MensBeam
/
HTML-DOM
3
0
Fork 0
Code Issues Pull Requests Releases Activity
Browse Source

Minor simplifications

split-manual
J. King 3 years ago
parent
bfa7352713
commit
8d5079deae
2 changed files with 60 additions and 91 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Split View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 110
      lib/Data.php
  2. 41
      lib/TreeBuilder.php

110
lib/Data.php
View File

@ -85,69 +85,62 @@ class Data {
}
// unless we're peeking, track line and column position, and whether we've hit EOF
if ($this->track) {
$this->checkChar($char);
}
return $char;
}
protected function checkChar(string $char): bool {
// track line and column number, and EOF
if ($char === "\n") {
$this->newlines[$this->data->posChar()] = $this->_column;
$this->_column = 0;
$this->_line++;
} elseif ($char === '') {
$this->eof = true;
return false;
} else {
$this->_column++;
$len = strlen($char);
$here = $this->data->posChar();
if ($this->lastError < $here) {
// look for erroneous characters
if ($len === 1) {
$ord = ord($char);
if (($ord < 0x20 && !in_array($ord, [0x0, 0x9, 0xA, 0xC])) || $ord === 0x7F) {
$this->error(ParseError::CONTROL_CHARACTER_IN_INPUT_STREAM);
$this->lastError = $here;
}
} elseif ($len === 2) {
if (ord($char[0]) == 0xC2) {
$ord = ord($char[1]);
if ($ord >= 0x80 && $ord <= 0x9F) {
if ($char === "\n") {
$this->newlines[$this->data->posChar()] = $this->_column;
$this->_column = 0;
$this->_line++;
} elseif ($char === '') {
$this->eof = true;
} else {
$this->_column++;
$len = strlen($char);
$here = $this->data->posChar();
if ($this->lastError < $here) {
// look for erroneous characters
if ($len === 1) {
$ord = ord($char);
if (($ord < 0x20 && !in_array($ord, [0x0, 0x9, 0xA, 0xC])) || $ord === 0x7F) {
$this->error(ParseError::CONTROL_CHARACTER_IN_INPUT_STREAM);
$this->lastError = $here;
}
}
} elseif ($len === 3) {
$head = ord($char[0]);
if ($head === 0xED) {
$tail = (ord($char[1]) << 8) + ord($char[2]);
if ($tail >= 0xA080 && $tail <= 0xBFBF) {
$this->error(ParseError::SURROGATE_IN_INPUT_STREAM);
$this->lastError = $here;
} elseif ($len === 2) {
if (ord($char[0]) == 0xC2) {
$ord = ord($char[1]);
if ($ord >= 0x80 && $ord <= 0x9F) {
$this->error(ParseError::CONTROL_CHARACTER_IN_INPUT_STREAM);
$this->lastError = $here;
}
}
} elseif ($head === 0xEF) {
$tail = (ord($char[1]) << 8) + ord($char[2]);
if (($tail >= 0xB790 && $tail <= 0xB7AF) || $tail >= 0xBFBE) {
} elseif ($len === 3) {
$head = ord($char[0]);
if ($head === 0xED) {
$tail = (ord($char[1]) << 8) + ord($char[2]);
if ($tail >= 0xA080 && $tail <= 0xBFBF) {
$this->error(ParseError::SURROGATE_IN_INPUT_STREAM);
$this->lastError = $here;
}
} elseif ($head === 0xEF) {
$tail = (ord($char[1]) << 8) + ord($char[2]);
if (($tail >= 0xB790 && $tail <= 0xB7AF) || $tail >= 0xBFBE) {
$this->error(ParseError::NONCHARACTER_IN_INPUT_STREAM);
$this->lastError = $here;
} elseif ($tail === 0xBFBD && $this->data->posErr === $here) {
$this->error(ParseError::NONCHARACTER_IN_INPUT_STREAM, $this->data->posByte);
$this->lastError = $here;
}
}
} elseif ($len === 4) {
$tail = (ord($char[2]) << 8) + ord($char[3]);
if ($tail >= 0xBFBE) {
$this->error(ParseError::NONCHARACTER_IN_INPUT_STREAM);
$this->lastError = $here;
} elseif ($tail === 0xBFBD && $this->data->posErr === $here) {
$this->error(ParseError::NONCHARACTER_IN_INPUT_STREAM, $this->data->posByte);
$this->lastError = $here;
}
$this->astrals[$here] = true;
}
} elseif ($len === 4) {
$tail = (ord($char[2]) << 8) + ord($char[3]);
if ($tail >= 0xBFBE) {
$this->error(ParseError::NONCHARACTER_IN_INPUT_STREAM);
$this->lastError = $here;
}
$this->astrals[$here] = true;
}
}
}
return true;
return $char;
}
public function unconsume(int $length = 1, bool $retreatPointer = true): void {
@ -194,20 +187,17 @@ class Data {
if ($this->track) {
// control characters produce parse errors
$match .= "\x01\x02\x03\x04\x05\x06\x07\x08\x0B\x0D\x0E\x0F\x10\x11\x12\x13\x14\x15\x16\x17\x18\x19\x1A\x1B\x1C\x1D\x1E\x1F\x7F";
}
$out = $this->data->asciiSpanNot($match."\r\n", $limit);
if ($this->track) {
$out = $this->data->asciiSpanNot($match."\r\n", $limit);
$this->_column += ($this->data->posChar() - $start);
return $out;
} else {
return $this->data->asciiSpanNot($match."\r\n", $limit);
}
return $out;
}
public function peek(int $length = 1): string {
assert($length > 0, new Exception(Exception::DATA_INVALID_DATA_CONSUMPTION_LENGTH, $length));
$string = $this->data->peekChar($length);
return $string;
return $this->data->peekChar($length);
}
/** Returns an indexed array with the line and column positions of the requested offset from the current position */

41
lib/TreeBuilder.php
View File

@ -3876,7 +3876,7 @@ class TreeBuilder {
];
}
public function insertCharacterToken(CharacterToken $token) {
public function insertCharacterToken(CharacterToken $token): void {
# 1. Let data be the characters passed to the algorithm, or, if no characters
# were explicitly specified, the character of the character token being
# processed.
@ -3890,7 +3890,8 @@ class TreeBuilder {
assert($adjustedInsertionLocation instanceof \DOMNode, new Exception(Exception::TREEBUILDER_INVALID_INSERTION_LOCATION));
# 3. If the adjusted insertion location is in a Document node, then abort these
# steps.
if ((($insertBefore === false) ? $adjustedInsertionLocation : $adjustedInsertionLocation->parentNode) instanceof \DOMDocument) {
// NOTE: foster parenting will never point to before the root element
if ($adjustedInsertionLocation instanceof \DOMDocument) {
return;
}
@ -3915,42 +3916,23 @@ class TreeBuilder {
}
}
public function insertCommentToken(CommentToken $token, \DOMNode $position = null) {
public function insertCommentToken(CommentToken $token, \DOMNode $position = null): void {
# When the steps below require the user agent to insert a comment while
# processing a comment token, optionally with an explicitly insertion position
# position, the user agent must run the following steps:
# 1. Let data be the data given in the comment token being processed.
// Already provided through the token object.
# 2. If position was specified, then let the adjusted insertion location be
# position. Otherwise, let adjusted insertion location be the appropriate place
# for inserting a node.
if (!is_null($position)) {
$adjustedInsertionLocation = $position;
$insertBefore = false;
} else {
$location = $this->appropriatePlaceForInsertingNode();
$adjustedInsertionLocation = $location['node'];
$insertBefore = $location['insert before'];
}
// OPTIMIZATION: Comments are never foster-parented
$position = $position ?? $this->appropriatePlaceForInsertingNode()['node'];
# 3. Create a Comment node whose data attribute is set to data and whose node
# document is the same as that of the node in which the adjusted insertion
# location finds itself.
if ($adjustedInsertionLocation instanceof \DOMDocument) {
$nodeDocument = $adjustedInsertionLocation;
} else {
$nodeDocument = $adjustedInsertionLocation->ownerDocument;
}
$commentNode = $nodeDocument->createComment($token->data);
# 4. Insert the newly created node at the adjusted insertion location.
if ($insertBefore === false) {
$adjustedInsertionLocation->appendChild($commentNode);
} else {
$adjustedInsertionLocation->parentNode->insertBefore($commentNode, $adjustedInsertionLocation);
}
$position->appendChild($this->DOM->createComment($token->data));
}
public function insertStartTagToken(StartTagToken $token, \DOMNode $intendedParent = null, string $namespace = null): Element {
@ -3963,8 +3945,6 @@ class TreeBuilder {
# Let the adjusted insertion location be the appropriate place for inserting
# a node.
$location = $this->appropriatePlaceForInsertingNode($intendedParent);
$adjustedInsertionLocation = $location['node'];
$insertBefore = $location['insert before'];
# Let element be the result of creating an element for the token in the given
# namespace, with the intended parent being the element in which the adjusted
# insertion location finds itself.
@ -3974,12 +3954,11 @@ class TreeBuilder {
# - 1. Push a new element queue onto the custom element reactions stack.
// DEVIATION: Unnecessary because there is no scripting in this implementation.
# - 2. Insert element at the adjusted insertion location.
if ($insertBefore === false) {
$adjustedInsertionLocation->appendChild($element);
if ($location['insert before'] === false) {
$location['node']->appendChild($element);
} else {
$adjustedInsertionLocation->parentNode->insertBefore($element, $adjustedInsertionLocation);
$location['node']->parentNode->insertBefore($element, $location['node']);
}
# - 3. Pop the element queue from the custom element reactions stack, and
# invoke custom element reactions in that queue.
// DEVIATION: Unnecessary because there is no scripting in this implementation.

Write Preview
Loading…
Cancel
Save