exp/html: fix foster-parenting when elements are implicitly closed When an element (like <nobr> or <p>) was implicitly closed by another start tag, it would keep foster parenting from working because the check for what was on top of the stack of open elements was in the wrong place. Move the check to addChild. Pass 2 additional tests. R=nigeltao CC=golang-dev https://golang.org/cl/6460045
diff --git a/src/pkg/exp/html/parse.go b/src/pkg/exp/html/parse.go index afdfc99..6b1f40c 100644 --- a/src/pkg/exp/html/parse.go +++ b/src/pkg/exp/html/parse.go
@@ -208,7 +208,15 @@ // addChild adds a child node n to the top element, and pushes n onto the stack // of open elements if it is an element node. func (p *parser) addChild(n *Node) { + fp := false if p.fosterParenting { + switch p.top().DataAtom { + case a.Table, a.Tbody, a.Tfoot, a.Thead, a.Tr: + fp = true + } + } + + if fp { p.fosterParent(n) } else { p.top().Add(n) @@ -222,7 +230,6 @@ // fosterParent adds a child node according to the foster parenting rules. // Section 12.2.5.3, "foster parenting". func (p *parser) fosterParent(n *Node) { - p.fosterParenting = false var table, parent *Node var i int for i = len(p.oe) - 1; i >= 0; i-- { @@ -1308,11 +1315,8 @@ return true } - switch p.top().DataAtom { - case a.Table, a.Tbody, a.Tfoot, a.Thead, a.Tr: - p.fosterParenting = true - defer func() { p.fosterParenting = false }() - } + p.fosterParenting = true + defer func() { p.fosterParenting = false }() return inBodyIM(p) }
diff --git a/src/pkg/exp/html/parse_test.go b/src/pkg/exp/html/parse_test.go index e3c69b6..e159b49 100644 --- a/src/pkg/exp/html/parse_test.go +++ b/src/pkg/exp/html/parse_test.go
@@ -389,6 +389,8 @@ `<a href="blah">aba<table><a href="foo">br<tr><td></td></tr>x</table>aoe`: true, `<a><table><a></table><p><a><div><a>`: true, `<a><table><td><a><table></table><a></tr><a></table><a>`: true, + // A similar reparenting situation involving <nobr>: + `<!DOCTYPE html><body><b><nobr>1<table><nobr></b><i><nobr>2<nobr></i>3`: true, // A <plaintext> element is reparented, putting it before a table. // A <plaintext> element can't have anything after it in HTML. `<table><plaintext><td>`: true,
diff --git a/src/pkg/exp/html/testlogs/tests26.dat.log b/src/pkg/exp/html/testlogs/tests26.dat.log index fa97b44..8f4d545 100644 --- a/src/pkg/exp/html/testlogs/tests26.dat.log +++ b/src/pkg/exp/html/testlogs/tests26.dat.log
@@ -1,6 +1,6 @@ PASS "<!DOCTYPE html><body><a href='#1'><nobr>1<nobr></a><br><a href='#2'><nobr>2<nobr></a><br><a href='#3'><nobr>3<nobr></a>" PASS "<!DOCTYPE html><body><b><nobr>1<nobr></b><i><nobr>2<nobr></i>3" -FAIL "<!DOCTYPE html><body><b><nobr>1<table><nobr></b><i><nobr>2<nobr></i>3" +PASS "<!DOCTYPE html><body><b><nobr>1<table><nobr></b><i><nobr>2<nobr></i>3" PASS "<!DOCTYPE html><body><b><nobr>1<table><tr><td><nobr></b><i><nobr>2<nobr></i>3" PASS "<!DOCTYPE html><body><b><nobr>1<div><nobr></b><i><nobr>2<nobr></i>3" PASS "<!DOCTYPE html><body><b><nobr>1<nobr></b><div><i><nobr>2<nobr></i>3"
diff --git a/src/pkg/exp/html/testlogs/tricky01.dat.log b/src/pkg/exp/html/testlogs/tricky01.dat.log index fb0390f..4d4cd0d 100644 --- a/src/pkg/exp/html/testlogs/tricky01.dat.log +++ b/src/pkg/exp/html/testlogs/tricky01.dat.log
@@ -4,6 +4,6 @@ PASS "<html>\n<dl>\n<dt><b>Boo\n<dd>Goo?\n</dl>\n</html>" PASS "<html><body>\n<label><a><div>Hello<div>World</div></a></label> \n</body></html>" PASS "<table><center> <font>a</center> <img> <tr><td> </td> </tr> </table>" -FAIL "<table><tr><p><a><p>You should see this text." +PASS "<table><tr><p><a><p>You should see this text." PASS "<TABLE>\n<TR>\n<CENTER><CENTER><TD></TD></TR><TR>\n<FONT>\n<TABLE><tr></tr></TABLE>\n</P>\n<a></font><font></a>\nThis page contains an insanely badly-nested tag sequence." PASS "<html>\n<body>\n<b><nobr><div>This text is in a div inside a nobr</nobr>More text that should not be in the nobr, i.e., the\nnobr should have closed the div inside it implicitly. </b><pre>A pre tag outside everything else.</pre>\n</body>\n</html>"