diff --git a/src/HTML5/Parser/DOMTreeBuilder.php b/src/HTML5/Parser/DOMTreeBuilder.php
index e9bca4e..8f3df88 100644
--- a/src/HTML5/Parser/DOMTreeBuilder.php
+++ b/src/HTML5/Parser/DOMTreeBuilder.php
@@ -322,6 +322,12 @@ public function startTag($name, $attributes = array(), $selfClosing = false)
break;
}
+ // Case when no exists, note section on 'Anything else' below.
+ // https://html.spec.whatwg.org/multipage/parsing.html#the-after-head-insertion-mode
+ if ($this->insertMode === static::IM_AFTER_HEAD && 'head' !== $name && 'body' !== $name) {
+ $this->startTag('body');
+ }
+
// Special case handling for SVG.
if ($this->insertMode === static::IM_IN_SVG) {
$lname = Elements::normalizeSvgElement($lname);
@@ -548,21 +554,20 @@ public function comment($cdata)
public function text($data)
{
- // XXX: Hmmm.... should we really be this strict?
+ // https://html.spec.whatwg.org/multipage/parsing.html#the-before-head-insertion-mode
if ($this->insertMode < static::IM_IN_HEAD) {
// Per '8.2.5.4.3 The "before head" insertion mode' the characters
- // " \t\n\r\f" should be ignored but no mention of a parse error. This is
- // practical as most documents contain these characters. Other text is not
- // expected here so recording a parse error is necessary.
+ // " \t\n\r\f" should be ignored .
$dataTmp = trim($data, " \t\n\r\f");
- if (!empty($dataTmp)) {
- // fprintf(STDOUT, "Unexpected insert mode: %d", $this->insertMode);
- $this->parseError('Unexpected text. Ignoring: ' . $dataTmp);
+ if (! empty($dataTmp)) {
+ $this->startTag('head');
+ $this->endTag('head');
+ $this->startTag('body');
+ } else {
+ return;
}
-
- return;
}
- // fprintf(STDOUT, "Appending text %s.", $data);
+
$node = $this->doc->createTextNode($data);
$this->current->appendChild($node);
}
diff --git a/test/HTML5/Html5Test.php b/test/HTML5/Html5Test.php
index 26d52fe..b181d53 100644
--- a/test/HTML5/Html5Test.php
+++ b/test/HTML5/Html5Test.php
@@ -480,4 +480,50 @@ public function testCDATA()
$res = $this->cycleFragment('a a test. ]]>b');
$this->assertRegExp('| a test\. \]\]>|', $res);
}
+
+ /**
+ * Test for issue #166.
+ *
+ * @param $input
+ * @param $expected
+ *
+ * @dataProvider tagOmissionProvider
+ */
+ public function testTagOmission($input, $expected)
+ {
+ $doc = $this->html5->loadHTML($input);
+
+ $out = $this->html5->saveHTML($doc);
+
+ $this->assertRegExp("|" . preg_quote($expected, "|") . "|", $out);
+ }
+
+ /**
+ * Tag omission test cases.
+ *
+ * @return \string[][]
+ */
+ public function tagOmissionProvider()
+ {
+ return $provider = array(
+ array(
+ 'Hello, This is a test.
Does it work this time?',
+ 'Hello, This is a test.
Does it work this time?',
+ ),
+ // test whitespace (\n)
+ array(
+ '
+
+
+
+
+
+',
+ '
+
+
+'
+ ),
+ );
+ }
}