feature #19430 [DomCrawler] Add support for XPath expression evaluation (jakzal)

fabpot · fabpot · commit cec494ceed62 · 2016-08-02T12:15:39.000+02:00
This PR was merged into the 3.2-dev branch. Discussion ---------- [DomCrawler] Add support for XPath expression evaluation | Q | A | ------------- | --- | Branch? | master | Bug fix? | no | New feature? | yes | BC breaks? | no | Deprecations? | no | Tests pass? | yes | Fixed tickets | #19162 | License | MIT | Doc PR | TODO Example usage: ```php <?php use Symfony\Component\DomCrawler\Crawler; use Symfony\Component\VarDumper\VarDumper; require_once __DIR__.'/vendor/autoload.php'; $html = '<html> <body> <span id="article-100" class="article">Article 1</span> <span id="article-101" class="article">Article 2</span> <span id="article-102" class="article">Article 3</span> </body> </html>'; $crawler = new Crawler(); $crawler->addHtmlContent($html); VarDumper::dump($crawler->filterXPath('//span[contains(@id, "article-")]')->evaluate('substring-after(@id, "-")')); // array:3 [ // 0 => "100" // 1 => "101" // 2 => "102" // ] VarDumper::dump($crawler->evaluate('substring-after(//span[contains(@id, "article-")]/@id, "-")')); // array:1 [ // 0 => "100" // ] VarDumper::dump($crawler->filterXPath('//span[@Class="article"]')->evaluate('count(@id)')); // array:3 [ // 0 => 1.0 // 1 => 1.0 // 2 => 1.0 // ] VarDumper::dump($crawler->evaluate('count(//span[@Class="article"])')); // array:1 [ // 0 => 3.0 // ] VarDumper::dump($crawler->evaluate('//span[1]')); // Symfony\Component\DomCrawler\Crawler { } ``` Commits ------- 3148fad [DomCrawler] Add support for XPath expression evaluation
diff --git a/Crawler.php b/Crawler.php
@@ -592,6 +592,36 @@ public function html()
         return $html;
     }
 
+    /**
+     * Evaluates an XPath expression.
+     *
+     * Since an XPath expression might evaluate to either a simple type or a \DOMDoneList,
+     * this method will return either an array of simple types or a new Crawler instance.
+     *
+     * @param string $xpath An XPath expression
+     *
+     * @return array|Crawler An array of evaluation results or a new Crawler instance
+     */
+    public function evaluate($xpath)
+    {
+        if (null === $this->document) {
+            throw new \LogicException('Cannot evaluate the expression on an uninitialized crawler.');
+        }
+
+        $data = array();
+        $domxpath = $this->createDOMXPath($this->document, $this->findNamespacePrefixes($xpath));
+
+        foreach ($this->nodes as $node) {
+            $data[] = $domxpath->evaluate($xpath, $node);
+        }
+
+        if (isset($data[0]) && $data[0] instanceof \DOMNodeList) {
+            return $this->createSubCrawler($data);
+        }
+
+        return $data;
+    }
+
     /**
      * Extracts information from the list of nodes.
      *
diff --git a/Tests/CrawlerTest.php b/Tests/CrawlerTest.php
@@ -1061,6 +1061,51 @@ public function testCountOfNestedElements()
         $this->assertCount(1, $crawler->filter('li:contains("List item 1")'));
     }
 
+    public function testEvaluateReturnsTypedResultOfXPathExpressionOnADocumentSubset()
+    {
+        $crawler = $this->createTestCrawler();
+
+        $result = $crawler->filterXPath('//form/input')->evaluate('substring-before(@name, "Name")');
+
+        $this->assertSame(array('Text', 'Foo', 'Bar'), $result);
+    }
+
+    public function testEvaluateReturnsTypedResultOfNamespacedXPathExpressionOnADocumentSubset()
+    {
+        $crawler = $this->createTestXmlCrawler();
+
+        $result = $crawler->filterXPath('//yt:accessControl/@action')->evaluate('string(.)');
+
+        $this->assertSame(array('comment', 'videoRespond'), $result);
+    }
+
+    public function testEvaluateReturnsTypedResultOfNamespacedXPathExpression()
+    {
+        $crawler = $this->createTestXmlCrawler();
+        $crawler->registerNamespace('youtube', 'http://gdata.youtube.com/schemas/2007');
+
+        $result = $crawler->evaluate('string(//youtube:accessControl/@action)');
+
+        $this->assertSame(array('comment'), $result);
+    }
+
+    public function testEvaluateReturnsACrawlerIfXPathExpressionEvaluatesToANode()
+    {
+        $crawler = $this->createTestCrawler()->evaluate('//form/input[1]');
+
+        $this->assertInstanceOf(Crawler::class, $crawler);
+        $this->assertCount(1, $crawler);
+        $this->assertSame('input', $crawler->first()->nodeName());
+    }
+
+    /**
+     * @expectedException \LogicException
+     */
+    public function testEvaluateThrowsAnExceptionIfDocumentIsEmpty()
+    {
+        (new Crawler())->evaluate('//form/input[1]');
+    }
+
     public function createTestCrawler($uri = null)
     {
         $dom = new \DOMDocument();
