mirror of
https://github.com/Llewellynvdm/php-ml.git
synced 2024-11-28 15:56:36 +00:00
02dab41830
* Issue #349: Provide a new NGramTokenizer. * Issue #349: Add tests. * Fixes from code review. * Implement NGramTokenizer with min and max gram support * Add missing tests for ngram * Add info about NGramTokenizer to docs and readme * Add performance test for tokenization
33 lines
1.2 KiB
PHP
33 lines
1.2 KiB
PHP
<?php
|
|
|
|
declare(strict_types=1);
|
|
|
|
namespace Phpml\Tests\Tokenization;
|
|
|
|
use Phpml\Tokenization\WhitespaceTokenizer;
|
|
|
|
class WhitespaceTokenizerTest extends TokenizerTest
|
|
{
|
|
public function testTokenizationOnAscii(): void
|
|
{
|
|
$tokenizer = new WhitespaceTokenizer();
|
|
|
|
$tokens = ['Lorem', 'ipsum-dolor', 'sit', 'amet,', 'consectetur/adipiscing', 'elit.',
|
|
'Cras', 'consectetur,', 'dui', 'et', 'lobortis;auctor.',
|
|
'Nulla', 'vitae', ',.,/', 'congue', 'lorem.', ];
|
|
|
|
self::assertEquals($tokens, $tokenizer->tokenize($this->getSimpleText()));
|
|
}
|
|
|
|
public function testTokenizationOnUtf8(): void
|
|
{
|
|
$tokenizer = new WhitespaceTokenizer();
|
|
|
|
$tokens = ['鋍鞎', '鳼', '鞮鞢騉', '袟袘觕,', '炟砏', '蒮', '謺貙蹖', '偢偣唲', '蒛', '箷箯緷', '鑴鱱爧', '覮轀,',
|
|
'剆坲', '煘煓瑐', '鬐鶤鶐', '飹勫嫢', '銪', '餀', '枲柊氠', '鍎鞚韕', '焲犈,',
|
|
'殍涾烰', '齞齝囃', '蹅輶', '鄜,', '孻憵', '擙樲橚', '藒襓謥', '岯岪弨', '蒮', '廞徲', '孻憵懥', '趡趛踠', '槏', ];
|
|
|
|
self::assertEquals($tokens, $tokenizer->tokenize($this->getUtf8Text()));
|
|
}
|
|
}
|