[snyk workflow] make path to node version file configurable - #12
Conversation
| type: string | ||
| required: false | ||
| default: "11" | ||
| NODE_VERSION_FILE: |
There was a problem hiding this comment.
I've followed the existing convention here (MY_INPUT), but in my experience GitHub Actions inputs are typically defined as my-input (see examples) - I wonder if we should make that change now before too many people pick this up.
There was a problem hiding this comment.
Does this mean we're able to use different node versions within the same repo? To provide a simple example against this: we've seen a few lambdas with a node 12 runtime, but build with node 14. That's pretty confusing IMO.
Are there any reasons one would want to use a different node version within a single repository?
There was a problem hiding this comment.
I guess there are a few reasons that we might want to specify the path to the node version file rather than using a hardcoded default:
- The repo uses a single node version (defined in
.nvmrc), but the file is not at the project root (this is my use case -cdkprojects typically seem to include this file under/cdk). - The repo uses a single node version, but it uses
.node-version(rather than.nvmrc) to manage the node version. - The repo uses multiple node versions e.g. version A for
cdkand version B for application code. I'm not 100% sure if this is desirable/acceptable and I'm not convinced it'd work properly with this workflow!
There was a problem hiding this comment.
- The repo uses a single node version (defined in
.nvmrc), but the file is not at the project root (this is my use case -cdkprojects typically seem to include this file under/cdk).
I think GuCDK is wrong here - it should place this file at the root (possibly tricky if it already exists).
- The repo uses a single node version, but it uses
.node-version(rather than.nvmrc) to manage the node version.
Not sure we do this (rough GH search), but a fair point.
- The repo uses multiple node versions e.g. version A for
cdkand version B for application code. I'm not 100% sure if this is desirable/acceptable and I'm not convinced it'd work properly with this workflow!
See 1. I think we should do everything we can to avoid this as it's pretty confusing!
There was a problem hiding this comment.
I've merged this PR as I think it's still sensible to allow users to configure this if it's necessary, but I agree with:
it should place this file at the root (possibly tricky if it already exists).
so I'll see if it can be easily moved for AMIgo.
There was a problem hiding this comment.
I've followed the existing convention here (MY_INPUT), but in my experience GitHub Actions inputs are typically defined as my-input (see examples) - I wonder if we should make that change now before too many people pick this up.
Definitely think we should do this sooner rather than later!
What does this change?
Follow-up to #11. In order to use this in my project I need to be able to configure the path to the
.nvmrcfile.How to test
I've tested this with Amigo: guardian/amigo#698.
How can we measure success?
More projects can benefit from this shared workflow.
Have we considered potential risks?
I don't think this PR introduces any new risks.