Skip to content

Latest commit

 

History

History
177 lines (113 loc) · 8.33 KB

README.md

File metadata and controls

177 lines (113 loc) · 8.33 KB
Raw

scala-lang.org

This repository contains the static source of scala-lang.org.

It does not contain the source of any content found under the docs.scala-lang.org subdomain (instead, visit the docs.scala-lang repo for that source).

This is a static site generated by Jekyll, and uses a whole host of open-source tools including a touch of Twitter's Bootstrap.

Dependencies

This site uses a Jekyll, a Ruby framework. The required Jekyll version is 3.4.0.

Building the site

There are two ways to run Jekyll to build the site:

  1. Using Bundler, so Jekyll and accompanying gems are installed only inside this directory.
  2. Using globally-installed Jekyll and accompanying gems.

The latter method is the one currently actually used on scala-lang.org. The former method is likely most convenient for users who already have a different version of Jekyll installed, or who are comfortable using Bundler and who don't want anything else installed system-wide.

Option 1) Building with Bundler

cd into the directory where you cloned this repository, then install the required gems with bundle install. This will automatically put the gems into ./bundle-vendor/bundle.

Start the server in the context of the bundle:

bundle exec jekyll serve

That's it.

If that doesn't work, to guarantee that your version of Ruby, etc, completely matches the production environment, you can also use rbenv. Start by cding into the directory where you cloned this repository, then execute the following:

rbenv install 2.3.1
rbenv local 2.3.1
rbenv rehash
gem install bundle
bundle install  # This will automatically put the gems into `./bundle-vendor/bundle`
bundle exec jekyll serve  # Start the server in the context of the bundle

From this point, everything else should be the same, regardless of which method you used to run Jekyll.

Option 2) Building with global Jekyll

Install Jekyll 3.4.0 on your system using RubyGems:

gem install jekyll -v 3.4.0

After cloning, cding into the directory where you cloned this repository and run:

jekyll serve

and watch the output. You should see something like:

 Configuration file: /Users/ben/src/scala-lang/_config.yml
             Source: /Users/ben/src/scala-lang
        Destination: /Users/ben/src/scala-lang/_site
  Incremental build: enabled
       Generating... done.
  Auto-regeneration: enabled for '/Users/ben/src/scala-lang'

Windows and UTF-8

If you get incompatible encoding errors when generating the site under Windows, then ensure that the console in which you are running jekyll can work with UTF-8 characters. As described in the blog Solving UTF problem with Jekyll on Windows you have to execute chcp 65001. This command is best added to the jekyll.bat-script.

Viewing the site

Regardless of your method of running Jekyll, the generated site is available at http://localhost:4000.

If you add --watch to your Jekyll command line, Jekyll will automatically watch for changes on the filesystem. When you change a file, the console will show that jekyll is regenerating. Wait until it says done to refresh your browser.

YAML Front Matter

The "YAML Front Matter" is nothing more than the header on each page that you intend for Jekyll to parse. It contains information such as the name of the HTML template (layout) chosen for the specific document, and the title of the document. An example YAML front matter might look like:

---
layout: page
title: My page title
---

You can use these fields in the YAML front matter later in your document. For example, to make a header with the title of the document, in Markdown you would write:

---
layout: page
title: My page title
---

# {{ page.title }}

Body text here...

# {{ page.title }} would be rendered in HTML as, <h1>My page title</h1>.

Markdown

There are dozens of guides and cheatsheets that cover Markdown syntax out there, though this screenshot from the free OS X Markdown editor, Mou, is an excellent and concise reference:

Mou screen shot

Linking to internal pages

The least error-prone way to link between documents, to link to local images, or anything else: [link text]({{ site.baseurl }}/path/to/page/page.html)

Here, {{ site.baseurl }} is a site-wide variable that represents the root directory of the static site. So, to display the Scala logo image, located in img/scala-logo.png, one must simply write: ![Img alt text]({{ site.baseurl }}/resources/img/scala-logo.png)

## Permalinks

In this new version of the scala-lang site we've tried to follow a pretty permalink style, so that any generated page will have an permalink finishing in a slash character (/). This will tell Jekyll to build that particular page as an index.html inside a folder with a name as specified in the provided permalink. i.e.: if a page has a permalink as follows:

permalink: /what-is-scala/

This will tell Jekyll to create a what-is-scala folder, with an index.html file inside. Links to this page will refer to the {{site.baseurl}}/what-is-scala/.

## Custom collections and data

In the previous version of the site, data used in different pages was contained in categorized blogs. This has been changed to use custom collections. Every custom collection is a folder starting with an underscore character (_), containing a markdown file for each member of the collection. As any markdown containing a page in the site, it starts with a YAML front matter containing the data for this item, and can optionally contain markdown text to be rendered as html.

Right now there are no collections being rendered as specific pages in the site (they're only consumed internally as static data), but in the future this can be changed by specifying the global output: true variable in the _config.yml custom collections section. You'll also need to specify a layout by using the defaults settings in the _config.yml file. i.e.:

defaults:
  - scope:
      path: ""
      type: collection_name
    values:
      layout: layout_name

To access data from a custom collection just refer to site.<collection_name>. The collection's name will be the name of its folder sans the underscore character. i.e.: to access the data inside _downloads, you can do it as follows:

site.downloads

Some of the data has been also modelled as YAML files inside the _data folder. Generally for data that is used throughout the site (i.e: the navigation bar links).

Resources and Workflow

On every commit to the scala/scala-lang repository a jenkins job will generate the site using jekyll and copy the resulting files to the webserver. NOTE: the rsync of this job also deletes whatever is in the webserver directory with explicit exceptions: we need to keep the files listed below. Kind of a hack.

There are additional files on the webserver:

  • Subdirectory scala-lang.org/old is a static copy of the old website. It was generated once and copied there, and it stays like that.
  • Most of the files in /home/linuxsoft/archives/scala/ (on chara, accessible through ssh with your LAMP account) are synchronized to the subdirectory scala-lang.org/files/archive by another hourly jenkins job. This folder is used by the nightly and release jenkins jobs to publish scala releases:
    • distribution files (tarballs etc) in /
      • older distribution files, RCs in /old/ (not sure how exactly this is split up..)
    • api docs for distributions in /api/
    • nightly builds in /nightly/distributions/
    • nightly api builds in /nightly/docs-xxx/
    • nightly pdf builds (spec etc) in /nightly/pdfs

Templates

We have the following (general) templates: (Note that this is not an exhaustive list.)

page.html

Example YAML front matter with all possible fields:

---
layout: page
title: I Haz Build: An Autobiography of the Build Kitten
by: Scala Jenkins (Build Kitty)
---