# si-fs
A FUSE filesystem for editing System Initiative assets in your favorite editor.
1. Get your System Initiative bearer token:
- Head to https://auth.systeminit.com/workspaces
- Pick the workspace you want to mount as a fuse filesystem.
- Click the settings "gear" icon and choose the API Tokens option.
- Generate an API token for use with the fuse filesystem.
- Add the token to your env with `export SI_BEARER_TOKEN='<the token here>'` (the single quotes will ensure the base64 parses on your command line).
2. Mount the filesystem:
```bash
buck2 run //bin/si-fs -- --foreground --workspace-id <WORKSPACE_ID> --endpoint <ENDPOINT> /mountpoint
```
At this point, we recommend running in the foreground, so you can see any errors
returned by the API client. In the future we will improve telemetry and provide
a file on the filesystem that streams errors and warnings.
3. You can now browse and edit assets at `/mountpoint`
## Features
- Browse assets in a filesystem hierarchy
- Edit assets with your preferred editor
- Unlock and install assets
- Create new assets and functions
- Edit asset bindings
- Changes sync on write to System Initiative
## Commands
### Create a change set
To create a new change set to work in, do:
```
cd /mountpoint/change-sets;
mkdir new_change_set_name
```
A change set will be forked from `HEAD` in that directory.
### Install a schema
To install an uinstalled schema, run:
```
cd /mountpoint/change-sets/<YOUR CHANGE SET>/schemas/<YOUR SCHEMA>;
touch INSTALLED
```
On the next read of the schema directory, you will see the `definition` and `functions` folders.
### Unlock a schema
To unlock a schema, create the `unlocked` directory in the schema's definition directory:
```
cd /mountpoint/change-sets/<YOUR CHANGE SET>/schemas/<YOUR SCHEMA>/definition;
mkdir unlocked
```
### Unlock a function
To unlock a function, create the `unlocked` directory in the function's definition directory:
```
cd /mountpoint/change-sets/<YOUR CHANGE SET>/schemas/<YOUR SCHEMA>/functions/<FUNCTION KIND>/<FUNCTION NAME>;
mkdir unlocked
```
Unlocking a function will automatically unlock the schema that function is attached to.
NOTE: Only functions in the schema directory can be unlocked. Functions in the
change-set functions folder are read-only.
#### Create a new function for a schema
You can make a new function with `mkdir`.
```
cd /mountpoint/change-sets/<YOUR CHANGE SET>/schemas/<YOUR SCHEMA>/functions/<FUNCTION KIND>/;
mkdir new_func_name
```
Authorization, Qualification, Code Generation and Management functions will be
created with default bindings and automatically attached to the unlocked schema.
Attribute and Action functions will be in a pending state. Edit the
`PENDING_BINDINGS_EDIT_ME.json` file to configure the initial bindings for these
functions. If the bindings parse and are accepted by the backend, the pending
bindings file will disappear, and an `unlocked` folder will appear in their
place, containing the initial function code and other attributes.
### Attach an existing function
An existing function can be attached with `mv`.
Find the function you want to attach in another schemas functions folder, or in
the change-set level functions folder.
Then cd to the function kind folder for the schema you want to attach that
function to.
If attaching a function from another schema, do the following:
```
cd /mountpoint/change-sets/<CHANGE SET>/schemas/<YOUR SCHEMA>/functions/<FUNCTION KIND>
mv /mountpoint/change-sets/<CHANGE SET>/schemas/<OTHER SCHEMA>/functions/<FUNCTION KIND>/<FUNCTION NAME>/locked .
```
The `unlocked` can also be attached.
If attaching a function from the change-set level functions folder:
```
cd /mountpoint/change-sets/<CHANGE SET>/schemas/<YOUR SCHEMA>/functions/<FUNCTION KIND>
mv /mountpoint/change-sets/<CHANGE SET>/functions/<FUNCTION KIND>/<FUNCTION NAME> .
```
The function must be on the same change set as the schema you are attaching to.
Like when creating a new function, Attribute and Action functions will be put
into a pending state.
