| Home | Search | About Us | Contact Us | Our Products | Documentation | Resources | Login |
May 16 2008 10:10 UTC
Tutorials
- CSV upload and download website
- Convert XML to CSV and back again
- Look up countries from IP addresses
Quick Links
Questions
Got a question for us?
Just Ask!
Quality
$15 Gift Certificate for every bug you find.
Payment Methods
Bookmarks
Previous Competition Winners
- Jose Ignacio Santa Cruz G. and Gustavo Valdés Aracena, Competition #3, June 26th, 2006
- Roman Nowak, Competition #2, March 20th, 2006
- Andre Luepke, Competition #1, March 6th, 2006
Back to the main competition page.
Competition #3, June 26th, 2006
XML Manager Winner:
Jose Ignacio Santa Cruz G. and Gustavo Valdés Aracena
The data to extract:
| http://www.example.com/first | First |
| http://www.example.com/ipsum | ipsum |
| http://www.example.com/amet | amet |
| http://www.example.com/semper | semper |
| http://www.example.com/second | Second |
| http://www.example.com/nunc | nunc |
...from this XML:
<html> <body> <div> <h1>First Paragraph</h1> <a href="http://www.example.com/first">First</a> <p>Lorem <a href="http://www.example.com/ipsum">ipsum</a> dolor sit amet, consectetuer adipiscing elit. Etiam sit <a href="http://www.example.com/amet">amet</a> diam vestibulum diam posuere suscipit. Nam convallis.</p> <p>Phasellus <a href="http://www.example.com/semper">semper</a>, nisl ac malesuada sodales, libero sapien vestibulum libero, quis cursus enim ligula sit amet justo.</p> </div> <div> <h1>Second Paragraph</h1> <a href="http://www.example.com/second">Second</a> <p>Donec iaculis, enim sit amet lobortis dapibus, felis <a href="http://www.example.com/nunc">nunc</a> pulvinar nibh, id varius eros ante et nunc.</p> </div> </body> </html>And the winning solutions were:
RECORD: //a FIELD 1: @href FIELD 2: .
Comments
This one was from Jose Ignacio Santa Cruz G. This is the most efficient solution. The principle idea is that you can ignore almost all the structure
and just go directly for the a elements. The //element syntax is extremely useful for this type of problem. Add this to XML Manager's ability to stream the XML, and you have a very powerful and fast data extraction engine.
RECORD: /html/body/div[*]//a FIELD 1: @href FIELD 2: .
This one was from Gustavo Valdés Aracena. In this solution, the record XPath is used to specify the exact subset of a elements to extract. This solution is useful when you need to exclude elements that are not relevant. No a elements outside of a div will be returned.
One other minor point: the predicate on the div element is not strictly necessary. The record XPath /html/body/div//a will work just as well.
Competition #2, March 20th, 2006
XML Manager Winner:
Roman Nowak
The challenge was to extract this data...
| Isaac | Newton | 051 8706351 | 1 Apple Avenue. Lincolnshire. |
| Micheal | Faraday | 051 8706352 | 2 Spark Street. Surrey. |
| Albert | Einstein | 051 8706353 | 3 Relative Road. Princeton. |
...from this XML:
<addressbook>
<person firstname="Isaac" lastname="Newton">
<phone>051 8706351</phone>
<address>
1 Apple Avenue.
Lincolnshire.
</address>
</person>
<person firstname="Micheal" lastname="Faraday">
<phone>051 8706352</phone>
<address>
2 Spark Street.
Surrey.
</address>
</person>
<person firstname="Albert" lastname="Einstein">
<phone>051 8706353</phone>
<address>
3 Relative Road.
Princeton.
</address>
</person>
</addressbook>
And the winning solution was:
RECORD: /addressbook/person FIELD 1: @firstname FIELD 2: @lastname FIELD 3: phone FIELD 4: address
Comments
This challenge demonstrates a very simple use case for XML Manager — the direct extract of data fields from
simple XML records. Some of the data is in attributes and some of data is in element text.
There is one additional refinement to this solution. In a field path, the name of an element returns all the text inside that element. In this case,
address, say, was used to return the address text. However, it returnds all the text between
the start and end address tags. This includes all the newlines and empty space. It would be nice to get rid of this,
and XML Manager provides a way to this: rb:trim(address). This performs a java.lang.String.trim() operation on
element text, removing all the surrounding whitespace. This function is not standard XPath,
but is provided by XML Manager to make things a bit easier.
Competition #1, March 6th, 2006
XML Manager Winner:
Andre Luepke
The challenge was to extract this data...
| Bugs | Bunny |
| Homer | Simpson |
| Fred | Flintstone |
...from this XML:
<table>
<tr>
<td>Bugs</td><td>Bunny</td>
</tr>
<tr>
<td>Homer</td><td>Simpson</td>
</tr>
<tr>
<td>Fred</td><td>Flintstone</td>
</tr>
</table>
And the winning solution was:
RECORD: /table/tr FIELD 1: td[1]/text() FIELD 2: td[2]/text()
Comments
This solution shows how you can use the
position of an element to get at the right data. The XML document uses
the elements of an HTML table, so td elements are
repeated. To get at both of them, you need to use the XPath position
predicate: [x]. In XPath positions start at one,
not zero, so you watch out for that as well.
The winning entry uses the text() function to get the
text of the td elements. In fact, you don't actually need
to use this function in this case, the expression td[1],
for example, will work just as well.
Copyright © 2004-2008 Ricebridge. All Rights Reserved.