Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
99.03% |
205 / 207 |
|
88.89% |
16 / 18 |
CRAP | |
0.00% |
0 / 1 |
| Ai_Crawlers | |
99.03% |
205 / 207 |
|
88.89% |
16 / 18 |
52 | |
0.00% |
0 / 1 |
| init | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| get_catalog | |
100.00% |
78 / 78 |
|
100.00% |
1 / 1 |
1 | |||
| has_data_sharing_opt_out | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| get_data_sharing_user_agents | |
100.00% |
17 / 17 |
|
100.00% |
1 / 1 |
1 | |||
| get_data_sharing_blocked_slugs | |
100.00% |
6 / 6 |
|
100.00% |
1 / 1 |
3 | |||
| get_blocked | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| sanitize_blocked | |
100.00% |
17 / 17 |
|
100.00% |
1 / 1 |
10 | |||
| register_legacy_migration | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| migrate_legacy_overrides | |
96.43% |
27 / 28 |
|
0.00% |
0 / 1 |
10 | |||
| is_blocked | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| get_blocked_slugs | |
100.00% |
10 / 10 |
|
100.00% |
1 / 1 |
5 | |||
| search_engines_allowed | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| is_crawl_restricted_subdomain | |
100.00% |
4 / 4 |
|
100.00% |
1 / 1 |
2 | |||
| has_static_robots_txt | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| is_path_based_multisite | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
2 | |||
| append_directives | |
94.44% |
17 / 18 |
|
0.00% |
0 / 1 |
8.01 | |||
| get_bootstrap_data | |
100.00% |
20 / 20 |
|
100.00% |
1 / 1 |
2 | |||
| privacy_settings_url | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| 1 | <?php |
| 2 | /** |
| 3 | * AI crawler access controls. |
| 4 | * |
| 5 | * Lets a site owner allow or block individual AI crawlers (training and |
| 6 | * answer-engine bots) by appending per-user-agent `Disallow` rules to the |
| 7 | * WordPress-generated robots.txt. |
| 8 | * |
| 9 | * The option `jetpack_seo_blocked_crawlers` holds the crawlers the owner blocked, per |
| 10 | * group. Every other crawler is allowed, so robots.txt doesn't change when the SEO page |
| 11 | * reaches a site (JETPACK-3135). |
| 12 | * |
| 13 | * Caveat: the `robots_txt` filter only feeds WordPress's *virtual* robots.txt. |
| 14 | * A detected static file in the WordPress installation is not modified — a |
| 15 | * limitation surfaced in the GEO tab (see {@see self::has_static_robots_txt()}). |
| 16 | * |
| 17 | * @package automattic/jetpack-seo-package |
| 18 | */ |
| 19 | |
| 20 | namespace Automattic\Jetpack\SEO; |
| 21 | |
| 22 | /** |
| 23 | * Emits robots.txt directives that block opted-out AI crawlers. |
| 24 | */ |
| 25 | class Ai_Crawlers { |
| 26 | |
| 27 | /** |
| 28 | * Option holding the blocked crawlers as `array( 'answer' => …, 'training' => … )`, each |
| 29 | * `'all'` (every crawler in the group, including ones added to the catalog later) or a |
| 30 | * list of slugs. A missing group or an empty list blocks nothing. Registered with |
| 31 | * `show_in_rest` by {@see Dashboard_Data::register_rest_settings()} so the GEO tab can save it. |
| 32 | * |
| 33 | * @var string |
| 34 | */ |
| 35 | const OPTION = 'jetpack_seo_blocked_crawlers'; |
| 36 | |
| 37 | /** |
| 38 | * Option the GEO tab saved before {@see self::OPTION}: `slug => bool`, holding only changes |
| 39 | * from the old defaults (training crawlers blocked, answer engines allowed). Deleted once |
| 40 | * {@see self::migrate_legacy_overrides()} has copied it. |
| 41 | * |
| 42 | * @var string |
| 43 | */ |
| 44 | const LEGACY_OPTION = 'jetpack_seo_ai_crawler_overrides'; |
| 45 | |
| 46 | /** |
| 47 | * Priority of {@see self::migrate_legacy_overrides()} on the `default_option_` filter. |
| 48 | * |
| 49 | * @var int |
| 50 | */ |
| 51 | const MIGRATION_PRIORITY = 20; |
| 52 | |
| 53 | /** |
| 54 | * The crawler groups, matching each catalog entry's `type`. |
| 55 | * |
| 56 | * @var string[] |
| 57 | */ |
| 58 | const GROUPS = array( 'answer', 'training' ); |
| 59 | |
| 60 | /** |
| 61 | * The `robots_txt` priority: after WordPress.com Simple appends bare `Disallow` |
| 62 | * lines at 11, which would otherwise land in the last AI crawler group. |
| 63 | * |
| 64 | * @var int |
| 65 | */ |
| 66 | const ROBOTS_TXT_PRIORITY = 12; |
| 67 | |
| 68 | /** |
| 69 | * Wire the robots.txt filter. |
| 70 | * |
| 71 | * @return void |
| 72 | */ |
| 73 | public static function init() { |
| 74 | add_filter( 'robots_txt', array( __CLASS__, 'append_directives' ), self::ROBOTS_TXT_PRIORITY, 2 ); |
| 75 | } |
| 76 | |
| 77 | /** |
| 78 | * The known AI crawler catalog: slug => [ label, user_agents, type ]. |
| 79 | * |
| 80 | * `slug` is the stable key persisted in {@see self::OPTION} and sent by the GEO |
| 81 | * tab; `user_agents` are every robots.txt token the operator uses for that |
| 82 | * crawler, each written to its own `User-agent:` group; |
| 83 | * `label` is the human name shown in the UI; `type` is `answer` (fetches to |
| 84 | * cite in live AI answers) or `training` (collects to train models). |
| 85 | * |
| 86 | * @return array<string, array{label: string, user_agents: string[], type: string}> |
| 87 | */ |
| 88 | public static function get_catalog() { |
| 89 | // Labels are the name a user is most likely to recognize. The GEO tab lists the |
| 90 | // exact tokens under each so a technical user can verify the emitted directives. |
| 91 | $catalog = array( |
| 92 | // Answer-engine crawlers fetch pages so AI assistants can cite them in |
| 93 | // live answers. Blocking them costs AI visibility. |
| 94 | 'oai-searchbot' => array( |
| 95 | 'label' => __( 'ChatGPT Search', 'jetpack-seo' ), |
| 96 | 'user_agents' => array( 'OAI-SearchBot' ), |
| 97 | 'type' => 'answer', |
| 98 | ), |
| 99 | 'claude-searchbot' => array( |
| 100 | 'label' => __( 'Claude Search', 'jetpack-seo' ), |
| 101 | 'user_agents' => array( 'Claude-SearchBot' ), |
| 102 | 'type' => 'answer', |
| 103 | ), |
| 104 | 'perplexitybot' => array( |
| 105 | 'label' => __( 'Perplexity', 'jetpack-seo' ), |
| 106 | 'user_agents' => array( 'PerplexityBot' ), |
| 107 | 'type' => 'answer', |
| 108 | ), |
| 109 | 'amzn-searchbot' => array( |
| 110 | 'label' => __( 'Amazon Alexa', 'jetpack-seo' ), |
| 111 | 'user_agents' => array( 'Amzn-SearchBot' ), |
| 112 | 'type' => 'answer', |
| 113 | ), |
| 114 | 'meta-webindexer' => array( |
| 115 | 'label' => __( 'Meta AI Search', 'jetpack-seo' ), |
| 116 | 'user_agents' => array( 'meta-webindexer' ), |
| 117 | 'type' => 'answer', |
| 118 | ), |
| 119 | 'duckassistbot' => array( |
| 120 | 'label' => __( 'DuckDuckGo Search Assist', 'jetpack-seo' ), |
| 121 | 'user_agents' => array( 'DuckAssistBot' ), |
| 122 | 'type' => 'answer', |
| 123 | ), |
| 124 | // Training crawlers collect content to train AI models. Blocking them doesn't |
| 125 | // affect search results or Google's AI Overviews. Google-Extended also covers |
| 126 | // Gemini app answers. |
| 127 | 'gptbot' => array( |
| 128 | 'label' => __( 'ChatGPT', 'jetpack-seo' ), |
| 129 | 'user_agents' => array( 'GPTBot' ), |
| 130 | 'type' => 'training', |
| 131 | ), |
| 132 | 'claudebot' => array( |
| 133 | 'label' => __( 'Claude', 'jetpack-seo' ), |
| 134 | 'user_agents' => array( 'ClaudeBot', 'anthropic-ai', 'Claude-Web' ), |
| 135 | 'type' => 'training', |
| 136 | ), |
| 137 | 'google-extended' => array( |
| 138 | 'label' => __( 'Google Gemini', 'jetpack-seo' ), |
| 139 | 'user_agents' => array( 'Google-Extended' ), |
| 140 | 'type' => 'training', |
| 141 | ), |
| 142 | 'applebot-extended' => array( |
| 143 | 'label' => __( 'Apple Intelligence', 'jetpack-seo' ), |
| 144 | 'user_agents' => array( 'Applebot-Extended' ), |
| 145 | 'type' => 'training', |
| 146 | ), |
| 147 | 'meta-externalagent' => array( |
| 148 | 'label' => __( 'Meta AI', 'jetpack-seo' ), |
| 149 | 'user_agents' => array( 'meta-externalagent', 'FacebookBot' ), |
| 150 | 'type' => 'training', |
| 151 | ), |
| 152 | 'bytespider' => array( |
| 153 | 'label' => __( 'ByteDance', 'jetpack-seo' ), |
| 154 | 'user_agents' => array( 'Bytespider' ), |
| 155 | 'type' => 'training', |
| 156 | ), |
| 157 | 'ccbot' => array( |
| 158 | 'label' => __( 'Common Crawl', 'jetpack-seo' ), |
| 159 | 'user_agents' => array( 'CCBot' ), |
| 160 | 'type' => 'training', |
| 161 | ), |
| 162 | 'amazonbot' => array( |
| 163 | // Amazon's own docs: Amazonbot "may be used to train Amazon AI |
| 164 | // models" (Amzn-SearchBot is the answer-engine bot above). |
| 165 | 'label' => __( 'Amazon', 'jetpack-seo' ), |
| 166 | 'user_agents' => array( 'Amazonbot' ), |
| 167 | 'type' => 'training', |
| 168 | ), |
| 169 | 'omgili' => array( |
| 170 | // Webz.io crawls to sell web data, including for AI training. |
| 171 | 'label' => __( 'Webz.io', 'jetpack-seo' ), |
| 172 | 'user_agents' => array( 'omgili', 'omgilibot', 'Webzio-Extended' ), |
| 173 | 'type' => 'training', |
| 174 | ), |
| 175 | ); |
| 176 | |
| 177 | return $catalog; |
| 178 | } |
| 179 | |
| 180 | /** |
| 181 | * Whether WordPress.com's site-wide data-sharing opt-out is enabled. |
| 182 | * |
| 183 | * Its `robots_txt` filter remains authoritative over per-bot settings. |
| 184 | * |
| 185 | * @return bool |
| 186 | */ |
| 187 | public static function has_data_sharing_opt_out() { |
| 188 | return (bool) get_option( 'wpcom_data_sharing_opt_out' ); |
| 189 | } |
| 190 | |
| 191 | /** |
| 192 | * The user agents WordPress.com's data-sharing opt-out blocks while it's on. |
| 193 | * |
| 194 | * A copy of the list in jetpack-mu-wpcom's blog-privacy `robots_txt` filter; |
| 195 | * AiCrawlersTest runs both filters together to keep them in step. |
| 196 | * |
| 197 | * @return string[] |
| 198 | */ |
| 199 | private static function get_data_sharing_user_agents() { |
| 200 | return array( |
| 201 | 'Amazonbot', |
| 202 | 'anthropic-ai', |
| 203 | 'Applebot-Extended', |
| 204 | 'Bytespider', |
| 205 | 'CCBot', |
| 206 | 'ClaudeBot', |
| 207 | 'FacebookBot', |
| 208 | 'Google-Extended', |
| 209 | 'GPTBot', |
| 210 | 'meta-externalagent', |
| 211 | 'omgili', |
| 212 | 'omgilibot', |
| 213 | 'PerplexityBot', |
| 214 | 'SentiBot', |
| 215 | 'sentibot', |
| 216 | ); |
| 217 | } |
| 218 | |
| 219 | /** |
| 220 | * The catalog slugs WordPress.com's data-sharing opt-out blocks while it's on: every |
| 221 | * crawler with a user agent on its list. {@see self::append_directives()} writes the |
| 222 | * crawler's other user agents, so the whole crawler is blocked. |
| 223 | * |
| 224 | * @return string[] |
| 225 | */ |
| 226 | public static function get_data_sharing_blocked_slugs() { |
| 227 | $user_agents = self::get_data_sharing_user_agents(); |
| 228 | $slugs = array(); |
| 229 | foreach ( self::get_catalog() as $slug => $info ) { |
| 230 | if ( array_intersect( $info['user_agents'], $user_agents ) ) { |
| 231 | $slugs[] = $slug; |
| 232 | } |
| 233 | } |
| 234 | return $slugs; |
| 235 | } |
| 236 | |
| 237 | /** |
| 238 | * The crawlers the owner blocked, per group. |
| 239 | * |
| 240 | * @return array<string, string|string[]> `'all'` or a list of slugs for each group. |
| 241 | */ |
| 242 | public static function get_blocked() { |
| 243 | return self::sanitize_blocked( get_option( self::OPTION, array() ) ); |
| 244 | } |
| 245 | |
| 246 | /** |
| 247 | * Normalizes a stored or submitted {@see self::OPTION} value: both groups present, each |
| 248 | * `'all'` or a list of slugs. Drops unknown slugs and slugs from the other group, and |
| 249 | * saves a list naming every crawler in its group as `'all'`. |
| 250 | * |
| 251 | * @param mixed $value The value to normalize. |
| 252 | * @param array|null $catalog Optional {@see self::get_catalog()} result, to avoid building it again. |
| 253 | * @return array<string, string|string[]> |
| 254 | */ |
| 255 | public static function sanitize_blocked( $value, $catalog = null ) { |
| 256 | $value = is_array( $value ) ? $value : array(); |
| 257 | $catalog = is_array( $catalog ) ? $catalog : self::get_catalog(); |
| 258 | $groups = array_fill_keys( self::GROUPS, array() ); |
| 259 | foreach ( $catalog as $slug => $info ) { |
| 260 | if ( isset( $groups[ $info['type'] ] ) ) { |
| 261 | $groups[ $info['type'] ][] = $slug; |
| 262 | } |
| 263 | } |
| 264 | |
| 265 | $blocked = array(); |
| 266 | foreach ( $groups as $type => $slugs ) { |
| 267 | $group = $value[ $type ] ?? array(); |
| 268 | if ( 'all' === $group ) { |
| 269 | $blocked[ $type ] = 'all'; |
| 270 | continue; |
| 271 | } |
| 272 | $list = is_array( $group ) |
| 273 | ? array_values( array_intersect( $slugs, array_filter( $group, 'is_string' ) ) ) |
| 274 | : array(); |
| 275 | $blocked[ $type ] = ( $slugs && count( $list ) === count( $slugs ) ) ? 'all' : $list; |
| 276 | } |
| 277 | |
| 278 | return $blocked; |
| 279 | } |
| 280 | |
| 281 | /** |
| 282 | * Run {@see self::migrate_legacy_overrides()} the first time {@see self::OPTION} is read while missing. |
| 283 | * |
| 284 | * @return void |
| 285 | */ |
| 286 | public static function register_legacy_migration() { |
| 287 | // After register_setting()'s default (priority 10), which would replace the migrated value. |
| 288 | add_filter( 'default_option_' . self::OPTION, array( __CLASS__, 'migrate_legacy_overrides' ), self::MIGRATION_PRIORITY ); |
| 289 | } |
| 290 | |
| 291 | /** |
| 292 | * `default_option_` callback: copies {@see self::LEGACY_OPTION} into {@see self::OPTION}, then deletes it. |
| 293 | * |
| 294 | * A training crawler saved as allowed means the owner worked on that group, so its unsaved |
| 295 | * siblings stay blocked. With no old option there is nothing to write. |
| 296 | * |
| 297 | * @param mixed $default_value The value `get_option()` would return for the missing option. |
| 298 | * @return mixed The migrated value, or `$default_value` when there is nothing to migrate. |
| 299 | */ |
| 300 | public static function migrate_legacy_overrides( $default_value ) { |
| 301 | // The catalog loads translations, which must wait for `init`. |
| 302 | if ( ! did_action( 'init' ) ) { |
| 303 | return $default_value; |
| 304 | } |
| 305 | |
| 306 | $legacy = get_option( self::LEGACY_OPTION ); |
| 307 | if ( ! is_array( $legacy ) ) { |
| 308 | return $default_value; |
| 309 | } |
| 310 | |
| 311 | $catalog = self::get_catalog(); |
| 312 | $answer_blocked = array(); |
| 313 | $training_allowed = array(); |
| 314 | foreach ( $legacy as $slug => $value ) { |
| 315 | $type = $catalog[ $slug ]['type'] ?? null; |
| 316 | if ( 'answer' === $type && $value ) { |
| 317 | $answer_blocked[] = $slug; |
| 318 | } elseif ( 'training' === $type && ! $value ) { |
| 319 | $training_allowed[] = $slug; |
| 320 | } |
| 321 | } |
| 322 | |
| 323 | $training = array_keys( wp_list_filter( $catalog, array( 'type' => 'training' ) ) ); |
| 324 | $blocked = self::sanitize_blocked( |
| 325 | array( |
| 326 | 'answer' => $answer_blocked, |
| 327 | 'training' => $training_allowed ? array_diff( $training, $training_allowed ) : array(), |
| 328 | ), |
| 329 | $catalog |
| 330 | ); |
| 331 | // add_option() applies this filter again. Re-armed after, for other sites read in this request. |
| 332 | remove_filter( 'default_option_' . self::OPTION, array( __CLASS__, 'migrate_legacy_overrides' ), self::MIGRATION_PRIORITY ); |
| 333 | $added = add_option( self::OPTION, $blocked, '', true ); |
| 334 | self::register_legacy_migration(); |
| 335 | if ( $added ) { |
| 336 | delete_option( self::LEGACY_OPTION ); |
| 337 | } |
| 338 | |
| 339 | return $blocked; |
| 340 | } |
| 341 | |
| 342 | /** |
| 343 | * Whether a given crawler is currently blocked. Unknown slugs return false. |
| 344 | * |
| 345 | * @param string $slug Catalog slug. |
| 346 | * @return bool |
| 347 | */ |
| 348 | public static function is_blocked( $slug ) { |
| 349 | return in_array( $slug, self::get_blocked_slugs(), true ); |
| 350 | } |
| 351 | |
| 352 | /** |
| 353 | * The catalog slugs currently blocked. While WordPress.com's data-sharing opt-out |
| 354 | * is on, its list replaces the owner's choices, as it does in robots.txt. |
| 355 | * |
| 356 | * @return string[] |
| 357 | */ |
| 358 | public static function get_blocked_slugs() { |
| 359 | if ( self::has_data_sharing_opt_out() ) { |
| 360 | return self::get_data_sharing_blocked_slugs(); |
| 361 | } |
| 362 | |
| 363 | $catalog = self::get_catalog(); |
| 364 | $blocked = self::sanitize_blocked( get_option( self::OPTION, array() ), $catalog ); |
| 365 | $slugs = array(); |
| 366 | foreach ( $catalog as $slug => $info ) { |
| 367 | $group = $blocked[ $info['type'] ] ?? array(); |
| 368 | if ( 'all' === $group || in_array( $slug, $group, true ) ) { |
| 369 | $slugs[] = $slug; |
| 370 | } |
| 371 | } |
| 372 | return $slugs; |
| 373 | } |
| 374 | |
| 375 | /** |
| 376 | * Whether the site allows search-engine (and therefore AI-crawler) indexing. |
| 377 | * |
| 378 | * When `blog_public` is off, WordPress disallows everything in robots.txt, so |
| 379 | * per-crawler controls are moot — the GEO tab surfaces this instead of showing |
| 380 | * toggles that can't take effect. |
| 381 | * |
| 382 | * @return bool |
| 383 | */ |
| 384 | public static function search_engines_allowed() { |
| 385 | return (int) get_option( 'blog_public', 1 ) === 1; |
| 386 | } |
| 387 | |
| 388 | /** |
| 389 | * Whether the site is served from a WordPress.com staging subdomain |
| 390 | * (`*.wpcomstaging.com`) where the platform blocks all crawling regardless of |
| 391 | * the site's own settings — so AI-crawler controls can't take effect. |
| 392 | * |
| 393 | * @return bool |
| 394 | */ |
| 395 | public static function is_crawl_restricted_subdomain() { |
| 396 | $host = (string) wp_parse_url( home_url(), PHP_URL_HOST ); |
| 397 | $suffix = '.wpcomstaging.com'; |
| 398 | return strlen( $host ) > strlen( $suffix ) |
| 399 | && substr( $host, -strlen( $suffix ) ) === $suffix; |
| 400 | } |
| 401 | |
| 402 | /** |
| 403 | * Whether a static `robots.txt` file exists in the WordPress installation |
| 404 | * directory. These controls only change WordPress's virtual robots.txt and |
| 405 | * cannot modify that file. |
| 406 | * |
| 407 | * @return bool |
| 408 | */ |
| 409 | public static function has_static_robots_txt() { |
| 410 | return Static_Files::exists( 'robots.txt' ); |
| 411 | } |
| 412 | |
| 413 | /** |
| 414 | * Whether this is a path-based multisite network. |
| 415 | * |
| 416 | * Such networks share one origin-level robots.txt, so per-site controls |
| 417 | * cannot be represented safely. |
| 418 | * |
| 419 | * @return bool |
| 420 | */ |
| 421 | public static function is_path_based_multisite() { |
| 422 | return is_multisite() && ! is_subdomain_install(); |
| 423 | } |
| 424 | |
| 425 | /** |
| 426 | * Append `Disallow: /` blocks for each blocked AI crawler to robots.txt. |
| 427 | * |
| 428 | * @param string $output The robots.txt content assembled so far. |
| 429 | * @param bool $public Whether the site is public (`blog_public`). Unused: the |
| 430 | * per-bot blocks apply either way. |
| 431 | * @return string The robots.txt content with AI-crawler directives appended. |
| 432 | */ |
| 433 | public static function append_directives( $output, $public ) { |
| 434 | unset( $public ); |
| 435 | |
| 436 | if ( self::is_path_based_multisite() ) { |
| 437 | return $output; |
| 438 | } |
| 439 | |
| 440 | // WordPress.com's privacy filter writes its own list while the opt-out is on, so |
| 441 | // write only the user agents it leaves out rather than duplicate its directives. |
| 442 | $skip = self::has_data_sharing_opt_out() ? self::get_data_sharing_user_agents() : array(); |
| 443 | $catalog = self::get_catalog(); |
| 444 | $user_agents = array(); |
| 445 | foreach ( self::get_blocked_slugs() as $slug ) { |
| 446 | foreach ( $catalog[ $slug ]['user_agents'] as $user_agent ) { |
| 447 | if ( ! in_array( $user_agent, $skip, true ) ) { |
| 448 | $user_agents[] = $user_agent; |
| 449 | } |
| 450 | } |
| 451 | } |
| 452 | if ( empty( $user_agents ) ) { |
| 453 | return $output; |
| 454 | } |
| 455 | |
| 456 | $lines = array( '', '# AI crawlers blocked via Jetpack SEO.' ); |
| 457 | foreach ( $user_agents as $user_agent ) { |
| 458 | $lines[] = 'User-agent: ' . $user_agent; |
| 459 | $lines[] = 'Disallow: /'; |
| 460 | $lines[] = ''; |
| 461 | } |
| 462 | |
| 463 | return $output . implode( "\n", $lines ) . "\n"; |
| 464 | } |
| 465 | |
| 466 | /** |
| 467 | * The GEO tab's crawler bootstrap payload: the catalog (camelCased for JS), the |
| 468 | * crawlers the owner blocked, the crawlers the data-sharing opt-out blocks, and the |
| 469 | * environment flags that decide whether the controls render or are replaced by an explanation. |
| 470 | * |
| 471 | * @return array |
| 472 | */ |
| 473 | public static function get_bootstrap_data() { |
| 474 | $catalog = array(); |
| 475 | foreach ( self::get_catalog() as $slug => $info ) { |
| 476 | $catalog[] = array( |
| 477 | 'slug' => $slug, |
| 478 | 'label' => $info['label'], |
| 479 | 'userAgents' => $info['user_agents'], |
| 480 | 'type' => $info['type'], |
| 481 | ); |
| 482 | } |
| 483 | |
| 484 | return array( |
| 485 | 'catalog' => $catalog, |
| 486 | 'blocked' => self::get_blocked(), |
| 487 | 'searchEnginesVisible' => self::search_engines_allowed(), |
| 488 | 'restrictedSubdomain' => self::is_crawl_restricted_subdomain(), |
| 489 | 'staticRobotsTxt' => self::has_static_robots_txt(), |
| 490 | 'dataSharingOptOut' => self::has_data_sharing_opt_out(), |
| 491 | 'dataSharingBlocked' => self::get_data_sharing_blocked_slugs(), |
| 492 | 'pathBasedMultisite' => self::is_path_based_multisite(), |
| 493 | 'privacySettingsUrl' => self::privacy_settings_url(), |
| 494 | 'robotsTxtUrl' => home_url( '/robots.txt' ), |
| 495 | ); |
| 496 | } |
| 497 | |
| 498 | /** |
| 499 | * URL of the site visibility settings (private, Coming Soon, "Prevent third-party |
| 500 | * sharing"), linked from the dashboard wherever one of them governs a toggle. |
| 501 | * |
| 502 | * Points at wp-admin → Settings → Reading, which is the same-origin home of those |
| 503 | * options on the sites where they exist (WordPress.com). Left as a package method |
| 504 | * so a Calypso deep link can replace it without touching the client. |
| 505 | * |
| 506 | * @return string |
| 507 | */ |
| 508 | public static function privacy_settings_url() { |
| 509 | return admin_url( 'options-reading.php' ); |
| 510 | } |
| 511 | } |