Singletons are generally bad, but it allows us to build the regexes once (and only once).
/**
* Attempt to parse link title (sans quotes)
*
* @param Cursor $cursor
*
* @return null|string The string, or null if no match
*/
public static function parseLinkTitle(Cursor $cursor)
{
if ($title = $cursor->match(RegexHelper::getInstance()->getLinkTitleRegex())) {
// Chop off quotes from title and unescape
return RegexHelper::unescape(substr($title, 1, strlen($title) - 2));
}
}
/**
* @param ContextInterface $context
* @param Cursor $cursor
*
* @return bool
*/
public function parse(ContextInterface $context, Cursor $cursor)
{
$tmpCursor = clone $cursor;
$indent = $tmpCursor->advanceWhileMatches(' ', 3);
$rest = $tmpCursor->getRemainder();
$data = new ListData();
if ($matches = RegexHelper::matchAll('/^[*+-]( +|$)/', $rest)) {
$spacesAfterMarker = strlen($matches[1]);
$data->type = ListBlock::TYPE_UNORDERED;
$data->delimiter = null;
$data->bulletChar = $matches[0][0];
} elseif ($matches = RegexHelper::matchAll('/^(\\d+)([.)])( +|$)/', $rest)) {
$spacesAfterMarker = strlen($matches[3]);
$data->type = ListBlock::TYPE_ORDERED;
$data->start = intval($matches[1]);
$data->delimiter = $matches[2];
$data->bulletChar = null;
} else {
return false;
}
$data->padding = $this->calculateListMarkerPadding($matches[0], $spacesAfterMarker, $rest);
$cursor->advanceToFirstNonSpace();
$cursor->advanceBy($data->padding);
// list item
$data->markerOffset = $indent;
// add the list if needed
$container = $context->getContainer();
if (!$container || !$context->getContainer() instanceof ListBlock || !$data->equals($container->getListData())) {
$context->addBlock(new ListBlock($data));
}
// add the list item
$context->addBlock(new ListItem($data));
return true;
}
/**
* Returns the position of the next non-space character
*
* @return int
*/
public function getFirstNonSpacePosition()
{
if ($this->firstNonSpaceCache === null) {
$match = RegexHelper::matchAt('/[^ ]/', $this->line, $this->currentPosition);
$this->firstNonSpaceCache = $match === null ? $this->length : $match;
}
return $this->firstNonSpaceCache;
}
/**
* @param ContextInterface $context
* @param InlineParserContext $inlineContext
*
* @return bool
*/
public function parse(ContextInterface $context, InlineParserContext $inlineContext)
{
$cursor = $inlineContext->getCursor();
if ($m = $cursor->match(RegexHelper::getInstance()->getHtmlTagRegex())) {
$inlineContext->getInlines()->add(new Html($m));
return true;
}
return false;
}
/**
* @param InlineParserContext $inlineContext
*
* @return bool
*/
public function parse(InlineParserContext $inlineContext)
{
$cursor = $inlineContext->getCursor();
if ($m = $cursor->match(RegexHelper::getInstance()->getHtmlTagRegex())) {
$inlineContext->getContainer()->appendChild(new HtmlInline($m));
return true;
}
return false;
}
/**
* @param ContextInterface $context
* @param Cursor $cursor
*
* @return bool
*/
public function parse(ContextInterface $context, Cursor $cursor)
{
$match = RegexHelper::matchAt(RegexHelper::getInstance()->getHtmlBlockOpenRegex(), $cursor->getLine(), $cursor->getFirstNonSpacePosition());
if ($match === null) {
return false;
}
$context->addBlock(new HtmlBlock());
$context->setBlocksParsed(true);
return true;
}
public static function parse(Cursor $cursor)
{
if (null === self::$regexp) {
$regex = RegexHelper::getInstance();
self::$regexp = sprintf('/^\\s*([.#][_a-z0-9-]+|%s%s)(?<!})\\s*/i', $regex->getPartialRegex(RegexHelper::ATTRIBUTENAME), $regex->getPartialRegex(RegexHelper::ATTRIBUTEVALUESPEC));
}
$state = $cursor->saveState();
$cursor->advanceToFirstNonSpace();
if ('{' !== $cursor->getCharacter()) {
$cursor->restoreState($state);
return [];
}
$cursor->advanceBy(1);
if (':' === $cursor->getCharacter()) {
$cursor->advanceBy(1);
}
$attributes = [];
while ($attribute = trim($cursor->match(self::$regexp))) {
if ('#' === $attribute[0]) {
$attributes['id'] = substr($attribute, 1);
continue;
}
if ('.' === $attribute[0]) {
$attributes['class'][] = substr($attribute, 1);
continue;
}
list($name, $value) = explode('=', $attribute, 2);
$first = $value[0];
$last = substr($value, -1);
if (('"' === $first && '"' === $last || "'" === $first && "'" === $last) && strlen($value) > 1) {
$value = substr($value, 1, -1);
}
if ('class' === strtolower(trim($name))) {
foreach (array_filter(explode(' ', trim($value))) as $class) {
$attributes['class'][] = $class;
}
} else {
$attributes[trim($name)] = trim($value);
}
}
if (0 === $cursor->advanceWhileMatches('}')) {
$cursor->restoreState($state);
return [];
}
if (isset($attributes['class'])) {
$attributes['class'] = implode(' ', $attributes['class']);
}
return $attributes;
}
/**
* @param ContextInterface $context
* @param Cursor $cursor
*
* @return bool
*/
public function parse(ContextInterface $context, Cursor $cursor)
{
if ($cursor->isIndented()) {
return false;
}
$match = RegexHelper::matchAt(RegexHelper::getInstance()->getThematicBreakRegex(), $cursor->getLine(), $cursor->getFirstNonSpacePosition());
if ($match === null) {
return false;
}
// Advance to the end of the string, consuming the entire line (of the thematic break)
$cursor->advanceBy(mb_strlen($cursor->getRemainder()));
$context->addBlock(new ThematicBreak());
$context->setBlocksParsed(true);
return true;
}
private function parseRow($line, array $columns, $type = TableCell::TYPE_BODY)
{
$cells = RegexHelper::matchAll(self::REGEXP_CELLS, $line);
if (null === $cells || !is_array($cells[0])) {
return;
}
$row = new TableRow();
foreach ($cells[0] as $i => $cell) {
$row->addChild(new TableCell(trim($cell), $type, isset($columns[$i]) ? $columns[$i] : null));
}
for ($j = count($columns) - 1; $j > $i; --$j) {
$row->addChild(new TableCell('', $type, null));
}
return $row;
}
/**
* @param ContextInterface $context
* @param Cursor $cursor
*
* @return bool
*/
public function parse(ContextInterface $context, Cursor $cursor)
{
$match = RegexHelper::matchAll('/^#{1,6}(?: +|$)/', $cursor->getLine(), $cursor->getFirstNonSpacePosition());
if (!$match) {
return false;
}
$cursor->advanceToFirstNonSpace();
$cursor->advanceBy(strlen($match[0]));
$level = strlen(trim($match[0]));
$str = $cursor->getRemainder();
$str = preg_replace('/^ *#+ *$/', '', $str);
$str = preg_replace('/ +#+ *$/', '', $str);
$context->addBlock(new Header($level, $str));
$context->setBlocksParsed(true);
return true;
}
/**
* @param ContextInterface $context
* @param Cursor $cursor
*
* @return bool
*/
public function parse(ContextInterface $context, Cursor $cursor)
{
if ($cursor->isIndented() && !$context->getContainer() instanceof ListBlock) {
return false;
}
$tmpCursor = clone $cursor;
$tmpCursor->advanceToFirstNonSpace();
$rest = $tmpCursor->getRemainder();
$data = new ListData();
$data->markerOffset = $cursor->getIndent();
if ($matches = RegexHelper::matchAll('/^[*+-]/', $rest)) {
$data->type = ListBlock::TYPE_UNORDERED;
$data->delimiter = null;
$data->bulletChar = $matches[0][0];
} elseif (($matches = RegexHelper::matchAll('/^(\\d{1,9})([.)])/', $rest)) && (!$context->getContainer() instanceof Paragraph || $matches[1] === '1')) {
$data->type = ListBlock::TYPE_ORDERED;
$data->start = intval($matches[1]);
$data->delimiter = $matches[2];
$data->bulletChar = null;
} else {
return false;
}
$markerLength = strlen($matches[0]);
// Make sure we have spaces after
$nextChar = $tmpCursor->peek($markerLength);
if (!($nextChar === null || $nextChar === "\t" || $nextChar === ' ')) {
return false;
}
// If it interrupts paragraph, make sure first line isn't blank
if ($context->getContainer() instanceof Paragraph && !RegexHelper::matchAt(RegexHelper::REGEX_NON_SPACE, $rest, $markerLength)) {
return false;
}
// We've got a match! Advance offset and calculate padding
$cursor->advanceToFirstNonSpace();
// to start of marker
$cursor->advanceBy($markerLength, true);
// to end of marker
$data->padding = $this->calculateListMarkerPadding($cursor, $markerLength);
// add the list if needed
$container = $context->getContainer();
if (!$container || !$context->getContainer() instanceof ListBlock || !$data->equals($container->getListData())) {
$context->addBlock(new ListBlock($data));
}
// add the list item
$context->addBlock(new ListItem($data));
return true;
}
/**
* @param Link $inline
* @param ElementRendererInterface $htmlRenderer
*
* @return HtmlElement
*/
public function render(AbstractInline $inline, ElementRendererInterface $htmlRenderer)
{
if (!$inline instanceof Link) {
throw new \InvalidArgumentException('Incompatible inline type: ' . get_class($inline));
}
$attrs = [];
foreach ($inline->getData('attributes', []) as $key => $value) {
$attrs[$key] = $htmlRenderer->escape($value, true);
}
if (!($this->config->getConfig('safe') && RegexHelper::isLinkPotentiallyUnsafe($inline->getUrl()))) {
$attrs['href'] = $htmlRenderer->escape($inline->getUrl(), true);
}
if (isset($inline->data['title'])) {
$attrs['title'] = $htmlRenderer->escape($inline->data['title'], true);
}
return new HtmlElement('a', $attrs, $htmlRenderer->renderInlines($inline->children()));
}
/**
* @param ContextInterface $context
* @param Cursor $cursor
*
* @return bool
*/
public function parse(ContextInterface $context, Cursor $cursor)
{
if ($cursor->isIndented()) {
return false;
}
if (!$context->getContainer() instanceof Paragraph) {
return false;
}
$match = RegexHelper::matchAll('/^(?:=+|-+)[ \\t]*$/', $cursor->getLine(), $cursor->getFirstNonSpacePosition());
if ($match === null) {
return false;
}
$level = $match[0][0] === '=' ? 1 : 2;
$strings = $context->getContainer()->getStrings();
$context->replaceContainerBlock(new Heading($level, $strings));
return true;
}
private function buildLink($inline, $htmlRenderer)
{
$attrs = [];
foreach ($inline->getData('attributes', []) as $key => $value) {
$attrs[$key] = $htmlRenderer->escape($value, true);
}
if (!($this->config->getConfig('safe') && RegexHelper::isLinkPotentiallyUnsafe($inline->getUrl()))) {
$attrs['href'] = $htmlRenderer->escape($inline->getUrl(), true);
}
if (isset($inline->data['title'])) {
$attrs['title'] = $htmlRenderer->escape($inline->data['title'], true);
}
if ($this->isExternalUrl($inline->getUrl())) {
$attrs['rel'] = 'nofollow';
$attrs['target'] = '_blank';
}
return new HtmlElement('a', $attrs, $htmlRenderer->renderInlines($inline->children()));
}
/**
* @param Image $inline
* @param ElementRendererInterface $htmlRenderer
*
* @return HtmlElement
*/
public function render(AbstractInline $inline, ElementRendererInterface $htmlRenderer)
{
if (!$inline instanceof Image) {
throw new \InvalidArgumentException('Incompatible inline type: ' . get_class($inline));
}
$attrs = [];
foreach ($inline->getData('attributes', []) as $key => $value) {
$attrs[$key] = $htmlRenderer->escape($value, true);
}
if ($this->config->getConfig('safe') && RegexHelper::isLinkPotentiallyUnsafe($inline->getUrl())) {
$attrs['src'] = '';
} else {
$attrs['src'] = $htmlRenderer->escape($inline->getUrl(), true);
}
$alt = $htmlRenderer->renderInlines($inline->children());
$alt = preg_replace('/\\<[^>]*alt="([^"]*)"[^>]*\\>/', '$1', $alt);
$attrs['alt'] = preg_replace('/\\<[^>]*\\>/', '', $alt);
if (isset($inline->data['title'])) {
$attrs['title'] = $htmlRenderer->escape($inline->data['title'], true);
}
return new HtmlElement('img', $attrs, '', true);
}
/**
* @param ContextInterface $context
* @param Cursor $cursor
*
* @return bool
*/
public function parse(ContextInterface $context, Cursor $cursor)
{
if ($cursor->isIndented()) {
return false;
}
if ($cursor->getFirstNonSpaceCharacter() !== '<') {
return false;
}
$savedState = $cursor->saveState();
$cursor->advanceToFirstNonSpace();
$line = $cursor->getRemainder();
for ($blockType = 1; $blockType <= 7; $blockType++) {
$match = RegexHelper::matchAt(RegexHelper::getHtmlBlockOpenRegex($blockType), $line);
if ($match !== null && ($blockType < 7 || !$context->getContainer() instanceof Paragraph)) {
$cursor->restoreState($savedState);
$context->addBlock(new HtmlBlock($blockType));
$context->setBlocksParsed(true);
return true;
}
}
$cursor->restoreState($savedState);
return false;
}
private function parseCaption($line)
{
$caption = RegexHelper::matchAll(self::REGEXP_CAPTION, $line);
if (null === $caption) {
return;
}
return new TableCaption($caption[1], $caption[2]);
}
/**
* @param ContextInterface $context
* @param Cursor $cursor
*/
public function handleRemainingContents(ContextInterface $context, Cursor $cursor)
{
$context->getTip()->addLine($cursor->getRemainder());
// Check for end condition
if ($this->type >= self::TYPE_1_CODE_CONTAINER && $this->type <= self::TYPE_5_CDATA) {
if ($cursor->match(RegexHelper::getHtmlBlockCloseRegex($this->type)) !== null) {
$this->finalize($context, $context->getLineNumber());
}
}
}
/**
* @param ContextInterface $context
* @param Cursor $cursor
*/
public function handleRemainingContents(ContextInterface $context, Cursor $cursor)
{
/** @var FencedCode $container */
$container = $context->getContainer();
// check for closing code fence
if ($cursor->getIndent() <= 3 && $cursor->getFirstNonSpaceCharacter() == $container->getChar()) {
$match = RegexHelper::matchAll('/^(?:`{3,}|~{3,})(?= *$)/', $cursor->getLine(), $cursor->getFirstNonSpacePosition());
if (strlen($match[0]) >= $container->getLength()) {
// don't add closing fence to container; instead, close it:
$this->setLength(-1);
// -1 means we've passed closer
return;
}
}
$context->getTip()->addLine($cursor->getRemainder());
}
