README

The PHP implementation of the rules for reading text that Raoh and Souther share: Unicode 18.0.0 default case conversion and normalization, the White_Space set, order and length in Unicode scalar values, the lexical grammar of temporal text, and the pattern language.

This package is developed in php/ of raoh-project/notation-199x, beside the Java, Go and Rust implementations, and is held to the same test vectors. Packagist reads it from raoh-project/notation-199x-php, a mirror that CI writes on each release and nobody commits to. Issues and pull requests go to notation-199x.

composer require raoh/notation-199x

It needs a 64-bit PHP 8 from 8.2 on, as composer.json requires it ("php-64bit": "^8.2"): a PHP 9 is taken once it has been tested. It needs no extension: no rule asks mbstring, intl, iconv or the date functions, whose answers follow the Unicode version and the parsers of the PHP they were built with. PCRE is asked only whether bytes are UTF-8.

What it has

Every rule is stated of text that is a sequence of Unicode scalar values. A PHP string is bytes, so ScalarValues::invalidUtf8At is the question a caller asks before it takes text in. The other functions take text that is valid UTF-8 and do not ask it, apart from Pattern::read, which refuses such text, and Pattern::matches, which accepts none.

use Raoh\Notation199x\CaseConversion;
use Raoh\Notation199x\Normalization;
use Raoh\Notation199x\NormalizationForm;
use Raoh\Notation199x\Pattern;
use Raoh\Notation199x\ScalarValues;
use Raoh\Notation199x\TemporalKind;
use Raoh\Notation199x\TemporalText;
use Raoh\Notation199x\WhiteSpace;

ScalarValues::invalidUtf8At("ab\xff");                          // 2
ScalarValues::count('日本語');                                   // 3
CaseConversion::lowercase('ΟΣ');                               // 'ος'
CaseConversion::uppercaseWithin('straße', 6);                  // null: 'STRASSE' is 7 long
Normalization::normalize(NormalizationForm::NFKD, 'ﬃ');        // 'ffi'
WhiteSpace::contains(0x3000);                                  // true
TemporalText::check(TemporalKind::Instant, '2016-12-31T23:59:60Z'); // TemporalAnswer::LeapSecond

$read = Pattern::read('[0-9]{3}-[0-9]{4}');  // a Pattern, a PatternRefused or a PatternBeyond
if ($read instanceof Pattern) {
    $read->matches('123-4567');               // true: the whole string is matched
}

A pattern is the language the specifications define, not PCRE's. It is matched against the whole string, and in time linear in the string whatever the pattern.

raoh-php is what this package is written for. A caller of raoh-php is not meant to call it, and raoh-php keeps its types out of its own API.

License

Apache License 2.0

raoh / notation-199x

Package info

Security

Statistics

README

What it has

License