My original Ada web site with links is here
Important note added June 2013:
This page is obsolete now, left here for archive and information only.
The Ada Lapack code is now housed at http://sourceforge.net/projects/ada-lapack/
I downloaded the original BLAS Ada binding written by Duncan Sands from http://topo.math.u-psud.fr/~sands/Programs/BLAS/index.html and the LAPACK Ada binding written by Wasu Chaopanon from ftp://ftp.cs.kuleuven.be/pub/Ada-Belgium/mirrors/gnu-ada/OLD/contrib/lapack-ada/
And made some minor improvments to the bindings.
This page describes the minor changes made and instructions how to use these bindings from Ada in order call LAPACK and BLAS Fortran functions.
A new tar file for LAPACK and for BLAS with all the changes can be downloaded from the link below.
The changes made to LAPACK binding involve streamlining the source tree structure,
writing new Makefiles, simplify the binding to use one package called lapack and also
adding the documentation shown below.
Changes for the BLAS binding were minimal. It involved changes to the source tree structure and writing Makefiles and adding the documentation shown below.
The Ada binding to LAPACK and BLAS is an Ada package which allows one to interface to the native lapack and blas libraries.
The native lapack and blas libraries need to be first installed on the system (on Linux, these
libraries will normally be found in /usr/lib/liblapack.so and /usr/lib/libblas.so
)
To use LAPACK from Ada, one needs to install both the native LAPACK and BLAS libraries since LAPACK depends on BLAS.
The Ada binding is a thin binding, meaning there is 1-1 mapping between the call to the Ada routine and the corresponding Fortran routine using the same function name in the Fortran libraries.
The following diagram illustrates the use of the LAPACK binding with the needed gnatmake
command to compile and link the client Ada program.
And a similar diagram for the BLAS binding interface
The source tree structure for LAPACK is described in this diagram
The full content of the LAPACK tree is listed here lapack_tree_listing.txt
The source tree structure for the BLAS binding is described in this diagram
These instructions explain how to use the Ada LAPACK and BLAS binding in the updated snapshot tar file.
/usr/lib/.
ada_lapack_073012.zip and ada_blas_073012.zip
common.mk where a Makefile
variable is set to point to the directory that contains the native LAPACK and
BLAS libraries. This is currently defined to point to /usr/lib. Edit this line to
change this location only if the location is different in your system.
make from the top of each tree. Make will build
the whole tree, including the bindings packages and the test programs.
make testing from the top of tree for LAPACK
and for BLAS.
tests/ directory of each
tree.
binding/ directory in each tree. This is the directory that
you need to set the -I option to point to when using gnatmake as shown in the
diagrams above.
ada_lapack/tests/pilot/
directory
$ gnatmake -I../pragmarc mine_detector.adb `gtkada-config` gcc-4.4 -c -I../pragmarc -I/usr/share/ada/adainclude/gtkada2 mine_detector.adb gcc-4.4 -c -I../pragmarc -I/usr/share/ada/adainclude/gtkada2 user_if.adb gcc-4.4 -c -I../pragmarc -I/usr/share/ada/adainclude/gtkada2 field.ads gcc-4.4 -c -I../pragmarc -I/usr/share/ada/adainclude/gtkada2 field-operations.adb gnatbind -I../pragmarc -aI/usr/share/ada/adainclude/gtkada2 -aO/usr/lib/ada/adalib/gtkada2 -x mine_detector.ali gnatlink mine_detector.ali -L/usr/lib -lgtkada2 -pthread -lgtk-x11-2.0 -lgdk-x11-2.0 -latk-1.0 -lgio-2.0 -lpangoft2-1.0 -lpangocairo-1.0 -lgdk_pixbuf-2.0 -lm -lcairo -lpango-1.0 -lfreetype -lfontconfig -lgobject-2.0 -lgmodule-2.0 -lgthread-2.0 -lrt -lglib-2.0 $
On Thu, 20 Nov 2008 12:09:41 +0100, Markus Schoepflin wrote: > is it possible to influence the behaviour of GNAT regarding the handling of > NANs? (Most importantly in the special case of division by zero.) > > We need to get exceptions whenever a NAN is generated, is this possible > somehow? (For example by setting Machine_Overflow to True and recompiling > the compiler itself.) You can scrap IEEE stuff in favor of Ada semantics by declaring your own floating-point [sub]type with a range specified. The compiler will be forced to check values: type Safe_Float is digits 6 range -10.0E10..+10.0E10; or subtype Safe_Float is Float range Float'Range; then X : Safe_Float := 1.0; Y : Safe_Float := 0.0; begin Y := X / Y; exception when Error : others => -- Should print "range check failed" Put_Line (Exception_Message (Error)); end; -- Regards, Dmitry A. Kazakov http://www.dmitry-kazakov.de
from http://en.wikibooks.org/wiki/Ada_Programming/Object_Orientation
package X is type Object is tagged null record; procedure do (This: in Object; That: in Boolean); end X; with X; procedure Main is Obj : X.Object; begin Obj.do (That => True); end Main;
If gnat is not installed, install it (on linux) using something similar to
sudo apt-get install gnat-4.6
write the following code in file called hello_world.adb
with ada.text_io; use ada.text_io; procedure hello_world is begin put_line("hello world"); end hello_world;
compile using
gnatmake hello_world.adb gcc-4.6 -c hello_world.adb gnatbind -x hello_world.ali gnatlink hello_world.ali
Run it using ./hello_world
This is an Ada implementation of decimal representation of \(e\) based on SPIGOT algorithm for \(\pi \) by S. Rabinowitz & S. Wagon, The American Mathematical Monthly, March 1995.
This source file is here
-- More e digits trivia. -- Feel free to copy, distribute as long as this header attached so -- original algorithm creators and implementors are known. -- -- This is an Ada implementation of decimal representation of 'e' -- based on SPIGOT algorithm for \pi by -- S. Rabinowitz & S. Wagon, _The_American_Mathematical_Monthly_, March 1995 -- -- A C implementation of the above was posted on the net by -- Ed Hook -- MRJ Technology Solutions, Inc. -- NAS, NASA Ames Research Center -- Internet: hook@nas.nasa.gov -- -- This is an Ada implementation of the above using GNAT (gnu Ada compiler), -- with the added feature is that it computes the frequency of each digit in e, -- and computes the largest consecutive sequences of each digit within the -- expression that represents digits of e. -- -- the following is the result. my PC is still running trying to find the -- frequency for 200,000 digits and more for e, and it's been several days -- and not finished. So this is a partial results. (PC is 200 MHz pentium, -- running Linux 2.0.36, and compiler is GNAT 3.11p -- -- offcourse as number of digits of e goes very large, each digit is expected -- to show as often as any other digit. -- -- by Nasser M. Abbasi nma@12000.org feb. 20, 1999. -- -- results: -- this is distribution table for digits in e as function of how many -- digits. -- for example, when looking at 5000 digits of e, we find 497 0's, -- 478 1's, etc.. (this is for digits after the decimal point of e) -- -- -- #digits in e -- -------------------------------------------------------------- -- 500 5,000 20,000 50,000 200,000 -- --------------------------------------------------------------- --how many 0's 51 497 1,949 4,948 19,916 --how many 1's 43 478 2,010 5,055 20,367 --how many 2's 50 492 2,020 4,969 19,794 --how many 3's 53 514 2,080 5,026 20,071 --how many 4's 52 470 1,989 4,966 20,082 --how many 5's 44 478 1,979 5,046 20,038 --how many 6's 51 545 2,057 5,133 20,221 --how many 7's 60 525 1,977 4,959 19,817 --how many 8's 40 509 1,966 4,972 19,939 --how many 9's 56 492 1,974 4,926 19,755 -- ------------------------------------------------------------------------ --most occurring '7' '7' '3' '6' '1' ------------------------------------------------------------------------ --least occurring '8' '4' '0' '9' '9' ------------------------------------------------------------------------ --difference --between largest 20 55 131 207 612 --and smallest --in frequency ------------------------------------------------------------------------ --difference --between largest 4% 1.1% 0.655% 0.414% 0.306% --and smallest --frequency in % -- -- --consecutive frequencies: under each column, there are 3 values, the first --is the number of digits that occurred next to each others for that digit, --and the start of this sub sequence, and its end, in position values. -- --for example, for 5,000 digits of e, we see that largest consecutive --sequence of digit '0' had length of 3, and it started at digit position --328 to position 330. Digit positions are counted from left to right at --the decimal point. for example e=2.718, here digit '7' is at position 1, --'1' is at position 2, etc.. -- -- #digits in e -- ----------------+-----------------+----------------------------------- -- 5,000 | 20,000 | 50,000 | 100,000 -- ----------------+------------------+------------------+--------------- -- 0's (3,328,330) | (4,7688,7691) | *no change* |(6,89296,89301) -- 1's (3,427,429) | (5,12220,12224) | *no change* | *no change* -- 2's (2,2744,2746) | (4,17309,17312) | (5,33483,33487) | *no change* -- 3's (4,3354,3375) | *no change* | *no change* | *no change* -- 4's (3,787,789) | (4,11806,11809) | *no change* | *no change* -- 5's (4,3620,3623) | *no change* | *no change* | *no change* -- 6's (5,4992,4996) | *no change* | *no change* | *no change* -- 7's (4,1071,1074) | *no change* | *no change* | *no change* -- 8's (4,723,726) | *no change* | *no change* | *no change* -- 9's (3,47,49) | *no change* | (4,29344,29347) | *no change* -- -- -- --Compiler: GNAT 3.11p , see http://www.adahome.com to download --To compile: save this file as dist_e_final.adb and type -- gnatmake dist_e_final.adb --system: Linux 2.0.36 --Date: feb. 17, 1999 --To Run: ./dist_e_final -- For example, to see e for 70 digits do: -- -- ./dist_e_final 70 -- 2.7182818284590452353602874713526624977572470936999595749669676277240766 -- frequency of 0 is 4 -- frequency of 1 is 3 -- frequency of 2 is 9 -- frequency of 3 is 4 -- frequency of 4 is 7 -- frequency of 5 is 7 -- frequency of 6 is 10 -- frequency of 7 is 12 -- frequency of 8 is 5 -- frequency of 9 is 9 -- -- performance note: On Pentium PRO 200 MHZ, using GNAT 3.11p, Linux 2.0.36, -- 128 MB RAM. No other activity on PC, and for 1,000,000 digits, this -- program will generate about 50 digits each minutes. So, for 1,000,000 -- digits it will take about 13 days. for larger than 1,000,000 you might -- encounter stack overrun, depending on amount of memory you have... -- -- notice the main algorithm is O(n^2). -- with Ada.Text_Io; use Ada.Text_Io; with ada.command_line; use ada.command_line; procedure Dist_E_final is type E_Type is array( Natural range <> ) of Natural; Distribution : array(0..9) of Natural := (others => 0); Num_Of_Digits : Natural; type Sequence_item is record Starts_At, Ends_At, Length : Natural; end record; Sequence: array(0..9) of Sequence_Item := (others=>(0,0,0)); current_Digit, Current_Sequence_Length, Current_Sequence_Start: Natural :=0; procedure Update_Sequence(Next_Digit_Position, next_digit: Natural) is begin if( next_Digit /= Current_Digit) then if( Sequence( current_Digit ).Length < Current_Sequence_Length) then Sequence( current_Digit ).Length := Current_Sequence_Length; Sequence( current_Digit ).Starts_At := Current_Sequence_start; Sequence( Current_Digit ).Ends_At := Next_Digit_Position -1; end if; Current_Digit := Next_Digit; Current_Sequence_Length := 1; Current_Sequence_Start := Next_Digit_Position; else Current_Sequence_Length := Current_Sequence_Length +1; end if; end Update_Sequence; procedure Done_Sequence( Current_Digit_Position: Natural) is begin if( Sequence( current_Digit ).Length < Current_Sequence_Length) then Sequence( current_Digit ).Length := Current_Sequence_Length; Sequence( current_Digit ).Starts_At := Current_Sequence_start; Sequence( Current_Digit ).Ends_At := current_Digit_Position ; end if; end Done_Sequence; begin if( Argument_Count /= 1 ) then Put_Line("usage: dist_e "); return; end if; begin Num_Of_Digits := natural'value( Argument(1)); if( Num_Of_Digits = 0 ) then Put_Line("value for number of digits must be larger than zero"); return; end if; exception when others => Put_Line("Exception. invalid value for number of digits"); return; end; declare -- the algorithm itself is in this block E: E_Type( 1 .. Num_Of_Digits+2 ) := (others=> 1); Carry : Natural; begin Put("2."); for I in E'first .. E'Last-2 loop Carry := 0; for J in reverse E'first .. E'Last loop E(J) := ( E(J) * 10 ) + Carry; Carry := E(J)/(J+1); E(J) := E(J) rem (J+1); end loop; Put(Natural'Image(Carry)(2)); -- print current digit of e Distribution( Carry ) := Distribution( Carry ) + 1; Update_Sequence(I,Carry); end loop; Done_Sequence(E'Last-2); end; New_Line; for I in Distribution'Range loop Put_line("frequency of " & Natural'Image(I) & " is " & natural'Image( Distribution(I) )); end loop; for I in sequence'Range loop if( Sequence(I).Length = 0 ) then Put_Line("Digit "& Natural'Image(I) & " was not seen."); else Put_line("largest concecutive seq of " & Natural'Image(I) &" started at digit " & natural'Image( sequence(I).Starts_at ) & " and ended at digit " & natural'Image( sequence(I).ends_at ) & " of length " & natural'Image( sequence(I).length )); end if; end loop; end Dist_E_final;
This package is an Ada implementation of getopt() as specified by the document "The Single UNIX Specification, Version 2", Copyright 1997 The Open Group
This source file is here
------------------------------------------------------------------------------ -- -- -- G E T O P T -- -- -- -- S p e c -- -- -- -- $Header: getopt.ads,v 1.1.1.1 1999/03/01 12:23:04 nabbasi Exp $ -- -- -- -- Copyright (C) 1998 Nasser Abbasi -- -- -- -- This is free software; you can redistribute it and/or modify it under -- -- terms of the GNU General Public License as published by the Free Soft- -- -- ware Foundation; either version 2, or (at your option) any later ver- -- -- sion. GETOPT is distributed in the hope that it will be useful, but WITH -- -- OUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY -- -- or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License -- -- for more details. Free Software Foundation, 59 Temple Place - Suite -- -- 330, Boston, MA 02111-1307, USA. -- -- -- -- As a special exception, if other files instantiate generics from this -- -- unit, or you link this unit with other files to produce an executable, -- -- this unit does not by itself cause the resulting executable to be -- -- covered by the GNU General Public License. This exception does not -- -- however invalidate any other reasons why the executable file might be -- -- covered by the GNU Public License. -- ------------------------------------------------------------------------------ -- change history: -- -- -- -- name changes -- -- ---------- -------------------------------------------------------------- -- NMA021899 created -- -- NMA030299 Made it modified GPL. chanegd header. -- -- -- -- description: -- -- -- -- This package is an Ada implementation of getopt() as specified by the -- -- document "The Single UNIX Specification, Version 2", Copyright 1997 The -- -- Open Group -- -- -- -- Compiler used: GNAT 3.11p -- -- Platform: Linux 2.0.36 ( Red hat 5.2) -- -- -- with Ada.Strings.Unbounded; use Ada.Strings.Unbounded; package Getopt is function Getopt (Optstring : String) return Integer; Optind : Positive; Optarg : Unbounded_String; Optopt : Character := ' '; Opterr : Integer := 1; end Getopt; ------------------------------------------------------------------------------ -- -- -- G E T O P T -- -- -- -- BODY -- -- -- -- $Header: getopt.adb,v 1.2 1999/03/01 12:54:03 nabbasi Exp $ -- -- -- -- -- -- -- -- Copyright (C) 1998 Nasser Abbasi -- -- -- -- -- -- This is free software; you can redistribute it and/or modify it under -- -- terms of the GNU General Public License as published by the Free Soft- -- -- ware Foundation; either version 2, or (at your option) any later ver- -- -- sion. GETOPT is distributed in the hope that it will be useful, but WITH -- -- OUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY -- -- or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License -- -- for more details. Free Software Foundation, 59 Temple Place - Suite -- -- 330, Boston, MA 02111-1307, USA. -- -- -- -- As a special exception, if other files instantiate generics from this -- -- unit, or you link this unit with other files to produce an executable, -- -- this unit does not by itself cause the resulting executable to be -- -- covered by the GNU General Public License. This exception does not -- -- however invalidate any other reasons why the executable file might be -- -- covered by the GNU Public License. -- -- -- ------------------------------------------------------------------------------ -- -- -- change history: -- -- -- -- name changes -- -- ---------- -------------------------------------------------------------- -- NMA021899 created -- -- NMA030299 Changed header to make it modified GPL -- -- -- -- description: -- -- -- -- This package is an Ada implementation of getopt() as specified by the -- -- document "The Single UNIX Specification, Version 2", Copyright 1997 The -- -- Open Group -- -- -- -- This describes the items involveed using example -- -- -- -- -- -- curopt -- -- | -- -- V -- -- "-f foo -dbc -k" -- -- ^ -- -- | -- -- optind -- -- -- -- optind is position (index) that tells which command line argument is -- -- being processed now. -- -- curopt tells which optchar is being processed within one command line -- -- argument. This is needed only if more that one optchar are stuck -- -- togother in one argument with no space, as in -df where both d and f -- -- are valid optchar and d takes no optarg. -- -- -- -- -- -- Compiler used: GNAT 3.11p -- -- Platform: Linux 2.0.36 ( Red hat 5.2) -- -- with Ada.Strings.Unbounded; use Ada.Strings.Unbounded; with Ada.Command_Line; use Ada.Command_Line; with Ada.Text_Io; use Ada.Text_Io; package body Getopt is Curopt : Natural := 2; -------------------- -- No_Optarg_Case -- -------------------- procedure No_Optarg_Case is begin if (Curopt < Argument (Optind)'Length) then Curopt := Curopt + 1; else Curopt := 2; Optind := Optind + 1; end if; end No_Optarg_Case; ------------ -- Getopt -- ------------ function Getopt (Optstring : String) return Integer is begin if (Argument_Count = 0 or else optind > Argument_Count or else (Argument (optind)(1) /= '-')) then return -1; end if; if (Argument (optind)'Length = 1) then return -1; end if; -- according to The Single UNIX Specification, Version 2, if "--" -- is found, return -1 after ++optind. if (Argument (Optind)(2) = '-') then Optind := Optind + 1; return -1; end if; -- if we get here, the command argument has "-X" for I in Optstring'Range loop if (Optstring (I) = Argument (optind)(Curopt)) then if (I < Optstring'Length) then if (Optstring (I + 1) = ':') then -- see if optarg stuck to optchar if (Argument (Optind)'Length - Curopt > 0) then Optarg := To_Unbounded_String (Argument (optind)(Curopt + 1 .. Argument (optind)'Length)); Curopt := Curopt + 1; optind := Optind + 1; return character'Pos (Optstring (I)); end if; -- see if optarg on separate argument if (Optind < Argument_Count) then Curopt := 2; optind := optind + 1; optarg := To_Unbounded_String (Argument (optind)); optind := optind + 1; return character'Pos (Optstring (I)); else Optind := Optind + 1; Optopt := Optstring (I); if (Opterr = 1 and Optstring (1) /= ':') then Put_Line (Standard_Error, "Argument expected for the -"& Optstring (I .. I) & " option"); end if; if (Optstring (1) = ':') then return Character'Pos (':'); else return Character'Pos ('?'); end if; end if; else -- current optchar matches and has no arg option No_Optarg_Case; return character'Pos (Optstring (I)); end if; else -- last char in optstring, can't have argument No_Optarg_Case; return character'Pos (Optstring (I)); end if; end if; end loop; Optopt := Argument (Optind)(Curopt); No_Optarg_Case; -- we get here if current command argument not found in optstring return character'Pos ('?'); end Getopt; begin Optarg := To_Unbounded_String (""); Optind := 1; end Getopt;
This is a test program of the above package. source code
-- Test example showing how to use GETOPT Ada package -- Nasser M. Abbasi with Ada.Text_Io; use Ada.Text_Io; with Ada.Command_Line; use Ada.Command_Line; with Ada.Strings.Unbounded; use Ada.Strings.Unbounded; with Getopt; procedure Test_Getopt is Test_String : String := "c:di:n:p:u:V"; Optchar : character; Value : Integer; begin Getopt.Opterr := 1; loop Value := Getopt.Getopt( Test_String ); exit when Value = -1; optchar := Character'Val( Value ); case optchar is when 'c' => Put_Line("commant is "& To_String(Getopt.Optarg)); when 'd' => Put_Line("debug on"); when 'i' => Put_line("got -i, its argument is:" & To_String(Getopt.Optarg) ); when 'n' => Put_line("got -n, its argument is:" & To_String(Getopt.Optarg)); when 'p' => Put_line("got -p, its argument is:" & To_String(Getopt.Optarg)); when 'u' => Put_line("got -u, its argument is:" & To_String(Getopt.Optarg)); when 'V' => Put_line("got -V"); when '?' => Put_Line("got ?, optopt is " & Getopt.Optopt); when ':' => Put_Line("get :, optopt is "& Getopt.optopt); when others => null; end case; end loop; -- now lets print the remaining arguments if any declare Index : positive; begin Index := Getopt.Optind; for I in Index..Argument_Count loop Put_Line( Argument(I) ); end loop; end; end Test_Getopt;