Function Oriented Design

Function-Oriented Design Structured Design Methodology Methodology Structured design methodology views every software system as having some inputs that are converted into the desired output by the software system. During design, structured design methodology aims to control and inuence the structure of the nal program. It aims at designing a system so that the programs implementing the system have a nice hierarchical structure, with functionally cohesive modules and as few interconnections between modules as possible. Factoring is the process of decomposing a module so that the bulk of its work will be done by its subordinates.  system is said to be completely factored if all the actual processing is performed by the bottom!level atomic modules and the non! atomic modules largely perform "obs of control and coordination.  #he overall strategy is to identify identify the input and output streams, and the primary primary transformations that have to be performed to produce the outputs. #his strategy involves four steps$ i. ii. ii. iii. iii. iv. iv.

%esta estate te the the pr proble oblem m as as a dat data a ow ow dia diagr gram am Iden Identi tify fy the the inpu inputt and and out outpu putt ele eleme ment nts s Firs irst!le t!lev vel Facto actori ring ng Facto actori ring ng of inp input ut,, outpu output, t, and and trans transfo form rm bran branch ches es

(i)Restate (i) Restate the Problem as a Data Flow Diagram  #he DFD shows the ma"or transforms that the software software will have and how the data will ow through di&erent transforms. For e'ample, consider the problem of determining the number of di&erent words in a le. In this problem, p roblem, there is one input data stream (the input le) and an output data stream (count of di&erent words). #o #o transform the input i nput into the desired output, rst we form a list of all the words in the le. #he list is then sorted so as to make identication of di&erent words easier. easier. #he sorted list is then used to count the number of di&erent words. #he output of this transform is the desired count, which is then printed.

(ii) (ii)

Identi Identify fy the the Most Most bstra bstract ct In!u In!utt and and Out Out!ut !ut Data Data "lem "lemen ents ts

 #he goal of this second step is to separate the transforms from the data ow diagram that convert the input or output into the desired format from the ones that perform actual transformation.  #he most abstract input (*I) data elements are those data elements that are the furthest removed from the physical inputs but can still be considered as inputs to the system.  #he most abstract input data elements are recogni+ed by starting from the physical inputs and traveling towards the outputs in the data ow diagram, until the data elements are reached that can longer be considered incoming.  #he most abstract output data elements can be identied by starting from the outputs and traveling towards the inputs. #hey are the data elements that are furthest removed from the actual outputs but can still be considered outgoing. entral transforms perform basic transformations for the system, taking the most abstract input and transforming it into the most abstract output.

(iii)

First-#e$el Factoring -e rst specify a main module whose purpose is to invoke subordinated. #he main module is therefore a coordinating module. For each mo st abstract input data item, we specify an immediate subordinate module to the main module. For each most abstract output data item, a subordinate module that is an output module that accepts data from the main module is specied. For each central transform, we specify a module subordinate to the main module. #hese modules are transform modules. #hey accept data from the main module and return the appropriate data back to the main module.

(i$)

Factoring of In!ut% Out!ut% and &ransform 'ranches  #o factor an input module, the transform in the data ow diagram that produced the data item, is now treated as the central transform. #he process

that was performed for rst!level factoring is repeated here with the central transform, with the input module being treated as the main module.