Project acronym: QUESTION-HOW
Project Full Title:Quality Engineering
Solutions via Tools, Information and Outreach for the New Highly-enriched
Offerings from W3C: Evolving the Web in Europe
Project/Contract No. IST-2000-28767
Workpackage 1, Deliverable D1.3
Project Manager: Daniel Dardailler
<danield@w3.org>
Author of this document: Michel Bercovier , Danny Dolev
Created: 29 August 2002. Last updated: 29 August 2002.
Forms are the current customer – organization interface. Forms are present wherever there is a need of dialog between a user and an organization. Most sites offering a service to users have multiple forms to allow for registration, information retrieval and gathering , requests, shopping, etc…etc… Hence we were able to gather a collection of existing forms , representing complex operations such as job request, bank operations and so on… Users request to have continuous access to the organizational interfaces by means of such forms.
The necessity to be clear and effective ( a user will drop out if the operations and requests are tedious and difficult to understand) means that such forms will have a complex layout .
Organizations need efficient and powerful tool to handle forms. This is sometimes done by special purpose form generators or by complex layout of the web pages.
As a consequence thecurrent design of Web forms doesn't separate the purpose / content / actions from the presentation of a form.
XForms, based on XML, in contrast, are comprised of separate sections that describe what the form does, and how the form looks. This should allow for flexible presentation options, including classic XHTML forms, to be attached to an XML form definition.
Since XForms intend to become the bridge across these expectations validation of XForms’ capabilities is important.
XFormis still a specification at the stage of a working draft XForms 1.0 now for public review. None of the available browsers have an XHTML implementation of Xforms. There exists a few spetial pupose browsers currently devellpoed for XML and Xforms, but all of them are still at a beta stage. Hence we had problem using most of the clients that are listed in XForms - the next generation of Web forms . The only one we did succeed to use is the X-Smile browser http://www.x-smiles.org/ We expect the same difficulties on the Web Server software side.
The number of available XForm examples is VERY limited , and in general one find only the few given in the Xform working draft. Hence we decided to start from a simple form taken from the W3.org site : the step4 of the registration to the AC meeting of the W3C. It can be found at http://www.w3c.org.il/registrationW3C1Simplifiee.html and a copy is provided as example in this report
This has the advantage of being written in HTML 4.01 , hence follows a standard and its rules.
The fist problem we encountered was the use of the break command, given in a format compatible with XHTML ie <br /> , it was generating fatal errors in the XForm browser! Hence before processing the form to build an XForm version some editing had to be made. Moreover the variables and values of the form entries are defined in the HTML body at the same time as the actual layout , hence no separation of content and forms.
To map such a form to XForm one had to construct first the xform:model, where the actual variables and actions are defined. This has to be down by analyzing both source and original web pages to extract the necessary information. The resulting source text file can be viewed as a postscript file The method could be automatic if the original web pages had been properly constructed, but still it would need at least a double pass parsing to construct the Xform file.
One can view and compare original HTML page and the resulting Xform page under the X-Smile browser .
A large collection of forms written in various versions of HTML have been collected. It can be downloaded from
http://www.w3c.org.il/English-Forms.zip
Take a typical example of a form generated by a Form Generator , in pdf format, the corresponding source shows that the HTML generated is non standard and the HTML body is made of interleaved tables . Starting from such a source necessitates its editing to have a compliant HTML file. Next the table imbedding must be understood to build s corresponding layout by style sheets and XForm.
It is essential to manipulate the equivalent of tables under Xforms. This is not an easy step ( no tutorial, no examples). Moreover the Xform enabled browser being beta releases , some of the problems we encounter can be due to the software's lack of maturity . And not all recommendations are yet implemented. We will analyse and test how such tables , and their imbedding could be done under Xform.
Using the forms we have collected together with some Form editor characteristics we will try and identify repetitive structures to create an initial XML based Xform dictionary and see how it can map to the most important form editing requirements.
Delay due to late contract signature.