wamania / php-stemmer Goto Github PK

View Code? Open in Web Editor NEW

126.0 13.0 29.0 2.97 MB

Native PHP Stemmer

License: MIT License

PHP 100.00%

php snowball stemmer

php-stemmer's Introduction

php-stemmer

PHP native implementation of Snowball stemmer https://snowballstem.org/

Accept only UTF-8

Languages
Installation
Usage

Languages

Available :

Catalan (by Orestes Sanchez Benavente [email protected])
Danish
Dutch
English
Finnish (by Mikko Saari)
French
German
Italian
Norwegian
Portuguese
Romanian
Russian
Spanish
Swedish

Installation

For PHP5, use 1.3

composer require wamania/php-stemmer "^1.3"

For PHP7 use 2.x (branch 2.x is backward compatible with 1.x)

composer require wamania/php-stemmer "^2.0"

For PHP^7.3 and PHP^8.0 use 3.x (backward compatible, but phpunit^9 don't work with php < 7.3)

composer require wamania/php-stemmer "^3.0"

Usage

For 2.x, you should use the factory

use Wamania\Snowball\StemmerFactory;

// use ISO_639 (2 or 3 letters) or language name in english
$stemmer = StemmerFactory::create('fr');
$stemmer = StemmerFactory::create ('spanish');

// then 
$stem = $stemmer->stem('automóvil');

Or the manager

use Wamania\Snowball\StemmerManager;

$manager = new StemmerManager();
$stem = $manager->stem('automóvil', 'es');

In 1.3, you must instantiate manually

use Wamania\Snowball\French;

$stemmer = new French();
$stem = $stemmer->stem('anticonstitutionnellement');

php-stemmer's People

Contributors

Stargazers

Watchers

php-stemmer's Issues

Returning wrong stem for English adjective

I'm about to fork this stemmer for my language but before I could start I just tried a word using the English stemmer.

I gave it the adjective unethical but to my surprise it returned uneth whereas it should have spitted ethic as
Can you please look into it.

How to try the basic example in the Usage instructions

Hi, sorry for the very basic issue, but I couldn't make it work.

I used composer (latest version of your stemmer) in a PHP8 environment.

Composer downloaded a 21 Mb vendor directory with autoload.php inside:
I uploaded all the vendor directory with its contents and put outside a PHP file with:

require 'vendor/autoload.php';

Then I tried to execute the basic example you provided in the front page usage section but I got a class not found error.

Is there something else I should add to the autoload? Should I directly require classes? How?

Thanks and happy new year to everybody reading.

voku/portable-utf8 dependency

Hello, thank you for your work with the stemmer. Joomla has gotten a bug report that voku/portable-utf8 has a problem on PHP 8.2. (joomla/joomla-cms#39512) It turns out that this is only used from your php-stemmer AND that voku already fixed this in their latest 6.0.x versions. From my perspective 6.0.12 should work perfectly fine with php-stemmer, since none of the code you are using has been changed in the 6.0 release. Would it be possible for you to update your composer.json to require voku/portable-utf8 to "^5.4|^6.0"? That would help us with Joomla 5.0.

The big question would be, if you could also do another release for 2.x with that change as well. Then we could fix this issue in Joomla 4.x as well. Thank you!

PHP 8.2 and voku/portable-utf8

Hi there! I'm in the process of upgrading to PHP 8.2. Getting the following warnings:

Deprecated: Using ${var} in strings is deprecated, use {$var} instead in /vendor/voku/portable-utf8/src/voku/helper/UTF8.php on line 4653

Deprecated: Using ${var} in strings is deprecated, use {$var} instead in /vendor/voku/portable-utf8/src/voku/helper/UTF8.php on line 4664

Deprecated: Using ${var} in strings is deprecated, use {$var} instead in /vendor/voku/portable-utf8/src/voku/helper/UTF8.php on line 5731

Deprecated: Using ${var} in strings is deprecated, use {$var} instead in /vendor/voku/portable-utf8/src/voku/helper/UTF8.php on line 5742

Deprecated: Using ${var} in strings is deprecated, use {$var} instead in /vendor/voku/portable-utf8/src/voku/helper/UTF8.php on line 13398

Deprecated: Using ${var} in strings is deprecated, use {$var} instead in /vendor/voku/portable-utf8/src/voku/helper/UTF8.php on line 13398

Deprecated: Using ${var} in strings is deprecated, use {$var} instead in /vendor/voku/portable-utf8/src/voku/helper/UTF8.php on line 13409

Deprecated: Using ${var} in strings is deprecated, use {$var} instead in /vendor/voku/portable-utf8/src/voku/helper/UTF8.php on line 13409

I checked the latest release of voku/portable-utf8 and it seems like this has been fixed on their end. Could it be possible that you could fix this by updating the dependency in your composer.json to the version that fixes this?

php-stemmer/composer.json

Line 14 in 8ea32b6

"voku/portable-utf8": "^5.4|^6.0"

Thanks a lot for your help!

The stemmer replaces the "y" with "i" in the names

For example if I pass the name "Lucy" it converts to "luci", can you tell me if this is intended to work like this.

Not all English step methods are private

stem() should be the only public method, but the following ones are also exposed:

public function step3()
public function step4()
public function step5()
public function finish()

Lithuanian stemmer

Any chance to include the Lithuanian stemmer? https://snowballstem.org/algorithms/lithuanian/stemmer.html

Question about php 5 and 7 compability

The 1.3 version (for PHP 5) is compatible with 7x versions?

Greek language stemmer?

Hello,
this isn't an issue, but a request. There is a stemming algorithm for greek language.
https://snowballstem.org/algorithms/greek/stemmer.html

Can it be added here?

php 7.4

We discovered joomla/joomla-cms#27329 that the code in the utf8.php file is not compatible with php7.4 as it produces several deprecated notices like this

PHP Deprecated: Array and string offset access syntax with curly braces is deprecated

Is this something that is resolved in v2? Is v2 backwards compatible?

Bugfix release request

Hello @wamania,

would it be possible to tag a 2.1.1 including the fix from @SharkyKZ ? We would need it to fix the issue joomla/joomla-cms#31564 reported to joomla ?

Thanks for you work

Harald