Skip to main content

Handling action outputs

Dagger tries to detect which fields are outputs in an action. Simple values like strings, numbers and booleans are printed directly to the console, as you can see when the todo app example finishes:

APP_NAME=dagger-todo dagger do deploy
[] actions.deps
[] actions.test.script
[] client.env
[] actions.build.run.script
[] actions.deploy.container.script
[] client.filesystem."./".read
[] actions.deploy
[] actions.test
[] actions.build.run
[] actions.build.contents
[] actions.deploy.container
[] client.filesystem."./_build".write
[] actions.deploy.container.export
Field Value
url "https://dagger-todo.netlify.app"
deployUrl "https://62698983ffe8661d60613431--dagger-todo.netlify.app"
logsUrl "https://app.netlify.com/sites/dagger-todo/deploys/62698983ffe8661d60613431"

This is very useful to get immediate feedback on an action's results.

Piping a result

Besides the plain format (the default), you can also use json or yaml. JSON is particularly useful if you want to pipe a result into another process:

tip

For this example, ensure you have a registry on localhost listening on port 5042:

docker run -d -p 5042:5000 --restart=always --name localregistry registry:2
package main

import (
"dagger.io/dagger"
"universe.dagger.io/docker"
)

dagger.#Plan & {
actions: {
pull: docker.#Pull & {
source: "alpine"
}
push: docker.#Push & {
image: pull.output
dest: "localhost:5042/alpine"
}
}
}
➜ dagger --output-format json do push | jq '.result'
"localhost:5042/alpine:latest@sha256:a777c9c66ba177ccfea23f2a216ff6721e78a662cd17019488c417135299cd89"
tip

You can silence the info logs by raising the log level (or redirecting stderr somewhere else):

➜ dagger -l error --output-format json do push | jq '.result'

Saving into a file

You can also save the output to a file using the --output flag. Let's do it in yaml this time:

➜ dagger --output-format yaml --output result.yaml do push
cat result.yaml
result: localhost:5042/alpine:latest@sha256:47a163eb7b572819d862b4a2c95a399829c8c79fab51f1d40c59708aa0e35331

Controlling the output

You're not limited to the outputs of an action because you can make your own in a wrapper action:

package main

import (
"dagger.io/dagger"
"universe.dagger.io/docker"
)

dagger.#Plan & {
actions: {
pull: docker.#Pull & {
source: "alpine"
}
// wrap docker.#Push to have more control over the outputs
push: {
_op: docker.#Push & {
image: pull.output
dest: "localhost:5042/alpine"
}

// The resulting digest
digest: _op.result

// The $PATH set in the image
path: _op.image.config.env.PATH
}
}
}
➜ dagger do push
[] actions.pull
[] actions.push
Field Value
digest "localhost:5042/alpine:latest@sha256:47a163eb7b572819d862b4a2c95a399829c8c79fab51f1d40c59708aa0e35331"
path "/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin"

Full output control

Since you can only output simple values, you may find the need for a solution where you can output more complex types such as structs and lists. As showcased in the interacting with the client docs, Dagger has the ability to write into the client filesystem through the client API.

Using this capability we can then have full control of what to output. The downsides are that these won't print to the console (only to a file), and you won't be able to pipe directly from the dagger command.

Let's leverage CUE's default integrations and marshal a more complex value into a single json or yaml file.

package main

import (
"encoding/yaml"
"dagger.io/dagger"
"universe.dagger.io/docker"
)

dagger.#Plan & {
client: filesystem: "config.yaml": write: contents: yaml.Marshal(actions.pull.image.config)
actions: pull: docker.#Pull & {
source: "alpine"
}
}
➜ dagger do pull
[] actions.pull
[] client.filesystem."config.yaml".write
cat config.yaml
env:
PATH: /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin
cmd:
- /bin/sh

Dagger Cloud

Dagger Cloud is a managed service that aims to help you troubleshoot your current dagger runs by storing operation history and enabling functionality such as storing outputs and a comprehensive detailed view of your executions.

If you're interested in trying out Dagger Cloud, you can find more information in our docs section. Furthermore, if you have any feedback or ideas that could help improve the product, there is an open discussion in GitHub where you can leave us your inputs.