Skip to content

GitLab

T
tangled
Commit c73ad41c authored by scottmg@chromium.org's avatar scottmg@chromium.org
Browse files
Options

Fix handling of large xml documents 

libxml 2.9.0 (we updated from 2.7.x to 2.9.2 for M45) added a
default-on abort on > 10,000,000 byte documents which is a bit
arbitrary and is causing problems for users. The _HUGE option
makes it not do this. My assumption is that this is reasonable
behaviour security-wise as: 1) we were doing it this way until
recently; and 2) it shouldn't be any worse in the renderer than
just doing `for (;;) x+='y';`

(There's a big trivial xml file in
LayoutTests/http/tests/xmlhttprequest/resources/big.xml
that goes with the test, but git cl upload won't upload it. I
guess I'll dcommit just that file first?)

R=dominicc@chromium.org
BUG=528078,463958
TEST=LayoutTests/http/tests/xmlhttprequest/xmlhttprequest-big-document.html

Review URL: https://codereview.chromium.org/1316673007

git-svn-id: svn://svn.chromium.org/blink/trunk@201992 bbb929c8-8fbe-4397-9dbb-9b2b20218538
parent a9a55e9c
Expand all Hide whitespace changes
Inline Side-by-side
Showing with 55 additions and 1 deletion
third_party/WebKit/LayoutTests/http/tests/xmlhttprequest/xmlhttprequest-big-document-expected.txt 0 → 100644
View file @ c73ad41c
CONSOLE MESSAGE: line 16: [object HTMLCollection]
Test case for bug 528078: XML files fail to parse if over 10 million bytes in size
You should see PASS.
PASS
third_party/WebKit/LayoutTests/http/tests/xmlhttprequest/xmlhttprequest-big-document.html 0 → 100644
View file @ c73ad41c
<html>
<head>
<script type="text/javascript">
if (window.testRunner) {
testRunner.dumpAsText();
testRunner.waitUntilDone();
}
var xhr = new XMLHttpRequest();
xhr.onreadystatechange = function() {
if (xhr.readyState == 4 && xhr.status == 200) {
var doc = xhr.responseXML;
var children = doc.getElementsByTagName("child");
console.log(children);
document.getElementById("page").textContent = "PASS";
if (window.testRunner)
testRunner.notifyDone();
}
}
xhr.open("GET", "resources/big.xml", true);
xhr.send(null);
</script>
</head>
<body>
<p> Test case for <a href="https://code.google.com/p/chromium/issues/detail?id=528078">bug 528078</a>: XML files fail to parse if over 10 million bytes in size</p>
<p> You should see PASS.</p>
<div id="page"/>
</body>
</html>
third_party/WebKit/Source/core/xml/parser/XMLDocumentParser.cpp
View file @ c73ad41c
......@@ -734,6 +734,7 @@ PassRefPtr<XMLParserContext> XMLParserContext::createStringParser(xmlSAXHandlerP
{
initializeLibXMLIfNecessary();
xmlParserCtxtPtr parser = xmlCreatePushParserCtxt(handlers, 0, 0, 0, 0);
xmlCtxtUseOptions(parser, XML_PARSE_HUGE);
parser->_private = userData;
parser->replaceEntities = true;
return adoptRef(new XMLParserContext(parser));
......@@ -756,7 +757,8 @@ PassRefPtr<XMLParserContext> XMLParserContext::createMemoryParser(xmlSAXHandlerP
// Set parser options.
// XML_PARSE_NODICT: default dictionary option.
// XML_PARSE_NOENT: force entities substitutions.
xmlCtxtUseOptions(parser, XML_PARSE_NODICT | XML_PARSE_NOENT);
// XML_PARSE_HUGE: don't impose arbitrary limits on document size.
xmlCtxtUseOptions(parser, XML_PARSE_NODICT | XML_PARSE_NOENT | XML_PARSE_HUGE);
// Internal initialization
parser->sax2 = 1;
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment