Download the PHP package nick-jones/php-ucd without Composer

PHP UCD

This project aims to present a PHP interface into the Unicode Character Database (UCD). It provides a means to lookup, filter, and interrogate the metadata & properties of unicode characters.

Installation

You can install this library via composer:

composer require nick-jones/php-ucd

Usage

The primary interface to utilise is UCD\Database. This provides a number of methods to interrogate "codepoint assigned" entities (i.e. Character, NonCharacter, and Surrogate instances) that reside within the UCD:

• Database::getByCodepoint(Codepoint $codepoint) - resolves a codepoint assigned entity
• Database::getCharacterByCodepoint(Codepoint $codepoint) - as above, but will only return Character instances
• Database::getByCodepoints(Codepoint\Collection $codepoints) - resolves multiple codepoint assigned entities
• Database::getCodepointsByBlock(Block $block) - resolves codepoints residing in the supplied block
• Database::getByBlock(Block $block) - resolves codepoint assigned entities residing in the supplied block
• Database::getCodepointsByCategory(GeneralCategory $category) - resolves codepoints residing in the supplied category
• Database::getByCategory(GeneralCategory $category) - resolves codepoint assigned entities residing in the supplied category
• Database::getCodepointsByScript(Script $script) - resolves codepoints residing in the supplied script
• Database::getByScript(Script $script) - resolves codepoint assigned entities residing in the supplied script
• Database::all() - returns a Collection instance containing everything assigned a codepoint within the database
• Database::onlyCharacters() - returns a Collection instance containing only Character instances
• Database::onlyNonCharacters() - returns a Collection instance containing only NonCharacter instances
• Database::onlySurrogates() - returns a Collection instance containing only Surrogate instances

The UCD\Unicode\Character\Collection class, returned by a number of methods, provides methods for filtering, traversal, codepoint extractions, amongst other things.

It is likely that you will want to leverage the default Character\Repository for resolution of characters, etc, in which case, calling UCD\Collection::fromDisk() will give you an instance backed by FileRepository. You can, of course, leverage a different Character\Repository implementation, if you so wish, by providing it to the constructor of UCD\Database.

Because this project makes good use of generators, the memory footprint of interrogating the dataset is fairly nominal.

Caveats

As of Unicode 8.0 there are > 260,000 items assigned codepoints. Reading, filtering and traversing all of these will take a few seconds. As such, if your intention is to identify items by filtering rules, you would be well advised to cache the output in some suitable form (e.g. build a regex, or PHP array of codepoints) which can then be interrogated, rather than always returning to filter and traverse the database. If your intention is to perform lookup by codepoint, then it is no problem to call into this library when and as required, as this is an efficient operation.

Examples

Manual Filtering + Traversal

Say you wish to dump all characters that hold a numeric property and reside outside of the Basic Latin (ASCII) block. You could simply leverage the Collection::filterWith(callable $filter) method to interrogate the properties of each Character instance. You could then perhaps dump their latin equivalent representation by calling ::getNumber() on the Numericity property. For example:

Codepoint Lookup

Locating an individual character by its codepoint value is trivial:

It is just as trivial to interrogate multiple codepoints. For example, you could print the name of every codepoint residing within a string:

Factory methods are available on the Codepoint and Codepoint\Collection classes to construct instances based on UTF-8, UTF-16BE, UTF-16LE, UTF-32BE and UTF-32LE encoded character(s).

Regex Building

The library provides a means to build regular expression characters classes based codepoints that have been extracted or aggregated from a collection of characters. For example, if you wanted to produce a regular expression that matched numeric flavour bengali characters, then you could use something along the lines of:

Map Building

This library can be used for building maps for various purposes. One such example is building a lowercase → uppercase character map. This is relatively simple to achieve; interrogate the properties of each character to check whether a mapping to a different character exists - if one does, print it out in PHP syntax:

This can then be leveraged as follow:

Executable

The primary intention of this project is to act as a library, however a small utility command is available for testing and database generation/manipulation purposes. bin/ucd search <codepoint> will dump character information, and bin/ucd repository-transfer <from> <to> will transfer characters from one repository implementation to another. Please run bin/ucd for more detailed help.

Properties

The intention of the most interesting of the available character properties, as described in Unicode Standard Annex #44, Unicode Character Database - Properties, available for interrogation. There are, however, a good quantity of them, so this remains work in progress. The following are currently covered:

• [x] Name
• [x] Block
• [x] Age
• [x] General Category
• [x] Numeric Value
• [x] Numeric Type
• [x] Normalization
• [x] Canonical Combining Class
• [x] Decomposition Mapping
• [x] Decomposition Type
• [x] Join Control
• [x] Joining Group
• [x] Joining Type
• [x] Bidi Class
• [x] Bidi Control
• [x] Bidi Mirrored
• [x] Bidi Mirroring Glyph
• [x] Bidi Paired Bracket
• [x] Bidi Paired Bracket Type

Tests

PhpSpec class specifications and PHPUnit backed integration tests are provided. The easiest way to run them is via the Makefile; simply run make test.