html: avoid using raw text mode even when nested noscript tags
Assuming "in head noscript" insertion mode, the scripting flag will be disabled.
Thus, even if nested noscript tags exist,
the tokenizer should not go into the raw text mode.
This change makes the following test happy:
<head><noscript><noscript class="foo"><!--foo--></noscript>
Change-Id: I2620e751d8be3d313c3a2e2f992b1e21ce2dc2ee
Reviewed-on: https://go-review.googlesource.com/c/net/+/263878
Trust: Kunpei Sakai <namusyaka@gmail.com>
Trust: Nigel Tao <nigeltao@golang.org>
Reviewed-by: Nigel Tao <nigeltao@golang.org>
diff --git a/html/parse.go b/html/parse.go
index 392327a..f91466f 100644
--- a/html/parse.go
+++ b/html/parse.go
@@ -728,7 +728,13 @@
return inBodyIM(p)
case a.Basefont, a.Bgsound, a.Link, a.Meta, a.Noframes, a.Style:
return inHeadIM(p)
- case a.Head, a.Noscript:
+ case a.Head:
+ // Ignore the token.
+ return true
+ case a.Noscript:
+ // Don't let the tokenizer go into raw text mode even when a <noscript>
+ // tag is in "in head noscript" insertion mode.
+ p.tokenizer.NextIsNotRawText()
// Ignore the token.
return true
}
diff --git a/html/testdata/webkit/noscript01.dat b/html/testdata/webkit/noscript01.dat
new file mode 100644
index 0000000..ec3496c
--- /dev/null
+++ b/html/testdata/webkit/noscript01.dat
@@ -0,0 +1,237 @@
+#data
+<head><noscript><!doctype html><!--foo--></noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+Line: 1 Col: 31 Unexpected DOCTYPE. Ignored.
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <!-- foo -->
+| <body>
+
+#data
+<head><noscript><html class="foo"><!--foo--></noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+Line: 1 Col: 34 html needs to be the first start tag.
+#script-off
+#document
+| <html>
+| class="foo"
+| <head>
+| <noscript>
+| <!-- foo -->
+| <body>
+
+#data
+<head><noscript></noscript>
+#errors
+(1,6): expected-doctype-but-got-tag
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <body>
+
+#data
+<head><noscript> </noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| " "
+| <body>
+
+#data
+<head><noscript><!--foo--></noscript>
+#errors
+(1,6): expected-doctype-but-got-tag
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <!-- foo -->
+| <body>
+
+#data
+<head><noscript><basefont><!--foo--></noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <basefont>
+| <!-- foo -->
+| <body>
+
+#data
+<head><noscript><bgsound><!--foo--></noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <bgsound>
+| <!-- foo -->
+| <body>
+
+#data
+<head><noscript><link><!--foo--></noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <link>
+| <!-- foo -->
+| <body>
+
+#data
+<head><noscript><meta><!--foo--></noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <meta>
+| <!-- foo -->
+| <body>
+
+#data
+<head><noscript><noframes>XXX</noscript></noframes></noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <noframes>
+| "XXX</noscript>"
+| <body>
+
+#data
+<head><noscript><style>XXX</style></noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <style>
+| "XXX"
+| <body>
+
+#data
+<head><noscript></br><!--foo--></noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+Line: 1 Col: 21 Element br not allowed in a inhead-noscript context
+Line: 1 Col: 21 Unexpected end tag (br). Treated as br element.
+Line: 1 Col: 42 Unexpected end tag (noscript). Ignored.
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <body>
+| <br>
+| <!-- foo -->
+
+#data
+<head><noscript><head class="foo"><!--foo--></noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+Line: 1 Col: 34 Unexpected start tag (head).
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <!-- foo -->
+| <body>
+
+#data
+<head><noscript><noscript class="foo"><!--foo--></noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+Line: 1 Col: 34 Unexpected start tag (noscript).
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <!-- foo -->
+| <body>
+
+#data
+<head><noscript></p><!--foo--></noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+Line: 1 Col: 20 Unexpected end tag (p). Ignored.
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <!-- foo -->
+| <body>
+
+#data
+<head><noscript><p><!--foo--></noscript>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+Line: 1 Col: 19 Element p not allowed in a inhead-noscript context
+Line: 1 Col: 40 Unexpected end tag (noscript). Ignored.
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <body>
+| <p>
+| <!-- foo -->
+
+#data
+<head><noscript>XXX<!--foo--></noscript></head>
+#errors
+Line: 1 Col: 6 Unexpected start tag (head). Expected DOCTYPE.
+Line: 1 Col: 19 Unexpected non-space character. Expected inhead-noscript content
+Line: 1 Col: 30 Unexpected end tag (noscript). Ignored.
+Line: 1 Col: 37 Unexpected end tag (head). Ignored.
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <body>
+| "XXX"
+| <!-- foo -->
+
+#data
+<head><noscript>
+#errors
+(1,6): expected-doctype-but-got-tag
+(1,6): eof-in-head-noscript
+#script-off
+#document
+| <html>
+| <head>
+| <noscript>
+| <body>
