return (MP_COPY(PFILT_F(o)->m));
else if (ECPT_PYCHECK(o)) {
ec p = EC_INIT;
+ if (EC_ATINF(ECPT_P(o))) return (0);
getecptout(&p, o);
x = MP_COPY(p.x);
EC_DESTROY(&p);
mp *z;
mp_pyobj *zz = 0;
int radix = 0;
- char *kwlist[] = { "x", "radix", 0 };
+ static const char *const kwlist[] = { "x", "radix", 0 };
- if (!PyArg_ParseTupleAndKeywords(arg, kw, "O|i:new", kwlist, &x, &radix))
+ if (!PyArg_ParseTupleAndKeywords(arg, kw, "O|i:new", KWLIST, &x, &radix))
goto end;
if (MP_PYCHECK(x)) RETURN_OBJ(x);
if (!good_radix_p(radix, 1)) VALERR("bad radix");
static PyObject *mpmeth_tostring(PyObject *me, PyObject *arg, PyObject *kw)
{
int radix = 10;
- char *kwlist[] = { "radix", 0 };
- if (!PyArg_ParseTupleAndKeywords(arg, kw, "|i:tostring", kwlist, &radix))
+ static const char *const kwlist[] = { "radix", 0 };
+ if (!PyArg_ParseTupleAndKeywords(arg, kw, "|i:tostring", KWLIST, &radix))
goto end;
if (!good_radix_p(radix, 0)) VALERR("bad radix");
return (mp_topystring(MP_X(me), radix, 0, 0, 0));
PyObject *arg, PyObject *kw) \
{ \
long len = -1; \
- char *kwlist[] = { "len", 0 }; \
+ static const char *const kwlist[] = { "len", 0 }; \
PyObject *rc = 0; \
\
if (!PyArg_ParseTupleAndKeywords(arg, kw, "|l:" #name, \
- kwlist, &len)) \
+ KWLIST, &len)) \
goto end; \
if (len < 0) { \
len = mp_octets##c(MP_X(me)); \
static PyObject *mpmeth_primep(PyObject *me, PyObject *arg, PyObject *kw)
{
grand *r = &rand_global;
- char *kwlist[] = { "rng", 0 };
+ static const char *const kwlist[] = { "rng", 0 };
PyObject *rc = 0;
- if (!PyArg_ParseTupleAndKeywords(arg, kw, "|O&", kwlist, convgrand, &r))
+ if (!PyArg_ParseTupleAndKeywords(arg, kw, "|O&", KWLIST, convgrand, &r))
goto end;
rc = getbool(pgen_primep(MP_X(me), r));
end:
METH (modsqrt, "X.modsqrt(Y) -> square root of Y mod X, if X prime")
METH (leastcongruent,
"X.leastcongruent(B, M) -> smallest Z >= B with Z == X (mod M)")
- KWMETH(primep, "X.primep(rng = rand) -> true/false if X is prime")
- KWMETH(tostring, "X.tostring(radix = 10) -> STR")
- KWMETH(storel, "X.storel(len = -1) -> little-endian bytes")
- KWMETH(storeb, "X.storeb(len = -1) -> big-endian bytes")
+ KWMETH(primep, "X.primep([rng = rand]) -> true/false if X is prime")
+ KWMETH(tostring, "X.tostring([radix = 10]) -> STR")
+ KWMETH(storel, "X.storel([len = -1]) -> little-endian bytes")
+ KWMETH(storeb, "X.storeb([len = -1]) -> big-endian bytes")
KWMETH(storel2c,
- "X.storel2c(len = -1) -> little-endian bytes, two's complement")
+ "X.storel2c([len = -1]) -> little-endian bytes, two's complement")
KWMETH(storeb2c,
- "X.storeb2c(len = -1) -> big-endian bytes, two's complement")
+ "X.storeb2c([len = -1]) -> big-endian bytes, two's complement")
METH (tobuf, "X.tobuf() -> buffer format")
#undef METHNAME
{ 0 }
implicit conversions from `PrimeFilter', and other objects which\n\
convert to `long'.\n\
\n\
-Constructor MP(X, radix = R) attempts to convert X to an `MP'. If\n\
+Constructor MP(X, [radix = R]) attempts to convert X to an `MP'. If\n\
X is a string, it's read in radix-R form, or we look for a prefix\n\
if R = 0. Other acceptable things are field elements, elliptic curve\n\
points, group elements, Python `int' and `long' objects, and anything\n\
\n\
Notes:\n\
\n\
- * Use `//' for division. MPs don't have `/' division.",
+ * Use `//' for integer division. `/' gives exact rational division.",
0, /* @tp_traverse@ */
0, /* @tp_clear@ */
PyObject *z = 0;
mp *zz;
mptext_stringctx sc;
- char *kwlist[] = { "class", "x", "radix", 0 };
+ static const char *const kwlist[] = { "class", "x", "radix", 0 };
if (!PyArg_ParseTupleAndKeywords(arg, kw, "Os#|i:fromstring",
- kwlist, &me, &p, &len, &r))
+ KWLIST, &me, &p, &len, &r))
goto end;
if (!good_radix_p(r, 1)) VALERR("bad radix");
sc.buf = p; sc.lim = p + len;
Py_TPFLAGS_BASETYPE,
/* @tp_doc@ */
-"An object for multiplying many small integers.",
+"MPMul(N_0, N_1, ....): an object for multiplying many small integers.",
0, /* @tp_traverse@ */
0, /* @tp_clear@ */
static PyObject *mpmont_pynew(PyTypeObject *ty, PyObject *arg, PyObject *kw)
{
mpmont_pyobj *mm = 0;
- char *kwlist[] = { "m", 0 };
+ static const char *const kwlist[] = { "m", 0 };
mp *xx = 0;
- if (!PyArg_ParseTupleAndKeywords(arg, kw, "O&:new", kwlist, convmp, &xx))
+ if (!PyArg_ParseTupleAndKeywords(arg, kw, "O&:new", KWLIST, convmp, &xx))
goto end;
if (!MP_POSP(xx) || !MP_ODDP(xx)) VALERR("m must be positive and odd");
mm = (mpmont_pyobj *)ty->tp_alloc(ty, 0);
Py_TPFLAGS_BASETYPE,
/* @tp_doc@ */
-"A Montgomery reduction context.",
+"MPMont(N): a Montgomery reduction context.",
0, /* @tp_traverse@ */
0, /* @tp_clear@ */
PyObject *arg, PyObject *kw)
{
mpbarrett_pyobj *mb = 0;
- char *kwlist[] = { "m", 0 };
+ static const char *const kwlist[] = { "m", 0 };
mp *xx = 0;
- if (!PyArg_ParseTupleAndKeywords(arg, kw, "O&:new", kwlist, convmp, &xx))
+ if (!PyArg_ParseTupleAndKeywords(arg, kw, "O&:new", KWLIST, convmp, &xx))
goto end;
if (!MP_POSP(xx)) VALERR("m must be positive");
mb = (mpbarrett_pyobj *)ty->tp_alloc(ty, 0);
Py_TPFLAGS_BASETYPE,
/* @tp_doc@ */
-"A Barrett reduction context.",
+"MPBarrett(N): a Barrett reduction context.",
0, /* @tp_traverse@ */
0, /* @tp_clear@ */
{
mpreduce_pyobj *mr = 0;
mpreduce r;
- char *kwlist[] = { "m", 0 };
+ static const char *const kwlist[] = { "m", 0 };
mp *xx = 0;
- if (!PyArg_ParseTupleAndKeywords(arg, kw, "O&:new", kwlist, convmp, &xx))
+ if (!PyArg_ParseTupleAndKeywords(arg, kw, "O&:new", KWLIST, convmp, &xx))
goto end;
if (!MP_POSP(xx)) VALERR("m must be positive");
if (mpreduce_create(&r, xx)) VALERR("bad modulus (must be 2^k - ...)");
Py_TPFLAGS_BASETYPE,
/* @tp_doc@ */
-"A reduction context for reduction modulo primes of special form.",
+"MPReduce(N): a reduction context for reduction modulo Solinas primes.",
0, /* @tp_traverse@ */
0, /* @tp_clear@ */
PyObject *q = 0, *x, *z = 0;
mp *xx;
mp **v = 0;
- int i = 0, n = c->k;
+ Py_ssize_t i = 0, n = c->k;
Py_INCREF(me);
if (PyTuple_Size(arg) == n)
goto end;
Py_INCREF(q);
if (!PySequence_Check(q)) TYERR("want a sequence of residues");
- if (PySequence_Size(q) != n) VALERR("residue count mismatch");
+ i = PySequence_Size(q); if (i < 0) goto end;
+ if (i != n) VALERR("residue count mismatch");
v = xmalloc(n * sizeof(*v));
for (i = 0; i < n; i++) {
if ((x = PySequence_GetItem(q, i)) == 0) goto end;
static PyObject *mpcrt_pynew(PyTypeObject *ty, PyObject *arg, PyObject *kw)
{
mpcrt_mod *v = 0;
- int n, i = 0;
- char *kwlist[] = { "mv", 0 };
+ Py_ssize_t n, i = 0, j;
+ static const char *const kwlist[] = { "mv", 0 };
PyObject *q = 0, *x;
- mp *xx;
+ mp *xx = MP_NEW, *y = MP_NEW, *g = MP_NEW;
+ mpmul mm;
mpcrt_pyobj *c = 0;
if (PyTuple_Size(arg) > 1)
q = arg;
- else if (!PyArg_ParseTupleAndKeywords(arg, kw, "O:new", kwlist, &q))
+ else if (!PyArg_ParseTupleAndKeywords(arg, kw, "O:new", KWLIST, &q))
goto end;
Py_INCREF(q);
if (!PySequence_Check(q)) TYERR("want a sequence of moduli");
- n = PySequence_Size(q);
- if (PyErr_Occurred()) goto end;
+ n = PySequence_Size(q); if (n < 0) goto end;
if (!n) VALERR("want at least one modulus");
v = xmalloc(n * sizeof(*v));
for (i = 0; i < n; i++) {
if ((x = PySequence_GetItem(q, i)) == 0) goto end;
xx = getmp(x); Py_DECREF(x); if (!xx) goto end;
- v[i].m = xx; v[i].n = 0; v[i].ni = 0; v[i].nni = 0;
+ if (MP_CMP(xx, <=, MP_ZERO)) VALERR("moduli must be positive");
+ v[i].m = xx; v[i].n = 0; v[i].ni = 0; v[i].nni = 0; xx = MP_NEW;
+ }
+ mpmul_init(&mm);
+ for (j = 0; j < i; j++) mpmul_add(&mm, v[j].m);
+ xx = mpmul_done(&mm);
+ for (j = 0; j < i; j++) {
+ mp_div(&y, 0, xx, v[j].m);
+ mp_gcd(&g, 0, 0, y, v[j].m);
+ if (!MP_EQ(g, MP_ONE)) VALERR("moduli must be pairwise coprime");
}
+
c = (mpcrt_pyobj *)ty->tp_alloc(ty, 0);
mpcrt_create(&c->c, v, n, 0);
Py_DECREF(q);
+ mp_drop(xx); mp_drop(y); mp_drop(g);
return ((PyObject *)c);
end:
xfree(v);
}
Py_XDECREF(q);
+ mp_drop(xx); mp_drop(y); mp_drop(g);
return (0);
}
Py_TPFLAGS_BASETYPE,
/* @tp_doc@ */
-"A context for the solution of Chinese Remainder Theorem problems.",
+"MPCRT(SEQ): a context for solving Chinese Remainder Theorem problems.",
0, /* @tp_traverse@ */
0, /* @tp_clear@ */
mp *z;
mp_pyobj *zz = 0;
int radix = 0;
- char *kwlist[] = { "x", "radix", 0 };
+ static const char *const kwlist[] = { "x", "radix", 0 };
- if (!PyArg_ParseTupleAndKeywords(arg, kw, "O|i:gf", kwlist, &x, &radix))
+ if (!PyArg_ParseTupleAndKeywords(arg, kw, "O|i:gf", KWLIST, &x, &radix))
goto end;
if (GF_PYCHECK(x)) RETURN_OBJ(x);
if (!good_radix_p(radix, 1)) VALERR("radix out of range");
METH (irreduciblep, "X.irreduciblep() -> true/false")
#undef METHNAME
#define METHNAME(func) mpmeth_##func
- KWMETH(tostring, "X.tostring(radix = 10) -> STR")
- KWMETH(storel, "X.storel(len = -1) -> little-endian bytes")
- KWMETH(storeb, "X.storeb(len = -1) -> big-endian bytes")
+ KWMETH(tostring, "X.tostring([radix = 10]) -> STR")
+ KWMETH(storel, "X.storel([len = -1]) -> little-endian bytes")
+ KWMETH(storeb, "X.storeb([len = -1]) -> big-endian bytes")
KWMETH(storel2c,
- "X.storel2c(len = -1) -> little-endian bytes, two's complement")
+ "X.storel2c([len = -1]) -> little-endian bytes, two's complement")
KWMETH(storeb2c,
- "X.storeb2c(len = -1) -> big-endian bytes, two's complement")
+ "X.storeb2c([len = -1]) -> big-endian bytes, two's complement")
METH (tobuf, "X.tobuf() -> buffer format")
#undef METHNAME
{ 0 }
"Binary polynomials. Support almost all the standard arithmetic\n\
operations.\n\
\n\
-Constructor GF(X, radix = R) attempts to convert X to a `GF'. If\n\
+Constructor GF(X, [radix = R]) attempts to convert X to a `GF'. If\n\
X is a string, it's read in radix-R form, or we look for a prefix\n\
if R = 0. Other acceptable things are field elements, elliptic curve\n\
points, group elements, Python `int' and `long' objects, and anything\n\
\n\
Notes:\n\
\n\
- * Use `//' for division. GFs don't have `/' division.",
+ * Use `//' for Euclidean division. `/' gives exact rational division.",
0, /* @tp_traverse@ */
0, /* @tp_clear@ */
PyObject *z = 0;
mp *zz;
mptext_stringctx sc;
- char *kwlist[] = { "class", "x", "radix", 0 };
+ static const char *const kwlist[] = { "class", "x", "radix", 0 };
if (!PyArg_ParseTupleAndKeywords(arg, kw, "Os#|i:fromstring",
- kwlist, &me, &p, &len, &r))
+ KWLIST, &me, &p, &len, &r))
goto end;
if (!good_radix_p(r, 1)) VALERR("bad radix");
sc.buf = p; sc.lim = p + len;
{
gfreduce_pyobj *mr = 0;
gfreduce r;
- char *kwlist[] = { "m", 0 };
+ static const char *const kwlist[] = { "m", 0 };
mp *xx = 0;
- if (!PyArg_ParseTupleAndKeywords(arg, kw, "O&:new", kwlist, convgf, &xx))
+ if (!PyArg_ParseTupleAndKeywords(arg, kw, "O&:new", KWLIST, convgf, &xx))
goto end;
if (MP_ZEROP(xx)) ZDIVERR("modulus is zero!");
gfreduce_create(&r, xx);
Py_TPFLAGS_BASETYPE,
/* @tp_doc@ */
-"A reduction context for reduction modulo sparse irreducible polynomials.",
+"GFReduce(N): a context for reduction modulo sparse polynomials.",
0, /* @tp_traverse@ */
0, /* @tp_clear@ */
{
mp *p = 0, *beta = 0;
gfn_pyobj *gg = 0;
- char *kwlist[] = { "p", "beta", 0 };
+ static const char *const kwlist[] = { "p", "beta", 0 };
- if (!PyArg_ParseTupleAndKeywords(arg, kw, "O&O&:new", kwlist,
+ if (!PyArg_ParseTupleAndKeywords(arg, kw, "O&O&:new", KWLIST,
convgf, &p, convgf, &beta))
goto end;
gg = PyObject_New(gfn_pyobj, ty);
+ gg->p = 0;
if (gfn_create(p, beta, &gg->ntop, &gg->pton)) {
- FREEOBJ(gg);
+ Py_DECREF(gg);
gg = 0;
VALERR("can't invert transformation matrix");
}
end: \
mp_drop(xx); \
if (!z) return (0); \
- return (mp_pywrap(z)); \
+ return (gf_pywrap(z)); \
}
XFORMOP(pton, PTON)
XFORMOP(ntop, NTOP)
static void gfn_pydealloc(PyObject *me)
{
- gfn_destroy(GFN_PTON(me));
- gfn_destroy(GFN_NTOP(me));
+ if (GFN_P(me)) {
+ MP_DROP(GFN_P(me));
+ gfn_destroy(GFN_PTON(me));
+ gfn_destroy(GFN_NTOP(me));
+ }
FREEOBJ(me);
}
Py_TPFLAGS_BASETYPE,
/* @tp_doc@ */
-"An object for transforming elements of binary fields between polynomial\n\
-and normal basis representations.",
+"GFN(P, BETA): an object for transforming elements of binary fields\n\
+ between polynomial and normal basis representations.",
0, /* @tp_traverse@ */
0, /* @tp_clear@ */
static PyMethodDef methods[] = {
#define METHNAME(func) meth_##func
KWMETH(_MP_fromstring, "\
-fromstring(STR, radix = 0) -> (X, REST)\n\
+fromstring(STR, [radix = 0]) -> (X, REST)\n\
\n\
Parse STR as a large integer, according to radix. If radix is zero,\n\
read a prefix from STR to decide radix: allow `0' for octal, `0x' for hex\n\
or `R_' for other radix R.")
KWMETH(_GF_fromstring, "\
-fromstring(STR, radix = 0) -> (X, REST)\n\
+fromstring(STR, [radix = 0]) -> (X, REST)\n\
\n\
Parse STR as a binary polynomial, according to radix. If radix is zero,\n\
read a prefix from STR to decide radix: allow `0' for octal, `0x' for hex\n\
~mdw / catacomb-python / blobdiff