I do not want to parse some tags in XML

Question

Currently this would be a sample XML that I am working on:

<smsq>
  <sms>
  <id>96</id>
  <to>03333560511</to>
  <msg>  danial says: hahaha <space> nothing.
  </msg>
  </sms>
</smsq>

Now please notice, that the tag can contain other tags (which should not be parsed) and I had to make a dtd for that. The dtd was something like this:

<!DOCTYPE smsq [
  <!ELEMENT sms (mID,to,msg,type)>
  <!ELEMENT mID (#PCDATA)>
  <!ELEMENT to (#PCDATA)>
  <!ELEMENT msg (CDATA)>
]>

But the problem is that XML parser still goes in the tag and says that the tag should be closed with a tag. I just want to fetch the data as it is from the XML and I do not want to parse msg further.

Please help me resolve the problem and tell me if this can be done with DTDs.

Thanks!

Answer 1

Firstly the sample xml is not really xml as the "space" tag is not closed.

Secondly, it looks like the reason for not wanting to parse the "space" tag is because it's not really xml - just text that looks like xml. The text should be either escaped/encoded or enclosed in CDATA tags.

Lastly - if what you want to parse really is xml and you only want to parse the first level tags. I wouldn't bother with a real XML parser - i'd create my own ultra-simple parser - all it has to do is parse 1st level nodes - that shouldn't be too hard.

Good luck!

Answer 2

DTD can't help you with this problem. DTD is by no means required (though it is quite handy to have it).

The document you posted above is not a valid XML document. Period. That's the way it is, and no reasonable XML parser will parse it for you without raising the error.

What you can do though is to substitute < symbol with a < XML entity.

Answer 3

You can't make a DTD that makes buggy XML magically not buggy. The XML is not well-formed, so it can never be valid as well-formedness is a prerequisite of validity (validity isn't even important here AFAICT). It's analogous to how the words in an English sentence have to all be English words before it can be a gramatically-correct English sentence.

<space> is not closed. It should either have a following </space> inside the <msg>, be replaced with <space/> or if by saying you don't want it to be paresed you mean you want the actual text "<space>" in there, then you should encode it as such (i.e. <space>).

Answer 4

All XML tags have to be closed, either like <tag></tag> or <tag />.

If you want the <space> tag to be parsed as the text value of a tag, and not as a child tag, use < and > instead of < and >:

<space>

Answer 5

I would isolate the solution to your problem into a method and deal with it simply for now. After all, you may not have control over the correctness of the message content.

private static String getMessage(String msg){
    return msg.substring(msg.indexOf("<msg>")+5, msg.lastIndexOf("</msg>"));
}//method

You may enhance it later, as more use cases become available.

Edit: If someone puts an "msg" element in the content, then it still works