PHP
6 days ago
class-convertedurl.php
6 days ago
class-cssurlprocessor.php
6 days ago
class-nativeurlintextprocessorwrapper.php
6 days ago
class-urlintextprocessor.php
6 days ago
class-urlrewritecache.php
6 days ago
class-wpurl.php
6 days ago
class-wpwhatwgurl.php
6 days ago
class-wpwhatwgurlsearchparams.php
6 days ago
functions.php
6 days ago
public-suffix-list.php
6 days ago
functions.php
215 lines
| 1 | <?php |
| 2 | |
| 3 | namespace WordPress\DataLiberation\URL; |
| 4 | |
| 5 | use Rowbot\URL\URL; |
| 6 | use WordPress\DataLiberation\BlockMarkup\BlockMarkupUrlProcessor; |
| 7 | |
| 8 | require_once __DIR__ . '/class-urlrewritecache.php'; |
| 9 | |
| 10 | /** |
| 11 | * Migrate URLs in post content. See WPRewriteUrlsTests for |
| 12 | * specific examples. TODO: A better description. |
| 13 | * |
| 14 | * Example: |
| 15 | * |
| 16 | * ```php |
| 17 | * php > wp_rewrite_urls([ |
| 18 | * 'block_markup' => '<!-- wp:image {"src": "http://legacy-blog.com/image.jpg"} -->', |
| 19 | * 'url-mapping' => [ |
| 20 | * 'http://legacy-blog.com' => 'https://modern-webstore.org' |
| 21 | * ] |
| 22 | * ]) |
| 23 | * <!-- wp:image {"src":"https:\/\/modern-webstore.org\/image.jpg"} --> |
| 24 | * ``` |
| 25 | * |
| 26 | * @TODO Use a proper JSON parser and encoder to: |
| 27 | * * Support UTF-16 characters |
| 28 | * * Gracefully handle recoverable encoding issues |
| 29 | * * Avoid changing the whitespace in the same manner as |
| 30 | * we do in WP_HTML_Tag_Processor. e.g. if we start with: |
| 31 | * |
| 32 | * ```html |
| 33 | * <!-- wp:block {"url":"https://w.org"}` --> |
| 34 | * ^ no space here |
| 35 | * ``` |
| 36 | * |
| 37 | * then it would be nice to re-encode that block markup also without the space character. This is similar |
| 38 | * to how the tag processor avoids changing parts of the tag it doesn't need to change. |
| 39 | */ |
| 40 | function wp_rewrite_urls( $options ) { |
| 41 | static $rewrite_cache = null; |
| 42 | |
| 43 | if ( null === $rewrite_cache ) { |
| 44 | $rewrite_cache = new URLRewriteCache(); |
| 45 | } |
| 46 | |
| 47 | if ( empty( $options['base_url'] ) ) { |
| 48 | // Use first from-url as base_url if not specified. |
| 49 | $from_urls = array_keys( $options['url-mapping'] ); |
| 50 | $options['base_url'] = $from_urls[0]; |
| 51 | } |
| 52 | |
| 53 | $base_url_object = WPURL::parse( $options['base_url'] ); |
| 54 | if ( false === $base_url_object ) { |
| 55 | return $options['block_markup']; |
| 56 | } |
| 57 | |
| 58 | $url_mapping = array(); |
| 59 | $mapping_key_parts = array( 'base=' . $options['base_url'] ); |
| 60 | foreach ( $options['url-mapping'] as $from_url_string => $to_url_string ) { |
| 61 | $url_mapping[] = array( |
| 62 | 'from_url' => WPURL::parse( $from_url_string ), |
| 63 | 'to_url' => WPURL::parse( $to_url_string ), |
| 64 | ); |
| 65 | $mapping_key_parts[] = $from_url_string . '=>' . $to_url_string; |
| 66 | } |
| 67 | $mapping_cache_key = sha1( implode( "\0", $mapping_key_parts ) ); |
| 68 | |
| 69 | $p = new BlockMarkupUrlProcessor( $options['block_markup'], $options['base_url'] ); |
| 70 | while ( $p->next_url() ) { |
| 71 | $token_type = $p->get_token_type(); |
| 72 | $raw_url = $p->get_raw_url(); |
| 73 | $cache_key = $mapping_cache_key . "\0" . $token_type . "\0" . $raw_url; |
| 74 | |
| 75 | $cached = $rewrite_cache->get( $cache_key ); |
| 76 | if ( null !== $cached ) { |
| 77 | if ( false !== $cached ) { |
| 78 | $p->set_url( $cached['raw_url'], $cached['parsed_url'] ); |
| 79 | } |
| 80 | continue; |
| 81 | } |
| 82 | |
| 83 | $parsed_url = $p->get_parsed_url(); |
| 84 | $converted = false; |
| 85 | foreach ( $url_mapping as $mapping ) { |
| 86 | if ( is_child_url_of( $parsed_url, $mapping['from_url'] ) ) { |
| 87 | $converted = WPURL::replace_base_url( |
| 88 | $parsed_url, |
| 89 | array( |
| 90 | 'old_base_url' => $base_url_object, |
| 91 | 'new_base_url' => $mapping['to_url'], |
| 92 | 'raw_url' => $raw_url, |
| 93 | 'is_relative' => ( |
| 94 | '#text' !== $token_type && |
| 95 | ! WPURL::can_parse( $raw_url ) |
| 96 | ), |
| 97 | ) |
| 98 | ); |
| 99 | break; |
| 100 | } |
| 101 | } |
| 102 | |
| 103 | $cache_value = false; |
| 104 | if ( false !== $converted ) { |
| 105 | $cache_value = array( |
| 106 | 'raw_url' => (string) $converted, |
| 107 | 'parsed_url' => $converted->new_url, |
| 108 | ); |
| 109 | $p->set_url( $cache_value['raw_url'], $cache_value['parsed_url'] ); |
| 110 | } |
| 111 | |
| 112 | $rewrite_cache->set( $cache_key, $cache_value ); |
| 113 | } |
| 114 | |
| 115 | return $p->get_updated_html(); |
| 116 | } |
| 117 | |
| 118 | /** |
| 119 | * Check if a given URL matches the current site URL. |
| 120 | * |
| 121 | * @param URL $child The URL to check. |
| 122 | * @param string $parent_url The current site URL to compare against. |
| 123 | * |
| 124 | * @return bool Whether the URL matches the current site URL. |
| 125 | */ |
| 126 | function is_child_url_of( $child, $parent_url ) { |
| 127 | $parent_url = is_string( $parent_url ) ? WPURL::parse( $parent_url ) : $parent_url; |
| 128 | $child = is_string( $child ) ? WPURL::parse( $child ) : $child; |
| 129 | $child_pathname_no_trailing_slash = rtrim( urldecode( $child->pathname ), '/' ); |
| 130 | |
| 131 | if ( false === $child || false === $parent_url ) { |
| 132 | return false; |
| 133 | } |
| 134 | |
| 135 | if ( $parent_url->hostname !== $child->hostname ) { |
| 136 | return false; |
| 137 | } |
| 138 | |
| 139 | if ( $parent_url->protocol !== $child->protocol ) { |
| 140 | return false; |
| 141 | } |
| 142 | |
| 143 | $parent_pathname = urldecode( $parent_url->pathname ); |
| 144 | |
| 145 | return ( |
| 146 | // Direct match. |
| 147 | $parent_pathname === $child_pathname_no_trailing_slash || |
| 148 | $parent_pathname === $child_pathname_no_trailing_slash . '/' || |
| 149 | // Path prefix. |
| 150 | 0 === strncmp( $child_pathname_no_trailing_slash . '/', $parent_pathname, strlen( $parent_pathname ) ) |
| 151 | ); |
| 152 | } |
| 153 | |
| 154 | /** |
| 155 | * Decodes the first n **encoded bytes** a URL-encoded string. |
| 156 | * |
| 157 | * For example, `urldecode_n( '%22is 6 %3C 6?%22 – asked Achilles', 1 )` returns |
| 158 | * '"is 6 %3C 6?%22 – asked Achilles' because only the first encoded byte is decoded. |
| 159 | * |
| 160 | * @param string $input The string to decode. |
| 161 | * @param int $decode_n The number of bytes to decode in $input |
| 162 | * |
| 163 | * @return string The decoded string. |
| 164 | */ |
| 165 | function urldecode_n( $input, $decode_n ) { |
| 166 | // Fast paths: nothing to do. |
| 167 | if ( $decode_n <= 0 || false === strpos( $input, '%' ) ) { |
| 168 | return $input; |
| 169 | } |
| 170 | |
| 171 | $result = ''; |
| 172 | $at = 0; |
| 173 | while ( true ) { |
| 174 | if ( $at + 3 > strlen( $input ) ) { |
| 175 | break; |
| 176 | } |
| 177 | |
| 178 | $last_at = $at; |
| 179 | $at += strcspn( $input, '%', $at ); |
| 180 | // Consume bytes except for the percent sign. |
| 181 | $result .= substr( $input, $last_at, $at - $last_at ); |
| 182 | |
| 183 | // If we've already decoded the requested number of bytes, stop. |
| 184 | if ( strlen( $result ) >= $decode_n ) { |
| 185 | break; |
| 186 | } |
| 187 | |
| 188 | ++$at; |
| 189 | if ( $at > strlen( $input ) ) { |
| 190 | break; |
| 191 | } |
| 192 | |
| 193 | $decodable_length = strspn( |
| 194 | $input, |
| 195 | '0123456789ABCDEFabcdef', |
| 196 | $at, |
| 197 | 2 |
| 198 | ); |
| 199 | |
| 200 | if ( 2 === $decodable_length ) { |
| 201 | // Decodes the urlencoded hex sequence from URL. |
| 202 | // Note: This decodes bytes, not characters. It will recover the original byte sequence, |
| 203 | // not necessarily any valid UTF-8 characters. |
| 204 | $result .= chr( hexdec( $input[ $at ] . $input[ $at + 1 ] ) ); |
| 205 | $at += 2; |
| 206 | } else { |
| 207 | // Consume the next byte and move on. |
| 208 | $result .= '%'; |
| 209 | } |
| 210 | } |
| 211 | $result .= substr( $input, $at ); |
| 212 | |
| 213 | return $result; |
| 214 | } |
| 215 |