Changeset 140441 in webkit

Timestamp:

Jan 22, 2013, 11:12:15 AM (13 years ago)

Author:

abarth@webkit.org

Message:

BackgroundHTMLParser should simulate tree building in a separate function
​https://bugs.webkit.org/show_bug.cgi?id=107569

Reviewed by Eric Seidel.

This patch cleans up BackgroundHTMLParser::pumpTokenizer to prepare for
implementing speculative parsing.

• html/parser/BackgroundHTMLParser.cpp:

(WebCore::BackgroundHTMLParser::simulateTreeBuilder):
(WebCore):
(WebCore::BackgroundHTMLParser::pumpTokenizer):

• html/parser/BackgroundHTMLParser.h:

(BackgroundHTMLParser):

• page/Settings.in:

Location:

trunk/Source/WebCore

Files:

• ChangeLog (modified) (1 diff)
• html/parser/BackgroundHTMLParser.cpp (modified) (2 diffs)
• html/parser/BackgroundHTMLParser.h (modified) (1 diff)

trunk/Source/WebCore/ChangeLog

@@ +1 @@
+2013-01-22  Adam Barth  <abarth@webkit.org>
+
+        BackgroundHTMLParser should simulate tree building in a separate function
+        https://bugs.webkit.org/show_bug.cgi?id=107569
+
+        Reviewed by Eric Seidel.
+
+        This patch cleans up BackgroundHTMLParser::pumpTokenizer to prepare for
+        implementing speculative parsing.
+
+        * html/parser/BackgroundHTMLParser.cpp:
+        (WebCore::BackgroundHTMLParser::simulateTreeBuilder):
+        (WebCore):
+        (WebCore::BackgroundHTMLParser::pumpTokenizer):
+        * html/parser/BackgroundHTMLParser.h:
+        (BackgroundHTMLParser):
+        * page/Settings.in:
+
 2013-01-22  Sheriff Bot  <webkit.review.bot@gmail.com>
 

trunk/Source/WebCore/html/parser/BackgroundHTMLParser.cpp

@@ -119 +119 @@
 }
 
+void BackgroundHTMLParser::simulateTreeBuilder(const CompactHTMLToken& token)
+{
+    if (token.type() == HTMLTokenTypes::StartTag) {
+        const String& tagName = token.data();
+        if (threadSafeMatch(tagName, SVGNames::svgTag)
+            || threadSafeMatch(tagName, MathMLNames::mathTag))
+            m_inForeignContent = true;
+
+        // FIXME: This is just a copy of Tokenizer::updateStateFor which uses threadSafeMatches.
+        if (threadSafeMatch(tagName, textareaTag) || threadSafeMatch(tagName, titleTag))
+            m_tokenizer->setState(HTMLTokenizerState::RCDATAState);
+        else if (threadSafeMatch(tagName, plaintextTag))
+            m_tokenizer->setState(HTMLTokenizerState::PLAINTEXTState);
+        else if (threadSafeMatch(tagName, scriptTag))
+            m_tokenizer->setState(HTMLTokenizerState::ScriptDataState);
+        else if (threadSafeMatch(tagName, styleTag)
+            || threadSafeMatch(tagName, iframeTag)
+            || threadSafeMatch(tagName, xmpTag)
+            || (threadSafeMatch(tagName, noembedTag) && m_options.pluginsEnabled)
+            || threadSafeMatch(tagName, noframesTag)
+            || (threadSafeMatch(tagName, noscriptTag) && m_options.scriptEnabled))
+            m_tokenizer->setState(HTMLTokenizerState::RAWTEXTState);
+    }
+
+    if (token.type() == HTMLTokenTypes::EndTag) {
+        const String& tagName = token.data();
+        if (threadSafeMatch(tagName, SVGNames::svgTag) || threadSafeMatch(tagName, MathMLNames::mathTag))
+            m_inForeignContent = false;
+        if (threadSafeMatch(tagName, scriptTag))
+            m_isPausedWaitingForScripts = true;
+    }
+
+    // FIXME: Need to set setForceNullCharacterReplacement based on m_inForeignContent as well.
+    m_tokenizer->setShouldAllowCDATA(m_inForeignContent);
+}
+
 void BackgroundHTMLParser::pumpTokenizer()
 {
@@ -126 +162 @@
     while (m_tokenizer->nextToken(m_input.current(), m_token)) {
         m_pendingTokens.append(CompactHTMLToken(m_token));
-
-        const CompactHTMLToken& token = m_pendingTokens.last();
-
-        if (token.type() == HTMLTokenTypes::StartTag) {
-            const String& tagName = token.data();
-            if (threadSafeMatch(tagName, SVGNames::svgTag)
-                || threadSafeMatch(tagName, MathMLNames::mathTag))
-                m_inForeignContent = true;
-
-            // FIXME: This is just a copy of Tokenizer::updateStateFor which doesn't use HTMLNames.
-            if (threadSafeMatch(tagName, textareaTag) || threadSafeMatch(tagName, titleTag))
-                m_tokenizer->setState(HTMLTokenizerState::RCDATAState);
-            else if (threadSafeMatch(tagName, plaintextTag))
-                m_tokenizer->setState(HTMLTokenizerState::PLAINTEXTState);
-            else if (threadSafeMatch(tagName, scriptTag))
-                m_tokenizer->setState(HTMLTokenizerState::ScriptDataState);
-            else if (threadSafeMatch(tagName, styleTag)
-                || threadSafeMatch(tagName, iframeTag)
-                || threadSafeMatch(tagName, xmpTag)
-                || (threadSafeMatch(tagName, noembedTag) && m_options.pluginsEnabled)
-                || threadSafeMatch(tagName, noframesTag)
-                || (threadSafeMatch(tagName, noscriptTag) && m_options.scriptEnabled))
-                m_tokenizer->setState(HTMLTokenizerState::RAWTEXTState);
-        }
-        if (token.type() == HTMLTokenTypes::EndTag) {
-            const String& tagName = token.data();
-            if (threadSafeMatch(tagName, SVGNames::svgTag) || threadSafeMatch(tagName, MathMLNames::mathTag))
-                m_inForeignContent = false;
-            if (threadSafeMatch(tagName, scriptTag)) {
-                m_isPausedWaitingForScripts = true;
-                m_token.clear();
-                break;
-            }
-        }
-        // FIXME: Need to set setForceNullCharacterReplacement based on m_inForeignContent as well.
-        m_tokenizer->setShouldAllowCDATA(m_inForeignContent);
         m_token.clear();
+
+        simulateTreeBuilder(m_pendingTokens.last());
+
+        if (m_isPausedWaitingForScripts)
+            break;
 
         if (m_pendingTokens.size() >= pendingTokenLimit)

trunk/Source/WebCore/html/parser/BackgroundHTMLParser.h

@@ -62 +62 @@
 
     void pumpTokenizer();
+    void simulateTreeBuilder(const CompactHTMLToken&);
 
     void sendTokensToMainThread();