Unlike URI templates, URL patterns support component-specific matching, wildcards, named groups, and regular-expression constraints. This makes them particularly useful for applications that need to recognize URL structures, including routers.<\/p>\n\n\n\n
For example, you can define a pattern that matches book URLs with numeric identifiers:<\/p>\n\n\n
\nuse League\\Uri\\UrlPattern;\n\n$pattern = UrlPattern::from(\n pattern: '\/books\/:id(\\\\d+)', \n baseUrl: 'https:\/\/example.com'\n);\n<\/pre><\/div>\n\n\nAs with URI templates, match()<\/code> lets you check whether a URI conforms to the pattern:<\/p>\n\n\n\n$pattern->match('https:\/\/example.com\/books\/123'); \/\/ true\n$pattern->match('https:\/\/example.com\/books\/abc'); \/\/ false\n<\/pre><\/div>\n\n\nAnd when you need the captured values, extract()<\/code> gives you access to them:<\/p>\n\n\n\n$result = $pattern->extract('https:\/\/example.com\/books\/123');\n\n$result->path['id']; \/\/ '123'\n<\/pre><\/div>\n\n\nHere, the regular-expression constraint ensures that the identifier contains only digits, while the extraction result exposes the captured value.<\/p>\n\n\n\n
Although both APIs provide matching and extraction, they follow different models. UriTemplate<\/code> implements RFC 6570 and extracts variables without being aware of the URI component they represent. UrlPattern<\/code>, on the other hand, follows the WHATWG URL Pattern model and processes values according to their URI components.<\/p>\n\n\n\nThis distinction also affects how extracted values can be converted. Consider the hostname label xn--fi8h<\/code>, which represents the Unicode character 🍅. With UriTemplate<\/code>, the extracted value remains in its original Punycode form until you explicitly interpret it as a hostname:<\/p>\n\n\n\n$uriTemplate = new UriTemplate('{host}.example.com');\n$result = $uriTemplate->extract('xn--fi8h.example.com');\n\n$result->string('host'); \/\/ 'xn--fi8h'\n$result->host('host'); \/\/ '🍅'\n<\/pre><\/div>\n\n\nWith UrlPattern<\/code>, hostname parsing and decoding are already handled by the URL Pattern algorithm:<\/p>\n\n\n\n$pattern = new UrlPatternBuilder(host: '{:foo}.example.com')->build();\n$result = $pattern->extract('xn--fi8h.example.com');\n\n$result->host->string('foo'); \/\/ '🍅'\n<\/pre><\/div>\n\n\nThe two APIs also differ in their extraction behavior. UriTemplate::extract()<\/code> returns an inspectable result even when extraction is incomplete, whereas UrlPattern::extract()<\/code> returns null<\/code> when the URI does not match.<\/p>\n\n\n\nTogether, these APIs let you match URIs, extract structured data, and convert values into the forms your application needs, while respecting the different semantics of URI templates and URL patterns.<\/p>\n\n\n\n
From extracted strings to typed PHP values<\/h2>\n\n\n\n Extracting values is only part of the problem. Once a value has been retrieved from a URI, it often needs to be converted before it can be used by an application.<\/p>\n\n\n\n
Consider a pagination URI:<\/p>\n\n\n
\nhttps:\/\/example.com\/search?limit=10\n<\/pre><\/div>\n\n\nThe extracted value of limit<\/code> is a string. An application may need it as an integer instead.<\/p>\n\n\n\nVersion 7.9 introduces typed accessors that make these conversions part of the extraction API:<\/p>\n\n\n
\nQuery::fromUri($uri)->integer('limit'); \/\/ 10\n<\/pre><\/div>\n\n\nDepending on the extraction API, you can retrieve values as strings, booleans, integers, floats, arrays, dates, or enums. Collection-oriented accessors also support converting lists of extracted values.<\/p>\n\n\n\n
For example, a URI template can extract a date and a status:<\/p>\n\n\n
\n$template = new UriTemplate('\/articles\/{date}{?status}');\n$result = $template->extract('\/articles\/2026-10-17?status=published');\n\n$result->date('date', '!Y-m-d');\n$result->enum('status', ArticleStatus::class);\n<\/pre><\/div>\n\n\nThis allows the application to work with a DateTimeImmutable<\/code> instance and an enum case rather than manually converting the extracted strings.<\/p>\n\n\n\nConversions are deliberately safe: when a value is missing or cannot be converted, the typed accessor returns null<\/code>. This gives applications an explicit way to handle invalid or unavailable values.<\/p>\n\n\n\nThe result is a more direct path from a URI to application-level data, with less repetitive conversion code.<\/p>\n\n\n\n
Keeping the URI polyfill up to date<\/h2>\n\n\n\n Alongside these new APIs, the uri-polyfill<\/code> has been updated to follow the functionality provided by PHP’s native ext\/uri<\/code> extension.<\/p>\n\n\n\nThe polyfill remains important for projects that need the newer URI parsing capabilities while supporting PHP versions where the extension is unavailable. Keeping the implementation aligned helps maintain consistent behavior across supported environments.<\/p>\n\n\n\n
Conclusion<\/h2>\n\n\n\n League URI Toolkit 7.9 expands the role of URI handling in PHP applications.<\/p>\n\n\n\n
With extraction support in both UriTemplate<\/code> and UrlPattern<\/code>, URIs can now serve not only as values to parse, construct, and manipulate, but also as structured sources of application data. Typed accessors take that one step further by making it easier to convert extracted values into the PHP types an application expects.<\/p>\n\n\n\nThe uri-polyfill<\/code> update complements these additions by keeping the toolkit aligned with PHP’s evolving native URI support.<\/p>\n\n\n\nTogether, these changes make working with URIs more expressive and reduce the amount of custom code needed to turn them into useful application data.<\/p>\n","protected":false},"excerpt":{"rendered":"
League URI Toolkit 7.9 introduces a new way to work with URIs: extracting data from them and converting that data into meaningful PHP values. This release also updates the uri-polyfill to keep pace with the URI functionality available in PHP’s native ext\/uri extension. But the most significant addition is the introduction of extraction APIs for […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_monsterinsights_skip_tracking":false,"footnotes":""},"categories":[5],"tags":[887,412,886,794,820,884,885],"class_list":["post-3097","post","type-post","status-publish","format-standard","hentry","category-web","tag-data-extraction","tag-php","tag-rfc6570","tag-uri","tag-uri-template","tag-url-pattern","tag-whatwg"],"aioseo_notices":[],"aioseo_head":"\n\t\t\n\t