These instructions will get you a copy of the project up and running on your local machine for development and testing purposes. See deployment for notes on how to deploy the project on a live system.
Note: Running locally will show unpublished content that uses the `published: false` convention in frontmatter. Content with `published: false` will not be available on stage or production.
When you run `yarn start` the CSS and JS is built in parallel with the eleventy build. Once up and running both eleventy and the JS and CSS build scripts watch for changes. When something changes the site is re-built.
In development Eleventy knows nothing about the CSS and JavaScript builds. For automatic reloading of the JS and CSS, each script uses a fetch to the public API to tell browserSync there is new code and it reloads it for you.
Building for production is slightly different. The Eleventy process and the JS and CSS builds happen in series. Then a 3rd `asset-pipeline` process happens which takes the the built content from `./build` directoty and runs it through various optimizations.
In these steps the following takes place:
* Binary files are versioned with hashes in the file names.
* References to these file in CSS and JS are updated.
* CSS and JS are minified
* The HTML is processed to update the references to the assets new hash-based filenames.
All of this means that we can serve the site with far-future expires headers. If the resource is in the browser's cache it won't even make a request for it. To break the cache the resource's url needs to change. When something ie updated and the script re-run the hash in the filename will change, which won't be ccached and the browser will know to fetch it. This helps the site be fast.
Whilst the `asset-pipline` script is custom, it leverages a lot of existing libs where possible, these include Terser, postHTML, postCSS and various plugins.
At some point it's likely 11ty will have its own mechanism for wrangling assets and at that point this will no longer be required.
For the asset-pipeline script to do it's thing, all you need to do is refer to all assets with a path beginning with `/assets/` if you do that, everything else is handled for you ✨
2. Add a node with appropriate attributes, in the appropriate location, for the new page. See below for [details on how to understand the pages.json structure](#understanding-the-pagejson-structure).
3. Create a new page, nested inside a folder struture that matches the url path. For example, for permalink `/documentation/develop/best-practices-for-collecting-user-data-consents/`, you would create a file called `best-practices-for-collecting-user-data-consents.md` and place it in `documentation` > `develop`.
4. For reference on how to create a page, review the `sidebar-master-template.md` file, which lists all available modules. Some notes:
-`published: false` will withhold this content from stage and production, to publish content, remove this line.
-`skip_index: true` is used for pages that shouldn't be indexed for search results.
- When creating page sections that will be listed in the table of contents, add an ID attribute to the section container that matches the subpageitem added to Pages.yaml. If your layout requires several sections for one table of contents entry, nest your sections inside a containing element which has the ID attribute.
- Rull for creating section IDs: use the `h2` title of the section, converted to lowercase, spaces replaced with dashes, all non-alphanumeric characters removed. For example, the section `h2` title "Know your privacy settings" would be converted to `know-your-privacy-settings` for the section ID
- The first section following the "Page Hero" module should be the "Table of Contents" module: `modules/column-w-toc.html`.
- Each new page has a title and url attribute _**Note:** The url attribute must match exactly the permalink attribute of the page's front matter (including leading and trailing slashes)_
- Pages can also have a subpageitems node for sections within the page to be referenced in the table of contents for that page:
- Each subpageitem node has a title and ID attribute, where the ID matches the ID attribute of the section container _(IDs need to be added to a containing element, rather than the heading element, of the section. This is so that highlighting for that section stays active even when the section title is out of view)_
- Overview pages have category nodes for each of the sub categories for that section
- Categories have a category attibute which denotes the category title, and a pages attribute where sub pages of the overview page are listed
- The Documentation Topics section pages are nested inside a `subfolderitems` node, which creates the dropdown panel
If you don't want the page to be labelled as a draft, as and when it's ready remove `draft-label: true` from the relevant entry in `data/content-guidelines-pages.json`
The site is auto-deployed on commits to master to https://extensionworkshop-dev.allizom.org/. You can check the version on -dev with [the dev version link](https://extensionworkshop-dev.allizom.org/__version__)
### Stage Deploys
Tags with a version ending in `-stage` will be deployed to https://extensionworkshop.allizom.org/. You can check the version on stage with [the stage version link](https://extensionworkshop.allizom.org/__version__)
Tags of the format `vN.N.N` will be deployed to https://extensionworkshop.com/. You can check the version on prod with [the prod version link](https://extensionworkshop.com/__version__)
A good example tag would be `v2.0.1`.
### Version numbers
Tag versions should aim to follow follow the [semver](https://semver.org/) format.
