Create Operator or Output with Groovy

darktemptation

My Problem is quite simple:
I want to load a Webpage with Groovy in a Text-Operator (i.e. Document) and then extract certain attributes (e.g. all

-Texts).

Now I can fetch the HTML from a page with

"http://rapid-i.com".toURL().text

But the Script Operator does not return anything to the output, even when I use the "return" from Groovy.

Can somebody give me a hint?

steffen

Hm ... maybe I didnt grasp the problem, but rapidminer cannot deal with arbitrary groovy types. You have to convert the output into an IO-type rapidminer does understand.

... and "no", I do not know which one and how :-\.

haddock

Hi there,

I think you were getting nothing back because you were being re-directed, the following ( with the '/' on the end of the URL ) produces words of infinite beauty, wisdom, etc..etc..

//R5 Forum/groovout











http://rapid-i.com/".toURL().text)"/>

darktemptation

Thanks for the hints so far.

But@haddock:

Even if I try your setting, the log tells me that there is nothing delivered to the Result 1 Port.

And if you like you can also take another page where you aren't redirected.

"http://www.aboutgroovy.com".toURL().text

So the problem is still how to get the fetched result to the the output-port, with a data type that is known by RM.
Maybe then the question must be, how can I create an IOObject with a String/Text Attribute in Groovy (as Steffen suggest)?

haddock

Ooops, I thought you weren't getting anything back. If you want to manipulate the contents you can with macros, and so to logs and example, like this...

HTML







http://rapid-i.com/" .toURL () .text.substring (0,10));
//defhtml="http://rapid-i.com/" .toURL () .text.substring (0,10)"/>

There's some Groovy stuff on the Wiki which may be helpful.

darktemptation

Thanks haddock

This goes into the right direction, for what I looked. After a few modifications I found the way how to do it.

HTML







http://rapid-i.com/".toURL().text);
//defhtml="http://rapid-i.com/" .toURL () .text.substring (0,10)"/>

So I just used the scripting Operator to define the macro and used the macro as parameter for the "Create Document" Operator. This is a nice solution I think.

Now there will be the next step, how to find all the "

land

Hi,
a detailed description how to use the Groovy Operator and how to return things there is given in the How to Extend RapidMiner tutorial.

Greetings,
Sebastian

darktemptation

Hi Sebastian

After a short search I found the tutorial in the Shop
http://rapid-i.com/component/page,shop.product_details/flypage,flypage.tpl/product_id,52/category_id,5/option,com_virtuemart/Itemid,180/

I guess you mean this.
But when I read there the detailed description there is written:

Together with the white paper you receive two projects for Eclipse. The one is an extension containing all examples covered in thebookand the other is a template for building own Extensions with Eclipse.

你们在谈论什么书?有啊nly a White Paper to purchase/download.
And may it be possible to get a short preview of the content, at least this White Paper costs €40 (resp. CHF 60.-), then I want to know exactly what is covered in there and maybe see an example.

Best regards

Darktemptation

land

Hi,
this "book" is the white paper. The whitepaper has 45 pages in DIN A4, so other people would format it differently and call it a book.
Sorry, but what examples do you mean? It covers everything you need to write your own extensions. In fact we are using it internally to teach new colleges...There's everything in I know about extensions and I wrote the published ones...

Greetings,
Sebastian