Add ability to parse JSON in io.kestra.plugin.scripts.groovy.FileTransform about kestra HOT 5 CLOSED

@aku the Jackson library is in the classpath, you may be able to use it to process JSON.

from kestra.

Built in support for JSON is certainly welcome. In my case I was disappointed with Jython performance (I do a simple transformation of array of data which includes parsing JSON string) so I've decided to try other FileTranform tasks/plugins.
Nashorn doesn't support returning multiple rows, Groovy can't parse JSON. So I stuck with Jython

from kestra.

@aku the Jackson library is in the classpath, you may be able to use it to process JSON.

Hi @loicmathieu
Thanks! It took me a while to figure out how to use it but I managed to parse JSON objects using Jackson lib.
Here is a snippet:

      import com.fasterxml.jackson.*

      def jsonStr = '{ "name": "John Doe", "age": 99, "embedded": {"foo": "bar"} }';

      def mapper = new databind.ObjectMapper();
      def typeRef = new core.type.TypeReference<HashMap<String,Object>>() {};

      data = mapper.readValue(jsonStr, typeRef);
      
      logger.info('json object: {}', data);
      logger.info('embedded field: {}', data.embedded.foo)

Maybe you can update plugin documentation to provide some examples for other users?
Also, it would be nice to know plugin's limitations

from kestra.

good catch, we'll get back to you on that on Monday

btw would this task be useful in your use case? we were thinking of adding support for JSONata #3148

from kestra.

Hey, I got the following response from the team: "In process (in JVM) script tasks will never be able to install any packages due to security reasons. They are more performant than scripts in external processes but with fewer features."

This means we are not able to install any external Groovy libraries including JSON. To use JSON packages, you will need to use Node.js or Python script task instead (or JSONata once we add it).

from kestra.