SUSHI (aka "SUSHI Unshortens Short Hand Inputs") is a reference implementation command-line interpreter/compiler for FHIR Shorthand (FSH).

FHIR Shorthand (FSH) is a specially-designed language for defining the content of FHIR Implementation Guides (IG). It is simple and compact, with tools to produce Fast Healthcare Interoperability Resources (FHIR) profiles, extensions and implementation guides (IG). Because it is a language, written in text statements, FHIR Shorthand encourages distributed, team-based development using conventional source code control tools such as Github.

For more information about the evolving FSH syntax see the FHIR Shorthand Reference Manual.

SUSHI requires Node.js to be installed on the user's system. Users should install Node.js 12 (LTS), although the previous LTS version (Node.js 10) is also expected to work.

Once Node.js is installed, run the following command to install or update SUSHI:

$ npm install -g fsh-sushi

After installation, the sushi commandline will be available on your path:

$ sushi --help
Usage: sushi [path-to-fsh-defs] [options]

Options:
  -o, --out <out>  the path to the output folder
  -d, --debug      output extra debugging information
  -s, --snapshot   generate snapshot in Structure Definition output (default: false)
  -v, --version    print SUSHI version
  -h, --help       output usage information

Additional information:
  [path-to-fsh-defs]
    Default: "."
    If fsh/ subdirectory present, it is included in [path-to-fsh-defs]
  -o, --out <out>
    Default: "build"
    If fsh/ subdirectory present, default output is one directory above fsh/

SUSHI supports publishing implementation guides via the new template-based IG Publisher. The template-based publisher is still being developed by the FHIR community. See the Guidance for HL7 IG Creation for more details.

If the input folder (i.e., "FSH Tank") contains a sub-folder named "ig-data", then SUSHI will generate a basic Implementation Guide project that can be built using the template-based IG Publisher. SUSHI currently supports limited customization of the IG via the following files:

• ig-data/ig.ini: If present, the user-provided igi.ini values will be merged with SUSHI-generated ig.ini.
• ig-data/package-list.json: If present, it will be used instead of a generated package-list.json.
• ig-data/input/ignoreWarnings.txt: If present, will be copied into the corresponding location in the IG input folder
• ig-data/input/images/*: If present, image files will be copied into the IG input and can be referenced by user-provided pages.
• ig-data/input/pagecontent/index.[md|xml]: If present, it will provide the content for the IG's main page.
• ig-data/input/pagecontent/*.[md|xml]: If present, these files will be generated as individual pages in the IG and will be present in the table of contents.
• ig-data/input/pagecontent/{name-of-resource-file}-[intro|notes].[md|xml]: If present, these files will place content directly on the relevant resource page. Intro files will place content before the resource definition; notes files will place content after.
• ig-data/input/pagecontent/*: If present, all other files of any type that do not match the above patterns will be copied into the IG input, but will not appear in the table of contents.
• ig-data/input/{supported-resource-input-folder}: JSON files found in supported subfolders will be processed as additional resources in the IG, added to the IG JSON file, and copied to the corresponding locations in the IG input. This allows for a "Bring Your Own JSON" capability when you already have resources that were generated via another method. If a provided JSON file has the same URL as a FSH-defined resource, the provided JSON file will overwrite the FSH-defined one. Supported folders match those documented in the Guidance for FHIR IG Creation: capabilities, examples, extensions, models, operations, profiles, resources, vocabulary.

After running SUSHI, change to the output folder and run the _updatePublisher script and enter y|Y to download latest publisher when prompted. Then run the _genonce or _gencontinuous script. _genonce will generate the IG one time. _gencontinuous will continuously run the genonce script, which will regenerate the IG with any changes that were made to the input of the IG build.

If the input folder does not contain a sub-folder named "ig-data", then only the resources (e.g., profiles, extensions, etc.) will be generated.

SUSHI is a TypeScript project. At a minimum, SUSHI requires Node.js to build, test, and run the CLI. Developers should install Node.js 12 (LTS), although the previous LTS version (Node.js 10) is also expected to work.

Once Node.js is installed, run the following command from this project's root folder:

$ npm install

The following NPM tasks are useful in development:

To run any of these tasks, use npm run. For example:

$ npm run check

For the best experience, developers should use Visual Studio Code with the following plugins:

• ESLint
• Prettier - Code formatter
  • Consider configuring the formatOnSave feature in VS Code settings:

    "[typescript]": {
        "editor.formatOnSave": true
    }

• vscode-language-fsh

Copyright 2019 Health Level Seven International

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.