3

I have been asked to write a utility to search a large number of XML files for elements with a missing attribute. The department responsible cannot just make the attribute mandatory in the DTD because it will break hundreds of files. They want to edit them manually over a period of days/weeks.

I am writing a small command line tool in Scala 2.8.1. I will be using a "pull" parser so that I can keep my code functional-programming pure and run it multi-threaded.

I need the location of XML events. The API provided in Java 6 (javax.xml.stream. XMLStreamReader) has a method (getLocation()) that returns the line number of an event. I can use this to write messages that inform the user where to look for the missing attribute.

I would prefer to use the pull parser in scala.xml.pull.XMLEventReader, but it does not appear to offer location information.

Am I missing something? Is it somewhere else in the Scala API?

Daniel C. Sobral
  • 295,120
  • 86
  • 501
  • 681
Ralph
  • 31,584
  • 38
  • 145
  • 282
  • possible duplicate of [How to track the source line of an XML element?](http://stackoverflow.com/questions/4446137/how-to-track-the-source-line-of-an-xml-element) – Daniel C. Sobral Apr 19 '11 at 13:23
  • I'm amazed at how low the (original) duplicate of this question is on the list of related links. I'll try editing its title to make it closer. – Daniel C. Sobral Apr 19 '11 at 13:25
  • Ugh. I failed to see that what I thought was a the original question was a full XML loader, while this is concerned about _pull_ reader. Sorry. I'd retract my vote to close if it was possible. – Daniel C. Sobral Apr 19 '11 at 13:49
  • It is not a duplicate. The other question deals specifically with SAX parsers -- the API for Locator is clear. I am trying to use the Scala StAX parser. I even know how to do it in the `javax.xml.stream` package, just not the `scala.xml.pull` version. – Ralph Apr 19 '11 at 13:58

1 Answers1

3

As far as I can see, this is not provided by the XMLEventReader api.

It relies on io.Source and could provide access to the location but does not. I don't see an easy way to go around that as the object that has access to the position is private.

You may want to make your own copy of XMLEventReader that produces a custom XMLEvent with the position. The method to modify would be override def elemStart which has access to the position and could generate another EvPos(line:Int, column:Int) after each EvElemStart.

You may also consider using 2.9.RC1 as certain performance related bugs were fixed.

huynhjl
  • 41,520
  • 14
  • 105
  • 158