Attributes split and parse error

[Diego Bardari]

I have an html document with the following tag:

<span style="font-size:19.5pt;font-family:&quot;Arial&quot;,sans-serif;color:#41637E;letter-spacing:-.25pt">New story request<o:p></o:p></span>

It seems that after the clean process the style attribute is split into multiple attributes (even if setAllowMultiWordAttributes is set to true), where one of these attributes is -.25pt=-.25pt.

The problem is that calling createDOM on the resulting TagNode raise the following exception:

Exception in thread "main" org.w3c.dom.DOMException: INVALID_CHARACTER_ERR: An invalid or illegal XML character is specified. 
at org.apache.xerces.dom.CoreDocumentImpl.createAttribute(Unknown Source)
at org.apache.xerces.dom.ElementImpl.setAttribute(Unknown Source)
at org.htmlcleaner.DomSerializer.createSubnodes(DomSerializer.java:208)
at org.htmlcleaner.DomSerializer.createSubnodes(DomSerializer.java:220)
at org.htmlcleaner.DomSerializer.createSubnodes(DomSerializer.java:220)
at org.htmlcleaner.DomSerializer.createSubnodes(DomSerializer.java:220)
at org.htmlcleaner.DomSerializer.createSubnodes(DomSerializer.java:220)
at org.htmlcleaner.DomSerializer.createSubnodes(DomSerializer.java:220)
at org.htmlcleaner.DomSerializer.createSubnodes(DomSerializer.java:220)
at org.htmlcleaner.DomSerializer.createSubnodes(DomSerializer.java:220)
at org.htmlcleaner.DomSerializer.createSubnodes(DomSerializer.java:220)
at org.htmlcleaner.DomSerializer.createSubnodes(DomSerializer.java:220)
at org.htmlcleaner.DomSerializer.createDOM(DomSerializer.java:136)

This is caused I think by the fact that an attribute name cannot start with the - character by XML specifications.

Any help appreciated.

[Scott Wilson]

Diego - have you tried using the latest version of HtmlCleaner? I've tried this with the current Trunk version and I get this:

    @Test   
    public void multiWord() throws IOException, ParserConfigurationException{
        String initial = "<span style=\"font-size:19.5pt;font-family:&quot;Arial&quot;,sans-serif;color:#41637E;letter-spacing:-.25pt\">New story request<o:p></o:p></span>";
        cleaner.getProperties().setAllowMultiWordAttributes(true);
        DomSerializer ser = new DomSerializer(cleaner.getProperties());
        Document doc = ser.createDOM(cleaner.clean(initial));
        for (int i=0;i<doc.getDocumentElement().getElementsByTagName("span").item(0).getAttributes().getLength(); i++){
            System.out.println(doc.getDocumentElement().getElementsByTagName("span").item(0).getAttributes().item(i));
        }
    }   

Results in:

style="font-size:19.5pt;font-family:"Arial",sans-serif;color:#41637E;letter-spacing:-.25pt"

[Diego Bardari]

Hi Scott, thank you for the answer.

My problem was that I was unescaping the string before processing which resulted in:

<span style="font-size:19.5pt;font-family:"Arial",sans-serif;color:#41637E;letter-spacing:-.25pt">New story request<o:p></o:p></span>

which I think it's not a valid html string.

[Scott Wilson]

Aha, that makes sense