Add function to parse from a URL

lib/Microformats.php

@@ -9,6 +9,7 @@ namespace MensBeam;
 
 use MensBeam\HTML\Parser as HTMLParser;
 use MensBeam\Microformats\Parser as MfParser;
+use MensBeam\Microformats\Url;
 
 /** A generic parser for microformats
  *
@@ -28,14 +29,46 @@ use MensBeam\Microformats\Parser as MfParser;
  * data may be supported in future.
  */
 class Microformats {
+    /** Parses a resource at a URL for microformats
+     * 
+     * If retrieving the resource fails `null` is returned.
+     * 
+     * @param string $file The resource to retrieve and parse
+     * @param array $options Options for the parser; please see the class documentetation for details
+     */
+    public static function fromUrl(string $url, array $options = []): ?array {
+        $stream = fopen($url, "r");
+        if ($stream) {
+            $location = null;
+            $type = null;
+            $data = stream_get_contents($stream);
+            if ($data !== false) {
+                $meta = stream_get_meta_data($stream);
+                if ($meta && $meta['wrapper_type'] === "http") {
+                    foreach ($meta['wrapper_data'] ?? [] as $h) {
+                        if (preg_match('/^HTTP\//i', $h)) {
+                            $type = null;
+                        } elseif (preg_match('/^Location\s*:\s*(.*)/is', $h, $match)) {
+                            $location = (string) URL::fromString($match[1], $location ?? $url);
+                        } elseif (preg_match('/^Content-Type\s*:\s*(.*)/is', $h, $match)) {
+                            $type = $match[1];
+                        }
+                    }
+                }
+                return static::fromString($data, $type ?? "", $location ?? $url, $options);
+            }
+        }
+        return null;
+    }
+
     /** Parses a file for microformats
      * 
      * If reading the file fails `null` is returned.
      * 
      * While fopen wrappers can be used to open remote resources over HTTP, no
      * effort is made to support this specially by reading the `Content-Type`
-     * header or deducing the URL. Using a proper HTTP client such as Guzzle
-     * is highly recommended instead.
+     * header or deducing the final URL. The `Microformats::fromUrl` method
+     * should be used for this purpose instead.
      * 
      * @param string $file The file to read and parse
      * @param string $contentType The HTTP Content-Type of the file if known, optionally with parameters

tests/cases/MicroformatsTest.php

@@ -21,4 +21,25 @@ class MicroformatsTest extends \PHPUnit\Framework\TestCase {
     public function testParseMissingFile(): void {
         $this->assertNull(@Microformats::fromFile("THIS FILE DOES NOT EXIST", "", ""));
     }
+
+    public function testParseRedirectedUrl(): void {
+        $exp = [
+            'items'    => [
+                [
+                    'type' => ["h-test"],
+                    'properties' => [
+                        'name' => ["Ça et là"],
+                        'url'  => ["http://localhost:8000/root.html"],
+                    ],
+                ],
+            ],
+            'rels'     => [],
+            'rel-urls' => [],
+        ];
+        $this->assertSame($exp, Microformats::fromUrl("http://localhost:8000/redir"));
+    }
+
+    public function testParseInvalidUrl(): void {
+        $this->assertNull(@Microformats::fromUrl("http://localhost:8000/404"));
+    }
 }

tests/docroot/404.php

@@ -0,0 +1,10 @@
+<?php return [
+    'code'    => 404,
+    'mime'    => "text/html; charset=utf-8",
+    'content' => <<< HTML_DOC
+<div class="h-test">
+    <span class="p-name">Ça et là</span>
+    <a class="u-url" href="root.html">Voir ici</a>
+</div>
+HTML_DOC,
+];

tests/docroot/redir.php

@@ -0,0 +1,8 @@
+<?php return [
+    'code'    => 302,
+    'mime'    => "text/plain",
+    'content' => "",
+    'fields'  => [
+        "Location: /utf8",
+    ],
+];

tests/docroot/utf8.php

@@ -0,0 +1,10 @@
+<?php return [
+    'code'    => 200,
+    'mime'    => "text/html; charset=utf-8",
+    'content' => <<< HTML_DOC
+<div class="h-test">
+    <span class="p-name">Ça et là</span>
+    <a class="u-url" href="root.html">Voir ici</a>
+</div>
+HTML_DOC,
+];

tests/server.php

@@ -0,0 +1,81 @@
+<?php
+/** @license MIT
+ * Copyright 2017 J. King, Dustin Wilson et al.
+ * See LICENSE and AUTHORS files for details */
+
+declare(strict_types=1);
+
+namespace MensBeam\Microformats;
+
+require_once __DIR__."/bootstrap.php";
+
+/*
+
+This is a so-called router for the the internal PHP Web server:
+<http://php.net/manual/en/features.commandline.webserver.php>
+
+It is used to test feed parsing in a controlled environment,
+answering specific requests used in tests with the data required
+to pass the test.
+
+The parameters of the responses are kept in separate files,
+which include the following data:
+
+- Response content
+- Response code
+- Content type
+- Whether to send cache headers
+- Last modified
+- Any other headers
+
+*/
+
+ignore_user_abort(false);
+ob_start();
+$defaults = [ // default values for response
+    'code'    => 200,
+    'content' => "",
+    'mime'    => "application/octet-stream",
+    'lastMod' => time(),
+    'cache'   => true,
+    'fields'  => [],
+];
+
+$url = explode("?", $_SERVER['REQUEST_URI'])[0];
+if ($url === "/") {
+    $url = "/index";
+}
+$base = BASE."tests".\DIRECTORY_SEPARATOR."docroot";
+$test = $base.str_replace("/", \DIRECTORY_SEPARATOR, $url).".php";
+if (!file_exists($test)) {
+    $response = [
+        'code'    => 499,
+        'content' => "Test '$test' missing.",
+        'mime'    => "application/octet-stream",
+        'lastMod' => time(),
+        'cache'   => true,
+        'fields'  => [],
+    ];
+} else {
+    $response = array_merge($defaults, (include $test));
+}
+// set the response code
+http_response_code((int) $response['code']);
+// if the response has a body, set the content type and (possibly) the ETag.
+if (strlen((string) $response['content'])) {
+    header("Content-Type: ".$response['mime']);
+    if ($response['cache']) {
+        header('ETag: "'.md5($response['content']).'"');
+    }
+}
+// if caching is enabled, set the last-modified date
+if ($response['cache']) {
+    header("Last-Modified: ".gmdate("D, d M Y H:i:s \G\M\T", $response['lastMod']));
+}
+// set any other specified fields verbatim
+foreach ($response['fields'] as $h) {
+    header($h);
+}
+// send the content
+echo $response['content'];
+ob_end_flush();