.makim.yaml

version: 1.0
groups:
  containers:
    tasks:
      get-ip:
        help: Get the IP from a container-ID
        args:
          container-id:
            help: The container-ID
            type: string
            required: true
        shell: bash
        run: |
          docker inspect \
            -f {% raw -%}"{{range .NetworkSettings.Networks}}{{.IPAddress}}{{end}}"{%- endraw %} \
            ${{ args.container_id }}

  develop:
    tasks:
      test-certs:
        env-file: .env
        help: Generate and copy certs
        shell: bash
        run: |
          mkdir -p containers/esconfig/certs
          if docker cp es:/usr/share/elasticsearch/config/certs/http_ca.crt ./containers/esconfig/certs/; then
            echo "Certificate copied successfully."
            curl --cacert ./containers/esconfig/certs/http_ca.crt -u elastic:${{ env.ES_PASSWORD }} https://localhost:9200/
          else
            echo "Error: Could not find the certificate in the container."
          fi

      count-docs-in-index:
        help: |
          Download data from BioRxiv/MedRxiv API within a specified date range.
        args:
          index-name:
            help: Specify the index name (e.g., 'biorxiv' or 'medrxiv')
            type: string
            required: true      
        shell: bash
        run: |
          python scripts/es_index_doc_counter.py ${{ args.index_name }}

      setup-dev-env:
        help: Setup development environment and prepare .env file if it doesn't exist.
        args:
          index-name:
            help: Specify the server name to prepare environment for (e.g., 'biorxiv' or 'medrxiv').
            type: string
            required: true
        shell: bash
        run: |
          set -e

          index_data_dir="/opt/services/data/elasticsearch"
      
          download_dir="${index_data_dir}/rxivx/${{ args.index_name }}/downloaded"

          mkdir -p "${download_dir}"
      
          last_date=$(date -d "yesterday" '+%Y-%m-%d')
          current_date=$(date -d "yesterday" '+%Y-%m-%d')
          output_filename="${{ args.index_name }}_${last_date}_${current_date}.json"
          touch "${download_dir}/${output_filename}"
          ls -la ${download_dir}/${output_filename}
          echo "Prepared development environment for index: ${{ args.index_name }}"

          if [ -f ".env" ]; then
              echo ".env file already exists. Exiting without modifying environment variables."
          else
              env_template="./.env.tpl"
              if [ ! -f "${env_template}" ]; then
                  echo "The template file ${env_template} does not exist."
                  exit 1
              fi
              envsubst < "${env_template}" > ".env"
              echo "Environment variables prepared and saved to .env"
          fi

  scheduler:
    tasks:
      download-rxivr:
        help: |
          Download data from BioRxiv/MedRxiv API within a specified date range.  
        args:
          server:
            help: Specify the server to download (e.g., 'biorxiv' or 'medrxiv')
            type: string
            required: true
          begin:
            help: Specify the start date in YYYY-MM-DD format
            type: string
            required: true
          end:
            help: Specify the end date in YYYY-MM-DD format
            type: string
            required: true
          target:
            help: Path to save the JSON file (e.g., 'data/')
            type: string
            required: true
        shell: Rscript
        run: |
          library(devtools)

          # Check if the package is installed
          is_package_installed <- function(pkg) {
            pkg %in% rownames(installed.packages())
          }

          package_name <- "medrxivr"
          repo <- "esloch/medrxivr@fix-numeric-type"

          if (!is_package_installed(package_name)) {
            install_github(repo, dependencies=TRUE)
            message("Package installed successfully.")
          } else {
            message("Package is already installed.")
          }

          library(medrxivr)
          library(jsonlite)

          print("Starting download data from: ${{ args.server }}")

          biorxiv_data <- mx_api_content(
            server = '${{ args.server }}',
            from_date = "${{ args.begin }}",
            to_date = "${{ args.end }}"
          )

          json_data <- toJSON(biorxiv_data, pretty = TRUE)

          json_file_path <- "${{ args.target }}${{ args.server }}_${{ args.begin }}_${{ args.end }}.json"
          writeLines(json_data, json_file_path)

          cat("Data object stored to:", "${{ args.server }}", "json file \n")
  
      setup-cron:
        help: Setup cron jobs for the develop user
        shell: bash
        run: |
          CRON_PATH="/opt/services/es-journals/scripts/cronjobs"

          # Check if cronjob file exists
          if [ ! -f "$CRON_PATH" ]; then
            echo "Cronjob file does not exist at $CRON_PATH. Please create it first."
            exit 1
          fi

          # Add cron taks to develop user
          crontab -u devops $CRON_PATH

          echo "Cron jobs for LiteRev Elasticsearch have been set up successfully for user develop."

  clean:
    tasks:
      all:
        help: Clean unnecessary temporary files
        run: |
          rm -fr build/
          rm -fr dist/
          rm -fr .eggs/
          find . -name '*.egg-info' -exec rm -fr {} +
          find . -name '*.egg' -exec rm -f {} +
          find . -name '*.pyc' -exec rm -f {} +
          find . -name '*.pyo' -exec rm -f {} +
          find . -name '__pycache__' -exec rm -fr {} +
          find . -name '*~' -exec rm -f {} +
          rm -f .coverage
          rm -fr htmlcov/
          rm -fr .pytest_cache
          rm -fr .mypy_cache
          rm -fr .ruff_cache

  docs:
    tasks:
      build:
        help: Build documentation
        run: |
          mkdocs build --config-file docs/mkdocs.yaml

      preview:
        help: Preview documentation page locally
        dependencies:
          - task: docs.build
        run: |
          mkdocs build --config-file docs/mkdocs.yaml

  tests:
    tasks:
      linter:
        help: Run linter tools
        run: |
          pre-commit install
          pre-commit run --all-files --verbose