Skip to content

Commit 5babe7b

Browse files
dmsnellSergeyBiryukovsirreal
committed
HTML API: Backport updates from Core
Updates from WordPress/wordpress-develop at f4dda54df785d0a6957dedda3648f7fab58b829f - Coding style changes. - WordPress/wordpress-develop#5762 Adds support for the "any other tag" sections in the HTML Processor. - WordPress/wordpress-develop#5539 Adds support for list elements in the HTML Processor. - WordPress/wordpress-develop#5897 Adds support for HR elements in the HTML Processor. - WordPress/wordpress-develop#5895 Adds support for the AREA, BR, EMBED, KEYGEN, and WBR elements in the HTML Processor. - WordPress/wordpress-develop#5903 Adds support for the PRE and LISTING elements in the HTML Processor. - WordPress/wordpress-develop#5913 Updates "all other tags" support in HTML Processor and updates list of void elements. - WordPress/wordpress-develop#5906 Adds support for the PARAM, SOURCE, and TRACK elements in the HTML Processor. - WordPress/wordpress-develop#5683 Provides mechanism to scan all tokens in an HTML document in the Tag Processor. The PHP files in the compatability layer are merged and maintained in the Core repo and all changes or updates need to happen first in Core and then be brought over to Gutenberg as built files. Co-authored-by: Sergey Biryukov <[email protected]> Co-authored-by: Jon Surrell <[email protected]>
1 parent 621a3cd commit 5babe7b

2 files changed

Lines changed: 879 additions & 158 deletions

File tree

‎lib/compat/wordpress-6.5/html-api/class-gutenberg-html-processor-6-5.php‎

Lines changed: 88 additions & 34 deletions
Original file line numberDiff line numberDiff line change
@@ -102,17 +102,17 @@
102102
* - Containers: ADDRESS, BLOCKQUOTE, DETAILS, DIALOG, DIV, FOOTER, HEADER, MAIN, MENU, SPAN, SUMMARY.
103103
* - Custom elements: All custom elements are supported. :)
104104
* - Form elements: BUTTON, DATALIST, FIELDSET, LABEL, LEGEND, METER, PROGRESS, SEARCH.
105-
* - Formatting elements: B, BIG, CODE, EM, FONT, I, SMALL, STRIKE, STRONG, TT, U.
105+
* - Formatting elements: B, BIG, CODE, EM, FONT, I, PRE, SMALL, STRIKE, STRONG, TT, U, WBR.
106106
* - Heading elements: H1, H2, H3, H4, H5, H6, HGROUP.
107107
* - Links: A.
108108
* - Lists: DD, DL, DT, LI, OL, LI.
109-
* - Media elements: AUDIO, CANVAS, FIGCAPTION, FIGURE, IMG, MAP, PICTURE, VIDEO.
110-
* - Paragraph: P.
111-
* - Phrasing elements: ABBR, BDI, BDO, CITE, DATA, DEL, DFN, INS, MARK, OUTPUT, Q, SAMP, SUB, SUP, TIME, VAR.
112-
* - Sectioning elements: ARTICLE, ASIDE, NAV, SECTION.
109+
* - Media elements: AUDIO, CANVAS, EMBED, FIGCAPTION, FIGURE, IMG, MAP, PARAM, PICTURE, SOURCE, VIDEO, TRACK.
110+
* - Paragraph: BR, P.
111+
* - Phrasing elements: AREA, ABBR, BDI, BDO, CITE, DATA, DEL, DFN, INS, MARK, OUTPUT, Q, SAMP, SUB, SUP, TIME, VAR.
112+
* - Sectioning elements: ARTICLE, ASIDE, HR, NAV, SECTION.
113113
* - Templating elements: SLOT.
114114
* - Text decoration: RUBY.
115-
* - Deprecated elements: ACRONYM, BLINK, CENTER, DIR, ISINDEX, MULTICOL, NEXTID, SPACER.
115+
* - Deprecated elements: ACRONYM, BLINK, CENTER, DIR, ISINDEX, KEYGEN, LISTING, MULTICOL, NEXTID, SPACER.
116116
*
117117
* ### Supported markup
118118
*
@@ -149,17 +149,6 @@ class Gutenberg_HTML_Processor_6_5 extends Gutenberg_HTML_Tag_Processor_6_5 {
149149
*/
150150
const MAX_BOOKMARKS = 100;
151151

152-
/**
153-
* Static query for instructing the Tag Processor to visit every token.
154-
*
155-
* @access private
156-
*
157-
* @since 6.4.0
158-
*
159-
* @var array
160-
*/
161-
const VISIT_EVERYTHING = array( 'tag_closers' => 'visit' );
162-
163152
/**
164153
* Holds the working state of the parser, including the stack of
165154
* open elements and the stack of active formatting elements.
@@ -424,6 +413,30 @@ public function next_tag( $query = null ) {
424413
return false;
425414
}
426415

416+
/**
417+
* Ensures internal accounting is maintained for HTML semantic rules while
418+
* the underlying Tag Processor class is seeking to a bookmark.
419+
*
420+
* This doesn't currently have a way to represent non-tags and doesn't process
421+
* semantic rules for text nodes. For access to the raw tokens consider using
422+
* WP_HTML_Tag_Processor instead.
423+
*
424+
* @since 6.5.0 Added for internal support; do not use.
425+
*
426+
* @access private
427+
*
428+
* @return bool
429+
*/
430+
public function next_token() {
431+
$found_a_token = parent::next_token();
432+
433+
if ( '#tag' === $this->get_token_type() ) {
434+
$this->step( self::REPROCESS_CURRENT_NODE );
435+
}
436+
437+
return $found_a_token;
438+
}
439+
427440
/**
428441
* Indicates if the currently-matched tag matches the given breadcrumbs.
429442
*
@@ -520,7 +533,9 @@ public function step( $node_to_process = self::PROCESS_NEXT_NODE ) {
520533
$this->state->stack_of_open_elements->pop();
521534
}
522535

523-
parent::next_tag( self::VISIT_EVERYTHING );
536+
while ( parent::next_token() && '#tag' !== $this->get_token_type() ) {
537+
continue;
538+
}
524539
}
525540

526541
// Finish stepping when there are no more tokens in the document.
@@ -684,10 +699,12 @@ private function step_in_body() {
684699
case '-FOOTER':
685700
case '-HEADER':
686701
case '-HGROUP':
702+
case '-LISTING':
687703
case '-MAIN':
688704
case '-MENU':
689705
case '-NAV':
690706
case '-OL':
707+
case '-PRE':
691708
case '-SEARCH':
692709
case '-SECTION':
693710
case '-SUMMARY':
@@ -732,6 +749,18 @@ private function step_in_body() {
732749
$this->insert_html_element( $this->state->current_token );
733750
return true;
734751

752+
/*
753+
* > A start tag whose tag name is one of: "pre", "listing"
754+
*/
755+
case '+PRE':
756+
case '+LISTING':
757+
if ( $this->state->stack_of_open_elements->has_p_in_button_scope() ) {
758+
$this->close_a_p_element();
759+
}
760+
$this->insert_html_element( $this->state->current_token );
761+
$this->state->frameset_ok = false;
762+
return true;
763+
735764
/*
736765
* > An end tag whose tag name is one of: "h1", "h2", "h3", "h4", "h5", "h6"
737766
*/
@@ -934,11 +963,47 @@ private function step_in_body() {
934963
$this->run_adoption_agency_algorithm();
935964
return true;
936965

966+
/*
967+
* > An end tag whose tag name is "br"
968+
* > Parse error. Drop the attributes from the token, and act as described in the next
969+
* > entry; i.e. act as if this was a "br" start tag token with no attributes, rather
970+
* > than the end tag token that it actually is.
971+
*/
972+
case '-BR':
973+
$this->last_error = self::ERROR_UNSUPPORTED;
974+
throw new WP_HTML_Unsupported_Exception( 'Closing BR tags require unimplemented special handling.' );
975+
937976
/*
938977
* > A start tag whose tag name is one of: "area", "br", "embed", "img", "keygen", "wbr"
939978
*/
979+
case '+AREA':
980+
case '+BR':
981+
case '+EMBED':
940982
case '+IMG':
983+
case '+KEYGEN':
984+
case '+WBR':
941985
$this->reconstruct_active_formatting_elements();
986+
$this->insert_html_element( $this->state->current_token );
987+
$this->state->frameset_ok = false;
988+
return true;
989+
990+
/*
991+
* > A start tag whose tag name is "hr"
992+
*/
993+
case '+HR':
994+
if ( $this->state->stack_of_open_elements->has_p_in_button_scope() ) {
995+
$this->close_a_p_element();
996+
}
997+
$this->insert_html_element( $this->state->current_token );
998+
$this->state->frameset_ok = false;
999+
return true;
1000+
1001+
/*
1002+
* > A start tag whose tag name is one of: "param", "source", "track"
1003+
*/
1004+
case '+PARAM':
1005+
case '+SOURCE':
1006+
case '+TRACK':
9421007
$this->insert_html_element( $this->state->current_token );
9431008
return true;
9441009
}
@@ -961,30 +1026,21 @@ private function step_in_body() {
9611026
*/
9621027
switch ( $tag_name ) {
9631028
case 'APPLET':
964-
case 'AREA':
9651029
case 'BASE':
9661030
case 'BASEFONT':
9671031
case 'BGSOUND':
9681032
case 'BODY':
969-
case 'BR':
9701033
case 'CAPTION':
9711034
case 'COL':
9721035
case 'COLGROUP':
973-
case 'DD':
974-
case 'DT':
975-
case 'EMBED':
9761036
case 'FORM':
9771037
case 'FRAME':
9781038
case 'FRAMESET':
9791039
case 'HEAD':
980-
case 'HR':
9811040
case 'HTML':
9821041
case 'IFRAME':
9831042
case 'INPUT':
984-
case 'KEYGEN':
985-
case 'LI':
9861043
case 'LINK':
987-
case 'LISTING':
9881044
case 'MARQUEE':
9891045
case 'MATH':
9901046
case 'META':
@@ -993,20 +1049,16 @@ private function step_in_body() {
9931049
case 'NOFRAMES':
9941050
case 'NOSCRIPT':
9951051
case 'OBJECT':
996-
case 'OL':
9971052
case 'OPTGROUP':
9981053
case 'OPTION':
999-
case 'PARAM':
10001054
case 'PLAINTEXT':
1001-
case 'PRE':
10021055
case 'RB':
10031056
case 'RP':
10041057
case 'RT':
10051058
case 'RTC':
10061059
case 'SARCASM':
10071060
case 'SCRIPT':
10081061
case 'SELECT':
1009-
case 'SOURCE':
10101062
case 'STYLE':
10111063
case 'SVG':
10121064
case 'TABLE':
@@ -1019,9 +1071,6 @@ private function step_in_body() {
10191071
case 'THEAD':
10201072
case 'TITLE':
10211073
case 'TR':
1022-
case 'TRACK':
1023-
case 'UL':
1024-
case 'WBR':
10251074
case 'XMP':
10261075
$this->last_error = self::ERROR_UNSUPPORTED;
10271076
throw new WP_HTML_Unsupported_Exception( "Cannot process {$tag_name} element." );
@@ -1675,14 +1724,19 @@ public static function is_void( $tag_name ) {
16751724
return (
16761725
'AREA' === $tag_name ||
16771726
'BASE' === $tag_name ||
1727+
'BASEFONT' === $tag_name || // Obsolete but still treated as void.
1728+
'BGSOUND' === $tag_name || // Obsolete but still treated as void.
16781729
'BR' === $tag_name ||
16791730
'COL' === $tag_name ||
16801731
'EMBED' === $tag_name ||
1732+
'FRAME' === $tag_name ||
16811733
'HR' === $tag_name ||
16821734
'IMG' === $tag_name ||
16831735
'INPUT' === $tag_name ||
1736+
'KEYGEN' === $tag_name || // Obsolete but still treated as void.
16841737
'LINK' === $tag_name ||
16851738
'META' === $tag_name ||
1739+
'PARAM' === $tag_name || // Obsolete but still treated as void.
16861740
'SOURCE' === $tag_name ||
16871741
'TRACK' === $tag_name ||
16881742
'WBR' === $tag_name

0 commit comments

Comments
 (0)