Indexing

Tab for creating and managing the search index.

Purpose

Building the index is resource-heavy, so it runs in batches via Ajax. On this tab you can create an index, watch progress, and clear it if needed.

Interface

Statistics

Metric	Description
Total resources	Total resources in the system
Indexed	Resources in the index
Words in index	Total word count

Controls

Control	Description
Batch size	Resources per request (1–500)
Create index	Start indexing
Clear index	Remove all index entries

Progress bar

Shows indexing progress in percent.

How indexing works

1. Data extraction — resource fields are loaded from DB per mse_index_fields
2. Splitting — text split by regex from mse_index_split_words
3. Filtering — words shorter than mse_index_min_word_length are dropped
4. Morphology — base forms generated via phpMorphy for each word
5. Save — words and weights written to index table

System settings

Setting	Default	Description
mse_index_fields	pagetitle:3,...	Indexed fields with weights
mse_index_min_word_length	3	Minimum word length
mse_index_split_words	#\s|[,.:;!?...]#u	Regex for splitting
mse_use_scheduler	false	Use Scheduler for deferred indexing

Field weights

mse_index_fields sets weight per field. Format:

field1:weight1,field2:weight2,...

Higher weight ranks matches in that field higher.

Example configs

Standard:

pagetitle:3,longtitle:2,description:2,introtext:2,content:1

With TV:

pagetitle:3,content:1,tv_subtitle:2,tv_keywords:3

MiniShop3 products:

pagetitle:3,description:2,content:1,article:5,vendor_name:2

When to reindex

The index updates automatically on resource save. Full reindex is needed:

• After first install
• After changing indexing settings
• After bulk content import

Recommendations

Batch size	Use case
50	Default for most servers
10–20	Weak servers or large documents
100+	Strong servers and fast hosting

Optimal value

If you get timeout errors during indexing, reduce batch size.

Scheduler integration

mSearch supports deferred indexing via Scheduler, so indexing runs in the background and does not slow down saves.

Enabling

1. Install Scheduler
2. Enable system setting mse_use_scheduler

Behavior

On resource save:

Without Scheduler	With Scheduler
Indexing runs immediately	Task added to queue
Slows save	Save is fast
Blocks interface	Indexing in background

Task mse_index_resource

The task is registered automatically when mSearch is installed (if Scheduler is already installed).

Parameters:

Parameter	Description
resource_id	Resource ID to index
(none)	Full reindex of all resources

Manual full reindex

1. Go to Extras → Scheduler
2. Find task mse_index_resource (namespace: msearch)
3. Run the task with no parameters

Regular reindex

Use Scheduler to run reindex on a schedule, e.g. weekly.

Logging

The task logs progress to MODX log:

[mSearch indexResource task] Starting full reindex...
[mSearch indexResource task] Progress: 100/500 (indexed: 100)
...
[mSearch indexResource task] Full reindex completed. Total indexed: 500

Fallback without Scheduler

If Scheduler is not installed or the task is missing, mSearch falls back to direct indexing.

Events

Event	Description
mseOnBeforeIndex	Before indexing a resource. Can add fields.
mseOnGetWorkFields	Change list of fields to index
mseOnAfterIndex	After indexing a resource

Example: custom words

<?php
switch ($modx->event->name) {
    case 'mseOnBeforeIndex':
        $msearch->fields['custom_field'] = 1;
        $resource->set('custom_field', 'Extra keywords');
        break;
}

Example: MiniShop3 product options

<?php
switch ($modx->event->name) {
    case 'mseOnBeforeIndex':
        if ($resource->get('class_key') !== 'msProduct') {
            return;
        }

        $options = $modx->getCollection('msProductOption', [
            'product_id' => $resource->get('id'),
        ]);

        foreach ($options as $option) {
            $key = 'option_' . $option->get('key');
            $msearch->fields[$key] = 1;
            $resource->set($key, $option->get('value'));
        }
        break;
}