[Htmlparser-cvs] htmlparser/src/org/htmlparser/tests/lexerTests AttributeTests.java,1.8,1.9 LexerTests.java,1.13,1.14

[<der...@us...>]

Update of /cvsroot/htmlparser/htmlparser/src/org/htmlparser/tests/lexerTests
In directory sc8-pr-cvs1:/tmp/cvs-serv27740/lexerTests

Modified Files:
	AttributeTests.java LexerTests.java 
Log Message:
Added testcases but was unable to reproduce the following bugs in the version 1.4 codebase:
839264 toHtml() parse error in Javascripts with "form" keyword
833592 DOCTYPE element is not parsed correctly
826764 ParserException occurs only when using setInputHTML() instea
825820 Words conjoined
825645 <input> not getting parsed inside table
813838 links not parsed correctly
and
#851882 zero length alt tag causes bug in ImageScanner
#832530 empty attribute causes parser to fail
#805598 attribute src in tag img sometimes not correctly parsed
(these 3 are all the same bug, duplicates of the following):
#753012 IMG SRC not parsed v1.3 & v1.4
#755929 Empty string attr. value causes attr parsing to be stopped
#778781 SRC-attribute suppression in IMG-tags
Also reviewed these test cases, again, with none reproducible in 1.4:
#788746 parser crashes on comments like <!-- foobar --!>
#772700 Jsp Tags are not parsed correctly when in quoted attributes.



Index: AttributeTests.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/tests/lexerTests/AttributeTests.java,v
retrieving revision 1.8
retrieving revision 1.9
diff -C2 -d -r1.8 -r1.9
*** AttributeTests.java	8 Dec 2003 01:31:53 -0000	1.8
--- AttributeTests.java	2 Jan 2004 05:01:28 -0000	1.9
***************
*** 38,41 ****
--- 38,42 ----
  import org.htmlparser.lexer.nodes.Attribute;
  import org.htmlparser.lexer.nodes.PageAttribute;
+ import org.htmlparser.tags.ImageTag;
  import org.htmlparser.tags.Tag;
  import org.htmlparser.tests.ParserTestCase;
***************
*** 519,522 ****
--- 520,604 ----
          assertTrue ("Attribute missing", table.containsKey ("OTHER"));
          assertEquals ("Attribute has wrong value", "fred", (String)table.get ("OTHER"));
+     }
+     
+     /**
+      * see bug #778781 SRC-attribute suppression in IMG-tags
+      * & #753012 IMG SRC not parsed v1.3 & v1.4
+      * & #755929 Empty string attr. value causes attr parsing to be stopped
+      * & #778781 SRC-attribute suppression in IMG-tags
+      * & #832530 empty attribute causes parser to fail
+      * & #851882 zero length alt tag causes bug in ImageScanner
+      *
+      *    HTML before parse:
+      *    <img src="images/first" alt="first">"
+      *    <img src="images/second" alt="">
+      *    <img alt="third" src="images/third">
+      *    <img alt="" src="images/fourth">
+      *
+      *    HTML after parse:
+      *    <IMG ALT="first" SRC="images/first">
+      *    <IMG ALT="" SRC="images/second">
+      *    <IMG ALT="third" SRC="images/third">
+      *    <IMG ALT="">
+      */
+     public void testSrcAndAlt () throws ParserException
+     {
+         String html = "<img src=\"images/first\" alt=\"first\">";
+ 
+         createParser (html);
+         parseAndAssertNodeCount (1);
+         assertTrue ("Node should be an ImageTag", node[0] instanceof ImageTag);
+         ImageTag img = (ImageTag)node[0];
+         assertTrue ("bad source", "images/first".equals (img.getImageURL ()));
+         assertTrue ("bad alt", "first".equals (img.getAttribute ("alt")));
+         assertStringEquals ("toHtml()", html, img.toHtml ());
+     }
+ 
+     /**
+      * see bug #778781 SRC-attribute suppression in IMG-tags
+      */
+     public void testSrcAndEmptyAlt () throws ParserException
+     {
+         String html = "<img src=\"images/second\" alt=\"\">";
+ 
+         createParser (html);
+         parseAndAssertNodeCount (1);
+         assertTrue ("Node should be an ImageTag", node[0] instanceof ImageTag);
+         ImageTag img = (ImageTag)node[0];
+         assertTrue ("bad source", "images/second".equals (img.getImageURL ()));
+         assertTrue ("bad alt", "".equals (img.getAttribute ("alt")));
+         assertStringEquals ("toHtml()", html, img.toHtml ());
+     }
+ 
+     /**
+      * see bug #778781 SRC-attribute suppression in IMG-tags
+      */
+     public void testAltAndSrc () throws ParserException
+     {
+         String html = "<img alt=\"third\" src=\"images/third\">";
+ 
+         createParser (html);
+         parseAndAssertNodeCount (1);
+         assertTrue ("Node should be an ImageTag", node[0] instanceof ImageTag);
+         ImageTag img = (ImageTag)node[0];
+         assertTrue ("bad source", "images/third".equals (img.getImageURL ()));
+         assertTrue ("bad alt", "third".equals (img.getAttribute ("alt")));
+         assertStringEquals ("toHtml()", html, img.toHtml ());
+     }
+ 
+     /**
+      * see bug #778781 SRC-attribute suppression in IMG-tags
+      */
+     public void testEmptyAltAndSrc () throws ParserException
+     {
+         String html = "<img alt=\"\" src=\"images/third\">";
+ 
+         createParser (html);
+         parseAndAssertNodeCount (1);
+         assertTrue ("Node should be an ImageTag", node[0] instanceof ImageTag);
+         ImageTag img = (ImageTag)node[0];
+         assertTrue ("bad source", "images/third".equals (img.getImageURL ()));
+         assertTrue ("bad alt", "".equals (img.getAttribute ("alt")));
+         assertStringEquals ("toHtml()", html, img.toHtml ());
      }
  }

Index: LexerTests.java
===================================================================
RCS file: /cvsroot/htmlparser/htmlparser/src/org/htmlparser/tests/lexerTests/LexerTests.java,v
retrieving revision 1.13
retrieving revision 1.14
diff -C2 -d -r1.13 -r1.14
*** LexerTests.java	31 Dec 2003 14:40:50 -0000	1.13
--- LexerTests.java	2 Jan 2004 05:01:28 -0000	1.14
***************
*** 697,700 ****
--- 697,750 ----
      }
  
+     /**
+      * See bug #825820 Words conjoined
+      */
+     public void testConjoined ()
+         throws
+             ParserException
+     {
+         StringBuffer buffer;
+         NodeIterator iterator;
+         Node node;
+         String expected;
+ 
+         expected = "The Title\nThis is the body.";
+         String html1 = "<html><title>The Title\n</title>" +
+             "<body>This is <a href=\"foo.html\">the body</a>.</body></html>";
+         createParser (html1);
+         buffer = new StringBuffer ();
+         for (iterator = parser.elements (); iterator.hasMoreNodes (); )
+         {
+             node = iterator.nextNode ();
+             String text = node.toPlainTextString ();
+             buffer.append (text);
+         }
+         assertStringEquals ("conjoined text", expected, buffer.toString ());
+ 
+         String html2 = "<html><title>The Title</title>\n" +
+             "<body>This is <a href=\"foo.html\">the body</a>.</body></html>";
+         createParser (html2);
+         buffer = new StringBuffer ();
+         for (iterator = parser.elements (); iterator.hasMoreNodes (); )
+         {
+             node = iterator.nextNode ();
+             String text = node.toPlainTextString ();
+             buffer.append (text);
+         }
+         assertStringEquals ("conjoined text", expected, buffer.toString ());
+         
+         String html3 = "<html><title>The Title</title>" +
+             "<body>\nThis is <a href=\"foo.html\">the body</a>.</body></html>";
+         createParser (html3);
+         buffer = new StringBuffer ();
+         for (iterator = parser.elements (); iterator.hasMoreNodes (); )
+         {
+             node = iterator.nextNode ();
+             String text = node.toPlainTextString ();
+             buffer.append (text);
+         }
+         assertStringEquals ("conjoined text", expected, buffer.toString ());
+     }
+ 
  }