freeCodeCamp · isAltoTech · Jun 17, 2026
@@ -0,0 +1,85 @@
+module Docs
+  class Coldfusion
+    class CleanHtmlFilter < Filter
+      def call
+        # Listing/category pages (Tags, Functions, a category, or a guide index)
+        # use a different layout; keep their main container as-is after cleanup.
+        @doc = build_root
+
+        # Remove site chrome and interactive widgets.
+        css('nav', 'footer', 'script', 'noscript', '#cfbreak', '.newsletter').remove
+        css('.modal', '.add-example-modal-lg', '.example-modal').remove
+        css('.example-btn', '.copy-btn', '.issuebutton', '.issuecount').remove
+        css('button').remove
+
+        # Drop the "Add An Example" / edit / fork affordances.
+        css('a[href*="github.com"]', '#forkme', '#foundeo').remove
+        css('a.label.label-danger').remove # Edit links
+
+        # Clean up the breadcrumb: keep the engine-version labels (they convey
+        # ColdFusion/Lucee/BoxLang availability) but drop navigation links and
+        # the issue tracker widget.
+        if (crumb = at_css('.breadcrumb'))
+          crumb.css('.label-warning').remove
+          crumb.css('.divider').remove
+          crumb.css('a[rel="nofollow"]').remove
+          # Remove navigation breadcrumb items (CFDocs > Functions > cf45 > …)
+          # that are not engine-availability labels.
+          crumb.css('li:not(.pull-right)').each do |li|
+            li.remove unless li.at_css('.label-acf, .label-lucee, .label-boxlang, .label-railo')
+          end
+        end
+
+        # Code blocks: tag them so DevDocs applies CFML syntax highlighting.
+        css('pre.prettyprint', 'pre').each do |node|
+          node.remove_attribute('class')
+          node['data-language'] = 'coldfusion'
+        end
+
+        # Inline code: nothing special needed, but strip prettyprint hints.
+        css('code').each { |node| node.remove_attribute('class') }
+
+        # Remove now-empty wrappers left behind by the source template's many
+        # conditional blank lines.
+        css('div', 'p', 'span', 'ul', 'ol').each do |node|
+          node.remove if node.inner_html.strip.empty? && node.element_children.empty?
+        end
+
+        doc
+      end
+
+      # cfdocs splits an entry's content across the `.jumbotron` header (name,
+      # description, syntax), the `.breadcrumb`, and the main `.container`
+      # (arguments, compatibility, links, examples). Merge them into one root.
+      #
+      # NOTE: between filters the document is re-parsed as an HTML *fragment*
+      # (there is no <body>), so selectors must not depend on `body >`.
+      def build_root
+        # First .jumbotron is the page header; #cfbreak is the trailing
+        # newsletter jumbotron, which we ignore.
+        header = css('.jumbotron').reject { |n| n['id'] == 'cfbreak' }
+                                  .map { |n| n.at_css('.container') || n }
+                                  .first
+        breadcrumb = at_css('.breadcrumb')
+
+        # The main content container holds the reference sections. It is a
+        # `.container` that is not the breadcrumb and not inside a jumbotron or
+        # nav. Identify it by the section headings it contains.
+        main = css('.container').find do |node|
+          next false if node.matches?('.breadcrumb')
+          next false if node.ancestors('.jumbotron').any? || node.ancestors('nav').any?
+          node.at_css('h2, .param, .panel') || node.at_css('#examples')
+        end
+
+        root = Nokogiri::HTML.fragment('<div></div>').at_css('div')
+        root << header.dup if header
+        root << breadcrumb.dup if breadcrumb
+        root << main.dup if main
+
+        # Fall back to the full document/fragment if the expected structure is
+        # missing (e.g. some guide pages).
+        root.element_children.any? ? root : (at_css('body') || doc)
+      end
+    end
+  end
+end
@@ -0,0 +1,66 @@
+module Docs
+  class Coldfusion
+    class EntriesFilter < Docs::EntriesFilter
+      # Category/listing slugs that aggregate other entries and must not appear
+      # as entries themselves. Categories generally end in "-functions" or
+      # "-tags", but list the fixed index pages explicitly.
+      INDEX_SLUGS = %w(index tags functions all).freeze
+
+      def include_default_entry?
+        entry_page?
+      end
+
+      def get_name
+        if (h1 = at_css('#docname'))
+          h1.content.strip
+        elsif (h1 = at_css('h1'))
+          # Guide pages: use the heading text without anchor noise.
+          h1.content.strip
+        else
+          super
+        end
+      end
+
+      def get_type
+        return 'Guides' if guide_page?
+
+        # Use the second breadcrumb link (Tags / Functions) as the category.
+        crumb = css('.breadcrumb a').map { |a| a.content.strip }
+        if crumb.include?('Tags')
+          'Tags'
+        elsif crumb.include?('Functions')
+          'Functions'
+        else
+          'Guides'
+        end
+      end
+
+      private
+
+      # A real reference entry: a tag or function page. These have a `data-doc`
+      # whose value matches the slug (no spaces) and usually a `#syntax` block.
+      def entry_page?
+        return false if index_slug?
+        return true if guide_page?
+
+        doc_name = at_css('[data-doc]').try(:[], 'data-doc')
+        return false if doc_name.nil?
+        # Category pages have human titles with spaces (e.g. "String Functions").
+        !doc_name.include?(' ')
+      end
+
+      def guide_page?
+        # Guides have no breadcrumb but do have content; they are neither tags
+        # nor functions nor category indexes.
+        return false if index_slug?
+        at_css('.breadcrumb').nil? && at_css('h1')
+      end
+
+      def index_slug?
+        s = slug.to_s.downcase
+        return true if INDEX_SLUGS.include?(s)
+        s.end_with?('-functions', '-tags')
+      end
+    end
+  end
+end
@@ -0,0 +1,58 @@
+module Docs
+  class Coldfusion < UrlScraper
+    self.name = 'ColdFusion'
+    self.slug = 'coldfusion'
+    self.type = 'simple'
+    self.base_url = 'https://cfdocs.org/'
+    self.root_path = 'index.cfm'
+    self.links = {
+      home: 'https://cfdocs.org/',
+      code: 'https://github.com/foundeo/cfdocs'
+    }
+
+    html_filters.push 'coldfusion/entries', 'coldfusion/clean_html'
+
+    options[:root_title] = 'ColdFusion'
+
+    # cfdocs links categories with an encoded dash (e.g. /array%2Dfunctions);
+    # decode and clean those so entry paths look like /array-functions.
+    options[:decode_and_clean_paths] = true
+
+    # cfdocs.org renders a page for every tag/function/guide at the site root,
+    # e.g. /hash or /cfhtmltopdf. Category "listing" pages (such as /tags,
+    # /functions and /array-functions) are crawled to discover entries, but the
+    # Entries filter excludes them from the index.
+    #
+    # Skip site chrome, utilities, reports and other non-reference pages.
+    options[:skip] = %w(
+      404.cfm contributors.cfm trycf.cfm ucase.cfm llms.cfm
+      how-to-contribute opensearch.xml robots.txt)
+
+    options[:skip_patterns] = [
+      /\Aassets\b/,
+      /\Areports\b/,
+      /\Autilities\b/,
+      /\Aslack\b/,
+      /openimage/,
+      /\.json\z/,
+      /\.png\z/,
+      /\.ico\z/,
+      /\.xml\z/,
+      /\.css\z/,
+      /\.js\z/
+    ]
+
+    options[:attribution] = <<-HTML
+      &copy; 2012&ndash;present Foundeo, Inc. and the CFDocs contributors.<br>
+      Licensed under the MIT License.<br>
+      ColdFusion is a trademark of Adobe Systems Incorporated.
+    HTML
+
+    def get_latest_version(opts)
+      # CFDocs is continuously updated and has no formal version number; use the
+      # date of the latest commit as a proxy version.
+      commits = get_github_commits('foundeo', 'cfdocs', opts)
+      commits[0]['commit']['committer']['date'][0...10]
+    end
+  end
+end
@@ -0,0 +1 @@
+https://cfdocs.org/apple-touch-icon.png