This repository has been archived on 2024-01-12. You can view files and clone it, but cannot push or open issues or pull requests.
WolKal3000/wolkal3000-worker.php

346 lines
14 KiB
PHP
Raw Permalink Normal View History

2019-03-22 20:48:56 +01:00
<?php
defined( 'ABSPATH' ) or die( 'No script kiddies please!' );
// we may need a http proxy for the fetch. Should be set from the admin page.
// define ('http_proxy', 'http://example.org:8080');
// we'll set this as category => proxy, link => link, active => 0;
// in the admin page, all entries will be displayed with a checkbox for activating, deactivating, deleting.
require_once __DIR__ . "/wolkal3000-geocode.php";
2019-03-22 20:48:56 +01:00
/**
* The worker gets called by the WP scheduler.
2019-03-22 20:48:56 +01:00
*
* @since 0.1.0
*
*/
function wolkal3000_worker() {
2019-03-22 20:48:56 +01:00
/*
* retrieve the proxy from the db, and if it exists, construct a context.
* TODO: USER:PASS in DB.
*
* http://www.pirob.com/2013/06/php-using-getheaders-and-filegetcontents-functions-behind-proxy.html
*/
wolkal3000_error_log (INFO, __FUNCTION__ . " started");
$options = get_option('wolkal3000_options');
$terms = get_terms( array(
'taxonomy' => 'termine_type',
'hide_empty' => false, )
);
foreach($terms as $term){
$unique_id = 'wolkal3000_feed_' . $term->name;
if ( empty ( $options[$unique_id] ) || $options[$unique_id] == '' ) {
wolkal3000_error_log (INFO, "link for event category $term->name is not known; next");
continue;
}
2019-03-29 14:52:44 +01:00
/*
2019-03-29 14:52:44 +01:00
The update and delete logic goes as follows:
- before we start updating, we mark all entries from this event category with recent = false. (i.e. outdated).
- when we scan the ICS feed,
- if an entry with the same UID exists and was not updated in the meantime: do nothing.
- if it was updated in the meantime, we update our entry and set recent to true
- if no such entry exists, we insert a new one and set recent to true
- when we're finished, we delete all entries from this event category which still have recent set to false.
Apparently, they were deleted on the remote end.
*/
// so we look for all published event posts in the calendar event category
2019-03-29 14:52:44 +01:00
$args = array (
'post_type' => 'termine',
'post_status' => 'publish',
'meta_key' => '_wolkal3000_category',
2019-03-29 14:52:44 +01:00
'meta_value' => $term->name,
);
$post_ids = get_posts( $args );
// and set their recent flag to false.
2019-03-29 19:13:14 +01:00
if(is_array($post_ids)) {
foreach( $post_ids as $post_id ) {
$id = $post_id->ID;
update_post_meta( $id, '_wolkal3000_recent', 'false' );
2019-03-29 19:13:14 +01:00
}
}
2019-03-29 14:52:44 +01:00
// now we process the current feed.
$link = WOLKAL_PREFIX . $options[$unique_id] . WOLKAL_SUFFIX;
wolkal3000_error_log (INFO, "importing event category $term->name");
wolkal3000_do_import($term->name, $link);
2019-03-29 14:52:44 +01:00
// look if there are any published event posts in the current event category which were not posted anew or updated (ie recent == false)
$args = array (
'post_type' => 'termine',
'post_status' => 'publish',
'meta_query' => array(
array(
'key' => '_wolkal3000_category',
2019-03-29 14:52:44 +01:00
'value' => $term->name,
),
array(
'key' => '_wolkal3000_recent',
'value' => 'false',
2019-03-29 14:52:44 +01:00
),
)
);
$post_ids = get_posts( $args );
// and trash them.
2019-03-29 19:13:14 +01:00
if(is_array($post_ids)) {
foreach( $post_ids as $post_id ) {
$id = $post_id->ID;
wp_trash_post( $id );
wolkal3000_error_log (INFO, "Event post $id gelöscht.");
2019-03-29 19:13:14 +01:00
}
}
2019-03-22 20:48:56 +01:00
}
wolkal3000_error_log (INFO, __FUNCTION__ . " finished");
2019-03-22 20:48:56 +01:00
}
add_action( 'wolkal3000_worker_hook', 'wolkal3000_worker' );
2019-03-27 18:13:25 +01:00
require_once __DIR__ . '/icalparser/src/IcalParser.php';
require_once __DIR__ . '/icalparser/src/Recurrence.php';
require_once __DIR__ . '/icalparser/src/Freq.php';
require_once __DIR__ . '/icalparser/src/WindowsTimezones.php';
function curl_get_remote($url) {
$ch = curl_init();
curl_setopt($ch, CURLOPT_URL, $url);
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_HEADER, false);
$response = curl_exec($ch);
if ( curl_errno($ch) ) {
// $info = curl_getinfo($ch);
$message = __FUNCTION__ . ": cURL error " . curl_error($ch);
// wolkal3000_error_log (WARN, $message);
curl_close($ch);
throw new \RuntimeException($message);
}
curl_close($ch);
// now we're sure we have a valid response:
return ($response);
}
function wolkal3000_do_import($category, $link) {
2019-04-06 07:45:31 +02:00
$my_latlon = array('', '');
$cal = new \om\IcalParser();
$results = $cal->parseString(curl_get_remote($link));
// TODO: Fehlerbehandlung, wenn der Link kaputt ist. Muss graceful passieren.
// icalparser nutzt intern file_get_contents, und da kommt man nicht ohne weiteres ran. Evtl. ändern auf curl?
// oder abfangen mit @file...
// we must set a current user because we may not be logged in.
$user_id = 1;
$user = get_user_by( 'id', $user_id );
if( $user ) {
wp_set_current_user( $user_id, $user->user_login );
wp_set_auth_cookie( $user_id );
}
foreach ($cal->getSortedEvents() as $r) {
// if DTEND lies in the past, this event has expired. Ignore.
$now = new DateTime();
// $dtend = new DateTime($r['DTEND']);
$summary = $r['SUMMARY'];
$dtstart = $r['DTSTART']->format('d.m.Y H:i');
if ($r['DTEND'] < $now) {
continue;
} else {
wolkal3000_error_log (INFO, "processing $summary on $dtstart");
}
// The zeitstempel. No idea what it's for, but kal3000 seems to use it.
$wpc_from = $r['DTSTART']->format('d.m.Y H:i');
// code borrowed from kal3000_termine_save_postdata which will not be invoked.
$zeitstempel = strftime( strToTime( $wpc_from ) );
if(!$zeitstempel) {
// strftime doesn't seem to work, so let's get creative
preg_match("/([0-9]{1,2}).\s(\w{1,})\s([0-9]{4})\s([0-9]{2}):([0-9]{2})/", $wpc_from, $zeitstempel);
$month_number = "";
for($i=1;$i<=12;$i++){
if(strtolower(date_i18n("F", mktime(0, 0, 0, $i, 1, 0))) == strtolower($zeitstempel[2])){
$month_number = $i;
break;
}
}
$zeit = mktime($zeitstempel[4], $zeitstempel[5], 0, $month_number, $zeitstempel[1], $zeitstempel[3]);
$zeitstempel = date_i18n('U', $zeit);
}
2019-03-22 20:48:56 +01:00
// geocoden
$options = get_option('wolkal3000_options');
if ( $options['wolkal3000_geocoding'] != "off" ) {
2020-12-05 18:47:25 +01:00
$location = urldecode ($r['LOCATION']);
$my_latlon = wolkal3000_geocode($location);
2020-12-05 18:47:25 +01:00
$file = dirname (__FILE__) . "/latlon-$hash.txt";
file_put_contents ($file, var_export ($my_latlon, TRUE));
}
// create a default form
// $post = get_default_post_to_edit ('termine', false);
$post_type = 'termine';
// why can't I simply call get_default_post_to_edit? This gives an undefined function error!
$post = new stdClass;
$post->ID = 0;
$post->post_author = '';
$post->post_date = '';
$post->post_date_gmt = '';
$post->post_password = '';
$post->post_name = '';
$post->post_type = $post_type;
$post->post_status = 'draft';
$post->to_ping = '';
$post->pinged = '';
$post->comment_status = get_default_comment_status( $post_type );
$post->ping_status = get_default_comment_status( $post_type, 'pingback' );
$post->post_pingback = get_option( 'default_pingback_flag' );
$post->post_category = get_option( 'default_category' );
$post->page_template = 'default';
$post->post_parent = 0;
$post->menu_order = 0;
$post = new WP_Post( $post );
2020-06-03 19:01:19 +02:00
if(!isset($post_content)){
$post->post_content = "";
}else{
$post->post_content = apply_filters( 'default_content', $post_content, $post );
2020-06-03 19:01:19 +02:00
}
$post->post_title = apply_filters( 'default_title', $post_title, $post );
$post->post_excerpt = apply_filters( 'default_excerpt', $post_excerpt, $post );
// TODO:
if ( isset($r['ATTACH']) ) {
// create image attachment and associate with new post
$attach = $r['ATTACH'];
$summary = $r['SUMMARY'];
wolkal3000_error_log (INFO, "found attachment $attach for $summary");
}
2019-04-06 07:45:31 +02:00
if ( isset ( $r['CLASS'] ) && 'PRIVATE' == $r['CLASS']) {
$secretevent = true;
} else {
$secretevent = false;
}
// and fill in the post form
$post->post_author = '1';
2020-06-03 19:01:19 +02:00
if(!isset($r['DESCRIPTION'])){
$post->post_content = "Keine Beschreibung" ;
}else{
$post->post_content = $r['DESCRIPTION'];
2020-06-03 19:01:19 +02:00
}
$post->post_title = $r['SUMMARY'];
// create an excerpt for the overview page ([wpcalendar kat=...])
if (strlen ($r['DESCRIPTION']) > 160) {
$post->post_excerpt = substr ($r['DESCRIPTION'], 0, 160) . ' ...'; // first 160 chars of DESCRIPTION plus ' ...'
2020-06-03 19:01:19 +02:00
}else{
if(!isset($r['DESCRIPTION'])){
$post->post_excerpt = "Keine Beschreibung" ;
}else{
$post->post_excerpt = $r['DESCRIPTION'];
2020-06-03 19:01:19 +02:00
}
}
$post->post_status = 'publish';
$post->post_category = array ($category,);
// sanitized title. We will add a timestamp to enable recurring events
// this is not handled properly by wp_insert_post - recurring events would all have the same post_name.
// $post->post_name = $r['DTSTART']->format('Y-m-d-H-i') . '-' . strtolower( urlencode($r['SUMMARY']) ) ;
$post->visibility = 'public';
// now the wpcalendar metas.
$post->meta_input = array(
'_wpcal_from' => $r['DTSTART']->format('d.m.Y H:i'),
'_bis' => $r['DTEND']->format('d.m.Y H:i'),
'_geostadt' => wolkal3000_geocity($r['LOCATION']),
'_geoshow' => wolkal3000_geoshow($r['LOCATION']),
'_lat' => $my_latlon[0],
'_lon' => $my_latlon[1],
2019-04-06 07:45:31 +02:00
'_zoom' => '7',
'_veranstalter' => '',
'_veranstalterlnk' => '',
'_zeitstempel' => $zeitstempel,
'_wolkal3000_uid' => $r['UID'],
'_wolkal3000_recent' => 'true',
// '_wolkal3000_created' => $r['LAST-MODIFIED']->format('U'),
// '_wolkal3000_created' => $r['LAST-MODIFIED']->format('d.m.Y H:i'),
// '_wolkal3000_created' => $r['LAST-MODIFIED']->format('U'),
'_wolkal3000_category' => $category,
2019-04-06 07:45:31 +02:00
'_secretevent' => $secretevent,
);
// so we have the new posts's attributes. Now we need to decide what to do with it.
// first, we try to find a published post with the same UID and zeitstempel. (due to recurring events having the same UID)
// Alas, this will lead to events that were shifted to be trashed and posted anew. We cannot tell shifted events from recurring events.
2019-03-29 14:52:44 +01:00
$args = array (
'post_type' => 'termine',
'post_status' => 'publish',
'meta_query' => array(
array(
'key' => '_wolkal3000_uid',
'value' => $r['UID'],
),
array(
'key' => '_zeitstempel',
'value' => $zeitstempel,
),
)
2019-03-29 14:52:44 +01:00
);
$post_ids = get_posts( $args );
// did we find one? (It should really be only one!)
if ( is_array ( $post_ids ) ) {
if ( empty ( $post_ids ) ) {
// ok, none found, so we insert the new one
$post_id = wp_insert_post( $post );
if ( is_wp_error( $post_id ) ) {
$message = $post_id->get_error_message();
wolkal3000_error_log ( WARN, $message );
} else {
update_post_meta( $post_id, '_edit_last', $user_id );
$now = time();
$lock = "$now:$user_id";
update_post_meta( $post_id, '_edit_lock', $lock );
// and assign the taxonomy type and event category.
wp_set_object_terms( $post_id, $category, 'termine_type' );
wolkal3000_error_log (INFO, "posted new post $post_id");
}
} else {
// good, the post exists already.
$id = $post_ids[0]->ID;
$created = get_post_meta( $id, '_wolkal3000_created', true );
$lastmodified = $r['LAST-MODIFIED']->format('U');
// was it updated on the remote calendar? (was if modified after it was created remotely?)
if ( $lastmodified > $created ) {
// yes, so we update the existing post. We don't care _what_ changed.
$post->ID = $id ;
$post_id = wp_update_post( $post, false );
// and update the _created field
update_post_meta ( $id, '_wolkal3000_created', $lastmodified );
wolkal3000_error_log (INFO, "updated post $post_id");
} elseif ( $lastmodified < $created ) {
// iiiiek! A time reversal or a secret time machine! That should not happen!
wolkal3000_error_log (WARN, "post $id last-modified : created $lastmodified < $created ");
} // else both are equal, and we do nothing except setting recent to true.
update_post_meta ( $id, '_wolkal3000_recent', 'true' );
2019-03-29 14:52:44 +01:00
}
} else {
$file = dirname (__FILE__) . '/get_posts-' . $post->post_name . '.txt';
wolkal3000_error_log (WARN, "hmmm, get_posts() did not return an array. Logging to $file");
file_put_contents ($file, var_export ($post_ids, TRUE));
}
2019-03-29 14:52:44 +01:00
// and on the next entry.
}
2019-03-29 14:52:44 +01:00
// foreach end. we're finished.
}
2019-03-22 20:48:56 +01:00
2020-06-03 19:01:19 +02:00