1 | /////////////////////////////////////////////////////////////////////////////// |
---|
2 | version="$Id: atkins.lib,v 1.4 2007-07-20 11:26:46 Singular Exp $"; |
---|
3 | category="Teaching"; |
---|
4 | info=" |
---|
5 | LIBRARY: atkins.lib Procedures for teaching cryptography |
---|
6 | AUTHOR: Stefan Steidel, Stefan.Steidel@gmx.de |
---|
7 | |
---|
8 | NOTE: The library contains auxiliary procedures to compute the elliptic |
---|
9 | curve primality test of Atkin and the Atkin's Test itself. |
---|
10 | The library is intended to be used for teaching purposes but not |
---|
11 | for serious computations. Sufficiently high printLevel allows to |
---|
12 | control each step, thus illustrating the algorithms at work. |
---|
13 | |
---|
14 | |
---|
15 | PROCEDURES: |
---|
16 | newTest(L,D) checks if number D already exists in list L |
---|
17 | bubblesort(L) sorts elements of the list L |
---|
18 | disc(N,k) generates a list of negative discriminants |
---|
19 | Cornacchia(d,p) computes solution (x,y) for x^2+d*y^2=p |
---|
20 | CornacchiaModified(D,p) computes solution (x,y) for x^2+|D|*y^2=4p |
---|
21 | maximum(L) computes the maximal number contained in L |
---|
22 | expo(z,k) computes exp(z) |
---|
23 | jOft(t,k) computes the j-invariant of t |
---|
24 | round(r) rounds r to the nearest number out of Z |
---|
25 | HilbertClassPoly(D,k) computes the Hilbert Class Polynomial |
---|
26 | rootsModp(p,P) computes roots of the polynomial P modulo p |
---|
27 | wUnit(D) computes the number of units in Q(sqr(D)) |
---|
28 | Atkin(N,K,B) tries to prove that N is prime |
---|
29 | |
---|
30 | "; |
---|
31 | |
---|
32 | LIB "crypto.lib"; |
---|
33 | LIB "general.lib"; |
---|
34 | LIB "ntsolve.lib"; |
---|
35 | LIB "inout.lib"; |
---|
36 | |
---|
37 | /////////////////////////////////////////////////////////////////////////////// |
---|
38 | |
---|
39 | proc newTest(list L, number D) |
---|
40 | "USAGE: newTest(L,D); |
---|
41 | RETURN: 1, if D does not already exist in L, |
---|
42 | -1, if D does already exist in L |
---|
43 | EXAMPLE:example new; shows an example |
---|
44 | " |
---|
45 | { |
---|
46 | number a=1; // a=1 bedeutet: D noch nicht in L vorhanden |
---|
47 | int i; |
---|
48 | for(i=1;i<=size(L);i++) |
---|
49 | { |
---|
50 | if(D==L[i]) |
---|
51 | { |
---|
52 | a=-1; // a=-1 bedeutet: D bereits in L vorhanden |
---|
53 | break; |
---|
54 | } |
---|
55 | } |
---|
56 | return(a); |
---|
57 | } |
---|
58 | example |
---|
59 | { "EXAMPLE:"; echo = 2; |
---|
60 | ring r = 0,x,dp; |
---|
61 | list L=8976,-223456,556,-778,3,-55603,45,766677; |
---|
62 | number D=-55603; |
---|
63 | newTest(L,D); |
---|
64 | } |
---|
65 | |
---|
66 | proc bubblesort(list L) |
---|
67 | "USAGE: bubblesort(L); |
---|
68 | RETURN: list L, sort in decreasing order |
---|
69 | EXAMPLE:example bubblesort; shows an example |
---|
70 | " |
---|
71 | { |
---|
72 | number b; |
---|
73 | int n,i,j; |
---|
74 | while(j==0) |
---|
75 | { |
---|
76 | i=i+1; |
---|
77 | j=1; |
---|
78 | for(n=1;n<=size(L)-i;n++) |
---|
79 | { |
---|
80 | if(L[n]<L[n+1]) |
---|
81 | { |
---|
82 | b=L[n]; |
---|
83 | L[n]=L[n+1]; |
---|
84 | L[n+1]=b; |
---|
85 | j=0; |
---|
86 | } |
---|
87 | } |
---|
88 | } |
---|
89 | return(L); |
---|
90 | } |
---|
91 | example |
---|
92 | { "EXAMPLE:"; echo = 2; |
---|
93 | ring r = 0,x,dp; |
---|
94 | list L=-567,-233,446,12,-34,8907; |
---|
95 | bubblesort(L); |
---|
96 | } |
---|
97 | |
---|
98 | proc disc(number N, int k) |
---|
99 | "USAGE: disc(N,k); |
---|
100 | RETURN: list L of negative discriminants D, sort in decreasing order |
---|
101 | ASSUME: D<0, D kongruent 0 or 1 modulo 4 and |D|<4N |
---|
102 | NOTE: D=b^2-4*a, where 0<=b<=k and intPart((b^2)/4)+1<=a<=k for each b |
---|
103 | EXAMPLE:example disc; shows an example |
---|
104 | " |
---|
105 | { |
---|
106 | list L=-3,-4,-7; |
---|
107 | number D; |
---|
108 | number B; |
---|
109 | int a,b; |
---|
110 | for(b=0;b<=k;b++) |
---|
111 | { |
---|
112 | B=b^2; |
---|
113 | for(a=int(intPart(B/4))+1;a<=k;a++) |
---|
114 | { |
---|
115 | D=-4*a+B; |
---|
116 | if((D<0)&&((D mod 4)!=2)&&((D mod 4)!=3)&&(absValue(D)<4*N)&&(newTest(L,D)==1)) |
---|
117 | { |
---|
118 | L[size(L)+1]=D; |
---|
119 | } |
---|
120 | } |
---|
121 | } |
---|
122 | L=bubblesort(L); |
---|
123 | return(L); |
---|
124 | } |
---|
125 | example |
---|
126 | { "EXAMPLE:"; echo = 2; |
---|
127 | ring R = 0,x,dp; |
---|
128 | disc(2003,50); |
---|
129 | } |
---|
130 | |
---|
131 | |
---|
132 | |
---|
133 | proc Cornacchia(number d, number p) |
---|
134 | "USAGE: Cornacchia(d,p); |
---|
135 | RETURN: x,y such that x^2+d*y^2=p with p prime, |
---|
136 | -1, if the Diophantine equation has no solution, |
---|
137 | 0, if the parameters are wrong selected |
---|
138 | ASSUME: 0<d<p |
---|
139 | EXAMPLE:example Cornacchia; shows an example |
---|
140 | " |
---|
141 | { |
---|
142 | if((d<0)||(p<d)) // (0)[Test if assumptions well-defined] |
---|
143 | { |
---|
144 | return(0); |
---|
145 | // ERROR("Parameters wrong selected! It has to be 0<d<p!"); |
---|
146 | } |
---|
147 | else |
---|
148 | { |
---|
149 | number k,x(0),a,b,l,r,c,i; |
---|
150 | |
---|
151 | k=Jacobi(-d,p); // (1)[Test if residue] |
---|
152 | if(k==-1) |
---|
153 | { |
---|
154 | return(-1); |
---|
155 | // ERROR("The Diophantine equation has no solution!"); |
---|
156 | } |
---|
157 | else |
---|
158 | { |
---|
159 | x(0)=squareRoot(-d,p); // (2)[Compute square root] |
---|
160 | if((p/2>=x(0))||(p<=x(0))) |
---|
161 | { |
---|
162 | x(0)=-x(0) mod p; |
---|
163 | } |
---|
164 | a=p; |
---|
165 | b=x(0); |
---|
166 | l=intRoot(p); |
---|
167 | while(b>l) // (3)[Euclidean algorithm] |
---|
168 | { |
---|
169 | r=a mod b; |
---|
170 | a=b; |
---|
171 | b=r; |
---|
172 | } |
---|
173 | c=(p-b^2)/d; // (4)[Test solution] |
---|
174 | i=intRoot(c); |
---|
175 | if((((p-b^2) mod d)!=0)||(c!=i^2)) |
---|
176 | { |
---|
177 | return(-1); |
---|
178 | } |
---|
179 | else |
---|
180 | { |
---|
181 | list L=b,i; |
---|
182 | return(L); |
---|
183 | } |
---|
184 | } |
---|
185 | } |
---|
186 | } |
---|
187 | example |
---|
188 | { "EXAMPLE:"; echo = 2; |
---|
189 | ring R = 0,x,dp; |
---|
190 | Cornacchia(55,9551); |
---|
191 | } |
---|
192 | |
---|
193 | proc CornacchiaModified(number D, number p) |
---|
194 | "USAGE: CornacchiaModified(D,p); |
---|
195 | RETURN: x,y such that x^2+|D|*y^2=p with p prime, |
---|
196 | -1, if the Diophantine equation has no solution, |
---|
197 | 0, if the parameters are wrong selected |
---|
198 | ASSUME: D<0, D kongruent 0 or 1 modulo 4 and |D|<4p |
---|
199 | EXAMPLE:example CornacchiaModified; shows an example |
---|
200 | " |
---|
201 | { |
---|
202 | if((D>=0)||((D mod 4)==2)||((D mod 4)==3)||(absValue(D)>=4*p)) |
---|
203 | {// (0)[Test if assumptions well-defined] |
---|
204 | return(0); |
---|
205 | // ERROR("Parameters wrong selected!"); |
---|
206 | } |
---|
207 | else |
---|
208 | { |
---|
209 | if(p==2) // (1)[Case p=2] |
---|
210 | { |
---|
211 | if((D+8)==intRoot(D+8)^2) |
---|
212 | { |
---|
213 | return(intRoot(D+8),1); |
---|
214 | } |
---|
215 | else |
---|
216 | { |
---|
217 | return(-1); |
---|
218 | // ERROR("The Diophantine equation has no solution!"); |
---|
219 | } |
---|
220 | } |
---|
221 | else |
---|
222 | { |
---|
223 | number k,x(0),a,b,l,r,c; |
---|
224 | k=Jacobi(D,p); // (2)[Test if residue] |
---|
225 | if(k==-1) |
---|
226 | { |
---|
227 | return(-1); |
---|
228 | // ERROR("The Diophantine equation has no solution!"); |
---|
229 | } |
---|
230 | else |
---|
231 | { |
---|
232 | x(0)=squareRoot(D,p); // (3)[Compute square root] |
---|
233 | if((x(0) mod 2)!=(D mod 2)) |
---|
234 | { |
---|
235 | x(0)=p-x(0); |
---|
236 | } |
---|
237 | a=2*p; |
---|
238 | b=x(0); |
---|
239 | l=intRoot(4*p); |
---|
240 | while(b>l) // (4)[Euclidean algorithm] |
---|
241 | { |
---|
242 | r=a mod b; |
---|
243 | a=b; |
---|
244 | b=r; |
---|
245 | } |
---|
246 | c=(4*p-b^2)/absValue(D);// (5)[Test solution] |
---|
247 | if((((4*p-b^2) mod absValue(D))!=0)||(c!=intRoot(c)^2)) |
---|
248 | { |
---|
249 | return(-1); |
---|
250 | // ERROR("The Diophantine equation has no solution!"); |
---|
251 | } |
---|
252 | else |
---|
253 | { |
---|
254 | list L=b,intRoot(c); |
---|
255 | return(L); |
---|
256 | } |
---|
257 | } |
---|
258 | } |
---|
259 | } |
---|
260 | } |
---|
261 | example |
---|
262 | { "EXAMPLE:"; echo = 2; |
---|
263 | ring R = 0,x,dp; |
---|
264 | CornacchiaModified(-107,1319); |
---|
265 | } |
---|
266 | |
---|
267 | proc pFactor1(number n,int B, list P) |
---|
268 | "USAGE: pFactor1(n,B,P); n to be factorized, B a bound , P a list of primes |
---|
269 | RETURN: a list of factors of n or the message: no factor found |
---|
270 | NOTE: Pollard's p-factorization |
---|
271 | creates the product k of powers of primes (bounded by B) from |
---|
272 | the list P with the idea that for a prime divisor p of n p-1|k |
---|
273 | then p devides gcd(a^k-1,n) for some random a |
---|
274 | EXAMPLE:example pFactor1; shows an example |
---|
275 | " |
---|
276 | { |
---|
277 | int i; |
---|
278 | number k=1; |
---|
279 | number w; |
---|
280 | while(i<size(P)) |
---|
281 | { |
---|
282 | i++; |
---|
283 | w=P[i]; |
---|
284 | if(w>B) {break;} |
---|
285 | while(w*P[i]<=B) |
---|
286 | { |
---|
287 | w=w*P[i]; |
---|
288 | } |
---|
289 | k=k*w; |
---|
290 | } |
---|
291 | number a=random(2,2147483629); |
---|
292 | number d=gcdN(powerN(a,k,n)-1,n); |
---|
293 | if((d>1)&&(d<n)){return(d);} |
---|
294 | return(n); |
---|
295 | } |
---|
296 | example |
---|
297 | { "EXAMPLE:"; echo = 2; |
---|
298 | ring R = 0,z,dp; |
---|
299 | list L=primList(1000); |
---|
300 | pFactor1(1241143,13,L); |
---|
301 | number h=10; |
---|
302 | h=h^30+25; |
---|
303 | pFactor1(h,20,L); |
---|
304 | } |
---|
305 | |
---|
306 | proc maximum(list L) |
---|
307 | "USAGE: maximum(list L); |
---|
308 | RETURN: the maximal number contained in list L |
---|
309 | EXAMPLE:example maximum; shows an example |
---|
310 | " |
---|
311 | { |
---|
312 | number max=L[1]; |
---|
313 | int i; |
---|
314 | for(i=2;i<=size(L);i++) |
---|
315 | { |
---|
316 | if(L[i]>max) |
---|
317 | { |
---|
318 | max=L[i]; |
---|
319 | } |
---|
320 | } |
---|
321 | return(max); |
---|
322 | } |
---|
323 | example |
---|
324 | { "EXAMPLE:"; echo = 2; |
---|
325 | ring r = 0,x,dp; |
---|
326 | list L=465,867,1233,4567,776544,233445,2334,556; |
---|
327 | maximum(L); |
---|
328 | } |
---|
329 | |
---|
330 | static proc cmod(number x, number y) |
---|
331 | "USAGE: cmod(x,y); |
---|
332 | RETURN: x mod y |
---|
333 | ASSUME: x,y out of Z and x,y<=2147483647 |
---|
334 | NOTE: this algorithm is a helping procedure to be able to calculate |
---|
335 | x mod y with x,y out of Z while working in the complex field |
---|
336 | EXAMPLE:example cmod; shows an example |
---|
337 | " |
---|
338 | { |
---|
339 | int rest=int(x-y*int(x/y)); |
---|
340 | if(rest<0) |
---|
341 | { |
---|
342 | rest=rest+int(y); |
---|
343 | } |
---|
344 | return(rest); |
---|
345 | } |
---|
346 | example |
---|
347 | { "EXAMPLE:"; echo = 2; |
---|
348 | ring r = (complex,30,i),x,dp; |
---|
349 | number x=-1004456; |
---|
350 | number y=1233; |
---|
351 | cmod(x,y); |
---|
352 | } |
---|
353 | |
---|
354 | proc sqr(number w, int k) |
---|
355 | "USAGE: sqr(w,k); |
---|
356 | RETURN: the square root of w |
---|
357 | ASSUME: w>=0 |
---|
358 | NOTE: k describes the number of decimals being calculated in the real numbers, |
---|
359 | k, intPart(k/5) are inputs for the procedure "nt_solve" |
---|
360 | EXAMPLE:example sqr; shows an example |
---|
361 | " |
---|
362 | { |
---|
363 | poly f=var(1)^2-w; |
---|
364 | def S=basering; |
---|
365 | ring R=(real,k),var(1),dp; |
---|
366 | poly f=imap(S,f); |
---|
367 | ideal I=nt_solve(f,1.1,list(k,int(intPart(k/5)))); |
---|
368 | number c=leadcoef(I[1]); |
---|
369 | setring S; |
---|
370 | number c=imap(R,c); |
---|
371 | return(c); |
---|
372 | } |
---|
373 | example |
---|
374 | { "EXAMPLE:"; echo = 2; |
---|
375 | ring R = (real,60),x,dp; |
---|
376 | number ww=288469650108669535726081; |
---|
377 | sqr(ww,60); |
---|
378 | } |
---|
379 | |
---|
380 | |
---|
381 | |
---|
382 | proc expo(number z, int k) |
---|
383 | "USAGE: expo(z,k); |
---|
384 | RETURN: e^z to the order k |
---|
385 | NOTE: k describes the number of summands being calculated in the exponential power series |
---|
386 | EXAMPLE:example e; shows an example |
---|
387 | " |
---|
388 | { |
---|
389 | number q=1; |
---|
390 | number e=1; |
---|
391 | int n; |
---|
392 | for(n=1;n<=k;n++) |
---|
393 | { |
---|
394 | q=q*z/n; |
---|
395 | e=e+q; |
---|
396 | } |
---|
397 | return(e); |
---|
398 | } |
---|
399 | |
---|
400 | example |
---|
401 | { "EXAMPLE:"; echo = 2; |
---|
402 | ring r = (real,30),x,dp; |
---|
403 | number z=40.35; |
---|
404 | expo(z,1000); |
---|
405 | } |
---|
406 | |
---|
407 | |
---|
408 | |
---|
409 | proc jOft(number t, int k) |
---|
410 | "USAGE: jOft(t,k); |
---|
411 | RETURN: the j-invariant of t |
---|
412 | ASSUME: t is a complex number with positive imaginary part |
---|
413 | NOTE: k describes the number of summands being calculated in the power series, |
---|
414 | 10*k is input for the procedure "e" |
---|
415 | EXAMPLE:example jot; shows an example |
---|
416 | " |
---|
417 | { |
---|
418 | number q1,q2,qr1,qi1,tr,ti,m1,m2,f,j; |
---|
419 | |
---|
420 | number pi=3.1415926535897932384626433832795028841971693993751058209749445923078164062862089986280348253421170679821480865132823066470938446095505822317253594081284811174502841027019385211055596446229489549303819644288109756659334461284756482337867831652712019091456485669234603486104543266482133936072602491412737245870066063155881748815209209628292540917153643678925903600113305305488204665213841469519415116094330572703657595919530921861173819326117931051185480744623799627495673518857527248912279381830119491298336733624406566430860213949463952247371907021798609437027705392171762931767523846748184676694051320005681271452635608277857713427577896091736371787214684409012249534301465495853710507922796892589235420199561121290219608640344181598136297747713099605187072113499999983729780499510597317328160963185950244594553469083026425223082533446850352619311881710100031378387528865875332083814206171776691473035982534904287554687311595628638823537875937519577818577805321712268066130019278766111959092164201989; |
---|
421 | |
---|
422 | tr=repart(t); |
---|
423 | ti=impart(t); |
---|
424 | if(tr==-1/2){qr1=-1;} |
---|
425 | if(tr==0){qr1=1;} |
---|
426 | if((tr!=-1/2)&&(tr!=0)) |
---|
427 | { |
---|
428 | tr=tr-round(tr); |
---|
429 | qr1=expo(2*i*pi*tr,10*k); |
---|
430 | } |
---|
431 | |
---|
432 | qi1=expo(-pi*ti,10*k); |
---|
433 | q1=qr1*qi1^2; |
---|
434 | q2=q1^2; |
---|
435 | |
---|
436 | int n=1; |
---|
437 | while(n<=k) |
---|
438 | { |
---|
439 | m1=m1+(-1)^n*(q1^(n*(3*n-1)/2)+q1^(n*(3*n+1)/2)); |
---|
440 | m2=m2+(-1)^n*(q2^(n*(3*n-1)/2)+q2^(n*(3*n+1)/2)); |
---|
441 | n++; |
---|
442 | } |
---|
443 | |
---|
444 | f=q1*((1+m2)/(1+m1))^24; |
---|
445 | |
---|
446 | j=(256*f+1)^3/f; |
---|
447 | return(j); |
---|
448 | } |
---|
449 | example |
---|
450 | { "EXAMPLE:"; echo = 2; |
---|
451 | ring r = (complex,30,i),x,dp; |
---|
452 | number t=(-7+i*sqr(7,250))/2; |
---|
453 | jOft(t,50); |
---|
454 | } |
---|
455 | |
---|
456 | proc round(number r) |
---|
457 | "USAGE: round(r); |
---|
458 | RETURN: the nearest number to r out of Z |
---|
459 | ASSUME: r should be a rational or a real number |
---|
460 | EXAMPLE:example round; shows an example |
---|
461 | " |
---|
462 | { |
---|
463 | number a=absValue(r); |
---|
464 | number v=r/a; |
---|
465 | |
---|
466 | number d=10; |
---|
467 | int e; |
---|
468 | while(1) |
---|
469 | { |
---|
470 | e=e+1; |
---|
471 | if(a-d^e<0) |
---|
472 | { |
---|
473 | e=e-1; |
---|
474 | break; |
---|
475 | } |
---|
476 | } |
---|
477 | |
---|
478 | number b=a; |
---|
479 | int k; |
---|
480 | for(k=0;k<=e;k++) |
---|
481 | { |
---|
482 | while(1) |
---|
483 | { |
---|
484 | b=b-d^(e-k); |
---|
485 | if(b<0) |
---|
486 | { |
---|
487 | b=b+d^(e-k); |
---|
488 | break; |
---|
489 | } |
---|
490 | } |
---|
491 | } |
---|
492 | |
---|
493 | if(b<1/2) |
---|
494 | { |
---|
495 | return(v*(a-b)); |
---|
496 | } |
---|
497 | else |
---|
498 | { |
---|
499 | return(v*(a+1-b)); |
---|
500 | } |
---|
501 | } |
---|
502 | example |
---|
503 | { "EXAMPLE:"; echo = 2; |
---|
504 | ring R = (real,50),x,dp; |
---|
505 | number r=7357683445788723456321.6788643224; |
---|
506 | round(r); |
---|
507 | } |
---|
508 | |
---|
509 | |
---|
510 | |
---|
511 | proc HilbertClassPoly(number D, int k) |
---|
512 | "USAGE: HilbertClassPoly(D,k); |
---|
513 | RETURN: the monic polynomial of degree h(D) in Z[X] of which jOft((D+sqr(D))/2) is a root |
---|
514 | ASSUME: D is a negative discriminant |
---|
515 | NOTE: k is input for the procedure "jot", |
---|
516 | 5*k is input for the procedure "sqr", |
---|
517 | 10*k describes the number of decimals being calculated in the complex numbers |
---|
518 | EXAMPLE:example HilbertClassPolynomial; shows an example |
---|
519 | " |
---|
520 | { |
---|
521 | if(D>=0) // (0)[Test if assumptions well-defined] |
---|
522 | { |
---|
523 | ERROR("Parameter wrong selected!"); |
---|
524 | } |
---|
525 | else |
---|
526 | { |
---|
527 | def S=basering; |
---|
528 | ring R=0,x,dp; |
---|
529 | |
---|
530 | string s1,s2,s3; |
---|
531 | number a1,b1,t1,g1; |
---|
532 | number D=imap(S,D); |
---|
533 | number B=intRoot(absValue(D)/3); |
---|
534 | |
---|
535 | ring C=(complex,10*k,i),x,dp; |
---|
536 | number D=imap(S,D); |
---|
537 | |
---|
538 | poly P=1; // (1)[Initialize] |
---|
539 | number b=cmod(D,2); |
---|
540 | number B=imap(R,B); |
---|
541 | |
---|
542 | number t,a,g,tau,j; |
---|
543 | list L; |
---|
544 | |
---|
545 | int step=2; |
---|
546 | while(1) |
---|
547 | { |
---|
548 | if(step==2) // (2)[Initialize a] |
---|
549 | { |
---|
550 | t=(b^2-D)/4; |
---|
551 | L=b,1; |
---|
552 | a=maximum(L); |
---|
553 | step=3; |
---|
554 | } |
---|
555 | |
---|
556 | if(step==3) // (3)[Test] |
---|
557 | { |
---|
558 | if((cmod(t,a)!=0)) |
---|
559 | { |
---|
560 | step=4; |
---|
561 | } |
---|
562 | else |
---|
563 | { |
---|
564 | s1=string(a); |
---|
565 | s2=string(b); |
---|
566 | s3=string(t); |
---|
567 | |
---|
568 | setring R; |
---|
569 | execute("a1="+s1+";"); |
---|
570 | execute("b1="+s2+";"); |
---|
571 | execute("t1="+s3+";"); |
---|
572 | g1=gcd(gcd(a1,b1),t1/a1); |
---|
573 | setring C; |
---|
574 | g=imap(R,g1); |
---|
575 | |
---|
576 | if(g!=1) |
---|
577 | { |
---|
578 | step=4; |
---|
579 | } |
---|
580 | else |
---|
581 | { |
---|
582 | tau=(-b+i*sqr(absValue(D),5*k))/(2*a); |
---|
583 | j=jOft(tau,k); |
---|
584 | if((a==b)||(a^2==t)||(b==0)) |
---|
585 | { |
---|
586 | P=P*(var(1)-repart(j)); |
---|
587 | step=4; |
---|
588 | } |
---|
589 | else |
---|
590 | { |
---|
591 | P=P*(var(1)^2-2*repart(j)*var(1)+repart(j)^2+impart(j)^2); |
---|
592 | step=4; |
---|
593 | } |
---|
594 | } |
---|
595 | } |
---|
596 | } |
---|
597 | |
---|
598 | if(step==4) // (4)[Loop on a] |
---|
599 | { |
---|
600 | a=a+1; |
---|
601 | if(a^2<=t) |
---|
602 | { |
---|
603 | step=3; |
---|
604 | continue; |
---|
605 | } |
---|
606 | else |
---|
607 | { |
---|
608 | step=5; |
---|
609 | } |
---|
610 | } |
---|
611 | |
---|
612 | if(step==5) // (5)[Loop on b] |
---|
613 | { |
---|
614 | b=b+2; |
---|
615 | if(b<=B) |
---|
616 | { |
---|
617 | step=2; |
---|
618 | } |
---|
619 | else |
---|
620 | { |
---|
621 | break; |
---|
622 | } |
---|
623 | } |
---|
624 | } |
---|
625 | |
---|
626 | matrix M=coeffs(P,var(1)); |
---|
627 | |
---|
628 | list liste; |
---|
629 | int n; |
---|
630 | for(n=1;n<=nrows(M);n++) |
---|
631 | { |
---|
632 | liste[n]=round(repart(number(M[n,1]))); |
---|
633 | } |
---|
634 | |
---|
635 | poly Q; |
---|
636 | int m; |
---|
637 | for(m=1;m<=size(liste);m++) |
---|
638 | { |
---|
639 | Q=Q+liste[m]*var(1)^(m-1); |
---|
640 | } |
---|
641 | |
---|
642 | string s=string(Q); |
---|
643 | setring S; |
---|
644 | execute("poly Q="+s+";"); |
---|
645 | return(Q); |
---|
646 | } |
---|
647 | } |
---|
648 | example |
---|
649 | { "EXAMPLE:"; echo = 2; |
---|
650 | ring r = 0,x,dp; |
---|
651 | number D=-23; |
---|
652 | HilbertClassPoly(D,50); |
---|
653 | } |
---|
654 | |
---|
655 | proc rootsModp(int p, poly P) |
---|
656 | "USAGE: rootsModp(p,P); |
---|
657 | RETURN: list of roots of the polynomial P modulo p with p prime |
---|
658 | ASSUME: p>=3 |
---|
659 | NOTE: this algorithm will be called recursively, and it is understood |
---|
660 | that all the operations are done in Z/pZ (excepting sqareRoot(d,p)) |
---|
661 | EXAMPLE:example rootsModp; shows an example |
---|
662 | " |
---|
663 | { |
---|
664 | if(p<3) // (0)[Test if assumptions well-defined] |
---|
665 | { |
---|
666 | ERROR("Parameter wrong selected, since p<3!"); |
---|
667 | } |
---|
668 | else |
---|
669 | { |
---|
670 | def S=basering; |
---|
671 | ring R=p,var(1),dp; |
---|
672 | |
---|
673 | poly P=imap(S,P); |
---|
674 | number d; |
---|
675 | int a; |
---|
676 | list L; |
---|
677 | |
---|
678 | poly A=gcd(var(1)^p-var(1),P); // (1)[Isolate roots in Z/pZ] |
---|
679 | if(subst(A,var(1),0)==0) |
---|
680 | { |
---|
681 | L[1]=0; |
---|
682 | A=A/var(1); |
---|
683 | } |
---|
684 | |
---|
685 | if(deg(A)==0) // (2)[Small degree?] |
---|
686 | { |
---|
687 | return(L); |
---|
688 | } |
---|
689 | |
---|
690 | if(deg(A)==1) |
---|
691 | { |
---|
692 | matrix M=coeffs(A,var(1)); |
---|
693 | L[size(L)+1]=-leadcoef(M[1,1])/leadcoef(M[2,1]); |
---|
694 | setring S; |
---|
695 | list L=imap(R,L); |
---|
696 | return(L); |
---|
697 | } |
---|
698 | |
---|
699 | if(deg(A)==2) |
---|
700 | { |
---|
701 | matrix M=coeffs(A,var(1)); |
---|
702 | d=leadcoef(M[2,1])^2-4*leadcoef(M[1,1])*leadcoef(M[3,1]); |
---|
703 | |
---|
704 | ring T=0,var(1),dp; |
---|
705 | number d=imap(R,d); |
---|
706 | number e=squareRoot(d,p); |
---|
707 | setring R; |
---|
708 | number e=imap(T,e); |
---|
709 | |
---|
710 | L[size(L)+1]=(-leadcoef(M[2,1])+e)/(2*leadcoef(M[3,1])); |
---|
711 | L[size(L)+1]=(-leadcoef(M[2,1])-e)/(2*leadcoef(M[3,1])); |
---|
712 | setring S; |
---|
713 | list L=imap(R,L); |
---|
714 | return(L); |
---|
715 | } |
---|
716 | |
---|
717 | poly B=1; // (3)[Random splitting] |
---|
718 | poly C; |
---|
719 | while((deg(B)==0)||(deg(B)==deg(A))) |
---|
720 | { |
---|
721 | a=random(0,p-1); |
---|
722 | B=gcd((var(1)+a)^((p-1)/2)-1,A); |
---|
723 | C=A/B; |
---|
724 | } |
---|
725 | |
---|
726 | setring S; // (4)[Recurse] |
---|
727 | poly B=imap(R,B); |
---|
728 | poly C=imap(R,C); |
---|
729 | list l=L+rootsModp(p,B)+rootsModp(p,C); |
---|
730 | return(l); |
---|
731 | } |
---|
732 | } |
---|
733 | example |
---|
734 | { "EXAMPLE:"; echo = 2; |
---|
735 | ring r = 0,x,dp; |
---|
736 | poly f=x4+2x3-5x2+x; |
---|
737 | rootsModp(7,f); |
---|
738 | poly g=x5+112x4+655x3+551x2+1129x+831; |
---|
739 | rootsModp(1223,g); |
---|
740 | } |
---|
741 | |
---|
742 | proc wUnit(number D) |
---|
743 | "USAGE: wUnit(D); |
---|
744 | RETURN: the number of roots of unity in the quadratic order of discriminant D |
---|
745 | ASSUME: D<0 a discriminant kongruent to 0 or 1 modulo 4 |
---|
746 | EXAMPLE:example w; shows an example |
---|
747 | " |
---|
748 | { |
---|
749 | if((D>=0)||((D mod 4)==2)||((D mod 4)==3)) |
---|
750 | { |
---|
751 | ERROR("Parameter wrong selected!"); |
---|
752 | } |
---|
753 | else |
---|
754 | { |
---|
755 | if(D<-4) {return(2);} |
---|
756 | if(D==-4){return(4);} |
---|
757 | if(D==-3){return(6);} |
---|
758 | } |
---|
759 | } |
---|
760 | example |
---|
761 | { "EXAMPLE:"; echo = 2; |
---|
762 | ring r = 0,x,dp; |
---|
763 | number D=-3; |
---|
764 | wUnit(D); |
---|
765 | } |
---|
766 | |
---|
767 | |
---|
768 | |
---|
769 | proc Atkin(number N, int K, int B) |
---|
770 | "USAGE: Atkin(N,K,B); |
---|
771 | RETURN: 1, if N is prime, |
---|
772 | -1, if N is not prime, |
---|
773 | 0, if the algorithm is not applicable, since there are too little discriminants |
---|
774 | ASSUME: N is coprime to 6 and different from 1 |
---|
775 | NOTE: - K/2 is input for the procedure "disc", |
---|
776 | K is input for the procedure "HilbertClassPolynomial", |
---|
777 | B describes the number of recursions being calculated |
---|
778 | - The basis of the the algorithm is the following theorem: |
---|
779 | Let N be an integer coprime to 6 and different from 1 and E be an |
---|
780 | ellipic curve modulo N. Assume that we know an integer m and a |
---|
781 | point P of E(Z/NZ) satisfying the following conditions. |
---|
782 | (1) There exists a prime divisor q of m such that q>(4-th root(N)+1)^2. |
---|
783 | (2) m*P=O(E)=(0:1:0). |
---|
784 | (3) (m/q)*P=(x:y:t) with t element of (Z/NZ)*. |
---|
785 | Then N is prime. |
---|
786 | EXAMPLE:example Atkin; shows an example |
---|
787 | " |
---|
788 | { |
---|
789 | if(N==1) {return(-1);} // (0)[Test if assumptions well-defined] |
---|
790 | if((N==2)||(N==3)) {return(1);} |
---|
791 | if(gcdN(N,6)!=1) |
---|
792 | { |
---|
793 | if(printlevel>=1) {"ggT(N,6)="+string(gcdN(N,6));pause();} |
---|
794 | return(-1); |
---|
795 | } |
---|
796 | else |
---|
797 | { |
---|
798 | int i; // (1)[Initialize] |
---|
799 | int n(i); |
---|
800 | number N(i)=N; |
---|
801 | if(printlevel>=1) {"Setze i=0, n=0 und N(i)=N(0)="+string(N(i))+".";pause();} |
---|
802 | |
---|
803 | // declarations: |
---|
804 | int j(0),j(1),j(2),j(3),j(4),k; // running indices |
---|
805 | list L; // all primes smaller than 1000 |
---|
806 | list H; // sequence of negative discriminants |
---|
807 | number D; // discriminant out of H |
---|
808 | list L1,L2,S,S1,S2,R; // lists of relevant elements |
---|
809 | list P,P1,P2; // elliptic points on E(Z/N(i)Z) |
---|
810 | number m,q; // m=|E(Z/N(i)Z)| and q|m |
---|
811 | number a,b,j,c; // characterize E(Z/N(i)Z) |
---|
812 | number g,u; // g out of Z/N(i)Z, u=Jacobi(g,N(i)) |
---|
813 | poly T; // T=HilbertClassPoly(D,K) |
---|
814 | matrix M; // M contains the coefficients of T |
---|
815 | |
---|
816 | if(printlevel>=1) {"Liste H der moeglichen geeigneten Diskriminanten wird berechnet.";} |
---|
817 | H=disc(N,K/2); |
---|
818 | if(printlevel>=1) {"H="+string(H);pause();} |
---|
819 | |
---|
820 | int step=2; |
---|
821 | while(1) |
---|
822 | { |
---|
823 | if(step==2) // (2)[Is N(i) small??] |
---|
824 | { |
---|
825 | L=5,7,11,13,17,19,23,29,31,37,41,43,47,53,59,61,67,71,73,79,83,89,97,101,103,107,109,113,127,131,137,139,149,151,157,163,167,173,179,181,191,193,197,199,211,223,227,229,233,239,241,251,257,263,269,271,277,281,283,293,307,311,313,317,331,337,347,349,353,359,367,373,379,383,389,397,401,409,419,421,431,433,439,443,449,457,461,463,467,479,487,491,499,503,509,521,523,541,547,557,563,569,571,577,587,593,599,601,607,613,617,619,631,641,643,647,653,659,661,673,677,683,691,701,709,719,727,733,739,743,751,757,761,769,773,787,797,809,811,821,823,827,829,839,853,857,859,863,877,881,883,887,907,911,919,929,937,941,947,953,967,971,977,983,991,997; |
---|
826 | for(j(0)=1;j(0)<=size(L);j(0)++) |
---|
827 | { |
---|
828 | if(N(i)==L[j(0)]){return(1);} |
---|
829 | if(((N(i) mod L[j(0)])==0)&&(N(i)!=L[j(0)])) |
---|
830 | { |
---|
831 | if(printlevel>=1) {"N("+string(i)+")="+string(N(i))+" ist durch "+string(L[j(0)])+" teilbar.";pause();} |
---|
832 | step=14; |
---|
833 | break; |
---|
834 | } |
---|
835 | } |
---|
836 | if(step==2) |
---|
837 | { |
---|
838 | step=3; |
---|
839 | } |
---|
840 | } |
---|
841 | |
---|
842 | if(step==3) // (3)[Choose next discriminant] |
---|
843 | { |
---|
844 | n(i)=n(i)+1; |
---|
845 | if(n(i)==size(H)+1) |
---|
846 | { |
---|
847 | if(printlevel>=1) {"Algorithmus nicht anwendbar, da zu wenige geeignete Diskriminanten existieren."; |
---|
848 | "Erhoehe den Genauigkeitsparameter K und starte den Algorithmus erneut.";pause();} |
---|
849 | return(0); |
---|
850 | } |
---|
851 | D=H[n(i)]; |
---|
852 | if(printlevel>=1) {"Naechste Diskriminante D wird gewaehlt. D="+string(D)+".";pause();} |
---|
853 | if(Jacobi(D,N(i))!=1) |
---|
854 | { |
---|
855 | if(printlevel>=1) {"Jacobi(D,N("+string(i)+"))="+string(Jacobi(D,N(i)));pause();} |
---|
856 | continue; |
---|
857 | } |
---|
858 | else |
---|
859 | { |
---|
860 | L1=CornacchiaModified(D,N(i)); |
---|
861 | if(size(L1)>1) |
---|
862 | { |
---|
863 | if(printlevel>=1) {"Die Loesung (x,y) der Gleichung x^2+|D|y^2=4N("+string(i)+") lautet";L1;pause();} |
---|
864 | step=4; |
---|
865 | } |
---|
866 | else |
---|
867 | { |
---|
868 | if(L1[1]==-1) |
---|
869 | { |
---|
870 | if(printlevel>=1) {"Die Gleichung x^2+|D|y^2=4N("+string(i)+") hat keine Loesung.";pause();} |
---|
871 | continue; |
---|
872 | } |
---|
873 | if(L1[1]==0) |
---|
874 | { |
---|
875 | if(printLevel>=1) {"Algorithmus fuer N("+string(i)+")="+string(N(i))+" nicht anwendbar,"; |
---|
876 | "da zu wenige geeignete Diskriminanten existieren.";pause();} |
---|
877 | step=14; |
---|
878 | } |
---|
879 | } |
---|
880 | } |
---|
881 | } |
---|
882 | |
---|
883 | if(step==4) // (4)[Factor m] |
---|
884 | { |
---|
885 | if(printlevel>=1) {"Die Liste L2 der moeglichen m=|E(Z/N("+string(i)+")Z)| wird berechnet.";} |
---|
886 | if(absValue(L1[1])^2<=4*N(i)) {L2=N(i)+1+L1[1],N(i)+1-L1[1];} |
---|
887 | if(D==-4) |
---|
888 | { |
---|
889 | if(absValue(2*L1[2])^2<=4*N(i)) {L2[size(L2)+1]=N(i)+1+2*L1[2]; |
---|
890 | L2[size(L2)+1]=N(i)+1-2*L1[2];} |
---|
891 | } |
---|
892 | // An dieser Stelle wurde "<=4*N(i)" durch "<=16*N(i)" ersetzt. |
---|
893 | if(D==-3) |
---|
894 | { |
---|
895 | if(absValue(L1[1]+3*L1[2])^2<=16*N(i)) {L2[size(L2)+1]=N(i)+1+(L1[1]+3*L1[2])/2; |
---|
896 | L2[size(L2)+1]=N(i)+1-(L1[1]+3*L1[2])/2;} |
---|
897 | if(absValue(L1[1]-3*L1[2])^2<=16*N(i)) {L2[size(L2)+1]=N(i)+1+(L1[1]-3*L1[2])/2; |
---|
898 | L2[size(L2)+1]=N(i)+1-(L1[1]-3*L1[2])/2;} |
---|
899 | } |
---|
900 | /////////////////////////////////////////////////////////////// |
---|
901 | if(size(L2)==0) |
---|
902 | { |
---|
903 | if(printlevel>=1) {"Nach dem Satz von Hasse wurden keine moeglichen m=|E(Z/N("+string(i)+")Z)|"; |
---|
904 | "fuer D="+string(D)+" gefunden.";} |
---|
905 | step=3; |
---|
906 | continue; |
---|
907 | } |
---|
908 | else |
---|
909 | { |
---|
910 | if(printlevel>=1) {"L2=";L2;pause();} |
---|
911 | } |
---|
912 | |
---|
913 | if(printlevel>=1) {"Die Liste S der Faktoren aller moeglichen m wird berechnet.";} |
---|
914 | S=list(); |
---|
915 | for(j(1)=1;j(1)<=size(L2);j(1)++) |
---|
916 | { |
---|
917 | m=L2[j(1)]; |
---|
918 | if(m!=0) |
---|
919 | { |
---|
920 | S1=PollardRho(m,10000,1,L); |
---|
921 | S2=pFactor(m,100,L); |
---|
922 | S[size(S)+1]=list(m,S1+S2); |
---|
923 | } |
---|
924 | } |
---|
925 | if(printlevel>=1) {"S=";S;pause();} |
---|
926 | step=5; |
---|
927 | } |
---|
928 | |
---|
929 | if(step==5) // (5)[Does a suitable m exist??] |
---|
930 | { |
---|
931 | for(j(2)=1;j(2)<=size(S);j(2)++) |
---|
932 | { |
---|
933 | m=L2[j(2)]; |
---|
934 | for(j(3)=1;j(3)<=size(S[j(2)][2]);j(3)++) |
---|
935 | { |
---|
936 | q=S[j(2)][2][j(3)]; |
---|
937 | // sqr(sqr(N(i),50),50) ersetzt intRoot(intRoot(N(i))) |
---|
938 | if((q>(sqr(sqr(N(i),50),50)+1)^2) && (MillerRabin(q,5)==1)) |
---|
939 | { |
---|
940 | step=6; |
---|
941 | break; |
---|
942 | } |
---|
943 | ////////////////////////////////////////////////////// |
---|
944 | } |
---|
945 | if(step==6) |
---|
946 | { |
---|
947 | if(printlevel>=1) {"Geeignetes Paar (m,q) gefunden, so dass q|m,"; |
---|
948 | "q>(4-th root(N("+string(i)+"))+1)^2 und q den Miller-Rabin-Test passiert."; |
---|
949 | "m="+string(m)+",";"q="+string(q);pause();} |
---|
950 | break; |
---|
951 | } |
---|
952 | else |
---|
953 | { |
---|
954 | step=3; |
---|
955 | } |
---|
956 | } |
---|
957 | if(step==3) |
---|
958 | { |
---|
959 | if(printlevel>=1) {"Kein geeignetes Paar (m,q), so dass q|m,"; |
---|
960 | "q>(4-th root(N("+string(i)+"))+1)^2 und q den Miller-Rabin-Test passiert, gefunden."; |
---|
961 | pause();} |
---|
962 | continue; |
---|
963 | } |
---|
964 | } |
---|
965 | |
---|
966 | if(step==6) // (6)[Compute elliptic curve] |
---|
967 | { |
---|
968 | if(D==-4) |
---|
969 | { |
---|
970 | a=-1; |
---|
971 | b=0; |
---|
972 | if(printlevel>=1) {"Da D=-4, setze a=-1 und b=0.";pause();} |
---|
973 | } |
---|
974 | if(D==-3) |
---|
975 | { |
---|
976 | a=0; |
---|
977 | b=-1; |
---|
978 | if(printlevel>=1) {"Da D=-3, setze a=0 und b=-1.";pause();} |
---|
979 | } |
---|
980 | if(D<-4) |
---|
981 | { |
---|
982 | if(printlevel>=1) {"Das Minimalpolynom T von j((D+sqr(D))/2) aus Z[X] fuer D="+string(D)+" wird berechnet.";} |
---|
983 | T=HilbertClassPoly(D,K); |
---|
984 | if(printlevel>=1) {"T="+string(T);pause();} |
---|
985 | |
---|
986 | M=coeffs(T,var(1)); |
---|
987 | T=0; |
---|
988 | |
---|
989 | for(j(4)=1;j(4)<=nrows(M);j(4)++) |
---|
990 | { |
---|
991 | M[j(4),1]=leadcoef(M[j(4),1]) mod N(i); |
---|
992 | T=T+M[j(4),1]*var(1)^(j(4)-1); |
---|
993 | } |
---|
994 | if(printlevel>=1) {"Setze T=T mod N("+string(i)+").";"T="+string(T);pause();} |
---|
995 | |
---|
996 | R=rootsModp(int(N(i)),T); |
---|
997 | if(deg(T)>size(R)) |
---|
998 | { |
---|
999 | ERROR("Das Polynom T zerfaellt modulo N("+string(i)+") nicht vollstaendig in Linearfaktoren." |
---|
1000 | "Erhoehe den Genauigkeitsparameter K und starte den Algorithmus erneut."); |
---|
1001 | } |
---|
1002 | if(printlevel>=1) {if(deg(T)>1) {"Die "+string(deg(T))+" Nullstellen von T modulo N("+string(i)+") sind"; |
---|
1003 | R;pause();} |
---|
1004 | if(deg(T)==1){"Die Nullstelle von T modulo N("+string(i)+") ist";R;pause();}} |
---|
1005 | |
---|
1006 | j=R[1]; |
---|
1007 | c=j*exgcdN(j-1728,N(i))[1]; |
---|
1008 | a=-3*c mod N(i); |
---|
1009 | b=2*c mod N(i); |
---|
1010 | if(printlevel>=1) {"Waehle die Nullstelle j="+string(j)+" aus und setze";"c=j/(j-1728) mod N("+string(i)+"), a=-3c mod N("+string(i)+"), b=2c mod N("+string(i)+")."; |
---|
1011 | "a="+string(a)+",";"b="+string(b);pause();} |
---|
1012 | } |
---|
1013 | step=7; |
---|
1014 | } |
---|
1015 | |
---|
1016 | if(step==7) // (7)[Find g] |
---|
1017 | { |
---|
1018 | if(D==-3) |
---|
1019 | { |
---|
1020 | while(1) |
---|
1021 | { |
---|
1022 | g=random(1,2147483647) mod N(i); |
---|
1023 | u=Jacobi(g,N(i)); |
---|
1024 | if((u==-1)&&(powerN(g,(N(i)-1)/3,N(i))!=1)) |
---|
1025 | { |
---|
1026 | if(printlevel>=1) {"g="+string(g);pause();} |
---|
1027 | break; |
---|
1028 | } |
---|
1029 | } |
---|
1030 | } |
---|
1031 | else |
---|
1032 | { |
---|
1033 | while(1) |
---|
1034 | { |
---|
1035 | g=random(1,2147483647) mod N(i); |
---|
1036 | u=Jacobi(g,N(i)); |
---|
1037 | if(u==-1) |
---|
1038 | { |
---|
1039 | if(printlevel>=1) {"g="+string(g);pause();} |
---|
1040 | break; |
---|
1041 | } |
---|
1042 | } |
---|
1043 | } |
---|
1044 | step=8; |
---|
1045 | } |
---|
1046 | |
---|
1047 | if(step==8) // (8)[Find P] |
---|
1048 | { |
---|
1049 | if(printlevel>=1) {"Ein zufaelliger Punkt P auf der Elliptischen Kurve"; |
---|
1050 | "mit der Gleichung y^2=x^3+ax+b fuer";"N("+string(i)+")="+string(N(i))+","; |
---|
1051 | " a="+string(a)+",";" b="+string(b);"wird gewaehlt.";} |
---|
1052 | P=ellipticRandomPoint(N(i),a,b); |
---|
1053 | if(printlevel>=1) {"P=("+string(P)+")";pause();} |
---|
1054 | |
---|
1055 | if(size(P)==1) |
---|
1056 | { |
---|
1057 | step=14; |
---|
1058 | } |
---|
1059 | else |
---|
1060 | { |
---|
1061 | step=9; |
---|
1062 | } |
---|
1063 | } |
---|
1064 | |
---|
1065 | if(step==9) // (9)[Find right curve] |
---|
1066 | { |
---|
1067 | if(printlevel>=1) {"Die Punkte P2=(m/q)*P und P1=q*P2 auf der Kurve werden berechnet.";} |
---|
1068 | P2=ellipticMult(N(i),a,b,P,m/q); |
---|
1069 | P1=ellipticMult(N(i),a,b,P2,q); |
---|
1070 | if(printlevel>=1) {"P1=("+string(P1)+"),";"P2=("+string(P2)+")";pause();} |
---|
1071 | |
---|
1072 | if((P1[1]==0)&&(P1[2]==1)&&(P1[3]==0)) |
---|
1073 | { |
---|
1074 | step=12; |
---|
1075 | } |
---|
1076 | else |
---|
1077 | { |
---|
1078 | if(printlevel>=1) {"Da P1!=(0:1:0), ist fuer die Koeffizienten a="+string(a)+" und b="+string(b)+" m!=|E(Z/N("+string(i)+")Z)|."; |
---|
1079 | "Waehle daher neue Koeffizienten a und b.";pause();} |
---|
1080 | step=10; |
---|
1081 | } |
---|
1082 | } |
---|
1083 | |
---|
1084 | if(step==10) // (10)[Change coefficients] |
---|
1085 | { |
---|
1086 | k=k+1; |
---|
1087 | if(k>=wUnit(D)) |
---|
1088 | { |
---|
1089 | if(printlevel>=1) {"Da k=wUnit(D)="+string(k)+", ist N("+string(i)+")="+string(N(i))+" nicht prim.";pause();} |
---|
1090 | step=14; |
---|
1091 | } |
---|
1092 | else |
---|
1093 | { |
---|
1094 | if(D<-4) {a=a*g^2 mod N(i); b=b*g^3 mod N(i); |
---|
1095 | if(printlevel>=1) {"Da D<-4, setze a=a*g^2 mod N("+string(i)+") und b=b*g^3 mod N("+string(i)+")."; |
---|
1096 | "a="+string(a)+",";"b="+string(b)+",";"k="+string(k);pause();}} |
---|
1097 | if(D==-4){a=a*g mod N(i); |
---|
1098 | if(printlevel>=1) {"Da D=-4, setze a=a*g mod N("+string(i)+").";"a="+string(a)+","; |
---|
1099 | "b="+string(b)+",";"k="+string(k);pause();}} |
---|
1100 | if(D==-3){b=b*g mod N(i); |
---|
1101 | if(printlevel>=1) {"Da D=-3, setze b=b*g mod N("+string(i)+").";"a="+string(a)+","; |
---|
1102 | "b="+string(b)+",";"k="+string(k);pause();}} |
---|
1103 | step=8; |
---|
1104 | continue; |
---|
1105 | } |
---|
1106 | } |
---|
1107 | |
---|
1108 | if(step==11) // (11)[Find a new P] |
---|
1109 | { |
---|
1110 | if(printlevel>=1) {"Ein neuer zufaelliger Punkt P auf der Elliptischen Kurve wird gewaehlt,"; |
---|
1111 | "da auch P2=(0:1:0).";} |
---|
1112 | P=ellipticRandomPoint(N(i),a,b); |
---|
1113 | if(printlevel>=1) {"P=("+string(P)+")";pause();} |
---|
1114 | |
---|
1115 | if(size(P)==1) |
---|
1116 | { |
---|
1117 | step=14; |
---|
1118 | } |
---|
1119 | else |
---|
1120 | { |
---|
1121 | if(printlevel>=1) {"Die Punkte P2=(m/q)*P und P1=q*P2 auf der Kurve werden berechnet.";} |
---|
1122 | P2=ellipticMult(N(i),a,b,P,m/q); |
---|
1123 | P1=ellipticMult(N(i),a,b,P2,q); |
---|
1124 | if(printlevel>=1) {"P1=("+string(P1)+"),";"P2=("+string(P2)+")";pause();} |
---|
1125 | |
---|
1126 | if((P1[1]!=0)||(P1[2]!=1)||(P1[3]!=0)) |
---|
1127 | { |
---|
1128 | if(printlevel>=1) {"Da P1!=(0:1:0), ist, fuer die Koeffizienten a="+string(a)+" und b="+string(b)+", m!=|E(Z/N("+string(i)+")Z)|."; |
---|
1129 | "Waehle daher neue Koeffizienten a und b.";pause();} |
---|
1130 | step=10; |
---|
1131 | continue; |
---|
1132 | } |
---|
1133 | else |
---|
1134 | { |
---|
1135 | step=12; |
---|
1136 | } |
---|
1137 | } |
---|
1138 | } |
---|
1139 | |
---|
1140 | if(step==12) // (12)[Check P] |
---|
1141 | { |
---|
1142 | if((P2[1]==0)&&(P2[2]==1)&&(P2[3]==0)) |
---|
1143 | { |
---|
1144 | step=11; |
---|
1145 | continue; |
---|
1146 | } |
---|
1147 | else |
---|
1148 | { |
---|
1149 | step=13; |
---|
1150 | } |
---|
1151 | } |
---|
1152 | |
---|
1153 | if(step==13) // (13)[Recurse] |
---|
1154 | { |
---|
1155 | if(i<B) |
---|
1156 | { |
---|
1157 | if(printlevel>=1) {string(i+1)+". Rekursion:";""; |
---|
1158 | "N("+string(i)+")="+string(N(i))+" erfuellt die Bedingungen des zugrunde liegenden Satzes,"; |
---|
1159 | "da P1=(0:1:0) und P2[3] aus (Z/N("+string(i)+")Z)*.";""; |
---|
1160 | "Untersuche nun, ob auch der gefundene Faktor q="+string(q)+" diese Bedingungen erfuellt."; |
---|
1161 | "Setze dazu i=i+1, N("+string(i+1)+")=q="+string(q)+" und beginne den Algorithmus von vorne.";pause();} |
---|
1162 | i=i+1; |
---|
1163 | int n(i); |
---|
1164 | number N(i)=q; |
---|
1165 | k=0; |
---|
1166 | step=2; |
---|
1167 | continue; |
---|
1168 | } |
---|
1169 | else |
---|
1170 | { |
---|
1171 | if(printlevel>=1) {"N(B)=N("+string(i)+")="+string(N(i))+" erfuellt die Bedingungen des zugrunde liegenden Satzes,"; |
---|
1172 | "da P1=(0:1:0) und P2[3] aus (Z/N("+string(i)+")Z)*."; |
---|
1173 | "Insbesondere ist N="+string(N)+" prim.";pause();} |
---|
1174 | return(1); |
---|
1175 | } |
---|
1176 | } |
---|
1177 | |
---|
1178 | if(step==14) // (14)[Backtrack] |
---|
1179 | { |
---|
1180 | if(i>0) |
---|
1181 | { |
---|
1182 | if(printlevel>=1) {"Setze i=i-1 und starte den Algorithmus fuer N("+string(i-1)+")="+string(N(i-1))+" mit"; |
---|
1183 | "neuer Diskriminanten von vorne.";pause();} |
---|
1184 | i=i-1; |
---|
1185 | k=0; |
---|
1186 | step=3; |
---|
1187 | } |
---|
1188 | else |
---|
1189 | { |
---|
1190 | if(printlevel>=1) {"N(0)=N="+string(N)+" und daher ist N nicht prim.";pause(n);} |
---|
1191 | return(-1); |
---|
1192 | } |
---|
1193 | } |
---|
1194 | } |
---|
1195 | } |
---|
1196 | } |
---|
1197 | example |
---|
1198 | { "EXAMPLE:"; echo = 2; |
---|
1199 | ring R = 0,x,dp; |
---|
1200 | printlevel=1; |
---|
1201 | Atkin(7691,100,5); |
---|
1202 | Atkin(8543,100,4); |
---|
1203 | Atkin(100019,100,5); |
---|
1204 | Atkin(10000079,100,2); |
---|
1205 | } |
---|
1206 | |
---|