Skip to content

Commit

Permalink
fix: apply upstream patch for in-context parsing (#3116)
Browse files Browse the repository at this point in the history
**What problem is this PR intended to solve?**

Apply upstream fix for #3112

- upstream bug report
https://gitlab.gnome.org/GNOME/libxml2/-/issues/672
- upstream fix
https://gitlab.gnome.org/GNOME/libxml2/-/commit/95f2a17440568694a6df6a326c5b411e77597be2

Fixes #3112


**Have you included adequate test coverage?**

Yes

**Does this change affect the behavior of either the C or the Java
implementations?**

Only affects the C implementation.
  • Loading branch information
flavorjones committed Jan 30, 2024
2 parents 08810c7 + da300b4 commit a26536d
Show file tree
Hide file tree
Showing 3 changed files with 50 additions and 0 deletions.
1 change: 1 addition & 0 deletions CHANGELOG.md
Expand Up @@ -15,6 +15,7 @@ Nokogiri follows [Semantic Versioning](https://semver.org/), please see the [REA

* [CRuby] `XML::Reader` defaults the encoding to UTF-8 if it's not specified in either the document or as a method parameter. Previously non-ASCII characters were serialized as NCRs in this case. [#2891] (@flavorjones)
* [CRuby] Restored support for compilation by GCC versions earlier than 4.6, which was broken in v1.15.0 (540e9aee). [#3090] (@adfoster-r7)
* [CRuby] Patched upstream libxml2 to allow parsing HTML5 in the context of a namespaced node (e.g., foreign content like MathML). [#3112, #3116] (@flavorjones)


## v1.16.0 / 2023-12-27
Expand Down
@@ -0,0 +1,33 @@
From 95f2a17440568694a6df6a326c5b411e77597be2 Mon Sep 17 00:00:00 2001
From: Nick Wellnhofer <wellnhofer@aevum.de>
Date: Tue, 30 Jan 2024 13:25:17 +0100
Subject: [PATCH] parser: Fix crash in xmlParseInNodeContext with HTML
documents

Ignore namespaces if we have an HTML document with namespaces added
manually.

Fixes #672.
---
parser.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/parser.c b/parser.c
index 1038d71b..f7842ed1 100644
--- a/parser.c
+++ b/parser.c
@@ -12415,8 +12415,10 @@ xmlParseInNodeContext(xmlNodePtr node, const char *data, int datalen,
}
xmlAddChild(node, fake);

- if (node->type == XML_ELEMENT_NODE) {
+ if (node->type == XML_ELEMENT_NODE)
nodePush(ctxt, node);
+
+ if ((ctxt->html == 0) && (node->type == XML_ELEMENT_NODE)) {
/*
* initialize the SAX2 namespaces stack
*/
--
2.42.0

16 changes: 16 additions & 0 deletions test/html5/test_api.rb
Expand Up @@ -238,6 +238,22 @@ def test_node_wrap_uses_parent_node_as_parsing_context_node
assert_equal("select", el.parent.parent.name)
end

def test_parse_in_context_of_foreign_namespace
if Nokogiri.uses_libxml?("~> 2.12.0")
skip_unless_libxml2_patch("0012-parser-Fix-crash-in-xmlParseInNodeContext-with-HTML.patch")
end

# https://github.com/sparklemotion/nokogiri/issues/3112
# https://gitlab.gnome.org/GNOME/libxml2/-/issues/672
doc = Nokogiri::HTML5::Document.parse("<html><body><math>")
math = doc.at_css("math")

nodes = math.parse("mrow") # segfaults in libxml 2.12 before 95f2a174

assert_kind_of(Nokogiri::XML::NodeSet, nodes)
assert_equal(1, nodes.length)
end

describe Nokogiri::HTML5::Document do
describe "#fragment" do
it "parses text nodes in a `body` context" do
Expand Down

0 comments on commit a26536d

Please sign in to comment.