percy-capybara 0.3.0 → 0.4.0

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: f6b29e232904e95b181e02e0381bd459dfd09a49
-  data.tar.gz: e6c802f2770ba1ab8201838b315945adc3b54cf3
+  metadata.gz: 0373826717cf4a83f083c9d8d71d92dfea1b595c
+  data.tar.gz: e32236928c98afe4613579a5a9db5869ea506f90
 SHA512:
-  metadata.gz: 634f54da9d0ca8f752dc4c3f22fe151072b6064cd8045cb4a7ae280e624dd1090fa5591c898fa57ef83eec37b92049ecafaf44adcd9ef4e84d5af2a0b1b326b7
-  data.tar.gz: 997c23094a26023115a2a3245d3429329829e9136c84f3370aa9d54c75e40f6ccc8a233ab8078ac1acfbbb07bcb764443a832bee2ae3c6e15b316bb1aad83506
+  metadata.gz: bbc8b68ae8f7df03bc575f64cd6efa4c9bcb7bfbde16e478735e57d43b1e5d3bb64bd7e380db3025f81bc8c3bfdd7d8f0283b21da075258f98357ac94c9a15d9
+  data.tar.gz: 053c1512eba240ff18904cb7790f5aa8093cc8356cfe086665be2b241ec9291679d2fb2d69a09ed790e055ee8f075ba652be3a26bed671ad23ffb2ec90eac5b6

data/lib/percy/capybara.rb

@@ -25,8 +25,8 @@ module Percy
     # multi-process tests where a single build must be created before forking.
     #
     # @see Percy::Capybara::Client::Builds#initialize_build
-    def self.initialize_build
-      capybara_client.initialize_build
+    def self.initialize_build(options = {})
+      capybara_client.initialize_build(options)
     end
 
     # Finalize the current build.

data/lib/percy/capybara/client.rb

@@ -1,5 +1,7 @@
 require 'percy/capybara/client/builds'
 require 'percy/capybara/client/snapshots'
+require 'percy/capybara/loaders/native_loader'
+require 'percy/capybara/loaders/sprockets_loader'
 
 module Percy
   module Capybara
@@ -13,15 +15,33 @@ module Percy
 
       attr_reader :client
 
+      attr_accessor :sprockets_environment
+      attr_accessor :sprockets_options
+
       def initialize(options = {})
         @client = options[:client] || Percy.client
         @enabled = options[:enabled]
+
+        if defined?(Rails)
+          @sprockets_environment = options[:sprockets_environment] || Rails.application.assets
+          @sprockets_options = options[:sprockets_options] || Rails.application.config.assets
+        end
       end
 
       def enabled?
         # Only enable if in supported CI environment or local dev with PERCY_ENABLE=1.
         @enabled ||= !Percy::Client::Environment.current_ci.nil? || ENV['PERCY_ENABLE'] == '1'
       end
+
+      def initialize_loader(options = {})
+        if sprockets_environment && sprockets_options
+          options[:sprockets_environment] = sprockets_environment
+          options[:sprockets_options] = sprockets_options
+          Percy::Capybara::Loaders::SprocketsLoader.new(options)
+        else
+          Percy::Capybara::Loaders::NativeLoader.new(options)
+        end
+      end
     end
   end
 end

data/lib/percy/capybara/client/builds.rb

@@ -2,12 +2,32 @@ module Percy
   module Capybara
     class Client
       module Builds
-        def current_build
+        def current_build(options = {})
           return if !enabled?  # Silently skip if the client is disabled.
-          @current_build ||= client.create_build(client.config.repo)
+          @current_build ||= client.create_build(client.config.repo, options)
         end
         alias_method :initialize_build, :current_build
 
+        def upload_missing_build_resources(build_resources)
+          # Upload any missing build resources.
+          new_build_resources = current_build['data'] &&
+            current_build['data']['relationships'] &&
+            current_build['data']['relationships']['missing-resources'] &&
+            current_build['data']['relationships']['missing-resources']['data']
+          return 0 if !new_build_resources
+
+          if !new_build_resources.empty?
+            puts "[percy] Uploading #{new_build_resources.length} new resources..."
+          end
+          new_build_resources.each do |missing_resource|
+            sha = missing_resource['id']
+            resource = build_resources.find { |r| r.sha == sha }
+            content = resource.content || File.read(resource.path)
+            client.upload_resource(current_build['data']['id'], content)
+          end
+          new_build_resources.length
+        end
+
         def build_initialized?
           !!@current_build
         end

data/lib/percy/capybara/client/snapshots.rb

@@ -2,48 +2,13 @@ require 'set'
 require 'faraday'
 require 'httpclient'
 require 'digest'
+require 'uri'
+require 'pathname'
 
 module Percy
   module Capybara
     class Client
       module Snapshots
-        # Modified version of Diego Perini's URL regex. https://gist.github.com/dperini/729294
-        URL_REGEX = Regexp.new(
-          # protocol identifier
-          "(?:(?:https?:)?//)" +
-          "(" +
-            # IP address exclusion
-            # private & local networks
-            "(?!(?:10|127)(?:\\.\\d{1,3}){3})" +
-            "(?!(?:169\\.254|192\\.168)(?:\\.\\d{1,3}){2})" +
-            "(?!172\\.(?:1[6-9]|2\\d|3[0-1])(?:\\.\\d{1,3}){2})" +
-            # IP address dotted notation octets
-            # excludes loopback network 0.0.0.0
-            # excludes reserved space >= 224.0.0.0
-            # excludes network & broacast addresses
-            # (first & last IP address of each class)
-            "(?:[1-9]\\d?|1\\d\\d|2[01]\\d|22[0-3])" +
-            "(?:\\.(?:1?\\d{1,2}|2[0-4]\\d|25[0-5])){2}" +
-            "(?:\\.(?:[1-9]\\d?|1\\d\\d|2[0-4]\\d|25[0-4]))" +
-          "|" +
-            # host name
-            "(?:(?:[a-z\\u00a1-\\uffff0-9]-*)*[a-z\\u00a1-\\uffff0-9]+)" +
-            # domain name
-            "(?:\\.(?:[a-z\\u00a1-\\uffff0-9]-*)*[a-z\\u00a1-\\uffff0-9]+)*" +
-          ")" +
-          # port number
-          "(?::\\d{2,5})?" +
-          # resource path
-          "(?:/[^\\s\"']*)?"
-        )
-        PATH_REGEX = /\A\/[^\\s\"']*/
-        DATA_URL_REGEX = /\Adata:/
-        LOCAL_HOSTNAMES = [
-          'localhost',
-          '127.0.0.1',
-          '0.0.0.0',
-        ].freeze
-
         # Takes a snapshot of the given page HTML and its assets.
         #
         # @param [Capybara::Session] page The Capybara page to snapshot.
@@ -54,11 +19,22 @@ module Percy
         def snapshot(page, options = {})
           return if !enabled?  # Silently skip if the client is disabled.
           name = options[:name]
+          loader = initialize_loader(page: page)
+
+          # If this is the first snapshot, create the build and upload build resources.
+          if !build_initialized?
+            build_resources = loader.build_resources
+            initialize_build(resources: build_resources)
+            upload_missing_build_resources(build_resources)
+          end
+
           current_build_id = current_build['data']['id']
-          resource_map = _find_resources(page)
-          snapshot = client.create_snapshot(current_build_id, resource_map.values, name: name)
+          resources = loader.snapshot_resources
+          resource_map = {}
+          resources.each { |r| resource_map[r.sha] = r }
 
-          # Upload the content for any missing resources.
+          # Create the snapshot and upload any missing snapshot resources.
+          snapshot = client.create_snapshot(current_build_id, resources, name: name)
           snapshot['data']['relationships']['missing-resources']['data'].each do |missing_resource|
             sha = missing_resource['id']
             client.upload_resource(current_build_id, resource_map[sha].content)
@@ -69,204 +45,6 @@ module Percy
 
           true
         end
-
-        # @private
-        def _find_resources(page)
-          resource_map = {}
-          resources = []
-          resources << _get_root_html_resource(page)
-          resources += _get_css_resources(page)
-          resources += _get_image_resources(page)
-          resources.each { |resource| resource_map[resource.sha] = resource }
-          resource_map
-        end
-        private :_find_resources
-
-        # @private
-        def _get_root_html_resource(page)
-          # Primary HTML.
-          script = <<-JS
-            var htmlElement = document.getElementsByTagName('html')[0];
-            return htmlElement.outerHTML;
-          JS
-          html = _evaluate_script(page, script)
-          resource_url = page.current_url
-          Percy::Client::Resource.new(
-            resource_url, is_root: true, mimetype: 'text/html', content: html)
-        end
-        private :_get_root_html_resource
-
-        # @private
-        def _get_css_resources(page)
-          resources = []
-          # Find all CSS resources.
-          # http://www.quirksmode.org/dom/w3c_css.html#access
-          script = <<-JS
-            function findStylesRecursively(stylesheet, css_urls) {
-              if (stylesheet.href) {  // Skip stylesheet without hrefs (inline stylesheets).
-                css_urls.push(stylesheet.href);
-
-                // Remote stylesheet rules cannot be accessed because of the same-origin policy.
-                // Unfortunately, if you touch .cssRules in Selenium, it throws a JavascriptError
-                // with 'The operation is insecure'. To work around this, skip reading rules of
-                // remote stylesheets but still include them for fetching.
-                //
-                // TODO: If a remote stylesheet has an @import, it will be missing because we don't
-                // notice it here. We could potentially recursively fetch remote imports in
-                // ruby-land below.
-                var parser = document.createElement('a');
-                parser.href = stylesheet.href;
-                if (parser.host != window.location.host) {
-                  return;
-                }
-              }
-              for (var i = 0; i < stylesheet.cssRules.length; i++) {
-                var rule = stylesheet.cssRules[i];
-                // Depth-first search, handle recursive @imports.
-                if (rule.styleSheet) {
-                  findStylesRecursively(rule.styleSheet, css_urls);
-                }
-              }
-            }
-
-            var css_urls = [];
-            for (var i = 0; i < document.styleSheets.length; i++) {
-              findStylesRecursively(document.styleSheets[i], css_urls);
-            }
-            return css_urls;
-          JS
-          resource_urls = _evaluate_script(page, script)
-
-          resource_urls.each do |url|
-            next if !_should_include_url?(url)
-            response = _fetch_resource_url(url)
-            next if !response
-            sha = Digest::SHA256.hexdigest(response.body)
-            resources << Percy::Client::Resource.new(
-              url, mimetype: 'text/css', content: response.body)
-          end
-          resources
-        end
-        private :_get_css_resources
-
-        # @private
-        def _get_image_resources(page)
-          resources = []
-          image_urls = Set.new
-
-          # Find all image tags on the page.
-          page.all('img').each do |image_element|
-            srcs = []
-            srcs << image_element[:src] if !image_element[:src].nil?
-
-            srcset_raw_urls = image_element[:srcset] || ''
-            temp_urls = srcset_raw_urls.split(',')
-            temp_urls.each do |temp_url|
-              srcs << temp_url.split(' ').first
-            end
-
-            srcs.each do |url|
-              image_urls << url
-            end
-          end
-
-          # Find all CSS-loaded images which set a background-image.
-          script = <<-JS
-            var raw_image_urls = [];
-
-            var tags = document.getElementsByTagName('*');
-            var el;
-            var rawValue;
-
-            for (var i = 0; i < tags.length; i++) {
-              el = tags[i];
-              if (el.currentStyle) {
-                rawValue = el.currentStyle['backgroundImage'];
-              } else if (window.getComputedStyle) {
-                rawValue = window.getComputedStyle(el).getPropertyValue('background-image');
-              }
-              if (!rawValue || rawValue === "none") {
-                continue;
-              } else {
-                raw_image_urls.push(rawValue);
-              }
-            }
-            return raw_image_urls;
-          JS
-          raw_image_urls = _evaluate_script(page, script)
-          raw_image_urls.each do |raw_image_url|
-            temp_urls = raw_image_url.scan(/url\(["']?(.*?)["']?\)/)
-            # background-image can accept multiple url()s, so temp_urls is an array of URLs.
-            temp_urls.each do |temp_url|
-              url = temp_url[0]
-              image_urls << url
-            end
-          end
-
-          image_urls.each do |image_url|
-            # If url references are blank, browsers will often fill them with the current page's
-            # URL, which makes no sense and will never be renderable. Strip these.
-            next if image_url == page.current_url || image_url.strip.empty?
-
-            # Make the resource URL absolute to the current page. If it is already absolute, this
-            # will have no effect.
-            resource_url = URI.join(page.current_url, image_url).to_s
-
-            next if !_should_include_url?(resource_url)
-
-            # Fetch the images.
-            # TODO(fotinakis): this can be pretty inefficient for image-heavy pages because the
-            # browser has already loaded them once and this fetch cannot easily leverage the
-            # browser's cache. However, often these images are probably local resources served by a
-            # development server, so it may not be so bad. Re-evaluate if this becomes an issue.
-            response = _fetch_resource_url(resource_url)
-            next if !response
-
-            sha = Digest::SHA256.hexdigest(response.body)
-            resources << Percy::Client::Resource.new(
-              resource_url, mimetype: response.content_type, content: response.body)
-          end
-          resources
-        end
-        private :_get_image_resources
-
-        # @private
-        def _should_include_url?(url)
-          # It is a URL or a path, but not a data URI.
-          url_match = URL_REGEX.match(url)
-          data_url_match = DATA_URL_REGEX.match(url)
-          result = (url_match || PATH_REGEX.match(url)) && !data_url_match
-
-          # Is not a remote URL.
-          if url_match && !data_url_match
-            host = url_match[1]
-            result = LOCAL_HOSTNAMES.include?(host)
-          end
-
-          !!result
-        end
-
-        # @private
-        def _fetch_resource_url(url)
-          response = Percy::Capybara::HttpFetcher.fetch(url)
-          if !response
-            STDERR.puts "[percy] Warning: failed to fetch page resource, this might be a bug: #{url}"
-            return nil
-          end
-          response
-        end
-        private :_fetch_resource_url
-
-        # @private
-        def _evaluate_script(page, script)
-          script = <<-JS
-            (function() {
-              #{script}
-            })();
-          JS
-          page.evaluate_script(script)
-        end
-        private :_evaluate_script
       end
     end
   end

data/lib/percy/capybara/loaders/base_loader.rb

@@ -0,0 +1,71 @@
+module Percy
+  module Capybara
+    module Loaders
+      class BaseLoader
+        # Modified version of Diego Perini's URL regex. https://gist.github.com/dperini/729294
+        URL_REGEX = Regexp.new(
+          # protocol identifier
+          "((?:https?:)?//)" +
+          "(" +
+            # IP address exclusion
+            # private & local networks
+            "(?!(?:10|127)(?:\\.\\d{1,3}){3})" +
+            "(?!(?:169\\.254|192\\.168)(?:\\.\\d{1,3}){2})" +
+            "(?!172\\.(?:1[6-9]|2\\d|3[0-1])(?:\\.\\d{1,3}){2})" +
+            # IP address dotted notation octets
+            # excludes loopback network 0.0.0.0
+            # excludes reserved space >= 224.0.0.0
+            # excludes network & broacast addresses
+            # (first & last IP address of each class)
+            "(?:[1-9]\\d?|1\\d\\d|2[01]\\d|22[0-3])" +
+            "(?:\\.(?:1?\\d{1,2}|2[0-4]\\d|25[0-5])){2}" +
+            "(?:\\.(?:[1-9]\\d?|1\\d\\d|2[0-4]\\d|25[0-4]))" +
+          "|" +
+            # host name
+            "(?:(?:[a-z\\u00a1-\\uffff0-9]-*)*[a-z\\u00a1-\\uffff0-9]+)" +
+            # domain name
+            "(?:\\.(?:[a-z\\u00a1-\\uffff0-9]-*)*[a-z\\u00a1-\\uffff0-9]+)*" +
+          ")" +
+          # port number
+          "(:\\d{2,5})?" +
+          # resource path
+          "(/[^\\s\"']*)?"
+        )
+
+        attr_reader :page
+
+        # @param [Capybara::Session] page The Capybara page.
+        def initialize(options = {})
+          @page = options[:page]
+        end
+
+        def build_resources
+          raise NotImplementedError.new('subclass must implement abstract method')
+        end
+
+        def snapshot_resources
+          raise NotImplementedError.new('subclass must implement abstract method')
+        end
+
+        # @private
+        def root_html_resource
+          Percy::Client::Resource.new(
+            current_path, is_root: true, mimetype: 'text/html', content: page.html)
+        end
+
+        # Transformed version of the current URL to be a relative path.
+        # This important because Rack::Test uses "www.example.com" as the actual current URL,
+        # which would force Percy to actually render example.com instead of the page. By always
+        # using a URL path as the resource URL, we guarantee that Percy will render what it's given.
+        #
+        # @private
+        def current_path
+          current_url = page.current_url
+          url_match = URL_REGEX.match(current_url)
+          return url_match[4] if url_match
+          current_url
+        end
+      end
+    end
+  end
+end