You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 
 

676 lines
22 KiB

<?php
// This file is part of Moodle - http://moodle.org/
//
// Moodle is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// Moodle is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU General Public License for more details.
//
// You should have received a copy of the GNU General Public License
// along with Moodle. If not, see <http://www.gnu.org/licenses/>.
/**
* Document representation.
*
* @package core_search
* @copyright 2015 David Monllao {@link http://www.davidmonllao.com}
* @license http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
*/
namespace core_search;
defined('MOODLE_INTERNAL') || die();
/**
* Represents a document to index.
*
* Note that, if you are writting a search engine and you want to change \core_search\document
* behaviour, you can overwrite this class, will be automatically loaded from \search_YOURENGINE\document.
*
* @package core_search
* @copyright 2015 David Monllao {@link http://www.davidmonllao.com}
* @license http://www.gnu.org/copyleft/gpl.html GNU GPL v3 or later
*/
class document implements \renderable, \templatable {
/**
* @var array $data The document data.
*/
protected $data = array();
/**
* @var array Extra data needed to render the document.
*/
protected $extradata = array();
/**
* @var \moodle_url Link to the document.
*/
protected $docurl = null;
/**
* @var \moodle_url Link to the document context.
*/
protected $contexturl = null;
/**
* @var \core_search\document_icon Document icon instance.
*/
protected $docicon = null;
/**
* @var int|null The content field filearea.
*/
protected $contentfilearea = null;
/**
* @var int|null The content field itemid.
*/
protected $contentitemid = null;
/**
* @var bool Should be set to true if document hasn't been indexed before. False if unknown.
*/
protected $isnew = false;
/**
* @var \stored_file[] An array of stored files to attach to the document.
*/
protected $files = array();
/**
* Change list (for engine implementers):
* 2017091700 - add optional field groupid
*
* @var int Schema version number (update if any change)
*/
const SCHEMA_VERSION = 2017091700;
/**
* All required fields any doc should contain.
*
* We have to choose a format to specify field types, using solr format as we have to choose one and solr is the
* default search engine.
*
* Search engine plugins are responsible of setting their appropriate field types and map these naming to whatever format
* they need.
*
* @var array
*/
protected static $requiredfields = array(
'id' => array(
'type' => 'string',
'stored' => true,
'indexed' => false
),
'itemid' => array(
'type' => 'int',
'stored' => true,
'indexed' => true
),
'title' => array(
'type' => 'text',
'stored' => true,
'indexed' => true,
'mainquery' => true
),
'content' => array(
'type' => 'text',
'stored' => true,
'indexed' => true,
'mainquery' => true
),
'contextid' => array(
'type' => 'int',
'stored' => true,
'indexed' => true
),
'areaid' => array(
'type' => 'string',
'stored' => true,
'indexed' => true
),
'type' => array(
'type' => 'int',
'stored' => true,
'indexed' => true
),
'courseid' => array(
'type' => 'int',
'stored' => true,
'indexed' => true
),
'owneruserid' => array(
'type' => 'int',
'stored' => true,
'indexed' => true
),
'modified' => array(
'type' => 'tdate',
'stored' => true,
'indexed' => true
),
);
/**
* All optional fields docs can contain.
*
* Although it matches solr fields format, this is just to define the field types. Search
* engine plugins are responsible of setting their appropriate field types and map these
* naming to whatever format they need.
*
* @var array
*/
protected static $optionalfields = array(
'userid' => array(
'type' => 'int',
'stored' => true,
'indexed' => true
),
'groupid' => array(
'type' => 'int',
'stored' => true,
'indexed' => true
),
'description1' => array(
'type' => 'text',
'stored' => true,
'indexed' => true,
'mainquery' => true
),
'description2' => array(
'type' => 'text',
'stored' => true,
'indexed' => true,
'mainquery' => true
)
);
/**
* Any fields that are engine specifc. These are fields that are solely used by a search engine plugin
* for internal purposes.
*
* Field names should be prefixed with engine name to avoid potential conflict with core fields.
*
* Uses same format as fields above.
*
* @var array
*/
protected static $enginefields = array();
/**
* We ensure that the document has a unique id across search areas.
*
* @param int $itemid An id unique to the search area
* @param string $componentname The search area component Frankenstyle name
* @param string $areaname The area name (the search area class name)
* @return void
*/
public function __construct($itemid, $componentname, $areaname) {
if (!is_numeric($itemid)) {
throw new \coding_exception('The itemid should be an integer');
}
$this->data['areaid'] = \core_search\manager::generate_areaid($componentname, $areaname);
$this->data['id'] = $this->data['areaid'] . '-' . $itemid;
$this->data['itemid'] = intval($itemid);
}
/**
* Add a stored file to the document.
*
* @param \stored_file|int $file The file to add, or file id.
* @return void
*/
public function add_stored_file($file) {
if (is_numeric($file)) {
$this->files[$file] = $file;
} else {
$this->files[$file->get_id()] = $file;
}
}
/**
* Returns the array of attached files.
*
* @return \stored_file[]
*/
public function get_files() {
// The files array can contain stored file ids, so we need to get instances if asked.
foreach ($this->files as $id => $listfile) {
if (is_numeric($listfile)) {
$fs = get_file_storage();
if ($file = $fs->get_file_by_id($id)) {
$this->files[$id] = $file;
} else {
unset($this->files[$id]); // Index is out of date and referencing a file that does not exist.
}
}
}
return $this->files;
}
/**
* Setter.
*
* Basic checkings to prevent common issues.
*
* If the field is a string tags will be stripped, if it is an integer or a date it
* will be casted to a PHP integer. tdate fields values are expected to be timestamps.
*
* @throws \coding_exception
* @param string $fieldname The field name
* @param string|int $value The value to store
* @return string|int The stored value
*/
public function set($fieldname, $value) {
if (!empty(static::$requiredfields[$fieldname])) {
$fielddata = static::$requiredfields[$fieldname];
} else if (!empty(static::$optionalfields[$fieldname])) {
$fielddata = static::$optionalfields[$fieldname];
} else if (!empty(static::$enginefields[$fieldname])) {
$fielddata = static::$enginefields[$fieldname];
}
if (empty($fielddata)) {
throw new \coding_exception('"' . $fieldname . '" field does not exist.');
}
// tdate fields should be set as timestamps, later they might be converted to
// a date format, it depends on the search engine.
if (($fielddata['type'] === 'int' || $fielddata['type'] === 'tdate') && !is_numeric($value)) {
throw new \coding_exception('"' . $fieldname . '" value should be an integer and its value is "' . $value . '"');
}
// We want to be strict here, there might be engines that expect us to
// provide them data with the proper type already set.
if ($fielddata['type'] === 'int' || $fielddata['type'] === 'tdate') {
$this->data[$fieldname] = intval($value);
} else {
// Remove disallowed Unicode characters.
$value = \core_text::remove_unicode_non_characters($value);
// Replace all groups of line breaks and spaces by single spaces.
$this->data[$fieldname] = preg_replace("/\s+/u", " ", $value);
if ($this->data[$fieldname] === null) {
if (isset($this->data['id'])) {
$docid = $this->data['id'];
} else {
$docid = '(unknown)';
}
throw new \moodle_exception('error_indexing', 'search', '', null, '"' . $fieldname .
'" value causes preg_replace error (may be caused by unusual characters) ' .
'in document with id "' . $docid . '"');
}
}
return $this->data[$fieldname];
}
/**
* Sets data to this->extradata
*
* This data can be retrieved using \core_search\document->get($fieldname).
*
* @param string $fieldname
* @param string $value
* @return void
*/
public function set_extra($fieldname, $value) {
$this->extradata[$fieldname] = $value;
}
/**
* Getter.
*
* Use self::is_set if you are not sure if this field is set or not
* as otherwise it will trigger a \coding_exception
*
* @throws \coding_exception
* @param string $field
* @return string|int
*/
public function get($field) {
if (isset($this->data[$field])) {
return $this->data[$field];
}
// Fallback to extra data.
if (isset($this->extradata[$field])) {
return $this->extradata[$field];
}
throw new \coding_exception('Field "' . $field . '" is not set in the document');
}
/**
* Checks if a field is set.
*
* @param string $field
* @return bool
*/
public function is_set($field) {
return (isset($this->data[$field]) || isset($this->extradata[$field]));
}
/**
* Set if this is a new document. False if unknown.
*
* @param bool $new
*/
public function set_is_new($new) {
$this->isnew = (bool)$new;
}
/**
* Returns if the document is new. False if unknown.
*
* @return bool
*/
public function get_is_new() {
return $this->isnew;
}
/**
* Returns all default fields definitions.
*
* @return array
*/
public static function get_default_fields_definition() {
return static::$requiredfields + static::$optionalfields + static::$enginefields;
}
/**
* Formats the timestamp preparing the time fields to be inserted into the search engine.
*
* By default it just returns a timestamp so any search engine could just store integers
* and use integers comparison to get documents between x and y timestamps, but search
* engines might be interested in using their own field formats. They can do it extending
* this class in \search_xxx\document.
*
* @param int $timestamp
* @return string
*/
public static function format_time_for_engine($timestamp) {
return $timestamp;
}
/**
* Formats a string value for the search engine.
*
* Search engines may overwrite this method to apply restrictions, like limiting the size.
* The default behaviour is just returning the string.
*
* @param string $string
* @return string
*/
public static function format_string_for_engine($string) {
return $string;
}
/**
* Formats a text value for the search engine.
*
* Search engines may overwrite this method to apply restrictions, like limiting the size.
* The default behaviour is just returning the string.
*
* @param string $text
* @return string
*/
public static function format_text_for_engine($text) {
return $text;
}
/**
* Returns a timestamp from the value stored in the search engine.
*
* By default it just returns a timestamp so any search engine could just store integers
* and use integers comparison to get documents between x and y timestamps, but search
* engines might be interested in using their own field formats. They should do it extending
* this class in \search_xxx\document.
*
* @param string $time
* @return int
*/
public static function import_time_from_engine($time) {
return $time;
}
/**
* Returns how text is returned from the search engine.
*
* @return int
*/
protected function get_text_format() {
return FORMAT_PLAIN;
}
/**
* Fills the document with data coming from the search engine.
*
* @throws \core_search\engine_exception
* @param array $docdata
* @return void
*/
public function set_data_from_engine($docdata) {
$fields = static::$requiredfields + static::$optionalfields + static::$enginefields;
foreach ($fields as $fieldname => $field) {
// Optional params might not be there.
if (isset($docdata[$fieldname])) {
if ($field['type'] === 'tdate') {
// Time fields may need a preprocessing.
$this->set($fieldname, static::import_time_from_engine($docdata[$fieldname]));
} else {
// No way we can make this work if there is any multivalue field.
if (is_array($docdata[$fieldname])) {
throw new \core_search\engine_exception('multivaluedfield', 'search_solr', '', $fieldname);
}
$this->set($fieldname, $docdata[$fieldname]);
}
}
}
}
/**
* Sets the document url.
*
* @param \moodle_url $url
* @return void
*/
public function set_doc_url(\moodle_url $url) {
$this->docurl = $url;
}
/**
* Gets the url to the doc.
*
* @return \moodle_url
*/
public function get_doc_url() {
return $this->docurl;
}
/**
* Sets document icon instance.
*
* @param \core_search\document_icon $docicon
*/
public function set_doc_icon(document_icon $docicon) {
$this->docicon = $docicon;
}
/**
* Gets document icon instance.
*
* @return \core_search\document_icon
*/
public function get_doc_icon() {
return $this->docicon;
}
public function set_context_url(\moodle_url $url) {
$this->contexturl = $url;
}
/**
* Gets the url to the context.
*
* @return \moodle_url
*/
public function get_context_url() {
return $this->contexturl;
}
/**
* Returns the document ready to submit to the search engine.
*
* @throws \coding_exception
* @return array
*/
public function export_for_engine() {
// Set any unset defaults.
$this->apply_defaults();
// We don't want to affect the document instance.
$data = $this->data;
// Apply specific engine-dependant formats and restrictions.
foreach (static::$requiredfields as $fieldname => $field) {
// We also check that we have everything we need.
if (!isset($data[$fieldname])) {
throw new \coding_exception('Missing "' . $fieldname . '" field in document with id "' . $this->data['id'] . '"');
}
if ($field['type'] === 'tdate') {
// Overwrite the timestamp with the engine dependant format.
$data[$fieldname] = static::format_time_for_engine($data[$fieldname]);
} else if ($field['type'] === 'string') {
// Overwrite the string with the engine dependant format.
$data[$fieldname] = static::format_string_for_engine($data[$fieldname]);
} else if ($field['type'] === 'text') {
// Overwrite the text with the engine dependant format.
$data[$fieldname] = static::format_text_for_engine($data[$fieldname]);
}
}
$fields = static::$optionalfields + static::$enginefields;
foreach ($fields as $fieldname => $field) {
if (!isset($data[$fieldname])) {
continue;
}
if ($field['type'] === 'tdate') {
// Overwrite the timestamp with the engine dependant format.
$data[$fieldname] = static::format_time_for_engine($data[$fieldname]);
} else if ($field['type'] === 'string') {
// Overwrite the string with the engine dependant format.
$data[$fieldname] = static::format_string_for_engine($data[$fieldname]);
} else if ($field['type'] === 'text') {
// Overwrite the text with the engine dependant format.
$data[$fieldname] = static::format_text_for_engine($data[$fieldname]);
}
}
return $data;
}
/**
* Apply any defaults to unset fields before export. Called after document building, but before export.
*
* Sub-classes of this should make sure to call parent::apply_defaults().
*/
protected function apply_defaults() {
// Set the default type, TYPE_TEXT.
if (!isset($this->data['type'])) {
$this->data['type'] = manager::TYPE_TEXT;
}
}
/**
* Export the document data to be used as a template context.
*
* Adding more info than the required one as people might be interested in extending the template.
*
* Although content is a required field when setting up the document, it accepts '' (empty) values
* as they may be the result of striping out HTML.
*
* SECURITY NOTE: It is the responsibility of the document to properly escape any text to be displayed.
* The renderer will output the content without any further cleaning.
*
* @param renderer_base $output The renderer.
* @return array
*/
public function export_for_template(\renderer_base $output) {
list($componentname, $areaname) = \core_search\manager::extract_areaid_parts($this->get('areaid'));
$searcharea = \core_search\manager::get_search_area($this->data['areaid']);
$title = $this->is_set('title') ? $this->format_text($searcharea->get_document_display_title($this)) : '';
$data = [
'componentname' => $componentname,
'areaname' => $areaname,
'courseurl' => course_get_url($this->get('courseid')),
'coursefullname' => format_string($this->get('coursefullname'), true, array('context' => $this->get('contextid'))),
'modified' => userdate($this->get('modified')),
'title' => ($title !== '') ? $title : get_string('notitle', 'search'),
'docurl' => $this->get_doc_url(),
'content' => $this->is_set('content') ? $this->format_text($this->get('content')) : null,
'contexturl' => $this->get_context_url(),
'description1' => $this->is_set('description1') ? $this->format_text($this->get('description1')) : null,
'description2' => $this->is_set('description2') ? $this->format_text($this->get('description2')) : null,
];
// Now take any attached any files.
$files = $this->get_files();
if (!empty($files)) {
if (count($files) > 1) {
$filenames = array();
foreach ($files as $file) {
$filenames[] = format_string($file->get_filename(), true, array('context' => $this->get('contextid')));
}
$data['multiplefiles'] = true;
$data['filenames'] = $filenames;
} else {
$file = reset($files);
$data['filename'] = format_string($file->get_filename(), true, array('context' => $this->get('contextid')));
}
}
if ($this->is_set('userid')) {
$data['userurl'] = new \moodle_url('/user/view.php', array('id' => $this->get('userid'), 'course' => $this->get('courseid')));
$data['userfullname'] = format_string($this->get('userfullname'), true, array('context' => $this->get('contextid')));
}
if ($docicon = $this->get_doc_icon()) {
$data['icon'] = $output->image_url($docicon->get_name(), $docicon->get_component());
}
return $data;
}
/**
* Formats a text string coming from the search engine.
*
* By default just return the text as it is:
* - Search areas are responsible of sending just plain data, the search engine may
* append HTML or markdown to it (highlighing for example).
* - The view is responsible of shortening the text if it is too big
*
* @param string $text Text to format
* @return string HTML text to be renderer
*/
protected function format_text($text) {
return format_text($text, $this->get_text_format(), array('context' => $this->get('contextid')));
}
}