Merge pull request #731 from rubytoolbox/co-readme-display

README display

Gemfile

@@ -61,6 +61,9 @@ gem "http"
 
 gem "sidekiq"
 
+gem "sanitize"
+gem "truncato"
+
 gem "redcarpet"
 gem "slim-rails"
 

Gemfile.lock

@@ -144,6 +144,7 @@ GEM
       ruby_parser (~> 3.10)
     hashdiff (1.0.1)
     high_voltage (3.1.2)
+    htmlentities (4.3.4)
     http (4.4.1)
       addressable (~> 2.3)
       http-cookie (~> 1.0)
@@ -199,6 +200,8 @@ GEM
     nio4r (2.5.3)
     nokogiri (1.10.10)
       mini_portile2 (~> 2.4.0)
+    nokogumbo (2.0.2)
+      nokogiri (~> 1.8, >= 1.8.4)
     notiffany (0.1.3)
       nenv (~> 0.1)
       shellany (~> 0.0)
@@ -330,6 +333,10 @@ GEM
     ruby_parser (3.14.1)
       sexp_processor (~> 4.9)
     rubyzip (2.3.0)
+    sanitize (5.2.1)
+      crass (~> 1.0.2)
+      nokogiri (>= 1.8.0)
+      nokogumbo (~> 2.0)
     sass (3.7.4)
       sass-listen (~> 4.0.0)
     sass-listen (4.0.0)
@@ -379,6 +386,9 @@ GEM
     thread_safe (0.3.6)
     tilt (2.0.10)
     timecop (0.9.1)
+    truncato (0.7.11)
+      htmlentities (~> 4.3.1)
+      nokogiri (>= 1.7.0, <= 2.0)
     turbolinks (5.2.1)
       turbolinks-source (~> 5.2)
     turbolinks-source (5.2.0)
@@ -470,6 +480,7 @@ DEPENDENCIES
   rubocop-performance
   rubocop-rails
   rubocop-rspec
+  sanitize
   sass-rails (~> 5.0)
   selenium-webdriver
   sidekiq
@@ -478,6 +489,7 @@ DEPENDENCIES
   spring
   spring-watcher-listen (~> 2.0.0)
   timecop
+  truncato
   turbolinks (~> 5)
   tzinfo-data
   uglifier (>= 1.3.0)

app/assets/stylesheets/components/readme.sass

@@ -0,0 +1,16 @@
+.readme
+  .label
+    @extend .column, .is-full, .is-2-widescreen
+    span
+      @extend .is-size-5
+      position: sticky
+      top: 90px
+      padding-bottom: 12px
+      border-bottom: 3px solid $primary
+      margin-bottom: 24px
+
+    i.fa
+      @extend .has-text-grey-light
+
+  .content
+    @extend .column

app/helpers/component_helpers.rb

@@ -55,6 +55,10 @@ def project_health_tag(health_status)
     render "components/project_health_tag", status: health_status
   end
 
+  def project_readme(readme)
+    render "components/project/readme", readme: readme
+  end
+
   def small_health_indicator(project)
     render "components/small_health_indicator", project: project
   end

app/models/github/readme.rb

@@ -1,11 +1,67 @@
 # frozen_string_literal: true
 
 class Github::Readme < ApplicationRecord
+  module Scrubber
+    class << self
+      #
+      # Sanitizes given html, drops named anchors and adjusts relative
+      # links to be based on given base_url (if passed, otherwise that step
+      # is skipped)
+      #
+      def scrub(html, base_url: nil)
+        return if html.blank?
+
+        sanitized = Sanitize.fragment(html, Sanitize::Config::RELAXED)
+
+        fix_links sanitized, base_url: base_url
+      end
+
+      private
+
+      def fix_links(sanitized, base_url:)
+        doc = Nokogiri::HTML.fragment sanitized
+
+        doc.css("a[href]").each do |a|
+          adjust_link a, base_url: base_url
+        end
+
+        doc.to_s
+      end
+
+      def adjust_link(link, base_url:)
+        href = link["href"]
+
+        # Scrub links to named anchors
+        if href.start_with?("#")
+          link.replace link.inner_html
+        # Relative links get aligned to base_url, depending on whether
+        # it's an absolute or relative path
+        elsif base_url && href.exclude?("://")
+          link["href"] = if href.start_with?("/")
+                           URI.join base_url, href
+                         else
+                           File.join base_url, href
+                         end
+        end
+      end
+    end
+  end
+
   self.primary_key = :path
   self.table_name = :github_readmes
 
   belongs_to :github_repo,
              primary_key: :path,
              foreign_key: :path,
              inverse_of:  :readme
+
+  def html=(html)
+    super Scrubber.scrub(html, base_url: github_repo&.blob_url)
+  end
+
+  def truncated_html(limit: 2000)
+    return if html.blank?
+
+    Truncato.truncate(html, max_length: limit)
+  end
 end

app/models/github_repo.rb

@@ -45,6 +45,12 @@ def url
     File.join "https://github.com", path
   end
 
+  def blob_url
+    return unless default_branch
+
+    File.join url, "blob", default_branch
+  end
+
   def issues_url
     File.join(url, "issues") if has_issues?
   end

app/models/project.rb

@@ -122,12 +122,15 @@ def self.search(query, order: Project::Order.new(directions: Project::Order::SEA
            :pull_request_acceptance_rate,
            :average_recent_committed_at,
            :sibling_gem_with_most_downloads,
+           :readme,
            to:        :github_repo,
            allow_nil: true,
            prefix:    :github_repo
 
   def self.find_for_show!(permalink)
-    includes_associations.find(Github.normalize_path(permalink))
+    includes_associations
+      .includes(github_repo: :readme)
+      .find Github.normalize_path(permalink)
   end
 
   def permalink=(permalink)

app/views/components/project/_readme.html.slim

@@ -0,0 +1,9 @@
+- if readme
+  section.readme.section: .container: .columns.is-multiline
+    .label
+      span
+        i.fa.fa-book  
+        strong Project Readme
+
+    .content
+      = readme.html.html_safe

app/views/pages/components/project_readme.html.slim

@@ -0,0 +1,12 @@
+.hero
+  section.section: .container
+    p.heading= link_to "Ruby Toolbox UI Components Styleguide", "/pages/components"
+    h2= current_page.split("/").last.humanize
+
+- readme = Github::Readme.new(html: "<h1>Hello World!</h1><p>Hello World</p>")
+
+= component_example "Project README" do
+  = project_readme readme
+
+= component_example "Project README is not present" do
+  = project_readme nil

app/views/projects/show.html.slim

@@ -21,7 +21,6 @@
 
   .columns: .column= project_health_tags @project
 
-
   .columns: .links.column
     = project_links @project
 
@@ -38,4 +37,4 @@
 section.section: .container: .project
   = project_metrics @project, expanded_view: true
 
-section.section: .container
+= project_readme @project.github_repo_readme

spec/features/project_spec.rb

@@ -0,0 +1,23 @@
+# frozen_string_literal: true
+
+require "rails_helper"
+
+RSpec.describe "Project Display", type: :feature do
+  let(:project) do
+    Factories.project "widgets"
+  end
+
+  it "can display Project README" do
+    visit project_path(project)
+    expect(page).not_to have_selector(".readme")
+
+    project.github_repo.create_readme! html: "<strong>some content</strong>", etag: "1234"
+    visit project_path(project)
+
+    expect(page).to have_selector(".readme")
+
+    within ".readme" do
+      expect(page).to have_text("some content")
+    end
+  end
+end

spec/models/github/readme_spec.rb

@@ -0,0 +1,82 @@
+# frozen_string_literal: true
+
+require "rails_helper"
+
+RSpec.describe Github::Readme, type: :model do
+  let(:repo) do
+    GithubRepo.new(
+      path:           "foo/bar",
+      default_branch: "main"
+    )
+  end
+
+  let(:model) do
+    described_class.new(
+      html:        "<p>Hello World</p>",
+      etag:        "123123",
+      github_repo: repo
+    )
+  end
+
+  describe "html=" do
+    it "passes the input through the scrubber" do
+      base_url = "https://example.com/foo"
+      allow(model.github_repo).to receive(:blob_url).and_return(base_url)
+      allow(Github::Readme::Scrubber).to receive(:scrub)
+        .with("input html", base_url: base_url)
+        .and_return("scrubbed")
+
+      model.html = "input html"
+
+      expect(model.html).to be == "scrubbed"
+    end
+  end
+
+  describe "#truncated_html" do
+    it "is nil when html is empty" do
+      model.html = " "
+      expect(model.truncated_html).to be nil
+    end
+
+    it "returns truncated html" do
+      model.html = '<a href="https://example.com">Hello</a><p>More</p>'
+      expect(model.truncated_html(limit: 40)).to be == "<a href='https://example.com'>Hello</a><p>...</p>"
+    end
+  end
+
+  describe Github::Readme::Scrubber do
+    describe ".scrub" do
+      it "returns nil if html is blank" do
+        expect(described_class.scrub(" \n ")).to be nil
+      end
+
+      it "cleans weird html content" do
+        html = %q{<a href="/" onclick="alert('lol');">Hello</a>}
+        expect(described_class.scrub(html)).to be == '<a href="/">Hello</a>'
+      end
+
+      it "removes links to named anchors" do
+        html = '<p><a href="#foobar">Hello</a></p>'
+        expect(described_class.scrub(html)).to be == "<p>Hello</p>"
+      end
+
+      # rubocop:disable RSpec/ExampleLength
+      it "exchanges relative links with base url when given" do
+        html = <<~HTML
+          <p><a href="https://example.com">Unchanged</a></p>
+          <p><a href="foo/relative">Changed</a></p>
+          <p><a href="/absolute">Changed too</a></p>
+        HTML
+
+        expected = <<~HTML
+          <p><a href="https://example.com">Unchanged</a></p>
+          <p><a href="https://example.com/subpath/foo/relative">Changed</a></p>
+          <p><a href="https://example.com/absolute">Changed too</a></p>
+        HTML
+
+        expect(described_class.scrub(html, base_url: "https://example.com/subpath")).to be == expected
+      end
+      # rubocop:enable RSpec/ExampleLength
+    end
+  end
+end

spec/models/github_repo_spec.rb

@@ -55,6 +55,16 @@ def create_repo!(path:, updated_at: 1.day.ago)
     end
   end
 
+  describe "#blob_url" do
+    it "is derived from the repo path and default_branch" do
+      expect(described_class.new(path: "foo/bar", default_branch: "main").blob_url).to be == "https://github.com/foo/bar/blob/main"
+    end
+
+    it "is nil when there is no default_branch" do
+      expect(described_class.new(path: "foo/bar", default_branch: nil).blob_url).to be nil
+    end
+  end
+
   describe "#wiki_url" do
     it "is nil when has_wiki is false" do
       expect(described_class.new(has_wiki: false).wiki_url).to be_nil

spec/models/project_spec.rb

@@ -26,6 +26,19 @@
     end
   end
 
+  describe ".find_for_show!" do
+    let(:project) do
+      Factories.project "sample"
+    end
+
+    it "eager-loads readme if present" do
+      project.github_repo.create_readme! html: "hello world", etag: "1234"
+
+      found_instance = described_class.find_for_show!(project.permalink)
+      expect { found_instance.github_repo_readme }.not_to make_database_queries
+    end
+  end
+
   describe ".with_bugfix_forks" do
     before do
       Factories.project "regular"