SWT - Diagrammatics

1
SWT - Diagrammatics

• Lecture 2/4 - Diagramming in Computer Science
• 27-April-2000

2
Review

• Definitions,
• Historical Facts,
• Maps, Geometry, Topological Diagrams, Science and
  Diagrams.
• Advantages of Diagramming,
• Diagram use across fields.

3
Overview

• Diagram Distinctions
• Diagram Taxonomies
• Diagram use in Computer Science
• Venn, Flowcharts, NSDs, Structure, Dataflow,
  ERDs, Cell and Arrows, State, Petri nets.
• Logic Gate Diagrams
• How to operate Theseus under CM

4
Diagram Distinctions

• Diagrams portray associations
• metric,
• topological and
• symbolic
• In computer systems, above the hardware level,
  Euclidean space is unimportant
• Much more common in software diagrams are
  associations in topological space.

5
Diagram Distinctions

• Within topological diagrams, associations can
  happen in three principal ways
• Adjoinment
• Linkage
• Containment

A
B
A
B
A
B
6
Diagram Taxonomies

• A Taxonomy is useful only to the ones that use
  it.
• A simple diagrammatic taxonomy can be based on
  the diagrammatic domain.
• Several researchers have focused on a variety of
  diagrammatic aspects and have proposed respective
  classifications.
• Efforts to categorise diagrams have created a
  large set of taxonomies.
• Alan. Blackwel has proposed a taxonomy of
  taxonomies
• http//www.mrc-cbu.cam.ac.uk/projects/twd/mypape
  rs/TwD98.html

7
A taxonomy of Diagram Taxonomies

• Dimensions of categorisation
• 1. The representation
• the graphic domain structure
• 2. The message
• the information domain
• 3. The relation between the representation and
  the message
• Pictorial correspondence
• 4. The process of interpreting and modifying
  representations
• Information processing tools
• 5. The context - convention
• Cultural conventions
• 6. The mental representation
• Interpersonal variation

8
Venn Diagrams

• Are related to circuits and logic gates
• Elements use containment to depict information
• Standard mathematical functions like a set of,
  a genuine subset of etc are depicted
• Easy to compare Venn Diagrams because of their
  visual representation

9
Venn Diagrams

• The two following Venn diagrams show that the
  next two functions are equal
• NOT(A OR B)
• (NOT(A)) AND (NOT(B))
• You can write a small java program to verify this
  if you want!

NOT(A OR B) (NOT(A)) AND
(NOT(B))
NOT A
A
NOT B
B
non shaded part is equal to double-shaded
part
10
Flowcharts

• They are topological, graph-based constructions
  that are often filled with program text.
• The control logic of the program is shown through
  simple branches and loops.
• They are usually generated by analysts as a
  specification to programmers, who then convert
  the charts into source code.
• However, for large systems, they can get messy,
  spanning in many pages as decisions have many
  branches.
• Goldstine claims he created the first flowchart
  for computers in 1947, while working with Von
  Neumann.

11
Flowcharts
start
get formatting parameters
Process
format disk
Display
If
format another?
yes
no
end
12
Nassi-Shneiderman (NS) Diagrams

• Hierarchy is shown using enclosure and
  adjacency
• Decisions are shown by splitting the lines
  into smaller, parallel boxes
• Loops are shown by enclosing a small box into
  a box labelled with the condition of the loop
• However, the early termination of loops (e.g.
  break) and multiple conditionals present problems
  for NS diagrams

13
Nassi-Shneiderman (NS) Diagrams
s1
if i1
true
false
while w1
s2
if i2
true
false
s4
s5
s3
s6
while w2
s7
s8
14
Structure Diagrams

• They are hierarchical, modular break downs of
  a program
• Between tree levels, links indicate what kind
  of information travels between levels
• They are usually represented by trees
• They are a part of the structural analysis
  activity, in which a system is partitioned in a
  top-down manner
• However, a multitude of labelled edges and nodes
  conveying lots of information can reduce
  readability

15
Structure Diagrams
calculate payroll
employee name
employee name
overtime
payment
record name
payment
date time
salary
records
get employee and pay record
calculate net pay
print pay cheque
employee category
expenses
tax
deductions
calculate deductions
calculate tax
16
Data-flow Diagrams

• Are oriented to flow-type operations
• Objects of data are shown in relationship to
  procedures
• No decision logic is shown
• The diagrams are most often used to model the
  flow of data
• However, they usually get large and complex and
  multiple-page spanning happens nearly always

17
Dataflow Diagrams
SALES DEPT
CUSTOMER FILE
CUSTOMER RECORD
ORDER FILE
GET CUSTOMER RECORD
NEW CUSTOMER INFORMATION
CREATE NEW CUSTOMER RECORD
NEW CUSTOMER RECORD
CHECK CUSTOMER CREDIT
ORDER INFORMATION
CUSTOMER INFORMATION
CUSTOMER FILE
COLLECT ALL ORDERS FOR CUSTOMER
INVALID CUSTOMER
CUSTOMER ORDER RECORDS
18
Entity-Relationship (ER) Diagrams
Attribute
Entity
Relationship

• The representation of data is often accomplished
  using diagrams
• ER diagrams are usually used to depict databases
• Extremely simple three types of nodes
• Entity and Relations form a graph and can have
  associated attributes and cardinality

19
Entity-Relationship (ER) Diagrams
Address
SSN
Salary
Name
Phone nr
Works-for
Name
Person
Company
Job title
HasMany
IsA
IsA
Manages
Department
Employee
Manager
Name
20
Software Level Charts

• At a higher level, the functions of a system are
  often thought as layers
• These diagrams work only on simple access schemes
• However, more complex schemes will result in
  a complex graph that cannot be
  represented with adjoining regions

Application
Motif
X-lib
DB API
Unix API
Unix Operating System
21
Cell and Arrow Diagrams

• In a combination of adjoinment and link-based
  conventions, data structures are often depicted
  as adjacent memory locations linked by pointers
• This is usually used for teaching purposes or for
  program documentation
• In programming, pointer manipulation of linked
  lists is shown as diagrams of Cell and Arrows

22
Cell and Arrow Diagrams
Insertion of a new element (element2) into a
linked list

element1
elementN
start
NULL
element1
elementN
(3)
start
(2)
NULL
element2
temp
(1)
23
State Transition Diagrams

• Well known in computer science as originating
  from the study of finite automata
• Are used for modelling a variety of event-based
  CS domains including parsing, user interface
  design, and circuit design
• At the application level, they represent
  transaction flows, appliance controls, marketing
  scripts etc
• With the exception of special symbols and
  terminal nodes, all nodes are treated the same.

24
State Transition Diagrams
Recognise if the pattern bc exists in string
aabaaabbabca
a
c
c
b
s1
s2
s3
a
b
25
Petri Nets

• Are closely related to data flow graphs
• The main distinction is that the graphs are
  bipartite, made up of a set of places and
  transitions
• Useful for concurrent, asynchronous, distributed,
  parallel, and nondeterministic systems.
• Each type of node can be further subdivided into
  subtypes

annihilator
terminal place
initial place
generator
trivial place
trivial transition
branching
splitting
collection
junction
26
Petri Nets
D available
Request D
Request D
D Ready
D Ready
Finished With D and P
Finished With D and P
Process
Process
Release D and P
Release D and P
P Ready
P Ready
Request P
Request P
P available
27
Logic Gates

• Logic gates are the components of logic circuits
• There are three main gates AND, OR and NOT
• An AND gate returns true (or 1) if both of its
  inputs are true
• An OR gate returns true (or 1) if either of its
  inputs is true
• A NOT gate returns the opposite of its input

AND OR
NOT
A B Q 0 0 0 0 1 0 1 0 0 1 1
1
A B Q 0 0 0 0 1 1 1 0 1 1 1
1
A Q 0 1 1 0
28
Logic Circuits

• Logic circuits are used in electronic devices.
• Formed by combining many logic gates
• More complex logic circuits are assembled from
  simpler ones which in turn are assembled from
  gates

A B Q 0 0 0 0 1 1 1 0 1 1 1
0
29
Combining gates Together

• A N input gate can be constructed by placing N
  gates in a special configuration

x
x
y
y
o
z
o
z
x
x
y
o
y
o
z
z
30
Other Logic Gates
nor
A B Q 0 0 1 0 1 0 1 0 0 1 1 0

• Include

nand
A B Q 0 0 1 0 1 1 1 0 1 1 1 0
xor
A B Q 0 0 0 0 1 1 1 0 1 1 1 0
buffer
xnor
A B Q 0 0 1 0 1 0 1 0 0 1 1 1
31
Analog Circuit Diagrams

• Represent electronic design schemata
• Strict Notation - Graph-Network topology
• No direction

32
The end of lecture 2