export.yml

name: Export Course Summaries

on:
  workflow_dispatch:
    inputs:
      course_id:
        description: "Course ID(s) to export, comma-separated (e.g. 30004 or 30004,30005)"
        required: true
        type: string
      export_type:
        description: "Choose export type,if use HTML email,there will be no attachment, just the email body."
        type: choice
        options:
          - HTML
          - PDF
          - Markdown
        default: 'HTML'
      sub_ids:
        description: "Optional comma-separated lecture sub_ids to filter (default: all summarized lectures)"
        required: false
        type: string
        default: ""

permissions:
  contents: read

jobs:
  export:
    runs-on: ubuntu-latest
    timeout-minutes: 30

    steps:
      - uses: actions/checkout@v4

      - name: Set up Python
        uses: actions/setup-python@v5
        with:
          python-version: "3.12"

      - name: Cache Python packages
        id: cache-pip
        uses: actions/cache@v4
        with:
          path: ${{ env.pythonLocation }}/lib/python*/site-packages
          key: pip-${{ hashFiles('requirements.txt') }}

      - name: Install Python dependencies
        if: steps.cache-pip.outputs.cache-hit != 'true'
        run: pip install -r requirements.txt

      - name: Install PDF dependencies
        if: inputs.export_type == 'PDF'
        run: |
          sudo apt-get update
          sudo apt-get install -y fonts-noto-cjk libpango-1.0-0 libcairo2
          pip install weasyprint pygments
          sudo mkdir -p /usr/local/share/fonts/custom
          sudo cp assets/MSYH.TTC /usr/local/share/fonts/custom/
          sudo cp assets/MSYHBD.TTC /usr/local/share/fonts/custom/
          sudo cp assets/MSYHL.TTC /usr/local/share/fonts/custom/
          sudo fc-cache -fv

      - name: Fetch database from data branch
        run: |
          mkdir -p data
          git fetch origin data --depth=1
          if git cat-file -e origin/data:data/icourse-index.enc 2>/dev/null; then
            mkdir -p data/sharded/shards
            git show origin/data:data/icourse-index.enc > data/sharded/icourse-index.enc
            shard_paths=$(git ls-tree -r --name-only origin/data | grep '^data/shards/' || true)
            for path in $shard_paths; do
              base=$(basename "$path")
              git show "origin/data:$path" > "data/sharded/shards/$base"
            done
            echo "Loaded $(ls data/sharded/shards 2>/dev/null | wc -l) shard(s)."
          elif git cat-file -e origin/data:data/icourse.db.gz.enc 2>/dev/null; then
            git show origin/data:data/icourse.db.gz.enc > data/icourse.db.gz.enc
            echo "Loaded legacy compressed database."
          elif git cat-file -e origin/data:data/icourse.db.enc 2>/dev/null; then
            git show origin/data:data/icourse.db.enc > data/icourse.db.enc
            echo "Loaded legacy uncompressed database."
          else
            echo "::error::No database found on data branch."
            exit 1
          fi

      - name: Decode database
        env:
          STUID: ${{ secrets.STUID }}
          UISPSW: ${{ secrets.UISPSW }}
          DASHSCOPE_API_KEY: ${{ secrets.DASHSCOPE_API_KEY }}
          SMTP_PASSWORD: ${{ secrets.SMTP_PASSWORD }}
        run: |
          if [ -f data/sharded/icourse-index.enc ]; then
            python scripts/db_shard.py reassemble data/sharded data/icourse.db
            echo "Reassembled sharded database."
            exit 0
          fi

          NEW_KEY=$(printf '%s' "ICSv2:${STUID}:${UISPSW}" | sha256sum | cut -d' ' -f1)
          LEGACY_KEY="${STUID}${UISPSW}${DASHSCOPE_API_KEY}${SMTP_PASSWORD}"

          decrypt_blob() {
            local in_path="$1"
            local out_path="$2"
            if openssl enc -aes-256-cbc -d -pbkdf2 -iter 100000 \
                 -in "$in_path" -out "$out_path" \
                 -pass pass:"$NEW_KEY" 2>/dev/null; then
              return 0
            fi
            openssl enc -aes-256-cbc -d -pbkdf2 \
              -in "$in_path" -out "$out_path" \
              -pass pass:"$LEGACY_KEY" 2>/dev/null
          }

          if [ -f data/icourse.db.gz.enc ]; then
            decrypt_blob data/icourse.db.gz.enc data/icourse.db.gz
            gzip -d data/icourse.db.gz
          else
            decrypt_blob data/icourse.db.enc data/icourse.db
          fi
          echo "Database decrypted."

      - name: Export course summaries
        env:
          SMTP_EMAIL: ${{ secrets.SMTP_EMAIL }}
          SMTP_PASSWORD: ${{ secrets.SMTP_PASSWORD }}
          RECEIVER_EMAIL: ${{ secrets.RECEIVER_EMAIL }}
          EXPORT_COURSE_ID: ${{ inputs.course_id }}
          EXPORT_SUB_IDS: ${{ inputs.sub_ids }}
        run: |
          case "${{ inputs.export_type }}" in
            "HTML")
              EXPORT_FLAG=""
              ;;
            "PDF")
              EXPORT_FLAG="--pdf"
              ;;
            "Markdown")
              EXPORT_FLAG="--md"
              ;;
          esac
          if [ -n "$EXPORT_SUB_IDS" ]; then
            python scripts/export_course.py --course-id "$EXPORT_COURSE_ID" --sub-ids "$EXPORT_SUB_IDS" $EXPORT_FLAG
          else
            python scripts/export_course.py --course-id "$EXPORT_COURSE_ID" $EXPORT_FLAG
          fi