Skip to content

Site-wide search with Laravel Scout & Sphinx

I needed to set up a simple indexer to increase site performance.  After looking at the different options, like Elastic Search, Meilisearch, TNT Search, ect.  I ended up on Sphinx.  Sphinx is a little dated, but it’s very fast and doesn’t have a lot of dependencies, since it’s written in C++.

To start I installed Sphinx

1
sudo apt-get install sphinxsearch

Then updated the config file

1
2
cd /etc/sphinxsearch/
sudo nano sphinx.conf

Here’s a very simple full-text search example

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
source database
{
  type          = mysql
 
  sql_host      = localhost
  sql_user      = user
  sql_pass      = secret
  sql_db        = database
  sql_port      = 3306 # optional, default is 3306
}
 
source src_articles : database
{
    sql_field_string = name
 
    sql_query   = SELECT id, name from articles
}
 
index articles
{
    type        = plain
    source      = src_articles
    path         = /var/lib/sphinxsearch/data/articles
 
    min_prefix_len = 3
 
    index_exact_words   = 1
    expand_keywords     = 1
 
    charset_type = utf-8
}
 
indexer
{
        mem_limit               = 256M
}
 
searchd
{
        listen                  = 9312
        listen                  = 9306:mysql41
 
        # log file, searchd run info is logged here
        # optional, default is 'searchd.log'
        log                     = /var/log/sphinxsearch/searchd.log
 
        # query log file, all search queries are logged here
        # optional, default is empty (do not log queries)
        query_log               = /var/log/sphinxsearch/query.log
 
        read_timeout        = 5
        max_children        = 30
        pid_file            = /var/run/sphinxsearch/searchd.pid
        seamless_rotate     = 1
        preopen_indexes     = 1
        unlink_old          = 1
        workers             = threads
        binlog_path         = /var/lib/sphinxsearch/data
        collation_server    = utf8_ci
        thread_stack        = 6144K
}

Then we need to start Sphinx, make sure to run it under the sphinxsearch user and run the index and create the cron to re-run the indexes once a day at midnight

1
2
3
4
5
6
7
8
sudo nano /etc/default/sphinxsearch and set START=yes
 
sudo su sphinxsearch -s /bin/bash
indexer --all
/etc/init.d/sphinxsearch start
 
sudo crontab -u sphinxsearch -e
0 0 * * * indexer --all --rotate

Then, install the plugins we need for Laravel

1
2
3
composer require laravel/scout:6.0.0
composer require constantable/laravel-scout-sphinx
php artisan vendor:publish --provider="Laravel\Scout\ScoutServiceProvider"

Then you need to add these into the providers in config/app.php

1
2
Constantable\SphinxScout\ServiceProvider::class,
Laravel\Scout\ScoutServiceProvider::class,

Alright! That should get Sphinx all set on the server, and ready to go in Laravel. In all models you want to be searchable, add this (also, they all need to have indexes defined in Sphinx settings)

1
2
3
use Laravel\Scout\Searchable;
 
use Searchable;

Now comes the controller method to intermingle the models :X This is still a WIP, scout has a very limited version of eloquent attached to it, so you can’t set the offset by hand. So this is as close as you can get with out-of-the-box Scout.

This supports a JSON version that intermingles 6 results from each source for an autocomplete field, as well as a paginated version.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
public function search(Request $request) {
        $rsx = $request->header('X-Requested-With');
 
        if (!($page = $request->get('page'))) {
            $page = 1;
        }
 
        $searchResults = [];
 
        if ($q = $request->get('q')) {
            $tables = ['article','video','book'];
 
            if ($rsx != 'XMLHttpRequest') {
                //get totals for pagination
                $total = 0;
                $totals = [];
                foreach ($tables as $k => $table) {
                    $totals[$k] = ('App\Models\\' . str_replace('_', '', ucwords($table, '_')))::search($q)->paginate(1);
                    $totals[$k] = $totals[$k]->total();
                    $total = $total + $totals[$k];
                }
 
                $pagination = new LengthAwarePaginator([], $total, $perPage);
                $pagination = str_replace('/?', '?q='.$q.'&', $pagination->render());
 
                $parts = sizeof($totals);
                foreach ($tables as $k => $table) {
                    $offset = ($page - 1) * 3;
                    if(($offset + 3) > $totals[$k]) {
                        $parts = $parts - 1;
                    }
                }
                if($parts==0)$parts=1;
                $perpart = 9 / $parts;
            }
 
            foreach($tables as $k => $table) {
                if ($rsx != 'XMLHttpRequest' && ($offset + 3) > $totals[$k] && $page != 1 && $totals[$k] != 0) {
                    continue;
                }
 
                $model = ('App\Models\\'.str_replace('_','',ucwords($table,'_')));
                $searchResults[$k] = $model::search($q);
                if ($rsx == 'XMLHttpRequest') {
                    $searchResults[$k] = $searchResults[$k]->take(6)->get()->all();
                } else {
                    $searchResults[$k] = $searchResults[$k]->paginate(floor($perpart)+($table == 'article' && is_float($perpart) ? 1 : 0))->all();
                }
 
                $searchResults[$k] = array_map(function ($i) use ($table) {$i->type = $table.'s';return $i;}, $searchResults[$k]);
            }
 
            $searchResults = $this->array_zip_merge($searchResults);
            $searchResults = array_filter($searchResults);
        }
 
        if ($rsx == 'XMLHttpRequest') {
            $rta = [];
            foreach ($searchResults as $s) {
                $rta[] = array(
                    'name' => $s['name'],
                    'date' => $s['updated_at'],
                    'link' => $s['link'],
                );
            }
            return json_encode($rta);
        }
 
        return view('search_results',['pagination'=>$pagination,'q'=>$q, 'searchResults'=>$searchResults]);
}
 
private function array_zip_merge($args) {
        $output = array();
        // The loop incrementer takes each array out of the loop as it gets emptied by array_shift().
        for ($args = $args; count($args); $args = array_filter($args)) {
            // &$arg allows array_shift() to change the original.
            foreach ($args as &$arg) {
                $output[] = array_shift($arg);
            }
        }
        return $output;
}

These resources were really helpful in getting everything set-up

https://laravel.com/docs/8.x/scout
https://www.digitalocean.com/community/tutorials/how-to-install-and-configure-sphinx-on-ubuntu-14-04
https://sam-ngu.medium.com/site-wide-search-in-laravel-with-scout-27bbdc75ae8
https://github.com/franksierra/OSMServer

Merge Excel Sheets with Laravel Excel

Laravel Excel allows you to create and format Excel documents with PHP within the Laravel framework.  I specifically needed to merge Excel sheets, and not just tack on the sheets to one document, but actually merge several files into a single sheet.  It’s pretty straightforward, but the documentation leaves a bit to be desired so it took me a while to narrow down the right way to do this.  The only downside is it will remove any formatting from the sheets.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
//first pull all the rows
$rows = [];
Excel::batch(storage_path('FOLDER/CONTAINING/EXCEL/FILES'), function($rows1, $file) use($rows) {
    $rows1->each(function($row) use($rows1) {
        global $rows;
 
        $rows[] = $row->toArray();
    });
 
});
 
//then add them to a new file
Excel::create('new_filename', function ($excel) use ($rows) {
    $excel->sheet('Sheet', function ($sheet) use ($rows) {
        global $rows;
 
        // Sheet manipulation
        foreach($rows as $row) {
            $sheet->appendRow($row);
        }
    });
})->store('xls', storage_path('excel/exports'));

Install a Magento extension from SSH

Magento is a popular open-source eCommerce platform, which has a lot of extensions to add to its basic functionality.  It has a method for installing extensions in its control panel, but in my experience, it hasn’t been very reliable.  So, I usually just install extensions from the shell.  And, this is my quick cheat sheet on how you can install an extension in the shell.

  1. cd into your Magento directory
  2. Get the component name for composer, in the Magento marketplace it will be located under “My Profile > My Purchases”
  3. composer require component/name
  4. run “bin/magento module:status” to get the module’s name, it’ll be under “List of disabled modules:”
  5. bin/magento module:enable ModuleName
  6. bin/magento setup:upgrade
  7. bin/magento setup:di:compile
  8. bin/magento cache:clean
  9. php bin/magento setup:static-content:deploy -f
  10. There’s a known issue with Magento that it often won’t generate a “js-translation.json” file and that will break its back-end, but you can simply create one with empty JSON to fix the issue
  11. nano pub/static/adminhtml/Magento/backend/en_US/js-translation.json
  12. []

And you’re done!  If everything went off without a hitch you should see that extension working in your Magento control panel.

Anime on Netflix with Japanese subtitles

In conjunction with the Language Learning with Netflix chrome app it’s easy to watch shows with Furigana!  However, only so many shows actually have Japanese subtitles on Netflix :(  So here’s a list of anime with Japanese subtitles available on Netflix.  You can also search Netflix by all shows with Japanese subtitles here.

 

Actually Good

Beast Stars

KuroMukuro

 

Not bad

Drifting Dragons

Devilman Crybaby

A.I.C.O.

Cagster of an Insect Cage

 

Total Garbo

Swordgai

 

Haven’t watched

Dorohedoro

Ultramarine Maxwell

Children of the Whales

Carol & Teusday

Forest of Piano

Kengan Ashura

Last Hope

Baki

7 Seeds

Scissor Seven

Ni no Kuni

Kake Guri

Ghost in Shell CG

Levious

Megalobox

revisions

Compare CSV files in Bash

I’ve been needing this forever!  I have huge CSV lists and have needed a way to quickly compare them.  For this I just need to make sure that the first field from one CSV does not exist anywhere in the second CSV.  It was actually pretty simple to write this in Bash and it runs really fast!

1
2
3
4
5
6
while IFS=, read -r field1 field2
  do
    if grep --F "$field1" list1.csv; then
      echo "$field1 found"
    fi
done < list2.csv