feat: add datalink create node #19
Conversation
ewels
left a comment
ewels
left a comment
There was a problem hiding this comment.
Really nice! Haven't tested yet, but a quick scan of the code looks great, couple of v. minor comments.
What kind of use cases do you see for this? Just thinking about the examples.
| "devDependencies": { | ||
| "nodemon": "^3.1.10" | ||
| }, | ||
| "scripts": { | ||
| "dev": "nodemon --exec 'node-red -u ./test-instance' --watch nodes --ext js" |
There was a problem hiding this comment.
Nice! Please could you document usage in CONTRIBUTING.md so that folks know how to use it? (and anywhere else that makes sense)
| * @param {object} options - Options object with baseUrl and workspaceId | ||
| * @returns {Promise<string|null>} The resource ID, or null if resourceName is empty | ||
| */ | ||
| async function resolveResource(RED, node, msg, resourceType, resourceName, { baseUrl, workspaceId = null } = {}) { |
There was a problem hiding this comment.
Do we need msg as an argument? Doesn't look like it's being used.
| /** | ||
| * Generic resource resolver for Seqera Platform API. | ||
| * Resolves a resource name to its ID by querying the appropriate endpoint. | ||
| * | ||
| * Supported resource types: | ||
| * - credentials: /credentials → credentials[].id | ||
| * - data-links: /data-links → dataLinks[].id | ||
| * - pipelines: /pipelines → pipelines[].pipelineId | ||
| * - compute-envs: /compute-envs → computeEnvs[].id | ||
| * - datasets: /datasets → datasets[].id |
There was a problem hiding this comment.
Super nice! I reckon we could delete a fair bit of code with some refactoring to use this, the lookup is done in quite a few places I think (doesn't have to be done in this PR but it'd be a nice follow-on).
| responseKey: "datasets", | ||
| idField: "id", | ||
| nameField: "name", | ||
| queryParams: { max: "100" }, |
There was a problem hiding this comment.
This seems a bit limiting, especially for datasets..
Could the function handle pagination do you think? I implemented pagination for datasets already here.
Ideally we would have an API endpoint to do this of course, rather than pulling thousands of dataset details just to do a simple lookup 👀
Adds a node for data-link creation and utility to resolve
resourceNametoresourceIdfor simpler HTTP request handling. The function works for credentials, pipelines, compute-envs, datasets, and data-links.Changes made:
resolveResource()functionmodule.exportsresolveResourcefrom_utils.jsUpdate
CLAUDE.mdwith guidanceAdded
nodemonconfig to auto-restart local instances for quicker development throughnpm run dev