diff options
author | Christophe Coevoet <[email protected]> | 2018-11-24 13:53:53 +0100 |
---|---|---|
committer | Christophe Coevoet <[email protected]> | 2018-11-26 12:28:27 +0100 |
commit | 88431be37966f89f2a96210ee855565212f34969 (patch) | |
tree | b7f7b443b0ac279a126637d73392352a72bb2b2d /src/HTML5/Parser/Tokenizer.php | |
parent | 54d066b1664a5fb749511c764e8d208f6bc80c3b (diff) |
Optimize the parsing of unquoted attributes
Diffstat (limited to 'src/HTML5/Parser/Tokenizer.php')
-rw-r--r-- | src/HTML5/Parser/Tokenizer.php | 38 |
1 files changed, 28 insertions, 10 deletions
diff --git a/src/HTML5/Parser/Tokenizer.php b/src/HTML5/Parser/Tokenizer.php index 6284733..49f5fc0 100644 --- a/src/HTML5/Parser/Tokenizer.php +++ b/src/HTML5/Parser/Tokenizer.php @@ -594,19 +594,37 @@ class Tokenizer protected function unquotedAttributeValue() { - $stoplist = "\t\n\f >"; $val = ''; $tok = $this->scanner->current(); - while (0 == strspn($tok, $stoplist) && false !== $tok) { - if ('&' == $tok) { - $val .= $this->decodeCharacterReference(true); - $tok = $this->scanner->current(); - } else { - if (strspn($tok, "\"'<=`") > 0) { + while (false !== $tok) { + switch ($tok) { + case "\n": + case "\f": + case ' ': + case "\t": + case '>': + break 2; + + case '&': + $val .= $this->decodeCharacterReference(true); + $tok = $this->scanner->current(); + + break; + + case "'": + case '"': + case '<': + case '=': + case '`': $this->parseError('Unexpected chars in unquoted attribute value %s', $tok); - } - $val .= $tok; - $tok = $this->scanner->next(); + $val .= $tok; + $tok = $this->scanner->next(); + break; + + default: + $val .= $this->scanner->charsUntil("\t\n\f >&\"'<=`"); + + $tok = $this->scanner->current(); } } |