Hi does anyone know hwo to remove an attrbute using xpath. In particular the rel attribute and its text from a link. i.e. <a href='http://google.com' rel='some text'>Link</a> and i want to remove rel='some text'.
There will be multiple links in the html i am parsing.
You can select items using xpath, but that's all it can do - it is a query language.
You need to use XSLT or an XML parser in order to remove attributes/elements.
As pointed out by Oded, Xpath merely identifies XML nodes. To remove/edit XML, you need some additional tooling.
One solution is the Ant-based plugin XMLTask (disclaimer - I wrote this). It provides a simple mechanism to read an XML file, identify parts of that using XPath, and change it (including removing nodes).
e.g.
<remove path="web/servlet/context[#id='redundant']"/>
Have you already tried using Javascript for this If that is applicable in your scenario:-
var allLinks=document.getElementsByTagName("a");
for(i=0;i<allLinks.length;i++)
{
allLinks[i].removeAttribute("rel");
}
Related
This is my XML
<my_xml>
<record>
<p>hello <b>world</b> this is some html</p>
</record>
</my_xml>
Can I use XPath to return the following?
<p>hello <b>world</b> this is some html</p>
my_xml/record/child::*
child::* selects all element children of the context node
see details
The quick answer is, no. You can't accomplish this with XPath, but, once you select the parent node (i.e. "record" in your example), you should be able to manipulate it in whichever language you are using to parse the XML. Unfortunately, it may not be "easy".
It sounds like you would want something like the innerHTML property, but for XML DOM instead of the HTML DOM. Unfortunately, nothing like this exists for the XML DOM. If you don't care about the nodes themselves, you could use the textContent property; in the case of your example, you would get "hello world this is some html", which doesn't seem to be what you want.
Check out this similar question, which includes a parsing algorithm in Java. It seems that you will need to write a similar algorithm in whichever language you're using to parse the XML.
For anyone looking for this in the future, this IS very much possible to do using a DOT, that will return the entire node content as text (at least in MSSQL xpath it does).
'(/my_xml/record/.)[1]'
I have a piece of HTML that I would like to parse with Nokogiri, but I do not know whether it is a full HTML document (with DOCTYPE, etc) or a fragment (e.g. just a div with some elements in it).
This makes a difference for Nokogiri, because it should use #fragment for parsing fragments but #parse for parsing full documents.
Is there a way to determine whether a given piece of text is a fragment or a full HTML document?
Denis
Depends on how trashed your page is, but
/^(?:\s*<!DOCTYPE)|(?:\s*<html)/
should work in most cases.
The simplest way would be to look for the mandatory <html> tag, using for instance a regular expression /<html[\s>])/ (allowing attributes).
Is this sufficient to solve your problem?
I'm trying to query all input nodes. All of the nodes that are not self-closing are being returned fine, but the nodes that are self-closing are not. Is there a way to address this that doesn't require me to changes the HTML?
Thanks!
This is the default behavior. If you want to change it, you need to play with the ElementFlags collection, and for example, just remove INPUT from it, just like I explained for OPTION on a similar question here on SO: XHTML Parsing with HTMLAgilityPack
Greetings,
I'm facing a problem with the following tech-stack: JWebUnit -> HtmlUnit -> Xalan.
I'm trying to find an element by XPATH, but the HTML document is pretty malformed.
Xalan stops finding elements when I reach the /body element on XPATH. I believe it's because the document contains two <body> tags and one being unclosed.
Everything works for /html/head or /html. But when I try /html/body (or /html/body[1], //body[1], or anything inside those tags) I get only null from Xalan.
Is there any way to get around with that? I just can't change the html document istself. Thank you kindly for your attention.
Best regards,
Thiago
HtmlUnit must be using something to convert HTML to XML. Perhaps you can tell it to use jsoup or tagsoup, which are very tolerant of messy HTML?
You might as well also write code to just dump the XML tree to a file so you can see what's in it.
I"m unsure about this. Would having PHP ( or I guess any template language like Django's or Mako or whatever ) inside an html file prevent me from making changes to it with XPath?
I'm very new to XPath. I would think that you could not, but as I said, I'm unsure.
Xpath is a query language. You use it to query XML content, not change it.
You can use Xpath in conjunction with other technologies (XSLT is the first one that comes to mind) in order to query you XML and then use the results of these queries to transform your XML.
XPath doesn't change the XML document.
Use XSLT or a any other XPath-hosting language that can produce a new XML document.