WWW.DISSERS.RU

БЕСПЛАТНАЯ ЭЛЕКТРОННАЯ БИБЛИОТЕКА

   Добро пожаловать!

Pages:     | 1 |   ...   | 45 | 46 || 48 | 49 |   ...   | 54 |

Finally, this led us to refactor all design, and instead generating all the mentioned resources, two meta-servlets where produced, so that one of them is capable to draw any form and the other one is capable to receive (after submitting) any kind of form. This reduces sensibly the publishing process, now only access to the database is required, and all resources are stored there. The database is accessed through the corporate intranet.

Case Study Probably the best way to describe how GEN works and the simplicity of the process to generate the new application to manage a survey is to follow an example.

First step: translating a PDF survey into HTML The first of all, we need the file containing the survey designed by the corresponding department. Usually surveys have been translated into PDF, anyway this is an immediate process, usually covered by Adobe Acrobat Writer or Distiller, as well as by any tool complaining the standard.

Information Systems This translation is based on an external tool, so we do not mind how the translation is done, the only important issue is the quality of the HTML we get after the translation. GEN only needs as input a HTML complaining the W3C standard. We always avoid HTML resulting from Microsoft Word exports, FrontPage, etc.

In this case study, we are using a translation tool which converts PDF into an HTML composed of layers. Each component of the document is set into a different layer, and the absolute position for each layer is written in the code. This is very useful for future processing of the template, while painting the components in the proper location over their relative texts and tables.

Second step: design the form for a survey Now GEN user shall define a form to capture the data corresponding to the survey. The HTML from the previous step is considered as a simple background to help the user drawing all components. This is the main task while using the editor, but also some other functions are to be completed: buttons to send, save or step through the forms, page breaks for viewing and generating PDF receipts, or dependencies and constraints specification.

First of all, the typical screen where user is going to define the working directory, survey name and type. GEN allows defining multilingual surveys, so that the language specification is also important for further publishing.

Each survey may be also converted in a template that may be used to create compositions in which a survey is composed of modules that are really survey templates. This is very useful to define headers or footers, or to include forms common to many applications.

Figure 2.- The survey is loaded in the editor to create the form Once the survey is loaded in the editor, the user may define the components and its location over the background. The layer information from the HTML obtained in the first step is very important now; when a user draws a component in the form, the editor calculates the position trying to associate the component to a field in the survey, matching them depending on the proximity.

Figure 3.- Drawing a component from the toolbar Fourth International Conference I.TECH 2006 Components are managed as usual in an IDE, including the possibility of copy, paste, align vertically or horizontally, size, resize or set equal size for several of them, defining default values for each component type, etc.

In some cases it is very interesting to define a domain of values for some components, especially in case of combo boxes or lists, in the figure below a example on how to define a fixed domain for states is shown. The possibility of recovering dynamically the domain from a table in a database has been also implemented.

Figure 4.- Constraint definition Another interesting functionality while defining surveys is to set some kind of constraint to the values of the different fields. The user may define both individual and group constraints. Individual constraints are to be applied to the value of a field, while group constraints are applied to the set of values of some fields. In any case, the constraints may obly to fill the field (mandatory) or define restrictions on the value of the field if any, through formula specification, as shown in the figure below. In the combo box the user may select some common predefined functions.

Figure 5.- Language selection. GEN generates multilingual applications Information Systems After some time defining the form associated to a survey, the user may have added and removed several components (in fact, usually, dozens of them). Each new component is named automatically by the editor and introduced in the tab order. The user may alter both, the given name to make the application comprehensive (very recommendable) and also the tab order, so that the final user may navigate through the fields with the tab control, which is one of the most accepted (and requested) features from final users.

Not only the form to publish the survey is important, but also to provide the users an identical aspect and information to reduce the difference to the hard copy surveys. In the same terms the background is the same that in that case, the designer provides the chance to add help buttons to the form, and then these buttons may be linked to a help file in HTML format that usually contains the same help provided to non Internet users.

The results: an external user fulfilling a survey When the process is finally finished, the user gets a result which may be considered in two perspectives:

- An application that supports the publication and collection of data from the web, on a surprisingly low cost and effort, and which is integrated in the portal of the organization.

- A service for the information providers, those are obliged by law to compliment the surveys, and may now carry out this task easily.

The applications generated by GEN are integrated in the web site through shared tables in an Oracle database, and the development of a section in which the user may select the survey in which he is interested. The access to the section is based on the typical user and password that has been replaced by an order number and control digit. This information is sent to each organization together with customized instructions and paper surveys via mail as official notifications, reusing the existing infrastructure and the old procedure, which is mandatory to maintain.

When a survey is generated, the user may define whether the survey shall exist also in test mode. This allows final users to practice with real surveys without writing results or submitting information. The only difference between tests and real surveys is a parameter switching between them.

The surveys can be displayed in all Spanish official languages as mandatory by law, but the data are considered to correspond to the same survey, this is logical until a problem arises: the staff in charge of collecting data and extracting statistical information do not use to speak all of them. It is considered a great lack not for the application but for the legal procedure.

In the figure 6 a fragment of the web showing a survey is presented. Note the final alignment of the fields, the different component types and the buttons in the bottom, offering the functions of Save, Send form, Cancel and Show errors.

Before sending the final answer to a survey, the user may save any number of times, and before submitting may query the errors in the different fields in order to make corrections. Sending a correct answer is mandatory by law.

The final reply of a user to a survey is stored in XML format into the database, and will be later exported to the analysis application.

This is the common process for most users, however, a number of enterprises, mainly holdings, do need to fill lots of surveys that they may automate by using their information systems. A complementary application has been developed allowing this kind of users replying massively to the surveys through bulk loads. This application, called G2G, receives XML files containing several surveys, validates the data using the XForms from GEN, after formatting incorrect data or truncating data too long for the precision in database.

Fourth International Conference I.TECH 2006 Figure 6. A piece of the final presentation of a survey on commerce Conclusion Sometimes the public organisations shall face political commitments with little help from the Government, same budgets and staff, allowing the development of beautiful projects in which technology makes life easier. Probably this project had lost the chance if there where enough staff to develop each application as usual software projects.

A tool has been developed that allows a non technical user, with no knowledge on software architectures, design, HTML, nor Java, struts or databases, carry out the task of creating applications. Of course this does not replaces technical staff, but get them to develop only technical high level tasks.

The fact is that a complex technology, that has been tested by Technical University of Madrid, and that shuns from the typical theoretical aspects from code generation, trying to open a new spectrum of possibilities, has been applied successfully to a specific application domain demonstrating that the approach is good and feasible.

For public administrations, the project allows a rise in the corporate image, and the observance of the Lisbon agreements in one of the most difficult aspects it was facing to.

The Spanish National Statistics Institute has been capable to meet the goals described in its services whitepaper, minimizing the impact in the staff and budget, and reducing sensitively the time to market for each survey, which is available in Internet in less than a week, so that the complete process increases a minimum percentage from the previous one, with no digital survey. Also the time in which all data has been collected is lower, because there is no need to transcript all data to a computer.

Information Systems Collaborators that fill in the forms, now have the chance to select the preferred choice: digital or paper surveys, knowing that digital ones provide the proper mechanisms to automate the calculations, report possible errors reducing the time to finish the job, etc.

The success in figures: in the first two months more than 10% of the surveys where submitted via Internet, more than 600 per day are being received, and 7 multilingual surveys have been generated; 168.433 organizations are working with the generated applications.

Bibliography [Alves2002] Alves, L., von Staa, A., 2002. A construction process for artifact generators using a CASE Tool. Proceedings of Workshop on Generative Programming 2002. Austin, Texas, USA, pp. 7-10.

[Arisholm1998] Arisholm, E. et al. 1998. Incorporating Rapid User Interface Prototyping in Object-Oriented Analysis and Design with Genova. The Eighth Nordic Workshop on Programming Environment Research.

[Batory 1994] Batory, et al. 1994. The GenVoca Model of Software-System Generators. IEEE Software, 0704-7459/94.

[Boehm2000] Jongmoon, B., Boehm, B., 2000. Empirical Analysis of CASE Tool Effects on Software Development Effort.

Technical Report University of South Carolina. Center for Software Engineering.

[Cuvillo2004a] Cuvillo, C. et al. 2004. Generation tool for DBMS focused applications. Applied Computing 2004. Lisbon.

Portugal.

[Cuvillo2004b] Cuvillo, C. et al. 2004. Multiplatform web applications generated from relational data models. (In spanish).

CISCI 2004. Florida, USA.

[Calejo2002] Calejo, M. et al. 2002. Web Application Maker - A model based approach to web database development. 6th International Conference on Enterprise Information Systems.

[Grnbk 1991] Grnbk, K.et al. 1991. ApplBuilder - an Object-Oriented Application Generator Supporting Rapid Prototyping. The 4th International Conference on Software Engineering & Its Applications.

[Schmidt2000] Schmidt, D., et al., 2000. POSA2: Pattern -Oriented Software Architecture: Patterns for Concurrent and Networked Objects.

Authors' Information Hector Garcia - Contrated Professor. Technical University of Madrid. E.U. Informtica. Ctra. de Valencia Km. 7.

E28031 Madrid. e-mail: hgarcia@eui.upm.es Carlos del Cuvillo - Associate Professor. Technical University of Madrid. E.U. Informtica. Ctra. de Valencia Km. 7. E28031 Madrid. e-mail: ccuvillo@eui.upm.es Diego Perez - Consultant. Technical University of Madrid. E.U. Informtica. Ctra. de Valencia Km. 7. EMadrid. e-mail: dperez@tdi.eui.upm.es Borja Lazaro - Technician, Group leader. Technical University of Madrid. E.U. Informtica. Ctra. de Valencia Km.

7. E28031 Madrid. e-mail: blazaro@eui.upm.es Alfredo Bermudez - Project manager. National Institute for Statistics. c/ Rosario Pino 14-16. E28071 Madrid.

e-mail: abc@ine.es Fourth International Conference I.TECH 2006 СИСТЕМА АВТОМАТИЗИРОВАННОЙ ПОДГОТОВКИ ВЫВОДОВ О ФИНАНСОВОМ СОСТОЯНИИ АКЦИОНЕРНЫХ ОБЩЕСТВ Григорий Н. Гнатиенко, Николай Н. Маляр Аннотация: Описывается программное обеспечение автоматизированной подготовки эксперессанализа финансового состояния предприятий. Система рассчитана на использование ее при осуществлении контрольно-аналитической деятельности по управлению государственными корпоративными правами. Приводятся области применения финансового анализа, структура и основные характеристики системы. Перечислены основные результаты применения описанного программного обеспечения.

Ключевые слова: анализ финансового состояния, государственные корпоративные прав, коэффициент, дивиденды.

Введение Для обоснованного принятия решений при управлении государственными корпоративными правами (ГКП) возникает необходимость в проведении оперативного анализа финансово-хозяйственной деятельности акционерных обществ. Важная роль финансового анализа при управлении ГКП объясняется тем, что в том числе и от качества предшествующего анализа зависит качество основанных на нем управленческих решений ([Гнатиенко, 1999а], [Гнатиенко, 1999б]).

Pages:     | 1 |   ...   | 45 | 46 || 48 | 49 |   ...   | 54 |



© 2011 www.dissers.ru - «Бесплатная электронная библиотека»

Материалы этого сайта размещены для ознакомления, все права принадлежат их авторам.
Если Вы не согласны с тем, что Ваш материал размещён на этом сайте, пожалуйста, напишите нам, мы в течении 1-2 рабочих дней удалим его.