ICGC-ARGO Doc Site

User documentation for the ICGC-ARGO platform.

Visit Site

ICGC-ARGO Platform - User Documentation

Contents

This docs site has been generated using Docusaurus.

Documentation is written in Markdown and can be found in /docs.

The website, including the Docusaurus library, custom pages and components, and all styling is kept in /website.

Dependencies

To run the docs site requires the following:

• NodeJS version 15
  • This is done best using nvm (node version manager). A good summary of this process can be found here for your reference: https://gist.github.com/d2s/372b5943bce17b964a79

Contributing

To contribute to the docs:

1. Clone the repository to your machine:

   git clone git@github.com:icgc-argo/argo-docs.git
   

2. Move into the website directory ...

   cd argo-docs/website
   

   ...and install dependencies via npm:

   npm ci
   

3. Setup Environment config:

   Copy the file argo-docs/website/.env.example to argo-docs/website/.env

   Note
   This file contains the configuration for the Algolia search used on the site. It is not necessary to set correct values for this config to get the site to run. If you would like to have working Algolia search on while working on your local dev instance, reach out to another team member for the account values.

4. Run the development server to see your edits live in the browser:

   npm start
   

   The site should open in a new page in your browser at the address: localhost:3000

5. Edit the markdown files in /docs! The content in the browser will update whenever you save the file.

   For other changes to the site, see the extended editing guide in /website:

   • add new documentation pages
   • modify the links in the Header or Footer
   • update the sidebar links

Updating Dictionary Content

The Data Dictionary content is not updated dynamically when the dictionary is updated. Instead, the dictionary content is included as static files included in this repository. This allows the docs site to be completely static and not have to load the dictionary from the source everytime it is requested.

To simplify updating the dictonary content, the process has been scripted so it can be run by following the commands below. Files will be saved in scripts/data folder and then moved to website directory by copy-data script.

1. Pull latest code and branch off develop

2. Firstly create .env file based on the .env.schema file in the scripts folder. Ensure LECTERN_ROOT and DICTIONARY_NAME variables are set.

3. From the argo-docs root directory:

cd scripts

4. Install script dependencies:

npm ci

5. Run the 'Add Dictionary' script:

npm run add

6. Follow the prompts in the script - Use keyboard arrow keys to highlight the versions you want to add. Hit enter when you have selected the desired versions.

7. The script has added files in the following places:

   • /website/static/data/schemas/{{version number}}
   • /website/static/data/schemas/diffs/{{version number}}

   And modified the following files:

   • /website/static/data/schemas/schema-versions.json
   • /website/src/pages/dictionary/data.json

   Commit all these added and modified files to git. The following command will work from the argo-docs root directory:

   git add ./website/static/data/schemas/ ./website/src/pages/dictionary/data.json
   
   git commit
   

8. Open a PR in the Argo Docs github using develop as the base branch.

Cleaning dictionary script data

npm run clean

Cleans scripts/data folder. Destructive action. Will remove everything. All versions will have to be added again.

Generate links

External projects eg. platform-ui link to docs site. Sometimes urls change so we need to update code. The following script is a convienence to help generate these new urls It outputs constants into a file called links.js

npm run links

Upgrading Docusaurus

• Increment each version and test for breaking changes
• nb. "swizzled" components are prone to breaking changes eg. prop name change
  • easy to find out new values by swizzling same component in new version of Docusaurus
• be sure to run npm run build to make sure build isn't broken
  • Docusaurus has some nice validators that run in the build step
• be sure to manual test site for breaking changes

Styling

ref: facebook/docusaurus#1763 docusaurus theme classic uses Infima stylesheets/component lib by default This is still currently a private Facebook repo so it can be hard to find out where styles are coming from/what they're doing

Best approach right now is to swizzle a component and use Emotion as we are developing our own components simiarly.