Changeset 243782 in webkit

Timestamp:

Apr 2, 2019, 6:55:38 PM (7 years ago)

Author:

Chris Dumez

Message:

HTML Parser: Remove conditional parsing of <noembed> content
​https://bugs.webkit.org/show_bug.cgi?id=196514

Reviewed by Geoffrey Garen.

LayoutTests/imported/w3c:

Resync WPT after ​https://github.com/web-platform-tests/wpt/pull/15471 to gain
test coverage. Both Gecko and Blink are passing the new check, only WebKit was
failing.

• web-platform-tests/domparsing/DOMParser-parseFromString-html-expected.txt:
• web-platform-tests/domparsing/DOMParser-parseFromString-html.html:

Source/WebCore:

Our HTML Parser has raw text handling for <noembed> content only if plugins are runnable.
However, the HTML specification doesn't ask such behavior [1], and it doesn't match to
our HTML serializer. We should always handle it as raw text.

Blink already made this change in ​https://chromium-review.googlesource.com/c/1477556.

[1] ​https://html.spec.whatwg.org/multipage/parsing.html#parsing-html-fragments:noembed

No new tests, updated existing test.

• html/parser/HTMLParserOptions.cpp:

(WebCore::HTMLParserOptions::HTMLParserOptions):

• html/parser/HTMLParserOptions.h:
• html/parser/HTMLTokenizer.cpp:

(WebCore::HTMLTokenizer::updateStateFor):

• html/parser/HTMLTreeBuilder.cpp:

(WebCore::HTMLTreeBuilder::processStartTagForInBody):

Location:

trunk

Files:

• LayoutTests/imported/w3c/ChangeLog (modified) (1 diff)
• LayoutTests/imported/w3c/web-platform-tests/domparsing/DOMParser-parseFromString-html-expected.txt (modified) (1 diff)
• LayoutTests/imported/w3c/web-platform-tests/domparsing/DOMParser-parseFromString-html.html (modified) (1 diff)
• Source/WebCore/ChangeLog (modified) (1 diff)
• Source/WebCore/html/parser/HTMLParserOptions.cpp (modified) (2 diffs)
• Source/WebCore/html/parser/HTMLParserOptions.h (modified) (1 diff)
• Source/WebCore/html/parser/HTMLTokenizer.cpp (modified) (1 diff)
• Source/WebCore/html/parser/HTMLTreeBuilder.cpp (modified) (1 diff)

trunk/LayoutTests/imported/w3c/ChangeLog

@@ +1 @@
+2019-04-02  Chris Dumez  <cdumez@apple.com>
+
+        HTML Parser: Remove conditional parsing of <noembed> content
+        https://bugs.webkit.org/show_bug.cgi?id=196514
+
+        Reviewed by Geoffrey Garen.
+
+        Resync WPT after https://github.com/web-platform-tests/wpt/pull/15471 to gain
+        test coverage. Both Gecko and Blink are passing the new check, only WebKit was
+        failing.
+
+        * web-platform-tests/domparsing/DOMParser-parseFromString-html-expected.txt:
+        * web-platform-tests/domparsing/DOMParser-parseFromString-html.html:
+
 2019-04-02  Chris Dumez  <cdumez@apple.com>
 

trunk/LayoutTests/imported/w3c/web-platform-tests/domparsing/DOMParser-parseFromString-html-expected.txt

@@ -9 +9 @@
 PASS Location value 
 PASS DOMParser parses HTML tag soup with no problems 
+PASS DOMParser should handle the content of <noembed> as raw text 
 PASS DOMParser throws on an invalid enum value 
 

trunk/LayoutTests/imported/w3c/web-platform-tests/domparsing/DOMParser-parseFromString-html.html

@@ -68 +68 @@
 
 test(function() {
+   const doc = new DOMParser().parseFromString('<noembed>&lt;a&gt;</noembed>', 'text/html');
+   assert_equals(doc.querySelector('noembed').textContent, '&lt;a&gt;');
+}, 'DOMParser should handle the content of <noembed> as raw text');
+
+test(function() {
     assert_throws(new TypeError(), function() {
         new DOMParser().parseFromString("", "text/foo-this-is-invalid");

trunk/Source/WebCore/ChangeLog

@@ +1 @@
+2019-04-02  Chris Dumez  <cdumez@apple.com>
+
+        HTML Parser: Remove conditional parsing of <noembed> content
+        https://bugs.webkit.org/show_bug.cgi?id=196514
+
+        Reviewed by Geoffrey Garen.
+
+        Our HTML Parser has raw text handling for <noembed> content only if plugins are runnable.
+        However, the HTML specification doesn't ask such behavior [1], and it doesn't match to
+        our HTML serializer. We should always handle it as raw text.
+
+        Blink already made this change in https://chromium-review.googlesource.com/c/1477556.
+
+        [1] https://html.spec.whatwg.org/multipage/parsing.html#parsing-html-fragments:noembed
+
+        No new tests, updated existing test.
+
+        * html/parser/HTMLParserOptions.cpp:
+        (WebCore::HTMLParserOptions::HTMLParserOptions):
+        * html/parser/HTMLParserOptions.h:
+        * html/parser/HTMLTokenizer.cpp:
+        (WebCore::HTMLTokenizer::updateStateFor):
+        * html/parser/HTMLTreeBuilder.cpp:
+        (WebCore::HTMLTreeBuilder::processStartTagForInBody):
+
 2019-04-02  Chris Dumez  <cdumez@apple.com>
 

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

@@ -38 +38 @@
 HTMLParserOptions::HTMLParserOptions()
     : scriptEnabled(false)
-    , pluginsEnabled(false)
     , usePreHTML5ParserQuirks(false)
     , maximumDOMTreeDepth(Settings::defaultMaximumHTMLParserDOMTreeDepth)
@@ -48 +47 @@
     RefPtr<Frame> frame = document.frame();
     scriptEnabled = frame && frame->script().canExecuteScripts(NotAboutToExecuteScript);
-    pluginsEnabled = frame && frame->loader().subframeLoader().allowPlugins();
 
     usePreHTML5ParserQuirks = document.settings().usePreHTML5ParserQuirks();

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

@@ -36 +36 @@
 
     bool scriptEnabled;
-    bool pluginsEnabled;
     bool usePreHTML5ParserQuirks;
     unsigned maximumDOMTreeDepth;

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

@@ -1417 +1417 @@
         || tagName == iframeTag
         || tagName == xmpTag
-        || (tagName == noembedTag && m_options.pluginsEnabled)
+        || (tagName == noembedTag)
         || tagName == noframesTag
         || (tagName == noscriptTag && m_options.scriptEnabled))

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

@@ -788 +788 @@
         return;
     }
-    if (token.name() == noembedTag && m_options.pluginsEnabled) {
+    if (token.name() == noembedTag) {
         processGenericRawTextStartTag(WTFMove(token));
         return;