function &traverse_dom_tree_pdf(&$root)
{
switch ($root->node_type()) {
case XML_DOCUMENT_NODE:
$child =& $root->first_child();
while ($child) {
$body =& traverse_dom_tree_pdf($child);
if ($body) {
return $body;
}
$child =& $child->next_sibling();
}
$null = null;
return $null;
case XML_ELEMENT_NODE:
if (strtolower($root->tagname()) == "body") {
return $root;
}
$child =& $root->first_child();
while ($child) {
$body =& traverse_dom_tree_pdf($child);
if ($body) {
return $body;
}
$child =& $child->next_sibling();
}
$null = null;
return $null;
default:
$null = null;
return $null;
}
}
function IFrameBox(&$root, $pipeline)
{
$this->InlineBlockBox();
// If NO src attribute specified, just return.
if (!$root->has_attribute('src') || trim($root->get_attribute('src')) == '') {
return;
}
// Determine the fullly qualified URL of the frame content
$src = $root->get_attribute('src');
$url = $pipeline->guess_url($src);
$data = $pipeline->fetch($url);
/**
* If framed page could not be fetched return immediately
*/
if (is_null($data)) {
return;
}
/**
* Render only iframes containing HTML only
*
* Note that content-type header may contain additional information after the ';' sign
*/
$content_type = $data->get_additional_data('Content-Type');
$content_type_array = explode(';', $content_type);
if ($content_type_array[0] != "text/html") {
return;
}
$html = $data->get_content();
// Remove control symbols if any
$html = preg_replace('/[\\x00-\\x07]/', "", $html);
$converter = Converter::create();
$html = $converter->to_utf8($html, $data->detect_encoding());
$html = html2xhtml($html);
$tree = TreeBuilder::build($html);
// Save current stylesheet, as each frame may load its own stylesheets
//
$pipeline->pushCSS();
$css =& $pipeline->getCurrentCSS();
$css->scan_styles($tree, $pipeline);
$frame_root = traverse_dom_tree_pdf($tree);
$box_child =& create_pdf_box($frame_root, $pipeline);
$this->add_child($box_child);
// Restore old stylesheet
//
$pipeline->popCSS();
$pipeline->pop_base_url();
}
function &process($html, &$pipeline, &$media)
{
// Run the XML parser on the XHTML we've prepared
$dom_tree = TreeBuilder::build($html);
// Check if parser returned valid document
if (is_null($dom_tree)) {
readfile(HTML2PS_DIR . '/templates/cannot_parse.html');
error_log(sprintf("Cannot parse document: %s", $pipeline->get_base_url()));
die("HTML2PS Error");
}
/**
* Detect the base URI for this document.
*
* According to the HTML 4.01 p. 12.4.1:
* User agents must calculate the base URI according to the following precedences (highest priority to lowest):
*
* 1. The base URI is set by the BASE element.
* 2. The base URI is given by meta data discovered during a protocol interaction, such as an HTTP header (see [RFC2616]).
* 3. By default, the base URI is that of the current document. Not all HTML documents have a base URI (e.g., a valid HTML document may appear in an email and may not be designated by a URI). Such HTML documents are considered erroneous if they contain relative URIs and rely on a default base URI.
*/
/**
* Check if BASE element present; use its first occurrence
*/
$this->_scan_base($dom_tree, $pipeline);
/**
* @todo fall back to the protocol metadata
*/
/**
* Parse STYLE / LINK nodes containing CSS references and definitions
* This should be done here, as the document body may include STYLE node
* (this violates HTML standard, but is rather often appears in Web)
*/
$css =& $pipeline->getCurrentCSS();
$css->scan_styles($dom_tree, $pipeline);
if (!is_null($media)) {
// Setup media size and margins
$pipeline->get_page_media(1, $media);
$pipeline->output_driver->update_media($media);
$pipeline->_setupScales($media);
}
$body =& traverse_dom_tree_pdf($dom_tree);
$box =& create_pdf_box($body, $pipeline);
return $box;
}
function FrameBox(&$root, &$pipeline)
{
$css_state =& $pipeline->getCurrentCSSState();
// Inherit 'border' CSS value from parent (FRAMESET tag), if current FRAME
// has no FRAMEBORDER attribute, and FRAMESET has one
$parent = $root->parent();
if (!$root->has_attribute('frameborder') && $parent->has_attribute('frameborder')) {
$parent_border = $css_state->getPropertyOnLevel(CSS_BORDER, CSS_PROPERTY_LEVEL_PARENT);
$css_state->setProperty(CSS_BORDER, $parent_border->copy());
}
$this->GenericContainerBox($root);
// If NO src attribute specified, just return.
if (!$root->has_attribute('src')) {
return;
}
// Determine the fullly qualified URL of the frame content
$src = $root->get_attribute('src');
$url = $pipeline->guess_url($src);
$data = $pipeline->fetch($url);
/**
* If framed page could not be fetched return immediately
*/
if (is_null($data)) {
return;
}
/**
* Render only iframes containing HTML only
*
* Note that content-type header may contain additional information after the ';' sign
*/
$content_type = $data->get_additional_data('Content-Type');
$content_type_array = explode(';', $content_type);
if ($content_type_array[0] != "text/html") {
return;
}
$html = $data->get_content();
// Remove control symbols if any
$html = preg_replace('/[\\x00-\\x07]/', "", $html);
$converter = Converter::create();
$html = $converter->to_utf8($html, $data->detect_encoding());
$html = html2xhtml($html);
$tree = TreeBuilder::build($html);
// Save current stylesheet, as each frame may load its own stylesheets
//
$pipeline->pushCSS();
$css =& $pipeline->getCurrentCSS();
$css->scan_styles($tree, $pipeline);
$frame_root = traverse_dom_tree_pdf($tree);
$box_child =& create_pdf_box($frame_root, $pipeline);
$this->add_child($box_child);
// Restore old stylesheet
//
$pipeline->popCSS();
$pipeline->pop_base_url();
}
function FrameBox(&$root, &$pipeline)
{
// Inherit 'border' CSS value from parent (FRAMESET tag), if current FRAME
// has no FRAMEBORDER attribute, and FRAMESET has one
$parent = $root->parent();
if (!$root->has_attribute('frameborder') && $parent->has_attribute('frameborder')) {
pop_border();
push_border(get_border());
}
$this->GenericContainerBox($root);
// If NO src attribute specified, just return.
if (!$root->has_attribute('src')) {
return;
}
// Determine the fullly qualified URL of the frame content
$src = $root->get_attribute('src');
$url = $pipeline->guess_url($src);
$data = $pipeline->fetch($url);
/**
* If framed page could not be fetched return immediately
*/
if (is_null($data)) {
return;
}
/**
* Render only iframes containing HTML only
*
* Note that content-type header may contain additional information after the ';' sign
*/
$content_type = $data->get_additional_data('Content-Type');
$content_type_array = explode(';', $content_type);
if ($content_type_array[0] != "text/html") {
return;
}
$html = $data->get_content();
// Remove control symbols if any
$html = preg_replace('/[\\x00-\\x07]/', "", $html);
$converter = Converter::create();
$html = $converter->to_utf8($html, $data->detect_encoding());
$html = html2xhtml($html);
$tree = TreeBuilder::build($html);
// Save current stylesheet, as each frame may load its own stylesheets
//
global $g_css;
$old_css = $g_css;
global $g_css_obj;
$old_obj = $g_css_obj;
scan_styles($tree, $pipeline);
// Temporary hack: convert CSS rule array to CSS object
$g_css_obj = new CSSObject();
foreach ($g_css as $rule) {
$g_css_obj->add_rule($rule, $pipeline);
}
// TODO: stinks. Rewrite
//
$frame_root = traverse_dom_tree_pdf($tree);
$box_child =& create_pdf_box($frame_root, $pipeline);
$this->add_child($box_child);
// Restore old stylesheet
//
$g_css = $old_css;
$g_css_obj = $old_obj;
$pipeline->pop_base_url();
}
function BoxPageMargin(&$pipeline, $at_rule)
{
$state =& $pipeline->get_current_css_state();
$state->pushDefaultState();
$root = null;
$at_rule->css->apply($root, $state, $pipeline);
$this->GenericContainerBox();
$this->readCSS($state);
$state->pushDefaultstate();
/**
* Check whether 'content' or '-html2ps-html-content' properties had been defined
* (if both properties are defined, -html2ps-html-content takes precedence)
*/
$raw_html_content =& $at_rule->get_css_property(CSS_HTML2PS_HTML_CONTENT);
$html_content = $raw_html_content->render($pipeline->get_counters());
if ($html_content !== '') {
// We should wrap html_content in DIV tag,
// as we treat only the very first box of the resulting DOM tree as margin box content
$html_content = html2xhtml("<div>" . $html_content . "</div>");
$tree = TreeBuilder::build($html_content);
$tree_root = traverse_dom_tree_pdf($tree);
$body_box =& create_pdf_box($tree_root, $pipeline);
$box =& $body_box->content[0];
} else {
$raw_content =& $at_rule->get_css_property(CSS_CONTENT);
$content = $raw_content->render($pipeline->get_counters());
$box =& InlineBox::create_from_text($content, WHITESPACE_PRE_LINE, $pipeline);
}
$this->add_child($box);
$state->popState();
$state->popState();
}
function &build(&$dom_tree, &$pipeline)
{
$body =& traverse_dom_tree_pdf($dom_tree);
$box =& create_pdf_box($body, $pipeline);
return $box;
}
