Accessing {data} in downstream jobs/events.

[greenaar]

I'm hoping to find a good example of using input.data inside a plugin.

Example, #142 talks about using data in a decision controller -- and indeed, I can see that chaining jobs I can track a record in data.

What I can't figure out is how to reference those from within a plugin (example the shell plugin, where i seem to be doing the majority of my testing). I'm not sure how you reference the data (or if it's possible). I've checked set and I don't see anything that looks like the data I'm putting in (I used the example from discussion 142 as my test data).

Is it possible to reference that information, or is it limited to use by xyops in inter-plugin operations? I assume this question is also relevant with bucket json data (ie if i wanted to reference the json data in a bucket programatically). Maybe I'm looking at the wrong place in the docs? I did a keyword search, but didn't come up with anything that stated 'data fields are not for use within plugins' or 'here's how to assign a data field/value to an environment variable'

Again, thanks for your help!

[jhuckaby]

The input data is available as part of the JSON piped to your process STDIN:

https://docs.xyops.io/plugins/job-input

So you need to consume (read) a line from STDIN, and parse the JSON. The input data is available as input.data inside that JSON record. This is a bit tricky with shell script, as it doesn't have native support for parsing JSON. However, you can use a 3rd party tool such as jq. Example:

#!/bin/bash

IFS= read -r line
foo=$(printf '%s\n' "$line" | jq -r '.input.data.foo')
echo "Foo is: $foo"

The JSON itself is a xyOps Job object, described in full detail here: https://docs.xyops.io/data/job

Example test run setup with pre-populated data:

Job output:

Hope this helps!

[greenaar]

Exactly what I needed. Thanks for your patience. You had the details in your docs, but the shell example was the missing glue i needed to tie it together.

I'll make sure to load the data stream as step 1 in future scripts, and that answers a handful of questions I'm sure I would've had. Fantastic!