diff -r c7c34916027a -r 177826044cd9 wp/wp-includes/class-wp-text-diff-renderer-table.php
--- a/wp/wp-includes/class-wp-text-diff-renderer-table.php Mon Oct 14 18:06:33 2019 +0200
+++ b/wp/wp-includes/class-wp-text-diff-renderer-table.php Mon Oct 14 18:28:13 2019 +0200
@@ -20,7 +20,7 @@
* @var int
* @since 2.6.0
*/
- public $_leading_context_lines = 10000;
+ public $_leading_context_lines = 10000;
/**
* @see Text_Diff_Renderer::_trailing_context_lines
@@ -56,6 +56,22 @@
protected $compat_fields = array( '_show_split_view', 'inline_diff_renderer', '_diff_threshold' );
/**
+ * Caches the output of count_chars() in compute_string_distance()
+ *
+ * @var array
+ * @since 5.0.0
+ */
+ protected $count_cache = array();
+
+ /**
+ * Caches the difference calculation in compute_string_distance()
+ *
+ * @var array
+ * @since 5.0.0
+ */
+ protected $difference_cache = array();
+
+ /**
* Constructor - Call parent constructor with params array.
*
* This will set class properties based on the key value pairs in the array.
@@ -66,8 +82,9 @@
*/
public function __construct( $params = array() ) {
parent::__construct( $params );
- if ( isset( $params[ 'show_split_view' ] ) )
- $this->_show_split_view = $params[ 'show_split_view' ];
+ if ( isset( $params['show_split_view'] ) ) {
+ $this->_show_split_view = $params['show_split_view'];
+ }
}
/**
@@ -86,7 +103,7 @@
* @param array $lines
* @param string $prefix
*/
- public function _lines( $lines, $prefix=' ' ) {
+ public function _lines( $lines, $prefix = ' ' ) {
}
/**
@@ -138,7 +155,7 @@
*/
public function _added( $lines, $encode = true ) {
$r = '';
- foreach ($lines as $line) {
+ foreach ( $lines as $line ) {
if ( $encode ) {
$processed_line = htmlspecialchars( $line );
@@ -153,7 +170,7 @@
*
* @param String $processed_line The processed diffed line.
* @param String $line The unprocessed diffed line.
- * @param string null The line context. Values are 'added', 'deleted' or 'unchanged'.
+ * @param string null The line context. Values are 'added', 'deleted' or 'unchanged'.
*/
$line = apply_filters( 'process_text_diff_html', $processed_line, $line, 'added' );
}
@@ -176,7 +193,7 @@
*/
public function _deleted( $lines, $encode = true ) {
$r = '';
- foreach ($lines as $line) {
+ foreach ( $lines as $line ) {
if ( $encode ) {
$processed_line = htmlspecialchars( $line );
@@ -188,7 +205,6 @@
} else {
$r .= '
' . $this->deletedLine( $line ) . "
\n";
}
-
}
return $r;
}
@@ -202,15 +218,15 @@
*/
public function _context( $lines, $encode = true ) {
$r = '';
- foreach ($lines as $line) {
+ foreach ( $lines as $line ) {
if ( $encode ) {
$processed_line = htmlspecialchars( $line );
/** This filter is documented in wp-includes/wp-diff.php */
$line = apply_filters( 'process_text_diff_html', $processed_line, $line, 'unchanged' );
}
- if ( $this->_show_split_view ) {
- $r .= '' . $this->contextLine( $line ) . $this->emptyLine() . $this->contextLine( $line ) . "
\n";
+ if ( $this->_show_split_view ) {
+ $r .= '' . $this->contextLine( $line ) . $this->emptyLine() . $this->contextLine( $line ) . "
\n";
} else {
$r .= '' . $this->contextLine( $line ) . "
\n";
}
@@ -248,58 +264,62 @@
// Compute word diffs for each matched pair using the inline diff
foreach ( $orig_matches as $o => $f ) {
- if ( is_numeric($o) && is_numeric($f) ) {
- $text_diff = new Text_Diff( 'auto', array( array($orig[$o]), array($final[$f]) ) );
- $renderer = new $this->inline_diff_renderer;
- $diff = $renderer->render( $text_diff );
+ if ( is_numeric( $o ) && is_numeric( $f ) ) {
+ $text_diff = new Text_Diff( 'auto', array( array( $orig[ $o ] ), array( $final[ $f ] ) ) );
+ $renderer = new $this->inline_diff_renderer;
+ $diff = $renderer->render( $text_diff );
// If they're too different, don't include any or
if ( preg_match_all( '!(.*?|.*?)!', $diff, $diff_matches ) ) {
// length of all text between or
- $stripped_matches = strlen(strip_tags( join(' ', $diff_matches[0]) ));
+ $stripped_matches = strlen( strip_tags( join( ' ', $diff_matches[0] ) ) );
// since we count lengith of text between or (instead of picking just one),
// we double the length of chars not in those tags.
- $stripped_diff = strlen(strip_tags( $diff )) * 2 - $stripped_matches;
- $diff_ratio = $stripped_matches / $stripped_diff;
- if ( $diff_ratio > $this->_diff_threshold )
+ $stripped_diff = strlen( strip_tags( $diff ) ) * 2 - $stripped_matches;
+ $diff_ratio = $stripped_matches / $stripped_diff;
+ if ( $diff_ratio > $this->_diff_threshold ) {
continue; // Too different. Don't save diffs.
+ }
}
// Un-inline the diffs by removing del or ins
- $orig_diffs[$o] = preg_replace( '|.*?|', '', $diff );
- $final_diffs[$f] = preg_replace( '|.*?|', '', $diff );
+ $orig_diffs[ $o ] = preg_replace( '|.*?|', '', $diff );
+ $final_diffs[ $f ] = preg_replace( '|.*?|', '', $diff );
}
}
- foreach ( array_keys($orig_rows) as $row ) {
+ foreach ( array_keys( $orig_rows ) as $row ) {
// Both columns have blanks. Ignore them.
- if ( $orig_rows[$row] < 0 && $final_rows[$row] < 0 )
+ if ( $orig_rows[ $row ] < 0 && $final_rows[ $row ] < 0 ) {
continue;
+ }
// If we have a word based diff, use it. Otherwise, use the normal line.
- if ( isset( $orig_diffs[$orig_rows[$row]] ) )
- $orig_line = $orig_diffs[$orig_rows[$row]];
- elseif ( isset( $orig[$orig_rows[$row]] ) )
- $orig_line = htmlspecialchars($orig[$orig_rows[$row]]);
- else
+ if ( isset( $orig_diffs[ $orig_rows[ $row ] ] ) ) {
+ $orig_line = $orig_diffs[ $orig_rows[ $row ] ];
+ } elseif ( isset( $orig[ $orig_rows[ $row ] ] ) ) {
+ $orig_line = htmlspecialchars( $orig[ $orig_rows[ $row ] ] );
+ } else {
$orig_line = '';
+ }
- if ( isset( $final_diffs[$final_rows[$row]] ) )
- $final_line = $final_diffs[$final_rows[$row]];
- elseif ( isset( $final[$final_rows[$row]] ) )
- $final_line = htmlspecialchars($final[$final_rows[$row]]);
- else
+ if ( isset( $final_diffs[ $final_rows[ $row ] ] ) ) {
+ $final_line = $final_diffs[ $final_rows[ $row ] ];
+ } elseif ( isset( $final[ $final_rows[ $row ] ] ) ) {
+ $final_line = htmlspecialchars( $final[ $final_rows[ $row ] ] );
+ } else {
$final_line = '';
+ }
- if ( $orig_rows[$row] < 0 ) { // Orig is blank. This is really an added row.
- $r .= $this->_added( array($final_line), false );
- } elseif ( $final_rows[$row] < 0 ) { // Final is blank. This is really a deleted row.
- $r .= $this->_deleted( array($orig_line), false );
+ if ( $orig_rows[ $row ] < 0 ) { // Orig is blank. This is really an added row.
+ $r .= $this->_added( array( $final_line ), false );
+ } elseif ( $final_rows[ $row ] < 0 ) { // Final is blank. This is really a deleted row.
+ $r .= $this->_deleted( array( $orig_line ), false );
} else { // A true changed row.
if ( $this->_show_split_view ) {
$r .= '' . $this->deletedLine( $orig_line ) . $this->emptyLine() . $this->addedLine( $final_line ) . "
\n";
} else {
- $r .= '' . $this->deletedLine( $orig_line ) . "
" . $this->addedLine( $final_line ) . "
\n";
+ $r .= '' . $this->deletedLine( $orig_line ) . '
' . $this->addedLine( $final_line ) . "
\n";
}
}
}
@@ -339,80 +359,81 @@
// Contains all pairwise string comparisons. Keys are such that this need only be a one dimensional array.
$matches = array();
- foreach ( array_keys($orig) as $o ) {
- foreach ( array_keys($final) as $f ) {
- $matches["$o,$f"] = $this->compute_string_distance( $orig[$o], $final[$f] );
+ foreach ( array_keys( $orig ) as $o ) {
+ foreach ( array_keys( $final ) as $f ) {
+ $matches[ "$o,$f" ] = $this->compute_string_distance( $orig[ $o ], $final[ $f ] );
}
}
- asort($matches); // Order by string distance.
+ asort( $matches ); // Order by string distance.
$orig_matches = array();
$final_matches = array();
foreach ( $matches as $keys => $difference ) {
- list($o, $f) = explode(',', $keys);
- $o = (int) $o;
- $f = (int) $f;
+ list($o, $f) = explode( ',', $keys );
+ $o = (int) $o;
+ $f = (int) $f;
// Already have better matches for these guys
- if ( isset($orig_matches[$o]) && isset($final_matches[$f]) )
+ if ( isset( $orig_matches[ $o ] ) && isset( $final_matches[ $f ] ) ) {
continue;
+ }
// First match for these guys. Must be best match
- if ( !isset($orig_matches[$o]) && !isset($final_matches[$f]) ) {
- $orig_matches[$o] = $f;
- $final_matches[$f] = $o;
+ if ( ! isset( $orig_matches[ $o ] ) && ! isset( $final_matches[ $f ] ) ) {
+ $orig_matches[ $o ] = $f;
+ $final_matches[ $f ] = $o;
continue;
}
// Best match of this final is already taken? Must mean this final is a new row.
- if ( isset($orig_matches[$o]) )
- $final_matches[$f] = 'x';
-
- // Best match of this orig is already taken? Must mean this orig is a deleted row.
- elseif ( isset($final_matches[$f]) )
- $orig_matches[$o] = 'x';
+ if ( isset( $orig_matches[ $o ] ) ) {
+ $final_matches[ $f ] = 'x';
+ } elseif ( isset( $final_matches[ $f ] ) ) {
+ // Best match of this orig is already taken? Must mean this orig is a deleted row.
+ $orig_matches[ $o ] = 'x';
+ }
}
// We read the text in this order
- ksort($orig_matches);
- ksort($final_matches);
+ ksort( $orig_matches );
+ ksort( $final_matches );
// Stores rows and blanks for each column.
- $orig_rows = $orig_rows_copy = array_keys($orig_matches);
- $final_rows = array_keys($final_matches);
+ $orig_rows = $orig_rows_copy = array_keys( $orig_matches );
+ $final_rows = array_keys( $final_matches );
// Interleaves rows with blanks to keep matches aligned.
// We may end up with some extraneous blank rows, but we'll just ignore them later.
foreach ( $orig_rows_copy as $orig_row ) {
- $final_pos = array_search($orig_matches[$orig_row], $final_rows, true);
- $orig_pos = (int) array_search($orig_row, $orig_rows, true);
+ $final_pos = array_search( $orig_matches[ $orig_row ], $final_rows, true );
+ $orig_pos = (int) array_search( $orig_row, $orig_rows, true );
if ( false === $final_pos ) { // This orig is paired with a blank final.
array_splice( $final_rows, $orig_pos, 0, -1 );
} elseif ( $final_pos < $orig_pos ) { // This orig's match is up a ways. Pad final with blank rows.
- $diff_pos = $final_pos - $orig_pos;
- while ( $diff_pos < 0 )
- array_splice( $final_rows, $orig_pos, 0, $diff_pos++ );
+ $diff_array = range( -1, $final_pos - $orig_pos );
+ array_splice( $final_rows, $orig_pos, 0, $diff_array );
} elseif ( $final_pos > $orig_pos ) { // This orig's match is down a ways. Pad orig with blank rows.
- $diff_pos = $orig_pos - $final_pos;
- while ( $diff_pos < 0 )
- array_splice( $orig_rows, $orig_pos, 0, $diff_pos++ );
+ $diff_array = range( -1, $orig_pos - $final_pos );
+ array_splice( $orig_rows, $orig_pos, 0, $diff_array );
}
}
// Pad the ends with blank rows if the columns aren't the same length
- $diff_count = count($orig_rows) - count($final_rows);
+ $diff_count = count( $orig_rows ) - count( $final_rows );
if ( $diff_count < 0 ) {
- while ( $diff_count < 0 )
- array_push($orig_rows, $diff_count++);
+ while ( $diff_count < 0 ) {
+ array_push( $orig_rows, $diff_count++ );
+ }
} elseif ( $diff_count > 0 ) {
$diff_count = -1 * $diff_count;
- while ( $diff_count < 0 )
- array_push($final_rows, $diff_count++);
+ while ( $diff_count < 0 ) {
+ array_push( $final_rows, $diff_count++ );
+ }
}
- return array($orig_matches, $final_matches, $orig_rows, $final_rows);
+ return array( $orig_matches, $final_matches, $orig_rows, $final_rows );
}
/**
@@ -425,19 +446,36 @@
* @return int
*/
public function compute_string_distance( $string1, $string2 ) {
- // Vectors containing character frequency for all chars in each string
- $chars1 = count_chars($string1);
- $chars2 = count_chars($string2);
+ // Use an md5 hash of the strings for a count cache, as it's fast to generate, and collisions aren't a concern.
+ $count_key1 = md5( $string1 );
+ $count_key2 = md5( $string2 );
- // L1-norm of difference vector.
- $difference = array_sum( array_map( array($this, 'difference'), $chars1, $chars2 ) );
+ // Cache vectors containing character frequency for all chars in each string.
+ if ( ! isset( $this->count_cache[ $count_key1 ] ) ) {
+ $this->count_cache[ $count_key1 ] = count_chars( $string1 );
+ }
+ if ( ! isset( $this->count_cache[ $count_key2 ] ) ) {
+ $this->count_cache[ $count_key2 ] = count_chars( $string2 );
+ }
+
+ $chars1 = $this->count_cache[ $count_key1 ];
+ $chars2 = $this->count_cache[ $count_key2 ];
+
+ $difference_key = md5( implode( ',', $chars1 ) . ':' . implode( ',', $chars2 ) );
+ if ( ! isset( $this->difference_cache[ $difference_key ] ) ) {
+ // L1-norm of difference vector.
+ $this->difference_cache[ $difference_key ] = array_sum( array_map( array( $this, 'difference' ), $chars1, $chars2 ) );
+ }
+
+ $difference = $this->difference_cache[ $difference_key ];
// $string1 has zero length? Odd. Give huge penalty by not dividing.
- if ( !$string1 )
+ if ( ! $string1 ) {
return $difference;
+ }
// Return distance per character (of string1).
- return $difference / strlen($string1);
+ return $difference / strlen( $string1 );
}
/**