Re: [Webtest] setInputField by ID - Lothar =?unknown-8bit?Q?M=E4rkle?= - com.canoo.lists.webtest

11 messages in com.canoo.lists.webtestRe: [Webtest] setInputField by ID

From	Sent On	Attachments
Maxim Annyuk	08 Dec 2005 06:44
Lothar Märkle	08 Dec 2005 07:59
Marc Guillemot	08 Dec 2005 08:36
Maxim Annyuk	09 Dec 2005 09:05
Dierk Koenig	09 Dec 2005 09:49
Edgar P Dollin	10 Dec 2005 06:32
Dierk Koenig	10 Dec 2005 07:34
Lothar =?unknown-8bit?Q?M=E4rkle?=	10 Dec 2005 13:13
Paul King	11 Dec 2005 02:59
Marc Guillemot	12 Dec 2005 01:36
Denis N. Antonioli	12 Dec 2005 06:48

Subject:	Re: [Webtest] setInputField by ID
From:	Lothar =?unknown-8bit?Q?M=E4rkle?= (loth...@netpioneer.de)
Date:	12/10/2005 01:13:57 PM
List:	com.canoo.lists.webtest

Hello Maxim,

On Fri, Dec 09, 2005 at 08:05:36PM +0300, Maxim Annyuk wrote:

LM> Following the html dtd, the name attribute is #IMPLIED, so it is valid LM> to omit it. But how useful is a input field without a name?

Sometimes job can be done by JavaScript and there is no necessity to specify "name" attribute. Especially if you don't want this value to be sent to server when the form is submitted.

I did not thought of this. Do all browsers omit the values if the form is send? That could be a useful trick. But you can't control your input, so the server side must be able to work with the (unwanted) values anyway.

Of course I can add "name" attribute without problems, but it's always bad when a test influences the program. It's like the tail wags the dog.

As Dirk already noted, the ideas of test driven development influence the code in a positive way. That your application is testable is a good thing tm:). In your case, improving webtest to support selection by id, or changing the code and add name attributes will both work to write the test.

I'd better improve the test. I hope once I'll find time for this.

setInputField by id would be cool, no question.