wpdb::strip_invalid_text( array $data ): array|WP_Error

3549	protected function strip_invalid_text( $data ) {3550		$db_check_string = false;3551 3552		foreach ( $data as &$value ) {3553			$charset = $value['charset'];3554 3555			if ( is_array( $value['length'] ) ) {3556				$length                  = $value['length']['length'];3557				$truncate_by_byte_length = 'byte' === $value['length']['type'];3558			} else {3559				$length = false;3560				/*3561				 * Since we have no length, we'll never truncate. Initialize the variable to false.3562				 * True would take us through an unnecessary (for this case) codepath below.3563				 */3564				$truncate_by_byte_length = false;3565			}3566 3567			// There's no charset to work with.3568			if ( false === $charset ) {3569				continue;3570			}3571 3572			// Column isn't a string.3573			if ( ! is_string( $value['value'] ) ) {3574				continue;3575			}3576 3577			$needs_validation = true;3578			if (3579				// latin1 can store any byte sequence.3580				'latin1' === $charset3581			||3582				// ASCII is always OK.3583				( ! isset( $value['ascii'] ) && $this->check_ascii( $value['value'] ) )3584			) {3585				$truncate_by_byte_length = true;3586				$needs_validation        = false;3587			}3588 3589			if ( $truncate_by_byte_length ) {3590				mbstring_binary_safe_encoding();3591				if ( false !== $length && strlen( $value['value'] ) > $length ) {3592					$value['value'] = substr( $value['value'], 0, $length );3593				}3594				reset_mbstring_encoding();3595 3596				if ( ! $needs_validation ) {3597					continue;3598				}3599			}3600 3601			// utf8 can be handled by regex, which is a bunch faster than a DB lookup.3602			if ( ( 'utf8' === $charset || 'utf8mb3' === $charset || 'utf8mb4' === $charset ) && function_exists( 'mb_strlen' ) ) {3603				$regex = '/3604					(3605						(?: [\x00-\x7F]                  # single-byte sequences   0xxxxxxx3606						|   [\xC2-\xDF][\x80-\xBF]       # double-byte sequences   110xxxxx 10xxxxxx3607						|   \xE0[\xA0-\xBF][\x80-\xBF]   # triple-byte sequences   1110xxxx 10xxxxxx * 23608						|   [\xE1-\xEC][\x80-\xBF]{2}3609						|   \xED[\x80-\x9F][\x80-\xBF]3610						|   [\xEE-\xEF][\x80-\xBF]{2}';3611 3612				if ( 'utf8mb4' === $charset ) {3613					$regex .= '3614						|    \xF0[\x90-\xBF][\x80-\xBF]{2} # four-byte sequences   11110xxx 10xxxxxx * 33615						|    [\xF1-\xF3][\x80-\xBF]{3}3616						|    \xF4[\x80-\x8F][\x80-\xBF]{2}3617					';3618				}3619 3620				$regex         .= '){1,40}                          # ...one or more times3621					)3622					| .                                  # anything else3623					/x';3624				$value['value'] = preg_replace( $regex, '$1', $value['value'] );3625 3626				if ( false !== $length && mb_strlen( $value['value'], 'UTF-8' ) > $length ) {3627					$value['value'] = mb_substr( $value['value'], 0, $length, 'UTF-8' );3628				}