Implement highlighting of Markdown pre blocks

src/packages/markdown-preview/lib/markdown-preview-view.coffee

@@ -3,8 +3,10 @@ _ = require 'underscore'
 ScrollView = require 'scroll-view'
 {$$$} = require 'space-pen'
 roaster = require 'roaster'
-LanguageMode = require 'language-mode'
-Buffer = require 'text-buffer'
+Editor = require 'editor'
+
+fenceNameToExtension =
+  "ruby": "rb"
 
 module.exports =
 class MarkdownPreviewView extends ScrollView
@@ -63,27 +65,32 @@ class MarkdownPreviewView extends ScrollView
   setLoading: ->
     @html($$$ -> @div class: 'markdown-spinner', 'Loading Markdown...')
 
+  tokenizeCodeBlocks: (html) =>
+    html = $(html)
+    preList = $(html.filter("pre"))
+
+    for codeBlock in preList.toArray()
+      codeBlock = $(codeBlock.firstChild)
+      if className = codeBlock.attr('class')
+        fenceName = className.replace(/^lang-/, '')
+
+        if extension = fenceNameToExtension[fenceName]
+          text = codeBlock.text()
+          syntax.selectGrammar("foo.#{extension}", text)
+          if grammar = syntax.selectGrammar("foo.#{extension}", text)
+            continue if grammar is syntax.nullGrammar
+            tokens = grammar.tokenizeLines(text)
+            grouping = ""
+            for token in tokens
+              grouping += Editor.buildLineHtml(token, text)
+            codeBlock.replaceWith(grouping)
+    html
+
   fetchRenderedMarkdown: ->
     @setLoading()
     roaster(@buffer.getText(), {}, (err, html) =>
       if err
         @setErrorHtml(err)
       else
-        result = @html(html)
-        preList = result.find("pre")
-        for pre in preList
-          grammar = _.find syntax.grammars, (grammar) ->
-            return "ruby" == grammar.scopeName.split(".").pop()
-
-          if grammar
-            languageMode = new LanguageMode(this, grammar)
-            console.log pre
-            code = pre.childNodes[0]
-            text = code.textContent.split("\n")
-
-            for line in text
-              tokens = languageMode.tokenizeLine(text)
-              console.log tokens
-          #codeBuffer = new Buffer("", text)
-          #x = 1
+        @html(@tokenizeCodeBlocks(html))
     )

src/packages/markdown-preview/spec/markdown-preview-view-spec.coffee

@@ -8,6 +8,7 @@ describe "MarkdownPreviewView", ->
   beforeEach ->
     project.setPath(project.resolve('markdown'))
     buffer = project.bufferForPath('file.markdown')
+    atom.activatePackage('ruby.tmbundle', sync: true)
     preview = new MarkdownPreviewView(buffer)
 
   afterEach ->
@@ -35,7 +36,7 @@ describe "MarkdownPreviewView", ->
   fdescribe "code block tokenization", ->
     describe "when the code block's fence name has a matching grammar", ->
       it "tokenizes the code block with the grammar", ->
-        expect(preview.find("pre code.lang-ruby .entity.name.function.ruby")).toExist()
+        expect(preview.find("pre span.entity.name.function.ruby")).toExist()
 
     describe "when the code block's fence name doesn't have a matching grammar", ->
       it "does not tokenize the code block", ->