From 29e33475621fd9e4be20ecdea45d017bbd4f65f8 Mon Sep 17 00:00:00 2001
From: Damian Mooyman <damian@silverstripe.com>
Date: Thu, 25 Sep 2014 16:04:48 +1200
Subject: [PATCH] API Convert::html2raw no longer wraps text automatically BUG
 Convert::html2raw now correctly decodes single quotes

---
 core/Convert.php            | 17 ++++++++-------
 docs/en/changelogs/3.2.0.md |  1 +
 tests/core/ConvertTest.php  | 43 +++++++++++++++++++++++++++++++------
 3 files changed, 46 insertions(+), 15 deletions(-)

diff --git a/core/Convert.php b/core/Convert.php
index d4e303805..c472f7c57 100644
--- a/core/Convert.php
+++ b/core/Convert.php
@@ -286,11 +286,11 @@ class Convert {
 	/**
 	 * Simple conversion of HTML to plaintext.
 	 *
-	 * @param $data string
-	 * @param $preserveLinks boolean
-	 * @param $wordwrap array
+	 * @param string $data Input data
+	 * @param bool $preserveLinks
+	 * @param int $wordwrap
 	 */
-	public static function html2raw($data, $preserveLinks = false, $wordWrap = 60, $config = null) {
+	public static function html2raw($data, $preserveLinks = false, $wordWrap = 0, $config = null) {
 		$defaultConfig = array(
 			'PreserveLinks' => false,
 			'ReplaceBoldAsterisk' => true,
@@ -343,9 +343,7 @@ class Convert {
 		$data = preg_replace('/<\/p>/i', "\n\n", $data );
 
 		// Replace HTML entities
-		//$data = preg_replace("/&#([0-9]+);/e", 'chr(\1)', $data);
-		//$data = str_replace(array("&lt;","&gt;","&amp;","&nbsp;"), array("<", ">", "&", " "), $data);
-		$data = html_entity_decode($data, ENT_COMPAT , 'UTF-8');
+		$data = html_entity_decode($data, ENT_QUOTES, 'UTF-8');
 		// Remove all tags (but optionally keep links)
 
 		// strip_tags seemed to be restricting the length of the output
@@ -355,7 +353,10 @@ class Convert {
 		} else {
 			$data = strip_tags($data, '<a>');
 		}
-		return trim(wordwrap(trim($data), $wordWrap));
+
+		// Wrap
+		if($wordWrap) $data = wordwrap(trim($data), $wordWrap);
+		return trim($data);
 	}
 
 	/**
diff --git a/docs/en/changelogs/3.2.0.md b/docs/en/changelogs/3.2.0.md
index e2524a987..a59680bc4 100644
--- a/docs/en/changelogs/3.2.0.md
+++ b/docs/en/changelogs/3.2.0.md
@@ -13,6 +13,7 @@
  * `ClassInfo::dataClassesFor` now returns classes which should have tables, regardless of whether those
    tables actually exist.
  * `SS_Filterable`, `SS_Limitable` and `SS_Sortable` now explicitly extend `SS_List`
+ * `Convert::html2raw` no longer wraps text by default and can decode single quotes.
 
 #### Deprecated classes/methods removed
 
diff --git a/tests/core/ConvertTest.php b/tests/core/ConvertTest.php
index a9c684b14..07f56c239 100644
--- a/tests/core/ConvertTest.php
+++ b/tests/core/ConvertTest.php
@@ -38,34 +38,63 @@ class ConvertTest extends SapphireTest {
 
 	public function testHtml2raw() {
 		$val1 = 'This has a <strong>strong tag</strong>.';
-		$this->assertEquals('This has a *strong tag*.', Convert::xml2raw($val1),
+		$this->assertEquals('This has a *strong tag*.', Convert::html2raw($val1),
 			'Strong tags are replaced with asterisks');
 
 		$val1 = 'This has a <b class="test" style="font-weight: bold">b tag with attributes</b>.';
-		$this->assertEquals('This has a *b tag with attributes*.', Convert::xml2raw($val1),
+		$this->assertEquals('This has a *b tag with attributes*.', Convert::html2raw($val1),
 			'B tags with attributes are replaced with asterisks');
 
 		$val2 = 'This has a <strong class="test" style="font-weight: bold">strong tag with attributes</STRONG>.';
-		$this->assertEquals('This has a *strong tag with attributes*.', Convert::xml2raw($val2),
+		$this->assertEquals('This has a *strong tag with attributes*.', Convert::html2raw($val2),
 			'Strong tags with attributes are replaced with asterisks');
 
 		$val3 = '<script type="text/javascript">Some really nasty javascript here</script>';
-		$this->assertEquals('', Convert::xml2raw($val3),
+		$this->assertEquals('', Convert::html2raw($val3),
 			'Script tags are completely removed');
 
 		$val4 = '<style type="text/css">Some really nasty CSS here</style>';
-		$this->assertEquals('', Convert::xml2raw($val4),
+		$this->assertEquals('', Convert::html2raw($val4),
 			'Style tags are completely removed');
 
 		$val5 = '<script type="text/javascript">Some really nasty
 		multiline javascript here</script>';
-		$this->assertEquals('', Convert::xml2raw($val5),
+		$this->assertEquals('', Convert::html2raw($val5),
 			'Multiline script tags are completely removed');
 
 		$val6 = '<style type="text/css">Some really nasty
 		multiline CSS here</style>';
-		$this->assertEquals('', Convert::xml2raw($val6),
+		$this->assertEquals('', Convert::html2raw($val6),
 			'Multiline style tags are completely removed');
+
+		$val7 = '<p>That&#39;s absolutely correct</p>';
+		$this->assertEquals(
+			"That's absolutely correct",
+			Convert::html2raw($val7),
+			"Single quotes are decoded correctly"
+		);
+
+		$val8 = 'Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed do eiusmod tempor '.
+				'incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud '.
+				'exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute '.
+				'irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla '.
+				'pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia '.
+				'deserunt mollit anim id est laborum.';
+		$this->assertEquals($val8, Convert::html2raw($val8), 'Test long text is unwrapped');
+		$this->assertEquals(<<<PHP
+Lorem ipsum dolor sit amet, consectetur adipiscing elit, sed
+do eiusmod tempor incididunt ut labore et dolore magna
+aliqua. Ut enim ad minim veniam, quis nostrud exercitation
+ullamco laboris nisi ut aliquip ex ea commodo consequat.
+Duis aute irure dolor in reprehenderit in voluptate velit
+esse cillum dolore eu fugiat nulla pariatur. Excepteur sint
+occaecat cupidatat non proident, sunt in culpa qui officia
+deserunt mollit anim id est laborum.
+PHP
+			,
+			Convert::html2raw($val8, false, 60),
+			'Test long text is wrapped'
+		);
 	}
 
 	/**