sasjs init¶

The sasjs init command turns a folder that already contains code into a SASjs project. Change into that folder and run:

sasjs init

It takes no arguments and works on the current working directory. Use sasjs create when you want a new project scaffolded from a template; use sasjs init when the SAS programs, jobs or other files are already in place and you want the SASjs tooling to work over them.

Prerequisites¶

The SASjs CLI installed (sasjs init works without a deployment target - it does not contact a SAS server)
Network access, for the @sasjs/core install

What it creates¶

Path	Purpose
`sasjs/sasjsconfig.json`	Project configuration, including a placeholder target
`sasjs/macros/`	The macro folder `macroFolders` points at - created with a `.gitkeep` so an empty folder survives a commit
`sasjs/doxy/`	Doxygen templates used by `sasjs doc`
`.sasjslint`	Lint rules used by `sasjs lint` - created only if absent
`.gitignore`	Adds `node_modules`, `sasjsbuild`, `sasjsresults` and `.env*` if those rules are missing; existing content is kept
`package.json`	Created with `npm init --yes` if the folder has none, then `@sasjs/core` is added as a dependency

node_modules and package-lock.json follow from that install.

It does not create sasjs/services, sasjs/jobs or the rest of the layout, and it does not add a README - sasjs create is the command that lays out a full project.

The generated sasjsconfig.json¶

{
  "$schema": "https://cli.sasjs.io/sasjsconfig-schema.json",
  "macroFolders": ["sasjs/macros"],
  "defaultTarget": "mytarget",
  "targets": [
    {
      "name": "mytarget",
      "serverType": "SASJS",
      "serverUrl": "",
      "appLoc": "/Public/apps/myapp"
    }
  ]
}

The single target is a placeholder. serverUrl is empty rather than a sample URL, because an empty value reads as "not configured yet" while a non-empty placeholder has to pass URL validation every time the configuration is read. Put the real serverUrl and appLoc in it, or add a target with sasjs add / by editing the file - see the configuration reference.

Running it in a folder that is already a SASjs project¶

If the folder already contains sasjs/sasjsconfig.json, sasjs init aborts without touching anything:

$ sasjs init
ERROR  This folder is already a SASjs project (sasjs/sasjsconfig.json exists). Aborting without making changes.
$ echo $?
4

Exit code 4 is the only code that means the folder is already a SASjs project, so a script can branch on it. Code 1 covers an invalid command or invalid arguments, and code 2 a genuine failure.

The check looks at the current folder only. Initialising inside a folder that belongs to a project higher up starts a new project in that folder, which is what a repository holding several separate SASjs apps needs.

The refusal is deliberate. A target carries user data - serverUrl, appLoc, authInfo, contextName - and an unconditional write would replace it with the placeholder above. sasjs create writes a sasjsconfig.json of its own, so sasjs init is not needed after sasjs create, and running it there aborts too.

To re-run the initialisation, delete or move sasjs/sasjsconfig.json first.

After initialisation¶

A freshly initialised project compiles straight away: the macro folder exists and the empty serverUrl passes validation, so no edit is needed before the first build step. Running sasjs compile in the folder compiles the project for the mytarget target.

Commands that talk to a SAS server - sasjs deploy, sasjs run, sasjs request and the authentication commands - need a real serverUrl first.

The other local commands work in the folder as they are:

sasjs build
sasjs doc
sasjs lint

The configuration, the build and deploy settings, and the target attributes are described in the sasjsconfig reference, and the full command list is on the home page.