lib.umich.edu is a headless Drupal site, built with Gatsby and React as the front-end.
- Clone
lib.umich.edu.
gh repo clone mlibrary/lib.umich.edu- Install dependencies.
Navigate into your site’s directory and install dependencies with NPM.
cd lib.umich.edu/
npm install- Configuration
ROBOTSTXT_MODE=production
If ROBOTSTXT_MODE is set to production, this tells the build to make a robots.txt that allows search engines to crawl the site. If your build is not for production, then don't set this variable or set it to development if needed.
CONTEXT=production
This env var is automatically available when building on Netlify, but if you're building the site not with Netlify set CONTEXT as production, otherwise ROBOTSTXT_MODE will not work as expected.
DRUPAL_URL=https://cms.staging.lib.umich.edu/
The site by default will build and pull data from https://cms.lib.umich.edu/, but you can change that with setting DRUPAL_URL.
DRUPAL_CONCURRENT_FILE_REQUESTS
You can optionally set DRUPAL_CONCURRENT_FILE_REQUESTS to adjust the gatsby-source-drupal concurrentFileRequests option (defaults to 20) to change how many simultaneous file requests are made to the server/service. This benefits build speed, however too many concurrent file request could cause memory exhaustion depending on the machines memory size so change with caution.
For example, run:
DRUPAL_CONCURRENT_FILE_REQUESTS=60 npm start
or
DRUPAL_CONCURRENT_FILE_REQUESTS=5 npm start
And see what works best for your network connection, memory, and general situation.
- Start developing.
Start it up.
npm startIf you wish to pull data from the staging CMS rather than the production CMS, then use this command:
npm run develop:stagingIf you wish to create a production build, then use this command:
npm run start:production⏳ Hang tight! Building the site can take an average of 20 minutes.
Sometimes when changing git branches for example you might need to run Gatsby clean.
npm run clean
npm start