From 6f5592ae60131a5c6fe8208c0ba87fb62661ec8b Mon Sep 17 00:00:00 2001
From: "Edward Z. Yang" <edwardzyang@thewritingpot.com>
Date: Mon, 25 Jun 2007 19:18:55 +0000
Subject: [PATCH] [2.0.1] Normalize newlines to \n for internal processing.

git-svn-id: http://htmlpurifier.org/svnroot/htmlpurifier/trunk@1235 48356398-32a2-884e-a903-53898d9a118a
---
 NEWS                                            |  3 +++
 library/HTMLPurifier/Generator.php              | 17 +++++++++++++++--
 library/HTMLPurifier/Injector/AutoParagraph.php |  8 ++++----
 library/HTMLPurifier/Lexer.php                  |  4 ++++
 library/HTMLPurifier/Lexer/DirectLex.php        |  2 +-
 tests/HTMLPurifier/ChildDef/TableTest.php       |  5 +++--
 tests/HTMLPurifier/GeneratorTest.php            |  1 +
 tests/HTMLPurifier/Lexer/DirectLexTest.php      | 17 ++++-------------
 8 files changed, 35 insertions(+), 22 deletions(-)

diff --git a/NEWS b/NEWS
index 225e65ca..3c6bfb7a 100644
--- a/NEWS
+++ b/NEWS
@@ -15,6 +15,9 @@ NEWS ( CHANGELOG and HISTORY )                                     HTMLPurifier
 ! Experimental AutoFormat functionality added: auto-paragraph and
   linkify your HTML input by setting %AutoFormat.AutoParagraph and
   %AutoFormat.Linkify to true
+! Newlines normalized internally, and then converted back to the
+  value of PHP_EOL. If this is not desired, set your newline format
+  using %Output.Newline.
 - Clean up special case code for <script> tags
 - Reorder includes for DefinitionCache decorators, fixes a possible
   missing class error
diff --git a/library/HTMLPurifier/Generator.php b/library/HTMLPurifier/Generator.php
index 3f39cd44..001b28f9 100644
--- a/library/HTMLPurifier/Generator.php
+++ b/library/HTMLPurifier/Generator.php
@@ -35,6 +35,15 @@ HTML
 );
 HTMLPurifier_ConfigSchema::defineAlias('Core', 'TidyFormat', 'Output', 'TidyFormat');
 
+HTMLPurifier_ConfigSchema::define('Output', 'Newline', null, 'string/null', '
+<p>
+    Newline string to format final output with. If left null, HTML Purifier
+    will auto-detect the default newline type of the system and use that;
+    you can manually override it here. Remember, \r\n is Windows, \r
+    is Mac, and \n is Unix. This directive was available since 2.0.1.
+</p>
+');
+
 /**
  * Generates HTML from tokens.
  * @todo Create a configuration-wide instance that all objects retrieve
@@ -114,6 +123,10 @@ class HTMLPurifier_Generator
                 $html = (string) $tidy;
             }
         }
+        // normalize newlines to system
+        $nl = $config->get('Output', 'Newline');
+        if ($nl === null) $nl = PHP_EOL;
+        $html = str_replace("\n", $nl, $html);
         return $html;
     }
     
@@ -153,11 +166,11 @@ class HTMLPurifier_Generator
      */
     function generateScriptFromToken($token) {
         if ($token->type != 'text') return $this->generateFromToken($token);
-        // return '<!--' . PHP_EOL . trim($token->data) . PHP_EOL . '// -->';
+        // return '<!--' . "\n" . trim($token->data) . "\n" . '// -->';
         // more advanced version:
         // thanks <http://lachy.id.au/log/2005/05/script-comments>
         $data = preg_replace('#//\s*$#', '', $token->data);
-        return '<!--//--><![CDATA[//><!--' . PHP_EOL . trim($data) . PHP_EOL . '//--><!]]>';
+        return '<!--//--><![CDATA[//><!--' . "\n" . trim($data) . "\n" . '//--><!]]>';
     }
     
     /**
diff --git a/library/HTMLPurifier/Injector/AutoParagraph.php b/library/HTMLPurifier/Injector/AutoParagraph.php
index 5811fd11..e2b17ba2 100644
--- a/library/HTMLPurifier/Injector/AutoParagraph.php
+++ b/library/HTMLPurifier/Injector/AutoParagraph.php
@@ -40,7 +40,7 @@ class HTMLPurifier_Injector_AutoParagraph extends HTMLPurifier_Injector
             $this->_splitText($text, $token);
         } elseif ($this->allowsElement('p')) {
             // case 3: we're in an element that allows paragraphs
-            if (strpos($text, PHP_EOL . PHP_EOL) !== false) {
+            if (strpos($text, "\n\n") !== false) {
                 // case 3.1: this text node has a double-newline
                 $token = array(new HTMLPurifier_Token_Start('p'));
                 $this->_splitText($text, $token);
@@ -57,7 +57,7 @@ class HTMLPurifier_Injector_AutoParagraph extends HTMLPurifier_Injector
                     }
                     if ($this->inputTokens[$i]->type == 'end') break;
                     if ($this->inputTokens[$i]->type == 'text') {
-                        if (strpos($this->inputTokens[$i]->data, PHP_EOL . PHP_EOL) !== false) {
+                        if (strpos($this->inputTokens[$i]->data, "\n\n") !== false) {
                             $ok = true;
                         }
                         if (!$this->inputTokens[$i]->is_whitespace) break;
@@ -104,7 +104,7 @@ class HTMLPurifier_Injector_AutoParagraph extends HTMLPurifier_Injector
                     if ($this->inputTokens[$i]->type == 'start') $j++;
                     if ($this->inputTokens[$i]->type == 'end') $j--;
                     if ($this->inputTokens[$i]->type == 'text') {
-                        if (strpos($this->inputTokens[$i]->data, PHP_EOL . PHP_EOL) !== false) {
+                        if (strpos($this->inputTokens[$i]->data, "\n\n") !== false) {
                             $ok = true;
                             break;
                         }
@@ -137,7 +137,7 @@ class HTMLPurifier_Injector_AutoParagraph extends HTMLPurifier_Injector
      * @private
      */
     function _splitText($data, &$result) {
-        $raw_paragraphs = explode(PHP_EOL . PHP_EOL, $data);
+        $raw_paragraphs = explode("\n\n", $data);
         
         // remove empty paragraphs
         $paragraphs = array();
diff --git a/library/HTMLPurifier/Lexer.php b/library/HTMLPurifier/Lexer.php
index 3147230e..76dfee7c 100644
--- a/library/HTMLPurifier/Lexer.php
+++ b/library/HTMLPurifier/Lexer.php
@@ -303,6 +303,10 @@ class HTMLPurifier_Lexer
             $html = $this->extractBody($html);
         }
         
+        // normalize newlines to \n
+        $html = str_replace("\r\n", "\n", $html);
+        $html = str_replace("\r", "\n", $html);
+        
         if ($config->get('HTML', 'Trusted')) {
             // escape convoluted CDATA
             $html = $this->escapeCommentedCDATA($html);
diff --git a/library/HTMLPurifier/Lexer/DirectLex.php b/library/HTMLPurifier/Lexer/DirectLex.php
index 7159fe64..508e8177 100644
--- a/library/HTMLPurifier/Lexer/DirectLex.php
+++ b/library/HTMLPurifier/Lexer/DirectLex.php
@@ -47,7 +47,7 @@ class HTMLPurifier_Lexer_DirectLex extends HTMLPurifier_Lexer
         if ($maintain_line_numbers) $current_line = 1;
         else $current_line = false;
         $context->register('CurrentLine', $current_line);
-        $nl = PHP_EOL;
+        $nl = "\n";
         // how often to manually recalculate. This will ALWAYS be right,
         // but it's pretty wasteful. Set to 0 to turn off
         $synchronize_interval = $config->get('Core', 'DirectLexLineNumberSyncInterval'); 
diff --git a/tests/HTMLPurifier/ChildDef/TableTest.php b/tests/HTMLPurifier/ChildDef/TableTest.php
index d5780a63..a8bf430b 100644
--- a/tests/HTMLPurifier/ChildDef/TableTest.php
+++ b/tests/HTMLPurifier/ChildDef/TableTest.php
@@ -38,10 +38,11 @@ class HTMLPurifier_ChildDef_TableTest extends HTMLPurifier_ChildDefHarness
         
         // whitespace sticks to the previous element, last whitespace is
         // stationary
-        $this->assertResult("\n   <tr />\n  <tr />\n ");
+        $this->assertResult("\n   <tr />\n  <tr />\n ", true, array('Output.Newline' => "\n"));
         $this->assertResult(
           "\n\t<tbody />\n\t\t<tfoot />\n\t\t\t",
-          "\n\t\t<tfoot />\n\t<tbody />\n\t\t\t"
+          "\n\t\t<tfoot />\n\t<tbody />\n\t\t\t",
+          array('Output.Newline' => "\n")
         );
         
     }
diff --git a/tests/HTMLPurifier/GeneratorTest.php b/tests/HTMLPurifier/GeneratorTest.php
index 7933fc42..f9040b5a 100644
--- a/tests/HTMLPurifier/GeneratorTest.php
+++ b/tests/HTMLPurifier/GeneratorTest.php
@@ -207,6 +207,7 @@ class HTMLPurifier_GeneratorTest extends HTMLPurifier_Harness
         
         $this->config = HTMLPurifier_Config::createDefault();
         $this->config->set('Core', 'TidyFormat', true);
+        $this->config->set('Output', 'Newline', "\n");
         
         // nice wrapping please
         $this->assertGeneration(
diff --git a/tests/HTMLPurifier/Lexer/DirectLexTest.php b/tests/HTMLPurifier/Lexer/DirectLexTest.php
index a3813fc5..7b15f05a 100644
--- a/tests/HTMLPurifier/Lexer/DirectLexTest.php
+++ b/tests/HTMLPurifier/Lexer/DirectLexTest.php
@@ -72,32 +72,23 @@ class HTMLPurifier_Lexer_DirectLexTest extends UnitTestCase
     
     function testLineNumbers() {
         
-        $html = '<b>Line 1</b>
-                 <i>Line 2</i>
-                 Still Line 2<br
-                 />Now Line 4
-                 
-                 <br />';
+        $html = "<b>Line 1</b>\n<i>Line 2</i>\nStill Line 2<br\n/>Now Line 4\n\n<br />";
         
         $expect = array(
             // line 1
             0 => new HTMLPurifier_Token_Start('b')
            ,1 => new HTMLPurifier_Token_Text('Line 1')
            ,2 => new HTMLPurifier_Token_End('b')
-           ,3 => new HTMLPurifier_Token_Text('
-                 ')
+           ,3 => new HTMLPurifier_Token_Text("\n")
             // line 2
            ,4 => new HTMLPurifier_Token_Start('i')
            ,5 => new HTMLPurifier_Token_Text('Line 2')
            ,6 => new HTMLPurifier_Token_End('i')
-           ,7 => new HTMLPurifier_Token_Text('
-                 Still Line 2')
+           ,7 => new HTMLPurifier_Token_Text("\nStill Line 2")
             // line 3
            ,8 => new HTMLPurifier_Token_Empty('br')
             // line 4
-           ,9 => new HTMLPurifier_Token_Text('Now Line 4
-                 
-                 ')
+           ,9 => new HTMLPurifier_Token_Text("Now Line 4\n\n")
             // line SIX
            ,10 => new HTMLPurifier_Token_Empty('br')
         );