shutterbug 0.5.4 → 0.5.5

checksums.yaml

@@ -1,15 +1,15 @@
 ---
 !binary "U0hBMQ==":
   metadata.gz: !binary |-
-    MzU4MGE0MGI2N2VkZjE5N2U4MGFmMTAyN2FiNGQ1OWJjOGUxMDMxNA==
+    MmI4MzNiYzkwMWQzYmJlY2I2NmY2NmNjMWE0NmMwNjEzY2RjNzU2ZQ==
   data.tar.gz: !binary |-
-    OGIxYzNmYjNhZDhjNDI2MmE2NWEwYmI2OWY2MTEzYjllNzAxZmEyYw==
+    YmNjNzEzYjkyNDViMDc1MzdmZmNiYTYwMWFhZTM1YmYwNzUxZTRmOA==
 SHA512:
   metadata.gz: !binary |-
-    MDc2YWQ3MmRiMmZlMzZiOTM2YmRlZmQ0ZWZlOTIzMzllMThmOGJmMDhmYzc5
-    MzUwODYwNDIyMmRjYmMyMzU1OWE2ODJlZDRhZmZkMzk3YTQyNTRkZTA1NGRh
-    ZTdjNGZjZGZkODE1YTkxMTc2MGEzMWE0MjgzOTM0NTE4YjhjYzk=
+    MTVhN2Q1MjhjYWNmYTQ3MDI2NGJkZjI4ZjM0NjAwNDBlM2RiZWU2OWM2ZDhh
+    OGIxZjlhNWZjZjFlMTVkZWJkNmY0MzliYWZkY2VmMTBhYTRjMGY0NDAyZGRl
+    NzhlZmVlZWJiMTRjOTRiYzA4ZjkyNGI0ZGQxNWMzYjk2NWE4NWE=
   data.tar.gz: !binary |-
-    MGZhNzI1MTQxMjg5YWZjMWQ3MGNjODNjMzlkOGJjOGYwZWY3MGMyODBhZTc3
-    YzYzNjMxNWZmNzcxMWUxMzFhYjU1MDAyYzE1YTMyYmIxNzE3YTBhNDkyOTZi
-    ZWM3NGQ0MDA1YTRmOWJlMmM2ZTMwYjU4YTQ2ZjFjMmQ4OTZkZWU=
+    OGUzOWNiMTllMGRlODFmOTJlOTZjMDM4OWE5OTM1MGYzNjkzNWEzN2FiMmQz
+    OTgzYWRkNzhlMjBkYjFjY2I3MDkzYzU1YjAxZjZiZTYyNjA1NjIyZDQ0N2Y4
+    Yzg3NGE3YzMwNDQ3N2Y2MzE5NjJmNGRmY2ZiOTRjZmY1YWU2ZDg=

data/README.md

@@ -91,6 +91,9 @@ And a Procfile which looks like this:
 
 ## Changes ##
 
+*  August 25, 2015 – v 0.5.5
+  * More aggressive single quote replacement in phantom_job.rb
+
 *  August 14, 2015 – v 0.5.4
   * Fix single quote encoding issue in phantom_job.rb. DM:
   > Phantom's html entity decoder borks on single quotes when you have an

data/lib/shutterbug.rb

@@ -1,5 +1,5 @@
 module Shutterbug
-  VERSION = "0.5.4"
+  VERSION = "0.5.5"
   autoload :Rackapp,        "shutterbug/rackapp"
   autoload :Configuration,  "shutterbug/configuration"
   autoload :Storage,        "shutterbug/storage"

data/lib/shutterbug/rasterize.js

@@ -51,6 +51,8 @@
 var system = require('system'),
     fs = require('fs'),
     baseRegEx = /<\s*base[^>]+href\s*=\s*['"]([^'"]*)['"][^>]*>/i,
+    baseFixupRegEx = /(&((amp;)*)lt;\s*base\s+href\s*=\s*)(['"])([^'"]*)(['"])([^&]*&((amp;)*)gt;)/gi,
+    metaFixupRegEx = /(&((amp;)*)lt;\s*meta\s+content\s*=\s*)(['"])([^'"]*)(['"])(\s+http-equiv\s*=\s*)(['"])([^'"]*)(['"])([^&]*&(((amp;))*)gt;)/gi,
     viewportSize = { width: 1000, height: 700 },
     filename, output, quality, size, base, html;
 
@@ -71,10 +73,56 @@ if (system.args.length > 3) {
 // let 'er rip!
 renderPage();
 
+// converts &quot; to &amp;quot; based on the number of already escaped amps found in the fixup regex
+function escapeQuote(allAmps) {
+  var quote = "&quot;",
+      numEscapes = allAmps.split("amp").length - 1,  // &('')lt; => 1 - 1 = 0 => &quot;, &(amp;)lt; => 2 - 1 = 1 => &amp;quot;, &(amp;amp;)lt; => 3 - 1 = 2 => &amp;amp;quot;
+      i;
+  for (i = 0; i < numEscapes; i++) {
+    quote = quote.replace('&', '&amp;');
+  }
+  return quote;
+}
+
+function fixupBase($0,$beforeQuote,$allAmps,$innerAmps,$leftQuote,$insideQuote,$rightQuote,$afterQuote) {
+  var escapedQuote = escapeQuote($allAmps);
+  return [$beforeQuote, escapedQuote, decodeURIComponent($insideQuote), escapedQuote, $afterQuote].join('');
+}
+
+function fixupMeta($0,$beforeQuotes,$allAmps,$innerAmps,$leftFirstQuote,$insideFirstQuote,$rightFirstQuote,$betweenQuotes,$leftSecondQuote,$insideSecondQuote,$rightSecondQuote,$afterQuotes) {
+  var escapedQuote = escapeQuote($allAmps);
+  return [$beforeQuotes, escapedQuote, decodeURIComponent($insideFirstQuote), escapedQuote, $betweenQuotes, escapedQuote, decodeURIComponent($insideSecondQuote), escapedQuote, $afterQuotes].join('');
+}
+
+function testFixup(regex, fixup, src, fixed) {
+  var result = src.replace(regex, fixup);
+  if (result !== fixed) {
+    console.log("Expected '" + fixed + "' got '" + result + "'")
+  }
+}
+
 // Main function, called once
 function renderPage() {
+    var contents = fs.read(filename);
+
+    // change to true for testing
+    if (false) {
+      console.log("Testing base fixup...");
+      testFixup(baseFixupRegEx, fixupBase, "&lt;base href='http://concord-consortium.github.io/lara-interactive-api/'&gt;", "&lt;base href=&quot;http://concord-consortium.github.io/lara-interactive-api/&quot;&gt;");
+      testFixup(baseFixupRegEx, fixupBase, "&amp;lt;base href='http://concord-consortium.github.io/lara-interactive-api/'&amp;gt;", "&amp;lt;base href=&amp;quot;http://concord-consortium.github.io/lara-interactive-api/&amp;quot;&amp;gt;");
+      testFixup(baseFixupRegEx, fixupBase, "&amp;amp;lt;base href='http://concord-consortium.github.io/lara-interactive-api/'&amp;amp;gt;", "&amp;amp;lt;base href=&amp;amp;quot;http://concord-consortium.github.io/lara-interactive-api/&amp;amp;quot;&amp;amp;gt;");
+      console.log("Testing meta fixup...");
+      testFixup(metaFixupRegEx, fixupMeta, "&lt;meta content='text/html;charset=utf-8' http-equiv='Content-Type'&gt;", "&lt;meta content=&quot;text/html;charset=utf-8&quot; http-equiv=&quot;Content-Type&quot;&gt;");
+      testFixup(metaFixupRegEx, fixupMeta, "&amp;lt;meta content='text/html;charset=utf-8' http-equiv='Content-Type'&amp;gt;", "&amp;lt;meta content=&amp;quot;text/html;charset=utf-8&amp;quot; http-equiv=&amp;quot;Content-Type&amp;quot;&amp;gt;");
+      testFixup(metaFixupRegEx, fixupMeta, "&amp;amp;lt;meta content='text/html;charset=utf-8' http-equiv='Content-Type'&amp;amp;gt;", "&amp;amp;lt;meta content=&amp;amp;quot;text/html;charset=utf-8&amp;amp;quot; http-equiv=&amp;amp;quot;Content-Type&amp;amp;quot;&amp;amp;gt;");
+    }
+
+    // fixup old clients that send single quoted escaped attributes in the metadata
+    // the single quoted escaped text causes phantom to incorrectly parse the html
+    contents = contents.replace(baseFixupRegEx, fixupBase).replace(metaFixupRegEx, fixupMeta);
+
     // this is the initial load of the html provided by phantom_job.rb
-    loadPage(fs.read(filename), function (page, base, hasChangedIframes) {
+    loadPage(contents, 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 (hasChangedIframes) {

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: shutterbug
 version: !ruby/object:Gem::Version
-  version: 0.5.4
+  version: 0.5.5
 platform: ruby
 authors:
 - Noah Paessel
@@ -10,7 +10,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2015-08-14 00:00:00.000000000 Z
+date: 2015-08-25 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: bundler