Metaprogramming from University to Industry

1
Metaprogrammingfrom University to Industry

• Zoltán Porkoláb
• gsd_at_elte.hu
• http//gsd.web.elte.hu
• Dept. of Programming Languages and Compilers,
• Faculty of Informatics
• Eötvös Loránd University, Budapest

2
Agenda

• I always knew C templates were the work of the
  Devil,
• and now I'm sure... - Cliff Click cited by Todd
  Veldhuisen
• Parameterized types
• C Template Metaprograms
• Power of generative metaprograms
• Sample usages
• Open questions

3
Parameterized types (Generics)

• Widely used in modern programming languages
• ADA generics
• Eiffel generics
• C templates
• Java generics Pizza, GJ, Java 1.5
• C generics
• Clean, Generic Haskell, other functional langs.

4
Why generics?
Conventional techniques are working only for
complete types

• int max( int a, int b)
• if ( a gt b ) return a else return b
• double max( double a, double b)
• if ( a gt b ) return a else return b
• //
• class date / /
• date d max ( d1, d2)

5
Preprocessor Macro

• define MAX(a,b) a gt b ? a b
• Works, because a macro is typeless.
• Processed not by the compiler, therefore there
  are a number of "secondary effects"
• MAX( x, y)2 -gt x gt y ? x y2
• MAX( x, y) -gt x gt y ? x y

6
Macro the limits

• void swap( int x, int y)
• int temp x x y y temp
• Does not work with macro a macro is typeless.
• We need a facility to use type parameters.

7
Templates

• template lttypename Tgt
• void swap( T x, T y)
• T temp x x y y temp
• Ada generics "a form of context-sensitive macro"
  
• C templates "a clever kind of macro that obeys
  the scope, naming, and type rules of C"
• Does not require different types used as
  arguments to be explicitly related. In
  particular, the argument types used as a template
  need not be from a single inheritance hierarchy.

8
C Function Templates

• Template is not a single function
• A schema to instantiate functions on request
• Parameter deduction
• Template instantiation
• Compilaton time
• template lttypename Tgt T max( T a, T b)
• if ( a gt b ) return a
• else return b
• int i 3, j 6, k
• double x 3.14, y 4.15, z
• k max(i,j)
• z max(x,y)

9
C Function Templates 2.

• Strong type system rules are applied
• int i 3 double y 3.14, z
• z max(i,y) // error
• template lttypename T, typename Sgt
• T max( T a, S b)
• if ( a gt b ) return a
• else return b
• z 3.0
• No deduction on return type
• No runtime information could be used

10
Explicit specialization

• Explicit specialization
• int i 3
• double y 3.14, z
• template lttypename R, typename T, typename Sgt
• R max( T a, S b)
• if ( a gt b ) return a
• else return b
• z maxltdoublegt(i,y)

11
User specialization

• const char s1 Hello, s2 world
• const char s max(s1,s2)
• template ltgt
• const char max(const char a, const char b)
• if ( strcmp(a,b) lt 0 ) return a
• else return b

12
Template overloading

• You can provide overloadd template definitions
• The compiler selects the most specific template
• cout ltlt max (4,5)
• cout ltlt maxltdoublegt(3.14,6)
• cout ltlt max (this, greater)

R max(S,T)
T max(T, T)
char max(char ,char )
13
C Class Templates 1.

• Similar way template classes could be defined
• template ltclass Tgt class matrix
• public
• matrix( int i, int j )
• matrix operator( const matrix other)
• T at(int i, int j)
• //
• private
• vectorltTgt v
• Created always with explicit specialisation
• matrixltintgt m(4,5)

14
C Class Templates 2.

• User specialization
• template ltgt class matrixltboolgt
• public
• matrix( int i, int j )
• matrix( const matrix other)
• bool at(int i, int j)
• private
• aBetterRepresentation v
• Used the same way
• matrixltboolgt m(4,5)

15
C Templates

• The C templates were first implemented in the
  early 90s
• Accepted as part of the ANSI/ISO in 1994
• Erwin Unruh 1994
• unruh.cpp 30 conversion from enum to Dlt2gt
  requested
• unruh.cpp 30 conversion from enum to Dlt3gt
  requested
• unruh.cpp 30 conversion from enum to Dlt5gt
  requested
• unruh.cpp 30 conversion from enum to Dlt7gt
  requested
• unruh.cpp 30 conversion from enum to Dlt11gt
  requested
• unruh.cpp 30 conversion from enum to Dlt13gt
  requested
• unruh.cpp 30 conversion from enum to Dlt17gt
  requested
• unruh.cpp 30 conversion from enum to Dlt19gt
  requested

16
Power of C templates

• The Compiler executes template metaprograms
• The result is a non-templated program
• executed in run-time
• In 1966 Böhm and Jacopini proved
• Turing machine implementation ltgt
  
  conditional and looping constructions
• The C templates are Turing-complete
• in compilation time

17
The Factorial example

• The run-time solution
• int factorial( int n)
• if ( n 1 ) return 1
• else return nfactorial(n-1)
• int main()
• cout ltlt factorial(15) ltlt endl
• return 0

18
The Factorial example

• The metaprogram solution
• template ltint Ngt struct Factorial
• enum value N FactorialltN-1gtvalue
• template ltgt struct Factoriallt1gt
• enum value 1
• int main()
• const int fact15 Factoriallt15gtvalue
• stdcout ltlt fact15 ltlt endl
• return 0

19
Conditional statement

• template ltbool condition, class Then, class Elsegt
  
• struct IF
• typedef Then RET
• template ltclass Then, class Elsegt
• struct IFltfalse, Then, Elsegt
• typedef Else RET
• template lttypename T, typename Sgt
• IFlt sizeof(T)ltsizeof(S), S, TgtRET max(T t, S s)
• if (t gt s) return t
• else return s

20
(Run-time) Programs vs. Metaprograms

• Function
• (runtime) Data
• Variable
• Condition
• Loop
• Assignment

• Class
• Type and constant
• Symbolic names
• Type selection
• Recursion
• No assignment

21
Data in Template Metaprograms

• Referential transparency No assignment
• Still possible to store, modify, and retrieve
  data
• Typelist
• struct NullType
• typedef Typelistlt char, Typelistltsigned char,
• Typelistltunsigned char, NullTypegt gt gt
  Charlist

char

signed char

unsigned char
NullType
22
Data handling

• template ltclass TListgt struct Length
• template ltgt
• struct LengthltNullTypegt
• enum value 0
• template ltclass T, class Ugt
• struct Length ltTypelistltT,Ugt gt
• enum value 1 LengthltUgtvalue

23
Motivation

• int main()
• const unsigned int di 12
• const unsigned int oi 014
• const unsigned int hi 0xc
• const unsigned int bi0 binary_value("1101")
  
• const unsigned int bi1 binarylt1100gtvalue
  

templateltunsigned long Ngt struct binary //
prepend higher bits to lowest bit static const
int valuebinaryltN/10gtvalue2N10
templateltgt struct binarylt0gt // specialization
static unsigned const value 0
24
Motivation

• Constant expression array size, case label,
  etc
• Better code compiled
• Faster in run-time
• Syntactically checked the language semantic is
  extended

Design time
Compilation time
Run-time
t
No change in type system Change in run-time values
Template metaprograms apply, automatic config
of the program
Decisions on Stategies, policies
25
Generative Metaprograms

• metaprogramming
• writing programs that represent and manipulate
  other programs or themselves (iereflection).
  Metaprograms are programs about programs.
• introspection
• the ability of a program to observe its own
  state
• intercession
• the ability to modify its own state
• Open compilers transformations on AST
• Hygenic macros (Scheme)
• Two level languages AspectJ, Template Haskell

26
Areas of Template Metaprogramming

• Expression templates
• Blitz, PETE
• Static interface checking
• Early catch of syntactical/semantical errors
• Extending the C type system
• Introspection
• Code adaption/optimalization
• Language embedding

27
Expression templates
Array a, b, c, d, e // Object-oriented way of
a b c d e double _t1 new doubleN
for ( int i0 iltN i) _t1i bi ci
double _t2 new doubleN for ( int i0
iltN i) _t2i _t1i di double _t3
new doubleNM for ( int i0 iltN i) _t3i
_t2i ei for ( int i0 iltN i) ai
_t3i delete _t3 delete _t2
delete _t1 // Fortran like solution for (
int i0 iltN i) ai bi ci di
ei

28
Language embedding

• SQL
• Gil, et.al. AraRat
• XML parsing
• Jarvi, et.al. type-safe XML library
• Regular expressions
• BoostXpressive
• Compiler embedding
• boostspirit

29
Language embedding

• SQL example
• string s select form tName where 11
• if ( cond1 )
• s and fName1 field1
• if ( cond2 )
• s and fName2 lt field2
• if ( cond3 )
• s and fName3 field3
• Run-time errors
• Injection attacks
• Conversion problems

30
Language embedding

• SQL example
• void f()
• const string s
• (
• (tName / (fName1 field1
• fName2 lt field2)
• )
• fName1, fName3
• ).asSQL()

31
Language embedding

• BoostXpressive example
• stdstring hello( "hello world!" )
• sregex rex1 sregexcompile( "(\\w)(\\w)!" )
• sregex rex2 (s1 _w) gtgt ' ' gtgt (s2 _w) gtgt
  '!'
• smatch what
• if( regex_match( hello, what, rex1 ) )
• stdcout ltlt what0 ltlt '\n' // whole match
• stdcout ltlt what1 ltlt '\n' // first capture
• stdcout ltlt what2 ltlt '\n' // second capture

32
Embedding alternatives

• If you are (accidently) not a C programmer
• .NET platform, VB, C
• LINQ project
• Java platform
• Eric Van Wijk et.al. Attribute Grammar-based
  Language Extensions for Java
• Stratego
• TU Delft
• Charles Simonyi Intentional programming

33
Intentional programming

• Charles Simonyi
• XEROX Palo Alto Bravo
• Microsoft Word, Excel
• Intentional software http//www.intentsoft.com

34
Open questions

• The real expressive power
• Standard tools
• but Loki from Andrei Alexandrescu
• boostmpl
• Garantees (number of instantiations, etc)
• How to design?
• How to debug?

35
Metaprogrammingfrom University to Industry

• Zoltán Porkoláb
• gsd_at_elte.hu
• http//gsd.web.elte.hu
• Dept. of Programming Languages and Compilers,
• Faculty of Informatics
• Eötvös Loránd University, Budapest

Thank you! Questions?