Ticket #38656: 38656.4.diff

src/wp-includes/formatting.php

@@ -492 +492 @@
         // Change multiple <br>s into two line breaks, which will turn into paragraphs.
         $pee = preg_replace( '|<br\s*/?>\s*<br\s*/?>|', "\n\n", $pee );
 
-        $allblocks = '(?:table|thead|tfoot|caption|col|colgroup|tbody|tr|td|th|div|dl|dd|dt|ul|ol|li|pre|form|map|area|blockquote|address|math|style|p|h[1-6]|hr|fieldset|legend|section|article|aside|hgroup|header|footer|nav|figure|figcaption|details|menu|summary)';
+        // Block level elements that can contain a <p>.
+        $peeblocks = 'caption|td|th|div|dd|dt|li|form|map|blockquote|address|fieldset|section|article|aside|header|footer|nav|figure|figcaption|details|menu';
+
+        // Block level elements that cannot contain a <p>.
+        $peefreeblocks = 'table|thead|tfoot|col|colgroup|tbody|tr|dl|ul|ol|pre|area|math|style|p|h[1-6]|hr|legend|hgroup|summary';
+
+        $allblocks     = "(?:$peeblocks|$peefreeblocks)";
+        $peeblocks     = "(?:$peeblocks)";
+        $peefreeblocks = "(?:$peefreeblocks)";
+
+        // Empty elements that cannot contain anything.
+        $emptyblocks = '(?:hr|br)';
+
+        // Media elements that will be processed later.
+        $mediablocks = '(?:object|param|embed|audio|video|source|track)';
+
+        // Split the HTML into a stream of elements and text nodes.
+        $stream = wp_html_split( $pee );
+
+        // Assume that the top level is pee-able.
+        $peeable = array( true );
+
+        foreach ( $stream as $id => $droplet ) {
+                if ( '<' === substr( $droplet, 0, 1 ) ) {
+                        if ( preg_match( '!<' . $emptyblocks . '(?: [^>]*)?>!s', $droplet ) ) {
+                                // If we encounter an empty element, ignore it.
+                                continue;
+                        } elseif ( preg_match( '!^<.+/>$!s', $droplet ) ) {
+                                // Ignore self-closing elements, too.
+                                continue;
+                        } elseif ( preg_match( '%^<!--.+-->$%s', $droplet ) ) {
+                                // Comments can be ignored.
+                                continue;
+                        } elseif ( preg_match( '%^<!\[CDATA\[.+\]\]>$%s', $droplet ) ) {
+                                // CDATA can totally be ignored.
+                                continue;
+                        } elseif ( preg_match( '!^</.+>$!s', $droplet ) ) {
+                                // Here's closing element, let's move back up a level.
+                                array_pop( $peeable );
+                        } elseif ( preg_match( '!^<' . $peeblocks . '(?: [^>]*)?>$!s', $droplet ) ) {
+                                // We've just entered a pee-able block, mark it so.
+                                $peeable[] = true;
+                        } elseif ( preg_match( '!^<' . $mediablocks . '(?: [^>]*)?>$!s', $droplet ) ) {
+                                // Media blocks can't really be pee'd, but they're handle later on.
+                                $peeable[] = true;
+                        } else {
+                                $peeable[] = false;
+                        }
+
+                        // We can't pee an element, so move to the next droplet.
+                        continue;
+                }
+
+                // If the current level can't be pee'd, protect it from being pee'd later.
+                if ( ! end( $peeable ) ) {
+                        $stream[ $id ] = str_replace( "\n", '<!-- wpnl -->', $droplet );
+                }
+        }
+
+        // If we accidentally marked the final newline as being unpee-able (for example, if there's
+        // some malformed HTML that didn't close properly), fix it.
+        if ( '<!-- wpnl -->' === $stream[ $id ] ) {
+                $stream[ $id ] = "\n";
+        }
+
+        // Reform the HTML stream.
+        $pee = implode( $stream );
 
         // Add a double line break above block-level opening tags.
         $pee = preg_replace( '!(<' . $allblocks . '[\s/>])!', "\n\n$1", $pee );
@@ -500 +566 @@
         // Add a double line break below block-level closing tags.
         $pee = preg_replace( '!(</' . $allblocks . '>)!', "$1\n\n", $pee );
 
+        // Add a double line break below block-level opening tags that are allowed to contain a <p>.
+        $pee = preg_replace( '%(<' . $peeblocks . '(?: [^>]*)?>)%', "$1\n\n", $pee );
+
+        // Add a double line break above block-level closing tags that are allowed to contain a <p>.
+        $pee = preg_replace( '%(</' . $peeblocks . '>)%', "\n\n$1", $pee );
+
         // Standardize newline characters to "\n".
         $pee = str_replace( array( "\r\n", "\r" ), "\n", $pee );
 
@@ -538 +610 @@
                 $pee = preg_replace( '|</figcaption>\s*|', '</figcaption>', $pee );
         }
 
+        // If there's only one paragraph inside a pee block, remove the newlines.
+        $pee = preg_replace( '%(<(' . $peeblocks . ")(?: [^>]*)?>)\n\n((?(?!\n\n).)*)\n\n(</\\2>)%s", '$1$3$4', $pee );
+
         // Remove more than two contiguous line breaks.
         $pee = preg_replace( "/\n\n+/", "\n\n", $pee );
 
@@ -569 +644 @@
         $pee = str_replace( '</blockquote></p>', '</p></blockquote>', $pee );
 
         // If an opening or closing block element tag is preceded by an opening <p> tag, remove it.
-        $pee = preg_replace( '!<p>\s*(</?' . $allblocks . '[^>]*>)!', '$1', $pee );
+        $pee = preg_replace( '%<p>(?:\s|<!-- wpnl -->)*(</?' . $allblocks . '[^>]*>)%', '$1', $pee );
 
         // If an opening or closing block element tag is followed by a closing <p> tag, remove it.
-        $pee = preg_replace( '!(</?' . $allblocks . '[^>]*>)\s*</p>!', '$1', $pee );
+        $pee = preg_replace( '%(</?' . $allblocks . '[^>]*>)(?:\s|<!-- wpnl -->)*</p>%', '$1', $pee );
 
         // Optionally insert line breaks.
         if ( $br ) {

tests/phpunit/tests/formatting/Autop.php

@@ -332 +332 @@
                 $content = array();
 
                 foreach ( $blocks as $block ) {
+                        if ( 'hr' === $block ) {
+                                continue;
+                        }
                         $content[] = "<$block>foo</$block>";
                 }
 
@@ -360 +363 @@
                 $content = array();
 
                 foreach ( $blocks as $block ) {
+                        if ( 'hr' === $block ) {
+                                continue;
+                        }
                         $content[] = "<$block attr='value'>foo</$block>";
                 }
 
@@ -506 +512 @@
         }
 
         /**
-         * wpautop() should convert multiple line breaks into a paragraph regarless of <br /> format
+         * wpautop() should convert multiple line breaks into a paragraph regardless of <br /> format
          *
          * @ticket 33377
          */
@@ -558 +564 @@
                 $this->assertEquals( $expected2, trim( wpautop( $content2 ) ) );
         }
 
+        /**
+         * @ticket 38656
+         */
+        function data_paragraphs_inside_blocks() {
+                $data = array(
+                        array(
+                                "<div>a\n\nb</div>",
+                                "<div>\n<p>a</p>\n<p>b</p>\n</div>",
+                        ),
+                        array(
+                                "<h1>a\n\nb</h1>",
+                                "<h1>a\n\nb</h1>",
+                        ),
+                        array(
+                                "<h1>a\nb</h1>",
+                                "<h1>a\nb</h1>",
+                        ),
+                        array(
+                                "<ul><li>a\n\nb</li></ul>",
+                                "<ul>\n<li>\n<p>a</p>\n<p>b</p>\n</li>\n</ul>",
+                        ),
+                        array(
+                                "<ul><li>a\n\n<ul><li>b\n\nc</li></ul></li></ul>",
+                                "<ul>\n<li>\n<p>a</p>\n<ul>\n<li>\n<p>b</p>\n<p>c</p>\n</li>\n</ul>\n</li>\n</ul>",
+                        ),
+                        array(
+                                "<div>a\n\n<hr>\n\nb</div>",
+                                "<div>\n<p>a</p>\n<hr>\n<p>b</p>\n</div>",
+                        ),
+                        array(
+                                "<table>\n\n<tr>\n\n<td>a\n\nb</td><td>\n\n<ul><li>c\n\n<ul><li>d\n\ne</li></ul>f\n\ng</li></ul>h</td>\n\n</tr>\n\n</table>",
+                                "<table>\n<tr>\n<td>\n<p>a</p>\n<p>b</p>\n</td>\n<td>\n<ul>\n<li>\n<p>c</p>\n<ul>\n<li>\n<p>d</p>\n<p>e</p>\n</li>\n</ul>\n<p>f</p>\n<p>g</p>\n</li>\n</ul>\n<p>h</p>\n</td>\n</tr>\n</table>",
+                        ),
+                );
+
+                return $data;
+        }
+
+        /**
+         * @ticket 38656
+         * @dataProvider data_paragraphs_inside_blocks
+         */
+        function test_paragraphs_inside_blocks( $content, $expected ) {
+                $this->assertEquals( $expected, trim( wpautop( $content ) ) );
+        }
 }