You are viewing a plain text version of this content. The canonical link for it is here.

Posted to dev@tika.apache.org by "Gerard Bouchar (JIRA)" <ji...@apache.org> on 2018/05/25 10:10:00 UTC

[jira] [Created] (TIKA-2652) HtmlParser generates incorrect meta tags

Gerard Bouchar created TIKA-2652:
------------------------------------

             Summary: HtmlParser generates incorrect meta tags
                 Key: TIKA-2652
                 URL: https://issues.apache.org/jira/browse/TIKA-2652
             Project: Tika
          Issue Type: Bug
            Reporter: Gerard Bouchar


Whatever the input HTML meta are, tika's HTML meta can only have a "name" and a "content"  attribute. This gives invalid HTML meta tags in the output.

For instance, the following valid HTML file

{code:html}
<!DOCTYPE html>
<html lang="en">
  <head>
    <title>Title</title>
    <meta http-equiv="refresh" content="0; url=http://example.com">
  </head>
  <body></body>
</html>
{code}

is transformed into a SAX stream corresponding to the following HTML :

{code:html}
<html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta name="dc:title" content="Title"/>
<meta name="Content-Encoding" content="ISO-8859-1"/>
<meta name="refresh" content="0; url=http://example.com"/>
<meta name="Content-Type" content="text/html; charset=ISO-8859-1"/>
<title>Title</title>
</head>
<body/></html>
{code}

The information that the original file had an "http-equiv" meta tag is lost, and replaced by a generic "meta name=" tag.

This is annoying when working with classes expecting valid meta redirection, such as Nutch's [HTMLMetaProcessot|https://github.com/apache/nutch/blob/master/src/plugin/parse-html/src/java/org/apache/nutch/parse/html/HTMLMetaProcessor.java], for instance.



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)