shutterbug 0.5.2 → 0.5.3

checksums.yaml

@@ -1,15 +1,15 @@
 ---
 !binary "U0hBMQ==":
   metadata.gz: !binary |-
-    YWM3Y2Q1MDlmMzI3MjVkYjkzNTYzYzEyMTUwNzdhNmFjZWVhMzM1OQ==
+    YzA4MzljOTJjNWJlZGIzY2ZhNzlkNDBmOGMwNjRhOGMwZjlhZmU3ZA==
   data.tar.gz: !binary |-
-    NTZhNGExZTc0OTNhOTFjMzE3MGE0N2NlMTQ1ZjQyZTVhNmRlYzkwNw==
+    NDJhM2E3YTY1NjVmMTc2ZjhmNDZhNzFlNDk0NmY2NzQ4MzJkNGFhMg==
 SHA512:
   metadata.gz: !binary |-
-    ODM2Nzc3MGVmZmQwOTVkZWZmNzg4NTllNjMzY2JiNjlhZTg5MzNhODNkNGJh
-    NTg2NmU4MGM3OGUwZDIyMTc3ZjQ5OWUxZGJlNmQ0ZjY0ZDBjYTRhOTEwZTZl
-    NGFjYzM2Y2M4N2IzODBmZTIzZjA4Y2Y4ZDg4YzNjMzk0YTY0Y2Q=
+    OTViY2RkMGU2MjE5ZDkzOWFlNmY2ZmFmZmE0ZTU0NGExZjQxYmMwMjA3NjBj
+    ZDE3YjliM2VlOTQ0ODUxYjNmNDljZTRkMzVjYmRlN2RlMmJjNTE3NTIzYjg0
+    YzQ2NGVlMDYzOTUwNDVhMzg0YTY1ZGRiNGViNzNhM2E3NTFmNTc=
   data.tar.gz: !binary |-
-    NWVhNDM2YzlkOWExNzJiMjIwNDgxZGZhZjZmYWY2ZTNkZWU4Y2U5NjE4M2Fh
-    NDliYTI4MDc3ZTdiMzUxNDY5MTZjMTViYzc4OWVmZTU3ZmIxYjZlMWFkOWM2
-    OTgyYmFmZmVkODA4NTcwMGM3MTg1ZjI1MzE0ZGQzY2RmNTIzNGY=
+    NjdiZjg0N2RiOWY1YjJjYTJmM2U3ZjJjZTY2YzUzMmVjODIwMzAyOWZiYTFi
+    YzFkY2FiNmZiODMyMDBlZDk1MzBlNDYzNTE4NjViYzA3OTg5N2RmOGY3MGIw
+    NWFlNDdjNmY2YTNkYWIwZmY1YTZiZTlmN2NhNGM4YzEzMWVjYzc=

data/Guardfile

@@ -1,7 +1,7 @@
 # A sample Guardfile
 # More info at https://github.com/guard/guard#readme
 
-guard :rspec do
+guard :rspec, cmd: "bundle exec rspec" do
   watch(%r{^spec/.+_spec\.rb$})
   watch(%r{^lib/(.+)\.rb$})     { |m| "spec/#{m[1]}_spec.rb" }
   watch('spec/spec_helper.rb')  { "spec" }
@@ -11,4 +11,4 @@ guard 'rack', :port => 9292 do
   watch('Gemfile.lock')
   watch('config.ru')
   watch(%r{^(js|lib)/.*})
-end
+end

data/README.md

@@ -91,8 +91,14 @@ And a Procfile which looks like this:
 
 ## Changes ##
 
+* June 4, 2105 – v 0.5.3
+    * `Rasterize.js`: Also loads non-base64 encoded iframe src content when walking
+    the iFrames. Also, better mechanics for waiting for page rendering.
+    * `s3_storage.rb`: Look for existing S3 bucket using `get`, to avoid 404 errors
+    when bucket limit of 100 is hit. (github issue #15)
+
 *  May 28, 2015 – v 0.5.2
-    *  Rasterize.js: "Loads an html page and does a depth-first walk of its
+    *  `Rasterize.js`: "Loads an html page and does a depth-first walk of its
     iframes. As the walk returns to the root each iframe's src is set to the
     base64 png rendering of its contents."  The purpose of this change was to  
     allow SVG document relative resource links (such as gradients) to work,  

data/lib/shutterbug/rasterize.js

@@ -29,16 +29,16 @@
   processIFrames(iframePage, cb6)
   getIFrameSrc()
   drainIFrameQueue() <- inner frame queue
-  cb6(hasIframes=false) <- no iframes found in inner iframe html so callback called immediately
-  cb4(iframePage, iframeBase, hasIframes)
+  cb6(hasChangedIframes=false) <- no iframes found in inner iframe html so callback called immediately
+  cb4(iframePage, iframeBase, hasChangedIframes)
 
   // replace page iframe src with rendered image
   createPage(iframeHTML, iframeBase, cb7) <-- loaded in separate page so the updated iframe src is rendered correctly
   cb7(finalIFrame)
   updateIframeAndContinue(finalIFrame)
   drainIFrameQueue() <- outer page queue
-  cb3(hasIframes=true)
-  cb1(page, base, hasIframes)
+  cb3(hasChangedIframes=true)
+  cb1(page, base, hasChangedIframes)
 
   // render page
   createPage(pageHtml, pageBase, cb8)
@@ -74,10 +74,10 @@ renderPage();
 // Main function, called once
 function renderPage() {
     // this is the initial load of the html provided by phantom_job.rb
-    loadPage(fs.read(filename), function (page, base, hasIframes) {
+    loadPage(fs.read(filename), function (page, base, hasChangedIframes) {
         // At this point the page has been "walked", and if it includes iframes, loadPage() has been called on each iframe.
         // The iframe srces have been updated to a data-uri that has an image taken of the iframe contents.
-        if (hasIframes) {
+        if (hasChangedIframes) {
           // phantom does not seem to want to render changed iframe content when the src parameter is modified
           // so we need to reload it in another page
           createPage(page.content, base, function (finalPage) {
@@ -103,25 +103,55 @@ function loadPage(html, callback) {
         baselessHtml = html.replace(baseRegEx, '');
 
     createPage(baselessHtml, base, function (page) {
-        processIframes(page, function (hasIframes) {
-          callback(page, base, hasIframes);
+        processIframes(page, function (hasChangedIframes) {
+          callback(page, base, hasChangedIframes);
         });
     });
 }
 
 // Creates the phantom page object using the passed in html and base and returns it in a callback
 function createPage(html, base, callback) {
-    var page = require('webpage').create();
+    var page = require('webpage').create(),
+        resourceCount = 0;
+
+    // onLoadFinished is signaled before all the contained iframe documents are loaded
+    // so we need to keep a count of the resources requested to make sure they load before we continue on
+    page.onResourceRequested = function() {
+      resourceCount++;
+    };
+    page.onResourceReceived = function(response) {
+      // phantom sends this for each chunk (4k in testing, the end is signaled with the stage variable)
+      if (response.stage == "end") {
+        resourceCount--;
+      }
+    };
+    page.onResourceError = function() {
+      resourceCount--;
+    };
 
     page.onLoadFinished = function(status) {
+        var loadTime = (new Date()).getTime();
+
         if (status !== 'success') {
             console.log('Unable to load html!');
             phantom.exit();
         }
 
+        // wait until the next tick for any contained iframes to finish making their requests
         window.setTimeout(function () {
-          callback(page);
-        }, 200);
+            // then wait for resourceCount to be zero or the timeout to hit
+            waitForAllResourcesToLoadOrTimeout();
+        }, 0);
+
+        function waitForAllResourcesToLoadOrTimeout() {
+          var now = (new Date()).getTime();
+          if ((resourceCount === 0) || (now - loadTime > 10000)) {
+              callback(page);
+          }
+          else {
+              window.setTimeout(waitForAllResourcesToLoadOrTimeout, 0)
+          }
+        }
     };
 
     page.viewportSize = viewportSize;
@@ -133,8 +163,9 @@ function createPage(html, base, callback) {
 // with its src being the base64 encoded png render of the iframe contents
 function processIframes(page, callback) {
 
-    var iframeSrc = getIframeSrc(page),
+    var iframeSrc = getIframeDataUriSrc(page),
         iframePng = [],
+        changedFrameCount = 0,
         queueIndex = 0;
 
     // Since this is async we can't loop but instead recursivly call drainIFrameQueue()
@@ -148,14 +179,24 @@ function processIframes(page, callback) {
     function drainIFrameQueue() {
       if (queueIndex < iframeSrc.length) {
 
+          // getIframeSrc returns null for frames that don't use data-uris in the src
+          // it doesn't just skip them so that we can maintain the queueIndex
+          if (iframeSrc[queueIndex] === null) {
+            queueIndex++;
+            drainIFrameQueue();
+            return;
+          }
+
+          changedFrameCount++;
+
           // This is the recursion down the iframes.  It restarts the entire process
           // of loading so we can handle an infinite number of embedded iframes.
           // The recursion stops when it completes a depth first search of all the iframes
           // starting at the root document.
 
-          loadPage(iframeSrc[queueIndex], function (iFrame, base, hasIFrames) {
+          loadPage(iframeSrc[queueIndex], function (iFrame, base, hasChangedIframes) {
               // convert the iframe contents to a png and then set the src to it
-              if (hasIFrames) {
+              if (hasChangedIframes) {
                   // reload the iframe so the src is used
                   createPage(iFrame.content, base, function (finalIFrame) {
                       updateIframeAndContinue(finalIFrame);
@@ -177,7 +218,7 @@ function processIframes(page, callback) {
           });
       }
       else {
-          callback(iframeSrc.length > 0);
+          callback(changedFrameCount > 0);
       }
     }
 }
@@ -190,7 +231,7 @@ function renderIframeToImage(iFrame) {
 // Walks all the iframes in the current page and pushes their src onto an array that is returned.
 // Phantom.js only allows simple types to cross the evaluate() boundary so the array is marshalled
 // as a JSON string.
-function getIframeSrc(page) {
+function getIframeDataUriSrc(page) {
     return JSON.parse(page.evaluate(function () {
         var iframes = document.getElementsByTagName('iframe'),
             src = [],
@@ -203,12 +244,8 @@ function getIframeSrc(page) {
                 src.push(iframe.src.substr(mimeType.length));
             }
             else {
-                try {
-                  src.push(serializer.serializeToString(iframe.contentDocument));
-                }
-                catch (e) {
-                  src.push('');
-                }
+                // null is used to signify that the iframe doesn't use data uris
+                src.push(null);
             }
         }
         // only simple types can be returned...

data/lib/shutterbug/storage/s3_storage.rb

@@ -26,8 +26,12 @@ module Shutterbug
           :public => true)
       end
 
+      def self.lookup_bin
+        self.connection.directories.get(Configuration.instance.s3_bin) || self.create_bin
+      end
+
       def self.s3_bin
-        @s3_bin ||= self.create_bin
+        @s3_bin ||= self.lookup_bin
       end
 
       def self.write(filename)

data/lib/shutterbug.rb

@@ -1,5 +1,5 @@
 module Shutterbug
-  VERSION = "0.5.2"
+  VERSION = "0.5.3"
   autoload :Rackapp,        "shutterbug/rackapp"
   autoload :Configuration,  "shutterbug/configuration"
   autoload :Storage,        "shutterbug/storage"

data/spec/shutterbug/s3_storage_spec.rb

@@ -2,11 +2,51 @@ require 'shared_examples_for_storage'
 require 'fog'
 
 describe Shutterbug::Storage::S3Storage do
-  let(:mock_write_result) { mock(:public_url => "http://amazon.cloud/url.png")}
+  let(:s3_bin)  { nil }
+  let(:storage) { Shutterbug::Storage::S3Storage}
+  let(:mock_write_result) { double(:public_url => "http://amazon.cloud/url.png")}
   before(:each) do
-    Shutterbug::Storage::S3Storage.stub!(:write => mock_write_result)
+    storage.stub(:write => mock_write_result)
   end
 
   it_behaves_like "a storage provider" do
   end
+ 
+  describe "some class methods" do
+    before(:each) do
+      Fog.mock!
+    end
+    describe "s3_bin" do
+      subject do
+        storage
+      end
+      describe "when the bin is already configured" do
+        let(:fake_bin) { "xyzzy" }
+        it "should not call #lookup_bin or #create_bin" do
+          subject.instance_variable_set(:@s3_bin, fake_bin)
+          subject.should_not_receive(:lookup_bin)
+          subject.should_not_receive(:create_bin)
+          subject.s3_bin.should eq fake_bin
+          # reset
+          subject.instance_variable_set(:@s3_bin,nil)
+        end
+      end
+      describe "when the bin can't be found" do
+        let(:fake_bin) { "xyzzy" }
+        it "should call #create_bin" do
+          subject.stub_chain(:connection,:directories,:get).and_return(nil)
+          subject.should_receive(:create_bin).and_return(fake_bin)
+          subject.s3_bin.should eq fake_bin
+        end
+      end
+      describe "when the bin exists" do
+        let(:fake_bin) { "xyzzy" }
+        it "should not call #create_bin" do
+          subject.stub_chain(:connection,:directories,:get).and_return(fake_bin)
+          subject.s3_bin.should eq fake_bin
+        end
+      end
+    end
+  end
 end
+

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: shutterbug
 version: !ruby/object:Gem::Version
-  version: 0.5.2
+  version: 0.5.3
 platform: ruby
 authors:
 - Noah Paessel
@@ -10,7 +10,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-05-28 00:00:00.000000000 Z
+date: 2015-06-05 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler