# FUNCTIONS
# ============================================================================

setup_s3() {
    echo "=== Creating S3 bucket ==="
    aws s3 mb "s3://${S3_BUCKET}" --region "${AWS_REGION}" 2>/dev/null || echo "Bucket already exists"

    # Block all public access
    aws s3api put-public-access-block \
        --bucket "${S3_BUCKET}" \
        --public-access-block-configuration \
        "BlockPublicAcls=true,IgnorePublicAcls=true,BlockPublicPolicy=true,RestrictPublicBuckets=true"

    echo "=== Creating IAM user for Space S3 access ==="
    aws iam create-user --user-name "${IAM_USER_NAME}" 2>/dev/null || echo "IAM user already exists"

    # S3-only policy scoped to this bucket
    S3_POLICY="{
        \"Version\": \"2012-10-17\",
        \"Statement\": [{
            \"Effect\": \"Allow\",
            \"Action\": [\"s3:GetObject\", \"s3:PutObject\", \"s3:ListBucket\"],
            \"Resource\": [
                \"arn:aws:s3:::${S3_BUCKET}\",
                \"arn:aws:s3:::${S3_BUCKET}/*\"
            ]
        }]
    }"
    aws iam put-user-policy \
        --user-name "${IAM_USER_NAME}" \
        --policy-name "S3DatasetAccess" \
        --policy-document "${S3_POLICY}"

    echo ""
    echo "=== Creating access keys ==="
    KEYS=$(aws iam create-access-key --user-name "${IAM_USER_NAME}" 2>/dev/null || echo "")
    if [ -n "${KEYS}" ]; then
        ACCESS_KEY=$(echo "${KEYS}" | python3 -c "import sys,json; print(json.load(sys.stdin)['AccessKey']['AccessKeyId'])")
        SECRET_KEY=$(echo "${KEYS}" | python3 -c "import sys,json; print(json.load(sys.stdin)['AccessKey']['SecretAccessKey'])")
        echo ""
        echo "Save these — you won't see the secret again:"
        echo "  AWS_ACCESS_KEY_ID     = ${ACCESS_KEY}"
        echo "  AWS_SECRET_ACCESS_KEY = ${SECRET_KEY}"
    else
        echo "Access keys already exist. Delete old ones in IAM console if you need new ones."
    fi

    echo ""
    echo "=== S3 setup complete ==="
}

upload_data() {
    echo "=== Uploading dataset to S3 ==="
    aws s3 sync "${DATASET_DIR}/wavs/" "s3://${S3_BUCKET}/${S3_PREFIX}/wavs/" --region "${AWS_REGION}"
    aws s3 sync "${DATASET_DIR}/transcripts/" "s3://${S3_BUCKET}/${S3_PREFIX}/transcripts/" --region "${AWS_REGION}"
    echo ""
    echo "=== Upload complete ($(aws s3 ls "s3://${S3_BUCKET}/${S3_PREFIX}/wavs/" --region "${AWS_REGION}" | wc -l) wav files) ==="
}

deploy_space() {
    echo "=== Deploying HuggingFace Space ==="

    if [ -z "${HF_TOKEN}" ]; then
        echo "Error: Set HF_TOKEN environment variable"
        exit 1
    fi

    # Create the Space repo if it doesn't exist
    huggingface-cli repo create "${HF_SPACE_NAME}" --type space --space-sdk gradio 2>/dev/null || true

    # Clone, copy files, push
    TMPDIR=$(mktemp -d)
    git clone "https://huggingface.co/spaces/${HF_SPACE_NAME}" "${TMPDIR}/space" 2>/dev/null || \
    git clone "https://huggingface.co/spaces/${HF_SPACE_NAME}" "${TMPDIR}/space"

    cp "${SPACE_DIR}/app.py" "${TMPDIR}/space/"
    cp "${SPACE_DIR}/init.py" "${TMPDIR}/space/"
    cp "${SPACE_DIR}/analytics.py" "${TMPDIR}/space/"
    cp "${SPACE_DIR}/ffas.py" "${TMPDIR}/space/"
    cp "${SPACE_DIR}/family_registry.py" "${TMPDIR}/space/"
    cp "${SPACE_DIR}/job_queue.py" "${TMPDIR}/space/"
    cp "${SPACE_DIR}/metrics_config.py" "${TMPDIR}/space/"
    cp "${SPACE_DIR}/remote_jobs.py" "${TMPDIR}/space/"
    cp "${SPACE_DIR}/storage.py" "${TMPDIR}/space/"
    cp "${SPACE_DIR}/utils_display.py" "${TMPDIR}/space/"
    cp "${SPACE_DIR}/constants.py" "${TMPDIR}/space/"
    cp "${SPACE_DIR}/requirements.txt" "${TMPDIR}/space/"
    cp "${SPACE_DIR}/README.md" "${TMPDIR}/space/"
    rm -rf "${TMPDIR}/space/backends" "${TMPDIR}/space/benchmark" "${TMPDIR}/space/evaluation" "${TMPDIR}/space/scripts"
    cp -R "${SPACE_DIR}/backends" "${TMPDIR}/space/backends"
    cp -R "${SPACE_DIR}/benchmark" "${TMPDIR}/space/benchmark"
    cp -R "${SPACE_DIR}/evaluation" "${TMPDIR}/space/evaluation"
    cp -R "${SPACE_DIR}/scripts" "${TMPDIR}/space/scripts"

    cd "${TMPDIR}/space"
    git add -A
    git commit -m "Update leaderboard app" || echo "No changes to commit"
    git push
    cd -

    rm -rf "${TMPDIR}"

    echo ""
    echo "=== Space deployed ==="
    echo "URL: https://huggingface.co/spaces/${HF_SPACE_NAME}"
    echo ""
    echo "Set these secrets in Space settings (Settings > Variables and Secrets):"
    echo "  AWS_ACCESS_KEY_ID     = <from setup-s3 output>"
    echo "  AWS_SECRET_ACCESS_KEY = <from setup-s3 output>"
    echo "  AWS_DEFAULT_REGION    = ${AWS_REGION}"
    echo "  DATASET_S3_BUCKET     = ${S3_BUCKET}"
    echo "  DATASET_S3_PREFIX     = ${S3_PREFIX}"
    echo "  RESULTS_S3_BUCKET     = ${S3_BUCKET}"
    echo "  RESULTS_S3_KEY        = results/leaderboard.csv"
}

# ============================================================================
# MAIN
# ============================================================================
case "${1:-help}" in
    setup-s3)     setup_s3 ;;
    upload-data)  upload_data ;;
    deploy-space) deploy_space ;;
    all)
        setup_s3
        upload_data
        deploy_space
        ;;
    *)
        echo "Usage: $0 {setup-s3|upload-data|deploy-space|all}"
        echo ""
        echo "  setup-s3      Create private S3 bucket + IAM user for the Space"
        echo "  upload-data   Upload private dataset to S3"
        echo "  deploy-space  Deploy HuggingFace Space"
        echo "  all           Run all steps in order"
        ;;
esac