added score increment if word count is greater than 30 when parsing content, for now commented out but will continue to test against various sites

git-svn-id: http://arc90labs-readability.googlecode.com/svn/trunk@71 d4e419ec-0920-11de-bbfd-a7c1bc4c261e

1 files changed, 5 insertions, 2 deletions

diff --git a/js/readability.js b/js/readability.js
index 57eb240..cde729d 100755
--- a/js/readability.js
+++ b/js/readability.js
@@ -129,6 +129,9 @@ function determineContentScore(score, parent, element)
 	if (element.tagName.toLowerCase() == "p" && getWordCount(element) > 20) //|| (score == 0 && getText(element).length > 10)) 

 		score++;

 	

+	//if (getWordCount(element) > 30) 

+	//	score++;

+	

 	// FIXME: not sure yet if this will be included, this would break 

 	// pages that use multiple containers for content, or we could tweak 

 	// the acceptable minimum... but that would have to be set quite 

@@ -148,8 +151,8 @@ function determineContentScore(score, parent, element)
 

 function parseContent() {

 	// replace all doubled-up <BR> tags with <P> tags, and remove inline fonts

-	document.body.innerHTML = document.body.innerHTML.replace(/<br[^>]*>\s*<br[^>]*>/gi, "<p />").replace(/<\/?font[^>]*>/gi, "");

-	

+	document.body.innerHTML = document.body.innerHTML.replace(/<br[^>]*>\s|&nbsp;*<br[^>]*>/gi, "<p />").replace(/<\/?font[^>]*>/gi, "");

+	console.log(document.body.innerHTML.substring(0, 600));

 	var articleContent = document.createElement("DIV");

 	var paragraphs = document.getElementsByTagName("P");

 	var contentBlocks = [];