pipeline.json

{
    "Name": "fruitstand",
    "Stages": [
        {
            "Name": "input",
            "Image": "ubuntu",
            "Cmd": [
                "sh", "-c",
                "cp /data/* /walrus/input/"
            ],
            "Volumes": ["data:/data"],
            "Cache": true
        }, 
        {
            "Name": "filter",
            "Image": "ubuntu",
            "Cmd": [
                "sh", "-c",
                "for fruit in apple orange banana; do grep $fruit /walrus/input/set1.txt | awk '{print $2}' >> /walrus/filter/$fruit; done"
            ],
            "Inputs" : [
                "input"
            ]
        }, 
        {
            "Name": "sum",
            "Image": "ubuntu",
            "Cmd": [
                "sh", "-c",
                "for fruit in apple orange banana; do { cat /walrus/filter/$fruit || echo 0; cat /walrus/filter/$fruit; } | awk '{s+=$1} END {print s}' > /walrus/sum/$fruit; done"
            ],
            "Inputs" : [
                "filter"
            ]
        }
    ],
    "Version": "",
    "Comment": "Fruit stand example pipeline from Pachyderm"
}