ba-sachsen-pica/alephino/Taskfile.yml

version: '3'

tasks:
  main:
    desc: Konvertierung von Alephino nach PICA3/CSV
    vars:
      DIR: '{{splitList ":" .TASK | first}}' # results in the task namespace, which is identical to the directory name
    cmds:
      - task: refine-pre
        vars: {PROJECT: leipzig}
      - task: refine-pre
        vars: {PROJECT: riesa}
      - task: refine-main

  refine-pre:
    dir: ./{{.DIR}}
    label: '{{.TASK}}-{{.PROJECT}}'
    vars:
      DIR: '{{splitList ":" .TASK | first}}'
      PORT: 3335 # assign a different port for each project
      RAM: 8192M # maximum RAM for OpenRefine java heap space
      LOG: '>(tee -a "log/{{.PROJECT}}.log") 2>&1'
    cmds:
      - echo "{{now | date "2006-01-02 15:04:05"}} {{.PROJECT}}"
      - task: :start # launch OpenRefine
        vars: {DIR: '{{.DIR}}/log', PROJECT: '{{.PROJECT}}', PORT: '{{.PORT}}', RAM: '{{.RAM}}'}
      - > # Import Titel
        "$CLIENT" -P {{.PORT}}
        --create "$(readlink -m input/{{.PROJECT}}-titel.txt)"
        --format fixed-width
        --encoding UTF-8
        --columnWidths 5
        --skipDataLines 0
        --storeBlankRows false
        --projectName titel
        > {{.LOG}}
      - > # Import Exemplare
        "$CLIENT" -P {{.PORT}}
        --create "$(readlink -m input/{{.PROJECT}}-exemplare.txt)"
        --format fixed-width
        --encoding UTF-8
        --columnWidths 5
        --skipDataLines 0
        --storeBlankRows false
        --projectName exemplare
        > {{.LOG}}
      - | # Titel: Korrekturen Einzelfälle
        "$CLIENT" -P {{.PORT}} titel --apply config/pre/titel-einzelfaelle.json > {{.LOG}}
      - | # Prefix M bzw. E für Feldnamen
        "$CLIENT" -P {{.PORT}} titel --apply config/pre/titel-prefix.json > {{.LOG}}
        "$CLIENT" -P {{.PORT}} exemplare --apply config/pre/exemplare-prefix.json > {{.LOG}}
      - | # Datensätze und Feldnamen sortieren
        "$CLIENT" -P {{.PORT}} titel --apply config/pre/titel-sortieren.json > {{.LOG}}
        "$CLIENT" -P {{.PORT}} exemplare --apply config/pre/exemplare-sortieren.json > {{.LOG}}
      - | # Mehrfachbelegungen zusammenführen
        "$CLIENT" -P {{.PORT}} titel --apply config/pre/titel-mehrfachbelegungen.json > {{.LOG}}
        "$CLIENT" -P {{.PORT}} exemplare --apply config/pre/exemplare-mehrfachbelegungen.json > {{.LOG}}
      - | # Felder löschen
        "$CLIENT" -P {{.PORT}} titel --apply config/pre/titel-loeschen.json > {{.LOG}}
        "$CLIENT" -P {{.PORT}} exemplare --apply config/pre/exemplare-loeschen.json > {{.LOG}}
      - | # Transponieren
        "$CLIENT" -P {{.PORT}} titel --apply config/pre/titel-transponieren.json > {{.LOG}}
        "$CLIENT" -P {{.PORT}} exemplare --apply config/pre/exemplare-transponieren.json > {{.LOG}}
      - | # Titel-ID separieren 
        "$CLIENT" -P {{.PORT}} titel --apply config/pre/titel-id-separieren.json > {{.LOG}}
        "$CLIENT" -P {{.PORT}} exemplare --apply config/pre/exemplare-id-separieren.json > {{.LOG}}
      - | # Titel: Exemplare anreichern
        "$CLIENT" -P {{.PORT}} titel --apply config/pre/titel-anreichern.json > {{.LOG}}
      - mkdir -p output
      - > # Export
        "$CLIENT" -P {{.PORT}} titel
        --output "$(readlink -m output/{{.PROJECT}}.tsv)"
        > {{.LOG}}
      - | # print allocated system resources
        PID="$(lsof -t -i:{{.PORT}})"
        echo "used $(($(ps --no-headers -o rss -p "$PID") / 1024)) MB RAM" > {{.LOG}}
        echo "used $(ps --no-headers -o cputime -p "$PID") CPU time" > {{.LOG}}
      - task: :kill # shut down OpenRefine immediately to save time and disk space
        vars: {DIR: '{{.DIR}}/log', PORT: '{{.PORT}}'}
      - task: :check # check OpenRefine log for any warnings and exit on error
        vars: {DIR: '{{.DIR}}'}
    sources:
      - input/{{.PROJECT}}.imp
      - config/pre/**
    generates:
      - output/{{.PROJECT}}.tsv
    ignore_error: true # workaround to avoid an orphaned Java process on error https://github.com/go-task/task/issues/141

  refine-main:
    dir: ./{{.DIR}}
    vars:
      DIR: '{{splitList ":" .TASK | first}}'
      PROJECT: alephino
      PORT: 3335 # assign a different port for each project
      RAM: 8192M # maximum RAM for OpenRefine java heap space
      LOG: '>(tee -a "log/{{.PROJECT}}.log") 2>&1'
    cmds:
      - echo "{{now | date "2006-01-02 15:04:05"}} {{.PROJECT}}"
      - task: :start # launch OpenRefine
        vars: {DIR: '{{.DIR}}/log', PROJECT: '{{.PROJECT}}', PORT: '{{.PORT}}', RAM: '{{.RAM}}'}
      - > # Zip-Archiv mit Output der Vorverarbeitung erstellen
        zip -j tmp.zip
        output/leipzig.tsv
        output/riesa.tsv
      - > # Import Zip-Archiv
        "$CLIENT" -P {{.PORT}}
        --create "$(readlink -m tmp.zip)"
        --format tsv
        --includeFileSources true
        --projectName {{.PROJECT}}
        > {{.LOG}}
        && rm tmp.zip
      - > # Spalten sortieren: Beginnen mit 1. M|001, 2. E|001, 3. File; damit Records-Mode erhalten bleibt
        "$CLIENT" -P {{.PORT}} {{.PROJECT}} --apply config/main/sortieren.json > {{.LOG}}
      - > # Bibliothekskürzel aus Import-Dateiname
        "$CLIENT" -P {{.PORT}} {{.PROJECT}} --apply config/main/file.json > {{.LOG}}
      - > # spec_A_E_01: Signatur 7100a
        "$CLIENT" -P {{.PORT}} {{.PROJECT}} --apply config/main/7100a.json > {{.LOG}}
#     - > # Export der PICA3-Spalten als CSV; Spalte 2199 muss vorne stehen, weil später für Sortierung benötigt
#       mkdir -p output &&
#       "$CLIENT" -P {{.PORT}} {{.PROJECT}}
#       --output "$(readlink -m output/{{.PROJECT}}.csv)"
#       --template "$(< config/main/template.txt)"
#       --rowSeparator ""
#       > {{.LOG}}
      - | # print allocated system resources
        PID="$(lsof -t -i:{{.PORT}})"
        echo "used $(($(ps --no-headers -o rss -p "$PID") / 1024)) MB RAM" > {{.LOG}}
        echo "used $(ps --no-headers -o cputime -p "$PID") CPU time" > {{.LOG}}
      - task: :stop # shut down OpenRefine and archive the OpenRefine project
        vars: {DIR: '{{.DIR}}/log', PROJECT: '{{.PROJECT}}', PORT: '{{.PORT}}'}
      - task: :check # check OpenRefine log for any warnings and exit on error
        vars: {DIR: '{{.DIR}}'}
    sources:
      - output/*.tsv
      - config/main/**
    generates:
      - log/{{.PROJECT}}.openrefine.tar.gz
#     - output/{{.PROJECT}}.csv
    ignore_error: true # workaround to avoid an orphaned Java process on error https://github.com/go-task/task/issues/141

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