<?php
/**
* Administration: Community Events class.
*
* @package WordPress
* @subpackage Administration
* @since 4.8.0
*/
/**
* Class WP_Community_Events.
*
* A client for api.wordpress.org/events.
*
* @since 4.8.0
*/
#[AllowDynamicProperties]
class WP_Community_Events {
/**
* ID for a WordPress user account.
*
* @since 4.8.0
*
* @var int
*/
protected $user_id = 0;
/**
* Stores location data for the user.
*
* @since 4.8.0
*
* @var false|array
*/
protected $user_location = false;
/**
* Constructor for WP_Community_Events.
*
* @since 4.8.0
*
* @param int $user_id WP user ID.
* @param false|array $user_location {
* Stored location data for the user. false to pass no location.
*
* @type string $description The name of the location
* @type string $latitude The latitude in decimal degrees notation, without the degree
* symbol. e.g.: 47.615200.
* @type string $longitude The longitude in decimal degrees notation, without the degree
* symbol. e.g.: -122.341100.
* @type string $country The ISO 3166-1 alpha-2 country code. e.g.: BR
* }
*/
public function __construct( $user_id, $user_location = false ) {
$this->user_id = absint( $user_id );
$this->user_location = $user_location;
}
/**
* Gets data about events near a particular location.
*
* Cached events will be immediately returned if the `user_location` property
* is set for the current user, and cached events exist for that location.
*
* Otherwise, this method sends a request to the w.org Events API with location
* data. The API will send back a recognized location based on the data, along
* with nearby events.
*
* The browser's request for events is proxied with this method, rather
* than having the browser make the request directly to api.wordpress.org,
* because it allows results to be cached server-side and shared with other
* users and sites in the network. This makes the process more efficient,
* since increasing the number of visits that get cached data means users
* don't have to wait as often; if the user's browser made the request
* directly, it would also need to make a second request to WP in order to
* pass the data for caching. Having WP make the request also introduces
* the opportunity to anonymize the IP before sending it to w.org, which
* mitigates possible privacy concerns.
*
* @since 4.8.0
* @since 5.5.2 Response no longer contains formatted date field. They're added
* in `wp.communityEvents.populateDynamicEventFields()` now.
*
* @param string $location_search Optional. City name to help determine the location.
* e.g., "Seattle". Default empty string.
* @param string $timezone Optional. Timezone to help determine the location.
* Default empty string.
* @return array|WP_Error A WP_Error on failure; an array with location and events on
* success.
*/
public function get_events( $location_search = '', $timezone = '' ) {
$cached_events = $this->get_cached_events();
if ( ! $location_search && $cached_events ) {
return $cached_events;
}
// Include an unmodified $wp_version.
require ABSPATH . WPINC . '/version.php';
$api_url = 'http://api.wordpress.org/events/1.0/';
$request_args = $this->get_request_args( $location_search, $timezone );
$request_args['user-agent'] = 'WordPress/' . $wp_version . '; ' . home_url( '/' );
if ( wp_http_supports( array( 'ssl' ) ) ) {
$api_url = set_url_scheme( $api_url, 'https' );
}
$response = wp_remote_get( $api_url, $request_args );
$response_code = wp_remote_retrieve_response_code( $response );
$response_body = json_decode( wp_remote_retrieve_body( $response ), true );
$response_error = null;
if ( is_wp_error( $response ) ) {
$response_error = $response;
} elseif ( 200 !== $response_code ) {
$response_error = new WP_Error(
'api-error',
/* translators: %d: Numeric HTTP status code, e.g. 400, 403, 500, 504, etc. */
sprintf( __( 'Invalid API response code (%d).' ), $response_code )
);
} elseif ( ! isset( $response_body['location'], $response_body['events'] ) ) {
$response_error = new WP_Error(
'api-invalid-response',
isset( $response_body['error'] ) ? $response_body['error'] : __( 'Unknown API error.' )
);
}
if ( is_wp_error( $response_error ) ) {
return $response_error;
} else {
$expiration = false;
if ( isset( $response_body['ttl'] ) ) {
$expiration = $response_body['ttl'];
unset( $response_body['ttl'] );
}
/*
* The IP in the response is usually the same as the one that was sent
* in the request, but in some cases it is different. In those cases,
* it's important to reset it back to the IP from the request.
*
* For example, if the IP sent in the request is private (e.g., 192.168.1.100),
* then the API will ignore that and use the corresponding public IP instead,
* and the public IP will get returned. If the public IP were saved, though,
* then get_cached_events() would always return `false`, because the transient
* would be generated based on the public IP when saving the cache, but generated
* based on the private IP when retrieving the cache.
*/
if ( ! empty( $response_body['location']['ip'] ) ) {
$response_body['location']['ip'] = $request_args['body']['ip'];
}
/*
* The API doesn't return a description for latitude/longitude requests,
* but the description is already saved in the user location, so that
* one can be used instead.
*/
if ( $this->coordinates_match( $request_args['body'], $response_body['location'] ) && empty( $response_body['location']['description'] ) ) {
$response_body['location']['description'] = $this->user_location['description'];
}
/*
* Store the raw response, because events will expire before the cache does.
* The response will need to be processed every page load.
*/
$this->cache_events( $response_body, $expiration );
$response_body['events'] = $this->trim_events( $response_body['events'] );
return $response_body;
}
}
/**
* Builds an array of args to use in an HTTP request to the w.org Events API.
*
* @since 4.8.0
*
* @param string $search Optional. City search string. Default empty string.
* @param string $timezone Optional. Timezone string. Default empty string.
* @return array The request args.
*/
protected function get_request_args( $search = '', $timezone = '' ) {
$args = array(
'number' => 5, // Get more than three in case some get trimmed out.
'ip' => self::get_unsafe_client_ip(),
);
/*
* Include the minimal set of necessary arguments, in order to increase the
* chances of a cache-hit on the API side.
*/
if ( empty( $search ) && isset( $this->user_location['latitude'], $this->user_location['longitude'] ) ) {
$args['latitude'] = $this->user_location['latitude'];
$args['longitude'] = $this->user_location['longitude'];
} else {
$args['locale'] = get_user_locale( $this->user_id );
if ( $timezone ) {
$args['timezone'] = $timezone;
}
if ( $search ) {
$args['location'] = $search;
}
}
// Wrap the args in an array compatible with the second parameter of `wp_remote_get()`.
return array(
'body' => $args,
);
}
/**
* Determines the user's actual IP address and attempts to partially
* anonymize an IP address by converting it to a network ID.
*
* Geolocating the network ID usually returns a similar location as the
* actual IP, but provides some privacy for the user.
*
* $_SERVER['REMOTE_ADDR'] cannot be used in all cases, such as when the user
* is making their request through a proxy, or when the web server is behind
* a proxy. In those cases, $_SERVER['REMOTE_ADDR'] is set to the proxy address rather
* than the user's actual address.
*
* Modified from https://stackoverflow.com/a/2031935/450127, MIT license.
* Modified from https://github.com/geertw/php-ip-anonymizer, MIT license.
*
* SECURITY WARNING: This function is _NOT_ intended to be used in
* circumstances where the authenticity of the IP address matters. This does
* _NOT_ guarantee that the returned address is valid or accurate, and it can
* be easily spoofed.
*
* @since 4.8.0
*
* @return string|false The anonymized address on success; the given address
* or false on failure.
*/
public static function get_unsafe_client_ip() {
$client_ip = false;
// In order of preference, with the best ones for this purpose first.
$address_headers = array(
'HTTP_CLIENT_IP',
'HTTP_X_FORWARDED_FOR',
'HTTP_X_FORWARDED',
'HTTP_X_CLUSTER_CLIENT_IP',
'HTTP_FORWARDED_FOR',
'HTTP_FORWARDED',
'REMOTE_ADDR',
);
foreach ( $address_headers as $header ) {
if ( array_key_exists( $header, $_SERVER ) ) {
/*
* HTTP_X_FORWARDED_FOR can contain a chain of comma-separated
* addresses. The first one is the original client. It can't be
* trusted for authenticity, but we don't need to for this purpose.
*/
$address_chain = explode( ',', $_SERVER[ $header ] );
$client_ip = trim( $address_chain[0] );
break;
}
}
if ( ! $client_ip ) {
return false;
}
$anon_ip = wp_privacy_anonymize_ip( $client_ip, true );
if ( '0.0.0.0' === $anon_ip || '::' === $anon_ip ) {
return false;
}
return $anon_ip;
}
/**
* Test if two pairs of latitude/longitude coordinates match each other.
*
* @since 4.8.0
*
* @param array $a The first pair, with indexes 'latitude' and 'longitude'.
* @param array $b The second pair, with indexes 'latitude' and 'longitude'.
* @return bool True if they match, false if they don't.
*/
protected function coordinates_match( $a, $b ) {
if ( ! isset( $a['latitude'], $a['longitude'], $b['latitude'], $b['longitude'] ) ) {
return false;
}
return $a['latitude'] === $b['latitude'] && $a['longitude'] === $b['longitude'];
}
/**
* Generates a transient key based on user location.
*
* This could be reduced to a one-liner in the calling functions, but it's
* intentionally a separate function because it's called from multiple
* functions, and having it abstracted keeps the logic consistent and DRY,
* which is less prone to errors.
*
* @since 4.8.0
*
* @param array $location Should contain 'latitude' and 'longitude' indexes.
* @return string|false Transient key on success, false on failure.
*/
protected function get_events_transient_key( $location ) {
$key = false;
if ( isset( $location['ip'] ) ) {
$key = 'community-events-' . md5( $location['ip'] );
} elseif ( isset( $location['latitude'], $location['longitude'] ) ) {
$key = 'community-events-' . md5( $location['latitude'] . $location['longitude'] );
}
return $key;
}
/**
* Caches an array of events data from the Events API.
*
* @since 4.8.0
*
* @param array $events Response body from the API request.
* @param int|false $expiration Optional. Amount of time to cache the events. Defaults to false.
* @return bool true if events were cached; false if not.
*/
protected function cache_events( $events, $expiration = false ) {
$set = false;
$transient_key = $this->get_events_transient_key( $events['location'] );
$cache_expiration = $expiration ? absint( $expiration ) : HOUR_IN_SECONDS * 12;
if ( $transient_key ) {
$set = set_site_transient( $transient_key, $events, $cache_expiration );
}
return $set;
}
/**
* Gets cached events.
*
* @since 4.8.0
* @since 5.5.2 Response no longer contains formatted date field. They're added
* in `wp.communityEvents.populateDynamicEventFields()` now.
*
* @return array|false An array containing `location` and `events` items
* on success, false on failure.
*/
public function get_cached_events() {
$transient_key = $this->get_events_transient_key( $this->user_location );
if ( ! $transient_key ) {
return false;
}
$cached_response = get_site_transient( $transient_key );
if ( isset( $cached_response['events'] ) ) {
$cached_response['events'] = $this->trim_events( $cached_response['events'] );
}
return $cached_response;
}
/**
* Adds formatted date and time items for each event in an API response.
*
* This has to be called after the data is pulled from the cache, because
* the cached events are shared by all users. If it was called before storing
* the cache, then all users would see the events in the localized data/time
* of the user who triggered the cache refresh, rather than their own.
*
* @since 4.8.0
* @deprecated 5.6.0 No longer used in core.
*
* @param array $response_body The response which contains the events.
* @return array The response with dates and times formatted.
*/
protected function format_event_data_time( $response_body ) {
_deprecated_function(
__METHOD__,
'5.5.2',
'This is no longer used by core, and only kept for backward compatibility.'
);
if ( isset( $response_body['events'] ) ) {
foreach ( $response_body['events'] as $key => $event ) {
$timestamp = strtotime( $event['date'] );
/*
* The `date_format` option is not used because it's important
* in this context to keep the day of the week in the formatted date,
* so that users can tell at a glance if the event is on a day they
* are available, without having to open the link.
*/
/* translators: Date format for upcoming events on the dashboard. Include the day of the week. See https://www.php.net/manual/datetime.format.php */
$formatted_date = date_i18n( __( 'l, M j, Y' ), $timestamp );
$formatted_time = date_i18n( get_option( 'time_format' ), $timestamp );
if ( isset( $event['end_date'] ) ) {
$end_timestamp = strtotime( $event['end_date'] );
$formatted_end_date = date_i18n( __( 'l, M j, Y' ), $end_timestamp );
if ( 'meetup' !== $event['type'] && $formatted_end_date !== $formatted_date ) {
/* translators: Upcoming events month format. See https://www.php.net/manual/datetime.format.php */
$start_month = date_i18n( _x( 'F', 'upcoming events month format' ), $timestamp );
$end_month = date_i18n( _x( 'F', 'upcoming events month format' ), $end_timestamp );
if ( $start_month === $end_month ) {
$formatted_date = sprintf(
/* translators: Date string for upcoming events. 1: Month, 2: Starting day, 3: Ending day, 4: Year. */
__( '%1$s %2$d–%3$d, %4$d' ),
$start_month,
/* translators: Upcoming events day format. See https://www.php.net/manual/datetime.format.php */
date_i18n( _x( 'j', 'upcoming events day format' ), $timestamp ),
date_i18n( _x( 'j', 'upcoming events day format' ), $end_timestamp ),
/* translators: Upcoming events year format. See https://www.php.net/manual/datetime.format.php */
date_i18n( _x( 'Y', 'upcoming events year format' ), $timestamp )
);
} else {
$formatted_date = sprintf(
/* translators: Date string for upcoming events. 1: Starting month, 2: Starting day, 3: Ending month, 4: Ending day, 5: Year. */
__( '%1$s %2$d – %3$s %4$d, %5$d' ),
$start_month,
date_i18n( _x( 'j', 'upcoming events day format' ), $timestamp ),
$end_month,
date_i18n( _x( 'j', 'upcoming events day format' ), $end_timestamp ),
date_i18n( _x( 'Y', 'upcoming events year format' ), $timestamp )
);
}
$formatted_date = wp_maybe_decline_date( $formatted_date, 'F j, Y' );
}
}
$response_body['events'][ $key ]['formatted_date'] = $formatted_date;
$response_body['events'][ $key ]['formatted_time'] = $formatted_time;
}
}
return $response_body;
}
/**
* Prepares the event list for presentation.
*
* Discards expired events, and makes WordCamps "sticky." Attendees need more
* advanced notice about WordCamps than they do for meetups, so camps should
* appear in the list sooner. If a WordCamp is coming up, the API will "stick"
* it in the response, even if it wouldn't otherwise appear. When that happens,
* the event will be at the end of the list, and will need to be moved into a
* higher position, so that it doesn't get trimmed off.
*
* @since 4.8.0
* @since 4.9.7 Stick a WordCamp to the final list.
* @since 5.5.2 Accepts and returns only the events, rather than an entire HTTP response.
* @since 6.0.0 Decode HTML entities from the event title.
*
* @param array $events The events that will be prepared.
* @return array The response body with events trimmed.
*/
protected function trim_events( array $events ) {
$future_events = array();
foreach ( $events as $event ) {
/*
* The API's `date` and `end_date` fields are in the _event's_ local timezone, but UTC is needed so
* it can be converted to the _user's_ local time.
*/
$end_time = (int) $event['end_unix_timestamp'];
if ( time() < $end_time ) {
// Decode HTML entities from the event title.
$event['title'] = html_entity_decode( $event['title'], ENT_QUOTES, 'UTF-8' );
array_push( $future_events, $event );
}
}
$future_wordcamps = array_filter(
$future_events,
static function ( $wordcamp ) {
return 'wordcamp' === $wordcamp['type'];
}
);
$future_wordcamps = array_values( $future_wordcamps ); // Remove gaps in indices.
$trimmed_events = array_slice( $future_events, 0, 3 );
$trimmed_event_types = wp_list_pluck( $trimmed_events, 'type' );
// Make sure the soonest upcoming WordCamp is pinned in the list.
if ( $future_wordcamps && ! in_array( 'wordcamp', $trimmed_event_types, true ) ) {
array_pop( $trimmed_events );
array_push( $trimmed_events, $future_wordcamps[0] );
}
return $trimmed_events;
}
/**
* Logs responses to Events API requests.
*
* @since 4.8.0
* @deprecated 4.9.0 Use a plugin instead. See #41217 for an example.
*
* @param string $message A description of what occurred.
* @param array $details Details that provide more context for the
* log entry.
*/
protected function maybe_log_events_response( $message, $details ) {
_deprecated_function( __METHOD__, '4.9.0' );
if ( ! WP_DEBUG_LOG ) {
return;
}
error_log(
sprintf(
'%s: %s. Details: %s',
__METHOD__,
trim( $message, '.' ),
wp_json_encode( $details )
)
);
}
}
ظاهرة تثير جدل بدأ يلاحظها مستخدمي تطبيق “شات جي بي تي” وهي أن التطبيق ينادي المستخدمين بأسمائهم الاولى اثناء المحادثات وتلقي المعلومات منه رغم انهم لم يخبروه بها مسبقا، نظرا لما قالته احد لمستخدمين لتطبيق الذكاء الاصطناعي “شات جي بي تي”، انها كانت تسأله عن معلومات بشكل طبيعي دون ان تذكر أي أسماء.
فوجئت بأن التطبيق رد عليها مستخدما اسمها واسم صديقتها، رغم أنها لم تصرح عنهما مطلقا في اي محادثه سابقه من محادثاتها. هذا الموقف أثار جدل حول مدي قدرة الذكاء الاصطناعي على الوصول إلى معلومات شخصية وخفيه، مما يثير مخاوف تعلق بالخصوصية واستخدام الذاكرة.
إن صنع نظام آلي قادر على تكوين ذاكرة اجتماعية عن البشر لم يعد مجرد تحد تقني… بل أصبح واقعًا مخيفًا، لأن ليس محور الكلام هنا عن مجرد جهاز بيخزن بيانات، نحن نتحدث عن آلة بتراقب، وبتحلل، وبتتعلم بصمت؛ الأهم من ذلك هو حقيقة أنه عندما نتحدث الي ذلك النظام ويرد على المستخدم بردود غير متوقعه وكأنه يعرفه ويتتبعه. هذه عملية لم يتمكن المستخدمين من تفسيرها.
تمييز الذكاء الاصطناعي
بما يتعلق بنظام الذكاء الاصطناعي أوضح هشام عبدالعال أستاذ في مجال التكنولوجيا، أن الذكاء الاصطناعي لم يعد مجرد أداة تقنية تقتصر على تحليل البيانات وتقديم ردود بشكل الي ولكنه أصبح مع مرور الوقت لديه قدرة على تكوين ذاكرة اجتماعية عن المستخدمين وذلك عن طريق تحليل سلوكهم واختياراتهم واسلوب حديثهم.
مضيفاً، أن نظام الذكاء الاصطناعي الذي يوجد داخل التطبيق لا يصنف الخلفيات الاجتماعية بشكل مباشر، لكنه يتعامل حسب اللغة المستخدمة عامية أو فصحى، لكن التطبيق لا يميز بين طالب أو ربة منزل أو حتى عامل بشكل مباشر، بل يتعامل مع نوعية الأسئلة والأسلوب المقدمة له.
وتابع عبدالعال، لكن في حاله وجود أكثر من شخص يستخدم نفس الجهاز أو الحساب الردود ستكون متأثرة بالسياق السابق، أو ممكن ذلك يعطي إحساس أن الآلة متذكره أو بتتفاعل بشكل مختلف، لكن بمجرد إنهاء المحادثة أو حذفها، فسيتم غلط المحادثة.
هل تتذكر الآلة؟
اتفق معه في الرأي الأستاذ عمر أشرف، أستاذ في التكنولوجيا، مضيفاً أنه رسميًا، الذكاء الاصطناعي مثل تطبيق شآت جي بي تي لا يحتفظ بتفاصيل شخصية عن المستخدم إلا إذا المستخدم فعل خاصية الذاكرة بشكل واضح.
لكن هناك مستخدمين لاحظوا أن النموذج يناديهم بأسمائهم بدون ما يقدموها بشكل مباشر، وذلك فتح باب للشك حول وجود نوع من التتبع الخفي أو آلية لتجميع الأنماط السلوكية من وراء الكواليس.
بيانات تجمع بصمت
وفيما يتعلق بالأمن السيبراني يري الطرفان، انه يوجد تحديات تواجه الأمن السيبراني في ذلك الموضوع فتكوين هذه الذاكرة يكون عن طريق جمع بيانات شخصية وسلوكية دقيقة للغاية وغالبًا ما يتم ذلك دون إدراك المستخدم الكامل كما تشمل هذه البيانات أنماط الكتابة، نوعية الأسئلة بل وقد تكشف ما بين السطور عن خلفيات نفسية أو اجتماعية
ضافوا، أنه على الرغم من أن بعض الشركات تبرر استخدام البيانات وذلك لتحسين تجربة المستخدم إلا إن غياب الشفافية يفتح الباب للتتبع غير المباشر للهوية.
وفي حالة استخدام الذكاء الاصطناعي بدون رقابة صارمة أو قوانين واضحة من الممكن ان تتحول هذه الذاكرة إلى ملف شخصي خفي وذلك قابل لاستغلاله في اغراض تجارية أو سياسية أو أمنيه.
الذكاء الاصطناعي لم يعد مجرد أداة، بل أصبح قادر على تكوين صورة آلية عن المستخدم دون علمه، وبينما يعد هذا تطور تقني كبير، فإن غياب الرقابة يثير مخاوف حقيقية حول الخصوصية واستخدام البيانات الشخصية، مما يتطلب ضوابط صارمة تحمي المستخدم قبل أن تتحول التقنية إلى مخاطر خفيه.
