version: '3'

tasks:
  main:
    desc: Powerhouse Museum Tutorial
    vars:
      DIR: '{{splitList ":" .TASK | first}}'
    cmds:
      - task: refine
      - task: :check # check OpenRefine log for any warnings and exit on error
        vars: {DIR: '{{.DIR}}'}

  refine:
    dir: ./{{.DIR}}
    vars:
      DIR: '{{splitList ":" .TASK | first}}'
      PROJECT: phm
      PORT: 3336 # assign a different port for each project
      RAM: 2048M # maximum RAM for OpenRefine java heap space
    deps: # will be executed each run independent of up-to-date check
      - task: download
    cmds:
      - task: :start # launch OpenRefine
        vars: {DIR: '{{.DIR}}', PORT: '{{.PORT}}', RAM: '{{.RAM}}'}
      - > # import file
        "$CLIENT" -P {{.PORT}}
        --create "$(readlink -m input/phm-collection.tsv)"
        --processQuotes false
        --guessCellValueTypes true
        --projectName {{.PROJECT}}
        > >(tee -a openrefine.log) 2>&1
      - > # apply transformation rules
        "$CLIENT" -P {{.PORT}} {{.PROJECT}}
        --apply config/phm-transform.json
        > >(tee -a openrefine.log) 2>&1
      - > # export to file
        mkdir -p output &&
        "$CLIENT" -P {{.PORT}} {{.PROJECT}}
        --output "$(readlink -m output/phm-results.tsv)"
        > >(tee -a openrefine.log) 2>&1
      - | # print allocated system resources
        PID="$(lsof -t -i:{{.PORT}})"
        echo "used $(($(ps --no-headers -o rss -p "$PID") / 1024)) MB RAM" \
          > >(tee -a openrefine.log)
        echo "used $(ps --no-headers -o cputime -p "$PID") CPU time" \
          > >(tee -a openrefine.log)
      - task: :stop # shut down OpenRefine and archive the OpenRefine project
        vars: {DIR: '{{.DIR}}', PORT: '{{.PORT}}', PROJECT: '{{.PROJECT}}'}
    sources:
      - Taskfile.yml
      - input/**
      - config/**
    generates:
      - openrefine.log
      - ./{{.PROJECT}}.openrefine.tar.gz
      - output/**
    ignore_error: true # workaround to avoid an orphaned Java process on error
                       # https://github.com/go-task/task/issues/141
  download:
    dir: ./{{.DIR}}
    vars:
      DIR: '{{splitList ":" .TASK | first}}'
    cmds:
      - mkdir -p input config
      - > # Download input
        wget --no-verbose -O input/phm-collection.tsv
        https://github.com/opencultureconsulting/openrefine-batch/raw/master/examples/powerhouse-museum/input/phm-collection.tsv
      - > # Download config
        wget --no-verbose -O config/phm-transform.json
        https://github.com/opencultureconsulting/openrefine-batch/raw/master/examples/powerhouse-museum/config/phm-transform.json

  default: # enable standalone execution (running `task` in project directory)
    cmds:
      - DIR="${PWD##*/}:main" && cd .. && task "$DIR"