bug #20235 [DomCrawler] Allow pipe (|) character in link tags when us…

…ing Xpath expressions (klausi, nicolas-grekas)

This PR was merged into the 2.7 branch.

Discussion
----------

[DomCrawler] Allow pipe (|) character in link tags when using Xpath expressions

| Q             | A
| ------------- | ---
| Branch?       | master
| Bug fix?      | yes
| New feature?  | no
| BC breaks?    | no
| Deprecations? | no
| Tests pass?   | yes
| Fixed tickets | #20229
| License       | MIT
| Doc PR        | -

@klausi could you please validate this patch? Is it an improvement over yours? (sorry I don't have the proper use case to test.)

Commits
-------

17757d8 [DomCrawler] Optimize DomCrawler::relativize()
5b26e33 [DomCrawler] Allow pipe (|) character in link tags when using Xpath expressions

src/Symfony/Component/DomCrawler/Crawler.php

@@ -856,22 +856,47 @@ private function relativize($xpath)
     {
         $expressions = array();
 
-        $unionPattern = '/\|(?![^\[]*\])/';
         // An expression which will never match to replace expressions which cannot match in the crawler
         // We cannot simply drop
         $nonMatchingExpression = 'a[name() = "b"]';
 
-        // Split any unions into individual expressions.
-        foreach (preg_split($unionPattern, $xpath) as $expression) {
-            $expression = trim($expression);
-            $parenthesis = '';
+        $xpathLen = strlen($xpath);
+        $openedBrackets = 0;
+        $startPosition = strspn($xpath, " \t\n\r\0\x0B");
 
-            // If the union is inside some braces, we need to preserve the opening braces and apply
-            // the change only inside it.
-            if (preg_match('/^[\(\s*]+/', $expression, $matches)) {
-                $parenthesis = $matches[0];
-                $expression = substr($expression, strlen($parenthesis));
+        for ($i = $startPosition; $i <= $xpathLen; ++$i) {
+            $i += strcspn($xpath, '"\'[]|', $i);
+
+            if ($i < $xpathLen) {
+                switch ($xpath[$i]) {
+                    case '"':
+                    case "'":
+                        if (false === $i = strpos($xpath, $xpath[$i], $i + 1)) {
+                            return $xpath; // The XPath expression is invalid
+                        }
+                        continue 2;
+                    case '[':
+                        ++$openedBrackets;
+                        continue 2;
+                    case ']':
+                        --$openedBrackets;
+                        continue 2;
+                }
+            }
+            if ($openedBrackets) {
+                continue;
+            }
+
+            if ($startPosition < $xpathLen && '(' === $xpath[$startPosition]) {
+                // If the union is inside some braces, we need to preserve the opening braces and apply
+                // the change only inside it.
+                $j = 1 + strspn($xpath, "( \t\n\r\0\x0B", $startPosition + 1);
+                $parenthesis = substr($xpath, $startPosition, $j);
+                $startPosition += $j;
+            } else {
+                $parenthesis = '';
             }
+            $expression = rtrim(substr($xpath, $startPosition, $i - $startPosition));
 
             // BC for Symfony 2.4 and lower were elements were adding in a fake _root parent
             if (0 === strpos($expression, '/_root/')) {
@@ -881,7 +906,7 @@ private function relativize($xpath)
             }
 
             // add prefix before absolute element selector
-            if (empty($expression)) {
+            if ('' === $expression) {
                 $expression = $nonMatchingExpression;
             } elseif (0 === strpos($expression, '//')) {
                 $expression = 'descendant-or-self::'.substr($expression, 2);
@@ -899,17 +924,24 @@ private function relativize($xpath)
                 // '.' is the fake root element in Symfony 2.4 and lower, which is excluded from results
                 $expression = $nonMatchingExpression;
             } elseif (0 === strpos($expression, 'descendant::')) {
-                $expression = 'descendant-or-self::'.substr($expression, strlen('descendant::'));
+                $expression = 'descendant-or-self::'.substr($expression, 12);
             } elseif (preg_match('/^(ancestor|ancestor-or-self|attribute|following|following-sibling|namespace|parent|preceding|preceding-sibling)::/', $expression)) {
                 // the fake root has no parent, preceding or following nodes and also no attributes (even no namespace attributes)
                 $expression = $nonMatchingExpression;
             } elseif (0 !== strpos($expression, 'descendant-or-self::')) {
                 $expression = 'self::'.$expression;
             }
             $expressions[] = $parenthesis.$expression;
+
+            if ($i === $xpathLen) {
+                return implode(' | ', $expressions);
+            }
+
+            $i += strspn($xpath, " \t\n\r\0\x0B", $i + 1);
+            $startPosition = $i + 1;
         }
 
-        return implode(' | ', $expressions);
+        return $xpath; // The XPath expression is invalid
     }
 
     /**

src/Symfony/Component/DomCrawler/Tests/CrawlerTest.php

@@ -387,6 +387,7 @@ public function testFilterXpathComplexQueries()
         $this->assertCount(5, $crawler->filterXPath('(//a | //div)//img'));
         $this->assertCount(7, $crawler->filterXPath('((//a | //div)//img | //ul)'));
         $this->assertCount(7, $crawler->filterXPath('( ( //a | //div )//img | //ul )'));
+        $this->assertCount(1, $crawler->filterXPath("//a[./@href][((./@id = 'Klausi|Claudiu' or normalize-space(string(.)) = 'Klausi|Claudiu' or ./@title = 'Klausi|Claudiu' or ./@rel = 'Klausi|Claudiu') or .//img[./@alt = 'Klausi|Claudiu'])]"));
     }
 
     public function testFilterXPath()
@@ -548,7 +549,7 @@ public function testFilterXPathWithSelfAxes()
 
         $this->assertCount(0, $crawler->filterXPath('self::a'), 'The fake root node has no "real" element name');
         $this->assertCount(0, $crawler->filterXPath('self::a/img'), 'The fake root node has no "real" element name');
-        $this->assertCount(9, $crawler->filterXPath('self::*/a'));
+        $this->assertCount(10, $crawler->filterXPath('self::*/a'));
     }
 
     public function testFilter()
@@ -969,6 +970,8 @@ public function createTestCrawler($uri = null)
 
                     <a href="?get=param">GetLink</a>
 
+                    <a href="/example">Klausi|Claudiu</a>
+
                     <form action="foo" id="FooFormId">
                         <input type="text" value="TextValue" name="TextName" />
                         <input type="submit" value="FooValue" name="FooName" id="FooId" />