695673f9b5
git-svn-id: http://cregut.svn.enseeiht.fr/2020/1sn/pim/projets/GH-05@210346 e13453a9-b01f-0410-a051-f404c4f0c485
355 lines
11 KiB
Ada
355 lines
11 KiB
Ada
with Ada.Text_IO; use Ada.Text_IO;
|
|
with Ada.Integer_Text_IO; use Ada.Integer_Text_IO;
|
|
with Ada.Command_Line; use Ada.Command_Line;
|
|
with Ada.Strings.Unbounded; use Ada.Strings.Unbounded;
|
|
with Ada.Text_IO.Text_Streams;
|
|
|
|
with Vector;
|
|
with Google_Naive;
|
|
with Google_Creux;
|
|
|
|
procedure pageRank is
|
|
|
|
ERROR_args: Exception;
|
|
ERROR_alpha: Exception;
|
|
ERROR_ite: Exception;
|
|
ERROR_unexpected: Exception;
|
|
INFO_help: Exception;
|
|
|
|
-- définition du type T_Double
|
|
Type T_Double is digits 18;
|
|
-- on utilise le module générique Float_IO pour pouvoir afficher T_Double directement
|
|
package Text_T_Double is
|
|
new Ada.Text_IO.Float_IO(Num => T_Double);
|
|
use Text_T_Double;
|
|
|
|
|
|
|
|
|
|
|
|
-- procédure pour récupérer les arguments de la ligne de commande
|
|
procedure get_args(filename: in out Unbounded_String;
|
|
ite_max: in out Natural;
|
|
alpha: in out T_Double;
|
|
naif: in out Boolean) is
|
|
i: Natural := 1;
|
|
begin
|
|
|
|
-- on affiche tous les arguments
|
|
put("Argument_Count = "); put(Argument_Count, 1); new_line;
|
|
for i in 1..Argument_Count loop
|
|
Put("Argument(");
|
|
Put(i, 0);
|
|
Put(") = ");
|
|
Put_line(Argument(i));
|
|
end loop;
|
|
|
|
new_line;
|
|
|
|
-- on vérifie d'abord que le nombre d'arguments est cohérent
|
|
if not(0 < Argument_Count and Argument_Count <= 7) then
|
|
raise ERROR_args;
|
|
else -- sinon on parse les arguments
|
|
loop
|
|
if Argument(i) = "-h" or Argument(i) = "--help" then
|
|
put_line("parsed help");
|
|
raise INFO_help;
|
|
|
|
elsif Argument(i) = "-n" or Argument(i) = "--naif" then
|
|
naif := True;
|
|
i := i + 1;
|
|
put_line("parsed naif");
|
|
|
|
elsif Argument(i) = "-a" or Argument(i) = "--alpha" then
|
|
alpha := T_Double'Value(Argument(i+1));
|
|
if alpha < 0.0 or alpha > 1.0 then
|
|
raise ERROR_alpha;
|
|
end if;
|
|
i := i + 2;
|
|
put_line("parsed alpha");
|
|
|
|
elsif Argument(i) = "-i" or Argument(i) = "--ite-max" then
|
|
ite_max := Natural'Value(Argument(i+1));
|
|
if ite_max > 150 then
|
|
raise ERROR_ite;
|
|
end if;
|
|
i := i + 2;
|
|
put_line("parsed ite_max");
|
|
|
|
elsif Argument(i)(Argument(i)'Last-3 .. Argument(i)'Last) = ".net" then
|
|
filename := To_Unbounded_String(Argument(i)(Argument(i)'First .. Argument(i)'Last-4));
|
|
i := i + 1;
|
|
put_line("parsed filename");
|
|
|
|
else
|
|
new_line;
|
|
put("Option: '");
|
|
put(Argument(i));
|
|
put("' non reconnu");
|
|
new_line;
|
|
raise ERROR_unexpected;
|
|
|
|
end if;
|
|
exit when i > Argument_Count;
|
|
end loop;
|
|
end if;
|
|
|
|
new_line;
|
|
put("alpha = "); put(alpha, Fore=>1, Aft=>10); new_line;
|
|
put("naif = "); put(Boolean'Pos(naif), 1); new_line;
|
|
put("ite_max = "); put(ite_max, 1); new_line;
|
|
put("filename = "); put_line(To_String(filename));
|
|
new_line;
|
|
|
|
exception
|
|
|
|
when CONSTRAINT_ERROR =>
|
|
if Argument(i) = "-a" or Argument(i) = "--alpha" then
|
|
raise ERROR_alpha;
|
|
elsif Argument(i) = "-i" or Argument(i) = "--ite-max" then
|
|
raise ERROR_ite;
|
|
end if;
|
|
|
|
end get_args;
|
|
|
|
|
|
|
|
|
|
|
|
-- procédure pour choisir le type d'algo, une fois N et les arguments récupérés
|
|
procedure type_algo(N: Positive;
|
|
N_links: Positive;
|
|
filename: in Unbounded_String;
|
|
file: in out Ada.Text_IO.File_Type;
|
|
alpha: T_Double;
|
|
ite_max: Natural;
|
|
naif: Boolean) is
|
|
|
|
-- on instancie le module générique Vecteur
|
|
package Vector_T_Double is
|
|
new Vector(T_Element => T_Double,
|
|
N => N);
|
|
use Vector_T_Double;
|
|
|
|
-- pour le retour chariot
|
|
stdout: constant Ada.Text_IO.File_Type := Ada.Text_IO.Standard_Output;
|
|
|
|
-- procédure qui effectue l'algorithme du pageRank avec Google_Naive
|
|
procedure algorithm_naif(file: in out Ada.Text_IO.File_Type;
|
|
alpha: in T_Double;
|
|
ite_max: in Natural;
|
|
pi: out T_Vecteur_Element) is
|
|
|
|
-- on instancie le module générique Naif de Google
|
|
package Google is
|
|
new Google_Naive(T_Element => T_Double,
|
|
N => N,
|
|
Vector_T_Element => Vector_T_Double);
|
|
use Google;
|
|
|
|
-- définition de la matrice Google
|
|
G: T_Google;
|
|
|
|
begin
|
|
|
|
initialize(pi, 1.0/T_Double(N));
|
|
put_line("initialized pi");
|
|
-- put(pi); new_line;
|
|
|
|
initialize(G);
|
|
put_line("initialized G");
|
|
-- put(G); new_line;
|
|
|
|
create_H(G, file);
|
|
put_line("created H");
|
|
close(file);
|
|
-- put(G); new_line;
|
|
|
|
create_S(G);
|
|
put_line("created S");
|
|
-- put(G); new_line;
|
|
|
|
create_G(G, alpha);
|
|
put_line("created G");
|
|
-- put(G); new_line;
|
|
|
|
-- on applique l'algorithme itératif
|
|
for i in 1..ite_max loop
|
|
pi := pi * G;
|
|
String'Write(Ada.Text_IO.Text_Streams.Stream(stdout),
|
|
"ite:" & Integer'Image(i) & " /" & Integer'Image(ite_max) & ASCII.CR);
|
|
end loop; new_line;
|
|
|
|
-- new_line;
|
|
-- put_line("final pi:");
|
|
-- put(pi);
|
|
|
|
end algorithm_naif;
|
|
|
|
-- procédure qui effectue l'algorithme du pageRank avec Google_Creux
|
|
procedure algorithm_creux(file: in out Ada.Text_IO.File_Type;
|
|
alpha: in T_Double;
|
|
ite_max: in Natural;
|
|
pi: out T_Vecteur_Element) is
|
|
|
|
-- on instancie le module générique Creux de Google
|
|
package Google is
|
|
new Google_Creux(T_Element => T_Double,
|
|
N => N,
|
|
N_links => N_links,
|
|
Vector_T_Element => Vector_T_Double);
|
|
use Google;
|
|
|
|
-- définition de la matrice Google
|
|
H: T_Google;
|
|
|
|
begin
|
|
|
|
initialize(pi, 1.0/T_Double(N));
|
|
put_line("initialized pi:");
|
|
-- put(pi); new_line;
|
|
|
|
create_H(H, file);
|
|
put_line("created H:");
|
|
-- put(H); new_line; new_line;
|
|
|
|
-- on applique l'algorithme itératif
|
|
for i in 1..ite_max loop
|
|
pi := calcul(pi, H, alpha);
|
|
String'Write(Ada.Text_IO.Text_Streams.Stream(stdout),
|
|
ASCII.CR & "ite:" & Integer'Image(i) & " /" & Integer'Image(ite_max));
|
|
end loop; new_line;
|
|
|
|
-- new_line;
|
|
-- put_line("final pi:");
|
|
-- put(pi);
|
|
|
|
end algorithm_creux;
|
|
|
|
|
|
|
|
|
|
|
|
-- procédure pour écrire les résultats dans les fichiers
|
|
procedure write_to_files(filename: in Unbounded_String;
|
|
pi_sorted: in T_Vecteur_Element;
|
|
pi_index: in T_Vecteur_Natural) is
|
|
|
|
file: Ada.Text_IO.File_Type;
|
|
|
|
begin
|
|
create(file, Out_File, To_String(filename & "_GH05.p"));
|
|
put(file, N, 1); put(file, ' ');
|
|
put(file, alpha, Fore=>1, Aft=>10); put(file, ' ');
|
|
put(file, ite_max, 1); new_line(file);
|
|
put(file, pi_sorted);
|
|
close(file);
|
|
|
|
create(file, Out_File, To_String(filename & "_GH05.ord"));
|
|
put(file, pi_index);
|
|
close(file);
|
|
end write_to_files;
|
|
|
|
|
|
|
|
|
|
|
|
-- définition des vecteurs
|
|
pi: T_Vecteur_Element;
|
|
pi_index: T_Vecteur_Natural;
|
|
|
|
begin
|
|
|
|
if naif then
|
|
algorithm_naif(file, alpha, ite_max, pi);
|
|
else
|
|
algorithm_creux(file, alpha, ite_max, pi);
|
|
end if;
|
|
|
|
-- on trie les poids par ordre décroissant, on tri en même temps les indices des pages
|
|
initialize(pi_index);
|
|
sort_with_index_desc(pi, pi_index);
|
|
|
|
-- new_line;
|
|
-- put_line("sorted pi:");
|
|
-- put(pi); new_line;
|
|
|
|
-- on écrit les resultats dans des fichiers
|
|
write_to_files(filename, pi, pi_index);
|
|
|
|
end type_algo;
|
|
|
|
|
|
|
|
|
|
|
|
-- définition des arguments
|
|
filename: Unbounded_String;
|
|
ite_max: Natural := 150;
|
|
naif: Boolean := False;
|
|
alpha: T_Double := 0.85;
|
|
|
|
-- définition des variables pour créer les matices/vecteurs
|
|
N: Positive;
|
|
N_links: Integer := -1;
|
|
file: Ada.Text_IO.File_Type;
|
|
|
|
begin
|
|
|
|
-- on récupère les arguments de la ligne de commande
|
|
get_args(filename, ite_max, alpha, naif);
|
|
put_line("args OK");
|
|
|
|
-- on ouvre le fichier .net
|
|
open(file, In_File, To_String(filename & ".net"));
|
|
put_line("file OK");
|
|
|
|
-- on récupère le nombre de liens
|
|
while not End_Of_File(file) loop
|
|
Skip_Line(file);
|
|
N_links := N_links + 1;
|
|
end loop;
|
|
reset(file, In_File);
|
|
put("N_links = "); put(N_links, 1); new_line;
|
|
|
|
-- on récupère le nombre de pages
|
|
get(file, N);
|
|
put("N = "); put(N, 1); new_line;
|
|
|
|
-- on peut maintenant choisir le type de matrice que l'on souhaite
|
|
type_algo(N, N_links, filename, file, alpha, ite_max, naif);
|
|
|
|
exception
|
|
|
|
-- si il y a une erreur lors du parsing des arguments, on rappelle l'usage.
|
|
when ERROR_args =>
|
|
new_line;
|
|
put_line("Erreur lors de la saisi de la commande.");
|
|
put_line("Usage: pagerank [-P] [-i max_iterations] [-a alpha] [-h] fichier_reseau.net");
|
|
|
|
when ERROR_unexpected =>
|
|
put_line("Essayez 'pagerank --help' pour plus d'informations.");
|
|
|
|
when ERROR_alpha =>
|
|
new_line;
|
|
put_line("Erreur lors de la saisi de alpha.");
|
|
put_line("alpha ∈ [0, 1]");
|
|
|
|
when ERROR_ite =>
|
|
new_line;
|
|
put_line("Erreur lors de la saisi de ite_max.");
|
|
put_line("ite_max ∈ ⟦0, 150⟧");
|
|
|
|
when INFO_help =>
|
|
new_line;
|
|
put_line("Usage: pagerank [OPTIONS] network.net");
|
|
put_line("Calcule le pagerank d'un réseau à partir de son réseau.net");
|
|
new_line;
|
|
put_line("Options:");
|
|
put_line(" -P, --naif specifies which type of matrix to use");
|
|
put_line(" -a, --alpha specifies the alpha constant (alpha ∈ [0, 1])");
|
|
put_line(" -i, --ite-max specifies the maximum number of iterations (ite_max ∈ ⟦0, 150⟧)");
|
|
put_line(" -h, --help display this help message and exit");
|
|
|
|
end pageRank;
|