This topic was automatically generated from Slack. You can find the original thread here.
is there a way to run the android command line tools within Pipedream’s workflows? With all working dependencies
This topic was automatically generated from Slack. You can find the original thread here.
is there a way to run the android command line tools within Pipedream’s workflows? With all working dependencies
I’m sorry, but currently Pipedream does not support running Android command line tools within its workflows. Pipedream workflows are primarily designed to work with HTTP APIs, code steps in Node.js or Python, and pre-built actions. If you have a specific use case or integration in mind, I recommend submitting an integration request at New Pipedream Integration Request. The Pipedream team regularly reviews these requests and it helps them prioritize new features and integrations.
Thanks, and what about running gradle?
I’m sorry, but currently Pipedream does not support running Gradle or other build tools directly within its workflows. Pipedream workflows are primarily designed to work with HTTP APIs, code steps in Node.js or Python, and pre-built actions. If you need to run Gradle builds as part of your workflow, you might consider using a continuous integration (CI) service that supports Gradle, such as Jenkins, CircleCI, or GitHub Actions. You could trigger a build on one of these services from a Pipedream workflow using their HTTP API, and then handle the results of the build within Pipedream once the CI service completes the build. If you have a specific use case or integration in mind, I recommend submitting an integration request at New Pipedream Integration Request. The Pipedream team regularly reviews these requests and it helps them prioritize new features and integrations.
what about downloading these deps to /tmp at runtime in a Bash step?
Yes, it’s possible to download dependencies to the /tmp directory at runtime in a Bash step within a Pipedream workflow. Here’s a basic example of how you might do this: export default defineComponent({ props: { bash: { type: "app", app: "bash", }, }, async run({ steps, $ }) { const command = ` cd /tmp wget https://example.com/my-dependency.tar.gz tar -xzf my-dependency.tar.gz `; const result = await this.bash.runCommand(command); console.log(result); }, }); In this example, we’re using the wget command to download a file from a URL, and then the tar command to extract the contents of the downloaded file. Replace https://example.com/my-dependency.tar.gz with the URL of the file you want to download. Please note that this is a basic example and might not work for all dependencies or systems. Depending on the specific dependencies you’re trying to install, you might need to modify the commands or add additional ones. Also, keep in mind that the /tmp directory is ephemeral and its contents are not persisted across different invocations of the workflow. If you need to use the downloaded dependencies in multiple steps or invocations, you might need to download them each time.
check the bash docs again. You do not construct a Node.js component to run bash code, you can just add a bash step and run standard bash code in there
You’re correct, I apologize for the confusion. In Pipedream, you can directly add a Bash script step in your workflow without needing to wrap it in a Node.js component. Here’s how you can do it: 1. Click on the “+ New Step” button in your workflow. 2. Choose “Run Bash Script”. 3. In the Bash script editor, you can write your Bash commands. Remember that any files you want to persist should be written to the /tmp directory, as other directories are read-only. Also, the environment is reset after each workflow run, so any changes made during a run (like installing packages) will not persist to future runs.
let me know if that works as a workaround. Right now, you’d have to wget or curl the binaries for amd64 / Fedora / CentOS, save them directly to /tmp, and then run them from there.
We’re planning to develop better ways to package custom deps like this in your workflow.
Thanks Dylan, I’ll give it a go now