A New CC Course

1
A New CC Course Based on Cooperation with Linz

• Vladimir Kurbalija, Mirjana Ivanovic
• Department of Mathematics and Informatics
• University of Novi Sad
• Serbia and Montenegro

2
Agenda

• Current CC Course in Novi Sad
• Lectures
• Exercises
• CC Course in Linz
• Lectures
• Exercises
• Conclusion

3
CC Course in Novi Sad

• One of core software courses at Computer Science
  directions
• 7th semester CC1 (obligatory), 8th semester CC2
  (elective), for students of
• Computer Science,
• Business Computer Science,
• Teaching of Computer Science

4
CC Course in Novi Sad - Lectures

• Practical approach
• Development of Pascal- compiler
• Subset of Pascal language
• data types Boolean and integer standard types,
  arrays and fixed records as structured types
• basic statements assignment statement, procedure
  call, if and while statements
• Standard input/output (read and write)
  procedures, user defined procedures including
  recursion.

5
CC Course in Novi Sad - Lectures

• The implementation
• recursive-descent manner for syntax analysis
• code generation for abstract P machine
• This approach was interesting 10 years ago
• Now, when new languages and tools are appeared we
  need modernisation!!!

6
CC Course in Novi Sad - Exercises

• Students repeat and train practical skils gained
  during lectures
• Several independent tasks small grammars
• Implementation language is Modula-2
• Compiler generator Coco/R

7
CC Course in Novi Sad - Exercises

• Tasks
• Lexical and syntax analysis and some parts of
  semantic analysis using Coco/R
• Hand written parsers (LA SA)
• Hand written parsers with semantic analysis and
  rarely with code generation or interpretation
• Some algorithms on grammars (memory organisation,
  checking consistency, computing first and follow
  sets)

8
Agenda

• Current CC Course in Novi Sad
• Lectures
• Exercises
• CC Course in Linz
• Lectures
• Exercises
• Conclusion

9
CC Course in Linz - Lectures

• One semester course held in English
• The same course at University of Oxford
• Goals of the course
• acquire the practical skills to write a simple
  compiler for an imperative programming language
• understand the concepts of scanning, parsing,
  name management in nested scopes, and code
  generation.
• learn to transfer the skills also to general
  software engineering tasks

10
CC Course in Linz - Lectures

• More theoretical approach
• Course goes through all phases of a compiler
  writing
• Shows the theoretical concepts underlying each
  phase as well as how to implement it efficiently
• Examples MicroJava compiler in Java, target
  language subset of Java byte code
• Example (05.SymbolTable.ppt, sl. 27-36)

11
5. Symbol Table 5.1 Overview 5.2 Objects 5.3 Sc
opes 5.4 Types 5.5 Universe
12
Types
Every object has a type with the following
properties

• size (in MicroJava always 4 bytes)
• structure (fields for classes, element type for
  arrays, ...)

13
Structure Nodes for Primitive Types
int a, b char c
kind name type next val adr level nPars locals
Var "a" - 0 0 - -
Var "b" - 1 0 - -
Var "c" - - 2 0 - -
object node
structure node
kind elemType nFields fields
Int - - -
Char - - -
There is just one structure node for int in the
whole symbol table. It is referenced by all
objects of type int. The same is true for
structure nodes of type char.
14
Structure Nodes for Arrays
int a int b
kind name type next val adr level nPars locals
Var "a" - 0 0 - -
Var "b" - - 1 0 - -
kind elemType nFields fields
Arr - -
Int - - -
The length of an array is statically unknown. It
is stored in the array at run time.
15
Structure Nodes for Classes
class C int x int y int z C v
kind name type next val adr level nPars locals
Type "C" - - - - -
Var "v" - - 0 0 - -
Int - - -
kind elemType nFields fields
Class - 3
kind name type next val adr level nPars locals
Var "x" - 0 - - -
Var "y" - 1 - - -
Var "z" - 2 - - -
16
Type Compatibility Name Equivalence
Two types are the same if they are represented by
the same type node (i.e. if they are denoted by
the same type name)
Type "T" ...
Var "a" ...
Var "b" ...
class T ... T a T b
Class - ... ...
The types of a and b are the same Name
equivalence is used in Java, C/C/C, Pascal,
..., MicroJava Exception In Java (and
MicroJava) two array types are the same if they
have the same element types!
17
Type Compatibility Structural Equivalence
Two types are the same if they have the same
structure (i.e. the same fields of the same
types, the same element type, ...)
Type "T1" ...
Var "x" ...
Type "T2" ...
Var "y" ...
class T1 int a, b class T2 int c, d T1
x T2 y
Class - 2
Class - 2
Var "a" ...
Var "b" ...
Var "c" ...
Var "d" ...
Int - - -
The types of x and y are the same (but not in
MicroJava!) Structural equivalence is used in
Modula-3 but not in MicroJava and in most other
languages!
18
Methods for Checking Type Compatibility
class Struct ... // checks if two types are
compatible (e.g. in compare operations) public
boolean compatibleWith (Struct other) return
(this.equals(other) this Tab.nullType
other.isRefType() other Tab.nullType
this.isRefType()) // checks if "this" is
assignable to "other" public boolean
assignableTo (Struct other) return
(this.equals(other) this Tab.nullType
other.isRefType()) // checks if two types
are the same (structural equivalence for arrays,
name equivalence otherwise) public boolean
equals (Struct other) if (kind
Arr) return (other.kind Arr
(other.elemType elemType other.elemType
Tab.noType)) else return (other
this) public boolean isRefType() return
kind Class kind Arr
necessary because of standard function len(arr)
19
Solving LL(1) Conflicts with the Symbol Table
Method syntax in MicroJava
void foo() int a a 0 ...
20
Solving the Conflict With Semantic Information
private static void Block() check(lbrace) for
() if (NextTokenIsType())
VarDecl() else if (sym ? First(Statement))
Statement() else if (sym ? rbrace, eof)
break else error("...") ... recover
... check(rbrace)
Block "" VarDecl Statement "".
private static boolean NextTokenIsType() if
(sym ! ident) return false Obj obj
Tab.find(la.str) return obj.kind Obj.Type
21
CC Course in Linz Exercises

• Students should acquire practical skills in
  compiler writing
• One (big) project divided in smaller subtasks
• Students should write a small compiler for a
  Java-like language - MicroJava
• The implementation language is also Java

22
CC Course in Linz Exercises

• The project consists of three levels
• Level 1 implementation of a scanner and a
  parser for the language MicroJava, error handling
• Level 2 - deals with symbol table handling and
  type checking
• Level 3 - deals with code generation for the
  MicroJava

23
CC Course in Linz Exercises

• Study material
• Teaching material (slides)
• Description of the project
• Specification of MicroJava language (tokens,
  language grammar, semantic and context
  constraints)
• Specification of MicroJava virtual machine
  (similar but simpler than Java VM) Memory
  layout and Instruction set
• Specification of object file format

24
CC Course in Linz Exercises

• Sources
• Frameworks of all classes needed for compiler
• Complete implementation of MicroJava VM
• Samples of MicroJava programs (for testing)

program P void main() int i i
0 while (i lt 5) print(i) i
i 1
25

• program Eratos
• char sieve
• int max // maximum prime to be found
• int npp // numbers per page
• void put(int x)
• if (npp 10) print(chr(13))
  print(chr(10)) npp 0
• print(x, 5)
• npp
• void found(int x)
• int i
• put(x)
• i x
• while (i lt max) sievei 'o' i i x
• void main()

26
Agenda

• Current CC Course in Novi Sad
• Lectures
• Exercises
• CC Course in Linz
• Lectures
• Exercises
• Conclusion

27
Conclusion

• Contents of both courses is similar
• Approach is slightly different. In new course
• Lectures are purely theoretical
• Exercises are more practical
• A newer programming language (Java) is used
  instead of Modula-2
• A target language (for code generation) is a
  subset of Java byte code instead of P-code

28
Conclusion

• Advantages of new course
• Concepts of compiler construction are shown on
  more formal way
• Modern and object-oriented language is used (Java
  and Java byte code)
• Students autonomously write (almost) whole
  compiler
• Disadvantages of new course
• Defining several language extensions for each
  student students like to corporate during task
  solution ?

29
Thank you for your attention