jekyll-theme-apcsp 1.1.0

data/lib/jekyll-theme-cs50/constants.rb

@@ -0,0 +1,59 @@
+module CS50
+
+  PLUGINS = {
+    "jekyll-algolia" => "1.7.1",
+    "jekyll-default-layout" => "0.1.5",
+    "jekyll-optional-front-matter" => "0.3.2",
+    "jekyll-redirect-from" => "0.16.0",
+    "jekyll-titles-from-headings" => "0.5.3"
+  }.freeze
+
+  DEFAULTS = {
+    "cs50" => {
+      "local" => {
+        "day" => "numeric",
+        "hour" => "numeric",
+        "minute" => "numeric",
+        "month" => "long",
+        "timeZoneName" => "short",
+        "weekday" => "long",
+        "year" => "numeric"
+      },
+      "locale" => "en",
+      "tz" => "America/New_York"
+    },
+    "exclude" => [
+      "Gemfile",
+      "Gemfile.lock",
+      "vendor"
+    ],
+    "include" => [
+      "license.md" # For OCW
+    ],
+    "optional_front_matter" => {
+      "remove_originals" => true
+    },
+    "plugins"  => CS50::PLUGINS.keys
+  }.freeze
+
+  OVERRIDES = {
+    "kramdown" => {
+      "gfm_quirks" => "paragraph_end",
+      "hard_wrap" => false,
+      "input" => "GFM",
+      "math_engine" => "mathjax",
+      "syntax_highlighter" => "rouge",
+      "template" => ""
+    },
+    "markdown" => "kramdown",
+    "permalink" => "pretty",
+    "redirect_from" => {
+      "json" => false
+    },
+    "sass" => {
+      "style" => "compressed"
+    },
+    "theme" => "jekyll-theme-cs50"
+  }.freeze
+
+end

data/lib/jekyll-theme-cs50.rb

@@ -0,0 +1,606 @@
+require "cgi"
+require "deep_merge"
+require "digest/md5"
+require "jekyll"
+require "jekyll-redirect-from"
+require "kramdown/parser/gfm"
+require "kramdown/parser/kramdown/link"
+require "pathname"
+require "sanitize"
+require "time"
+require "uri"
+
+require "jekyll-theme-cs50/constants"
+
+# https://stackoverflow.com/a/68039635/5156190
+class Hash 
+  def deep_compact
+    compact.transform_values{|vl| vl.is_a?(Hash) ? vl.deep_compact : vl }
+  end
+end
+
+# Inspired by http://www.glitchwrks.com/2017/07/25/jekyll-plugins, https://github.com/Shopify/liquid/wiki/Liquid-for-Programmers#create-your-own-tag-blocks
+
+module CS50
+
+  # Convert Markdown to HTML, preserving indentation (so that it can still be stripped elsewhere in pipeline)
+  def self.convert(s)
+    markdown, indentation = CS50::unindent(s)
+    html = $site.find_converter_instance(::Jekyll::Converters::Markdown).convert(markdown).strip
+    CS50::indent(html, indentation)
+  end
+
+  # Format Time for time.cs50.io
+  # https://stackoverflow.com/a/19329068/5156190
+  def self.format(t)
+    t.strftime("%Y%m%dT%H%M%S%z").sub(/\+0000/, "Z")
+  end
+
+  # Indent multiline string
+  def self.indent(s, n)
+    s.gsub(/^/, " " * n)
+  end
+
+  # Sanitize string, allowing only these tags, which are a (reasonable) subset of
+  # https://developer.mozilla.org/en-US/docs/Web/Guide/HTML/Content_categories#Phrasing_content
+  def self.sanitize(s)
+    Sanitize.fragment(s, :elements => ["b", "code", "em", "i", "img", "kbd", "span", "strong", "sub", "sup"]).strip
+  end
+
+  # Parse time
+  def self.strptime(s, now = nil)
+
+    # Try YYYY-MM-DD HH:MM:SS
+    begin
+      Time.strptime(s, "%Y-%m-%d %H:%M:%S")
+    rescue
+
+      # Try YYYY-MM-DD HH:MM
+      begin
+        Time.strptime(s, "%Y-%m-%d %H:%M")
+      rescue
+
+        # Try HH:MM:SS, relative to now
+        begin
+          t = Time.strptime(s, "%H:%M:%S", now)
+        rescue
+
+          # Try HH:MM, relative to now
+          begin
+            t = Time.strptime(s, "%H:%M", now)
+          rescue
+            raise "Invalid datetime: #{s}"
+          end
+        end
+
+        # Because Time.strptime parses relative to now's date,
+        # not now's date plus time, add one day if t is in past
+        if t < now
+          t += 24 * 60 * 60
+        end
+        t
+      end
+    end
+  end
+
+  # Unindent multiline string
+  # https://github.com/mynyml/unindent/blob/master/lib/unindent.rb
+  def self.unindent(s)
+    n = s.split("\n").select {|line| !line.strip.empty? }.map {|line| line.index(/[^\s]/) }.compact.min || 0
+    s = s.gsub(/^[[:blank:]]{#{n}}/, "")
+    return s, n
+  end
+
+  # https://github.com/pathawks/liquid-md5/blob/master/lib/liquid-md5.rb
+  module MDhash
+    def md5(input)
+      Digest::MD5.hexdigest input.strip
+    end
+  end
+  Liquid::Template.register_filter(MDhash)
+
+  module Mixins
+
+    def initialize(tag_name, markup, options)
+      @tag_name = tag_name
+      @markup = markup
+      super
+    end
+
+    def render(context)
+
+      # Interpolate any variables, a la render_variable in
+      # https://github.com/jekyll/jekyll/blob/master/lib/jekyll/tags/include.rb
+      output = @markup.gsub(/\{\{.*?\}\}/) do |s|
+        Liquid::Template.parse(s).render(context)
+      end
+
+      # Quote unquoted URLs
+      output = output.gsub(/(\S+\s*=\s*".*"|\S+\s*=\s*'.*'|".*"|'.*'|\S+)/) do |s|
+          if s =~ /^#{URI::regexp}$/
+            "\"#{s}\""
+          else
+            s
+          end
+      end
+
+      # Parse any arguments
+      @args, @kwargs = [], {}
+      output.scan(/"[^"]*"|'[^']'|\S+\s*=\s*"[^"]*"|\S+\s*=\s*'[^']*'|\S+\s*=\s*\S+|\S+/).each do |s|
+        if s.start_with?("'")
+          @args.push(s.gsub(/^'|'$/, ""))
+        elsif s.start_with?('"')
+          @args.push(s.gsub(/^"|"$/, ""))
+        else
+          key , value = s.split("=", 2)
+          if value.nil?
+            @args.push(key)
+          else
+            @kwargs[key] = value
+          end
+        end
+      end
+
+      # Return any content
+      super
+    end
+  end
+
+  class Tag < Liquid::Tag
+    include Mixins
+  end
+
+  class Block < Liquid::Block
+    include Mixins
+  end
+
+  class TestTag < Tag	
+    def render(context)	
+        super	
+        puts @args.inspect	
+    end	
+    Liquid::Template.register_tag("test", self)	
+  end
+
+  class AfterBeforeBlock < Block
+
+    def render(context)
+      html = CS50::convert(super)
+
+      # Parse timestamp
+      iso8601 = CS50::strptime(@args[0]).iso8601
+
+      # Render HTML
+      "<div data-#{@tag_name}='#{iso8601}'>#{html}</div>"
+    end
+
+    Liquid::Template.register_tag("after", self)
+    Liquid::Template.register_tag("before", self)
+
+  end
+
+  class AlertBlock < Block
+
+    def render(context)
+      html = CS50::convert(super)
+      alert = (["primary", "secondary", "success", "danger", "warning", "info", "light", "dark"].include? @args[0]) ? @args[0] : ""
+      "<div class='alert' data-alert='#{alert}' role='alert'>" \
+        "#{html}" \
+      "</div>"
+    end
+
+    Liquid::Template.register_tag("alert", self)
+
+  end
+
+  class CalendarTag < Tag
+
+    # https://gist.github.com/niquepa/4c59b7d52a15dde2367a
+    def initialize(tag_name, markup, options)
+      super
+
+    end
+
+    def render(context)
+      super
+
+      # Deprecated @ctz
+      if @kwargs.key?("ctz")
+        Jekyll.logger.warn "CS50 warning: no need for @ctz anymore"
+      end
+
+      # Calendar's height
+      height = @kwargs["height"] || "480"
+
+      # Default components
+      components = {
+        height: height,
+        hl: "en_US",
+        mode: @kwargs["mode"] || "AGENDA",
+        showCalendars: "0",
+        showDate: "0",
+        showNav: "0",
+        showPrint: "0",
+        showTabs: "0",
+        showTitle: "0",
+        showTz: "1",
+        src: @args[0]
+      }
+
+      # Build URL
+      src = URI::HTTPS.build(:host => "calendar.google.com", :path => "/calendar/embed", :query => URI.encode_www_form(components))
+
+      # Render HTML
+      "<iframe data-calendar='#{src}' #{@kwargs['ctz'] ? 'data-ctz' : ''} style='height: #{height}px;'></iframe>"
+    end
+
+    Liquid::Template.register_tag("calendar", self)
+
+  end
+
+  class LocalTag < Tag
+
+    @@regex = "\d{4}-\d{2}-\d{2} \d{1,2}:\d{2}(:\d{2})?"
+
+    def render(context)
+      super
+
+      # Parse required argument
+      if @args.length < 1
+        raise "Too few arguments"
+      elsif @args.length > 2
+        raise "Too many arguments: #{@markup}"
+      end
+      t1 = CS50::strptime(@args[0])
+      local = t1.iso8601
+      path = CS50::format(t1)
+
+      # Parse optional argument
+      if @args.length == 2
+        t2 = CS50::strptime(@args[1], t1)
+        if t2 < t1
+          raise "Invalid interval: #{@markup}"
+        end
+        local += "/" + t2.iso8601
+        path += "/" + CS50::format(t2)
+      end
+
+      # Return element
+      # Even though text node will be replaced by JavaScript, ensures Kramdown can
+      # assign an ID if this element is used as a heading
+      "<span data-local='#{local}'>#{local}</span>"
+    end
+
+    Liquid::Template.register_tag("local", self)
+
+  end
+
+  class NextTag < Tag
+
+    def render(context)
+      super
+      button = CS50::sanitize(CS50::convert((@args[0]) ? CGI.escapeHTML(@args[0]) : "Next"))
+      "<button class='btn btn-dark btn-sm' data-next type='button'>#{button}</button>"
+    end
+
+    Liquid::Template.register_tag("next", self)
+
+  end
+
+  class SpoilerBlock < Block
+
+    # https://stackoverflow.com/q/19169849/5156190
+    # https://developer.mozilla.org/en-US/docs/Web/HTML/Element/button (re phrasing, but not interactive, content)
+    def render(context)
+      html = CS50::convert(super)
+      summary = CS50::sanitize(CS50::convert((@args[0]) ? CGI.escapeHTML(@args[0]) : "Spoiler"))
+      "<details>" \
+        "<summary>#{summary}</summary>" \
+        "#{html}" \
+      "</details>"
+    end
+    Liquid::Template.register_tag("spoiler", self)
+
+  end
+
+  # Inspired by https://gist.github.com/niquepa/4c59b7d52a15dde2367a
+  class VideoTag < Tag
+
+    def render(context)
+      super
+
+      # Parse YouTube URL
+      if @args[0] 
+         
+        # If YouTube player
+        if @args[0] =~ /^https?:\/\/(?:www\.)?(?:youtube\.com\/(?:[^\/\n\s]+\/\S+\/|(?:v|e(?:mbed)?)\/|\S*?[?&]v=)|youtu\.be\/)([a-zA-Z0-9_-]{11})/
+
+          # Video's ID
+          v = $1
+
+          # Default components
+          components = {
+            "modestbranding" => "0",
+            "rel" => "0",
+            "showinfo" => "0"
+          }
+
+          # Supported components
+          params = CGI::parse(URI::parse(@args[0]).query || "")
+          ["autoplay", "controls", "end", "index", "list", "modestbranding", "mute", "playlist", "rel", "showinfo", "start", "t"].each do |param|
+
+            # If param was provided
+            if params.key?(param)
+
+              # Add to components, but map t= to start=
+              if param == "t" and !params.key?("start")
+                components["start"] = params["t"].first
+              else
+                components[param] = params[param].first
+              end
+            end
+          end
+
+          # Ensure playlist menu appears
+          if !params["list"].empty? or !params["playlist"].empty?
+            components["showinfo"] = "1"
+          end
+
+          # Build URL
+          # https://support.google.com/youtube/answer/171780?hl=en
+          src = URI::HTTPS.build(:host => "www.youtube.com", :path => "/embed/#{v}", :query => URI.encode_www_form(components))
+
+          # Return HTML
+          return "<div class='ratio ratio-16x9' data-video><iframe allow='accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture' allowfullscreen class='border' data-video src='#{src}'></iframe></div>"
+
+        # If CS50 Video Player
+        elsif @args[0] =~ /^https?:\/\/video\.cs50\.io\/([^?]+)/
+          return "<iframe allow='accelerometer; autoplay; encrypted-media; gyroscope; picture-in-picture' allowfullscreen class='border' data-video src='#{@args[0]}'></iframe>"
+        end
+      end
+
+      # Static
+      return "<div class='ratio ratio-16x9'><img alt='static' class='border' data-video src='https://i.imgur.com/xnZ5A2u.gif'></div>"
+    end
+
+    Liquid::Template.register_tag("video", self)
+
+  end
+
+  # Disable relative_url filter, since we prepend site.baseurl to all absolute paths,
+  # but we're not monkey-patching Jekyll::Filters::URLFilters::relative_url, since it's used by
+  # https://github.com/benbalter/jekyll-relative-links/blob/master/lib/jekyll-relative-links/generator.rb
+  module Filters
+    def relative_url(input)
+      Jekyll.logger.warn "CS50 warning: no need to use relative_url with this theme"
+      input
+    end
+  end
+  Liquid::Template.register_filter(CS50::Filters)
+
+end
+
+Jekyll::Hooks.register :site, :after_reset do |site|
+
+  # Override site.url so that jekyll-redirect-from doesn't prepend it
+  site.config["url"] = nil
+
+  # Strip trailing slashes from site.baseurl
+  unless site.config["baseurl"].nil?
+    site.config["baseurl"] = site.config["baseurl"].sub(/\/+$/, "")
+  end
+
+  # Disable jekyll-relative-links because it prepends site.baseurl to relative links
+  if site.config.key?("plugins") and site.config["plugins"].kind_of?(Array) and site.config["plugins"].include? "jekyll-relative-links"
+    site.config["plugins"] = site.config["plugins"] - ["jekyll-relative-links"]
+    Jekyll.logger.warn "CS50 warning: no need to use jekyll-relative-links with this theme"
+  end
+
+  # Merge in theme's configuration
+  site.config = CS50::DEFAULTS.dup.deep_merge!(site.config, {:merge_nil_values => true}).deep_merge!(CS50::OVERRIDES, {:merge_nil_values => true})
+  site.config["cs50"] = site.config["cs50"].deep_compact
+end
+
+Jekyll::Hooks.register :site, :pre_render do |site, payload|
+
+  # Expose site to Kramdown's monkey patches
+  $site = site
+
+  # Site's time zone
+  # https://stackoverflow.com/a/58867058/5156190
+  ENV["TZ"] = site.config["cs50"]["tz"]
+
+  # Promote site.cs50.assign.* to global variables
+  begin
+    site.config["cs50"]["assign"].each do |key, value|
+      payload[key] = value
+    end
+  rescue
+  end
+end
+
+Jekyll::Hooks.register [:site], :post_render do |site|
+
+  # Paths of pages
+  $paths = site.pages.map { |page| page.url }
+
+  def relative_path(from, to)
+
+    # Resolve to relative path
+    if !to.start_with?("/")
+      to = from + to
+    end
+    relative = Pathname.new(to).relative_path_from(Pathname.new(from)).to_s
+
+    # If path doesn't end with a trailing slash (before any fragment or query)
+    if match = relative.match(/\A([^#\?]*[^#\/\?])([#\?].*)?\z/)
+
+      # Construct absolute path
+      absolute = match.captures[0] + "/"
+      if !absolute.start_with?("/")
+        absolute = from + absolute
+      end
+      absolute = Pathname.new(absolute).cleanpath.to_s + "/"
+
+      # If it should have a trailing slash
+      if $paths.include?(absolute)
+
+        # Append trailing slash (plus any query or fragment)
+        relative = match.captures[0] + "/"
+        if !match.captures[1].nil?
+          relative += match.captures[1]
+        end
+      end
+    end
+
+    # Return path
+    relative
+  end
+
+  # For each page
+  site.pages.each do |page|
+
+    # If HTML
+    if page.output_ext == ".html"
+
+      # Parse page, including its layout
+      doc = Nokogiri::HTML5.parse(page.output)
+  
+      # For each node in DOM
+      doc.traverse do |node|
+
+        # If meta
+        if node.name == "meta"
+
+          # Parse url
+          if node["content"] =~ /^(\d+;\s*url=["']?)(.+)(["']?)$/i
+
+            # If a local path
+            if !/^#{URI::regexp}$/.match?($2) and !$2.start_with?("//")
+
+              # Resolve as relative path
+              node["content"] = $1 + relative_path(page.dir, $2) + $3
+            end
+          end
+        end
+
+        # If one of these elements
+        {"a" => "href", "img" => "src", "link" => "href", "script" => "src"}.each do |name, attribute|
+          if node.name == name
+
+            # With a non-nil attribute
+            if !node[attribute].nil?
+
+              # If a local path
+              if node[attribute] !~ /^#{URI::regexp}$/ and !node[attribute].start_with?("//")
+
+                # Resolve as relative path
+                node[attribute] = relative_path(page.dir, node[attribute])
+              end
+
+           end
+          end
+        end
+      end
+      page.output = doc.to_html
+
+    # If CSS
+    elsif page.output_ext == ".css"
+
+      # Resolve absolute paths in url() to relative paths
+      # https://developer.mozilla.org/en-US/docs/Web/CSS/url()
+      page.output = page.output.gsub(/url\(\s*([^\)]*)\s*\)/) do |s|
+        group = "#{$1}"
+        if match = group.match(/\A'(\/.*)'\z/) # url('/...')
+          "url('" + relative_path(page.dir, match.captures[0]).to_s + "')"
+        elsif match = group.match(/\A"(\/.*)"\z/) # url("/...")
+          'url("' + relative_path(page.dir, match.captures[0]).to_s + '")'
+        elsif match = group.match(/\A(\/(.*[^'"])?)\z/) # url(/...)
+          "url(" + relative_path(page.dir, match.captures[0]).to_s + ")"
+        else
+          s
+        end
+      end
+    end
+
+    # TODO: In offline mode, base64-encode images, embed CSS (in style tags) and JS (in script tags), a la
+    # https://github.com/jekyll/jekyll-mentions/blob/master/lib/jekyll-mentions.rb and
+    # https://github.com/jekyll/jemoji/blob/master/lib/jemoji.rb
+
+  end
+end
+
+# https://github.com/jekyll/jekyll-redirect-from/blob/master/lib/jekyll-redirect-from/redirect_page.rb
+module JekyllRedirectFrom
+  class RedirectPage < Jekyll::Page
+    def set_paths(from, to)
+      @context ||= context
+      from = ensure_leading_slash(from)
+      data.merge!(
+        "permalink" => from,
+        "redirect" => {
+          "from" => from,
+          "to" => to # Omit call to absolute_url, else redirect_to misinterprets relative URLs as absolute from /
+        }
+      )
+    end
+  end
+end
+
+module Kramdown
+  module Parser
+    class GFM < Kramdown::Parser::Kramdown
+
+      def parse_autolink
+        super
+
+        # Get autolink
+        current_link = @tree.children.select{ |element| [:a].include?(element.type) }.last
+        unless current_link.nil? 
+
+            # Hide scheme and trailing slash
+            current_link.children[0].value = current_link.children[0].value.gsub(/^https?:\/\/(www.)?|\/$/, "")
+        end
+      end
+
+      def parse_link
+        super
+
+        # Get link
+        current_link = @tree.children.select{ |element| [:a].include?(element.type) }.last
+        unless current_link.nil? 
+
+          # If inline link ends with .md
+          if match = current_link.attr["href"].match(/\A([^\s]*)\.md(\s*.*)\z/)
+
+            # Rewrite as /, just as jekyll-relative-links does
+            current_link.attr["href"] = match.captures[0] + "/" + match.captures[1]
+          end
+        end
+      end
+
+      # Remember list markers
+      def parse_list
+        super
+
+        # Get list
+        current_list = @tree.children.select{ |element| [:ul].include?(element.type) }.last
+        unless current_list.nil?
+
+          # For each li
+          current_list.children.each do |li|
+
+            # Line number of li
+            location = li.options[:location]
+
+            # Determine marker, might be nested inside of a blockquote
+            # https://kramdown.gettalong.org/syntax.html#blockquotes
+            li.attr["data-marker"] = @source.lines[location-1].sub(/^[\s>]*/, "")[0]
+          end
+        end
+        true
+      end
+
+    end
+  end
+end

data/lib/liquid-tag-parser.rb

@@ -0,0 +1 @@
+# Avoids `cannot load such file -- liquid-tag-parser`