Function: nffactor
Section: number_fields
C-Name: nffactor
Prototype: GG
Help: nffactor(nf,T): factor polynomial T in number field nf.
Doc: factorization of the univariate
 polynomial $T$ over the number field $\var{nf}$ given by \kbd{nfinit}; $T$
 has coefficients in $\var{nf}$ (i.e.~either scalar, polmod, polynomial or
 column vector). The factors are sorted by increasing degree.

 The main variable of $\var{nf}$ must be of \emph{lower}
 priority than that of $T$, see \secref{se:priority}. However if
 the polynomial defining the number field occurs explicitly  in the
 coefficients of $T$ as modulus of a \typ{POLMOD} or as a \typ{POL}
 coefficient, its main variable must be \emph{the same} as the main variable
 of $T$. For example,
 \bprog
 ? nf = nfinit(y^2 + 1);
 ? nffactor(nf, x^2 + y); \\@com OK
 ? nffactor(nf, x^2 + Mod(y, y^2+1)); \\ @com OK
 ? nffactor(nf, x^2 + Mod(z, z^2+1)); \\ @com WRONG
 @eprog

 It is possible to input a defining polynomial for \var{nf}
 instead, but this is in general less efficient since parts of an \kbd{nf}
 structure will then be computed internally. This is useful in two
 situations: when you do not need the \kbd{nf} elsewhere, or when you cannot
 initialize an \kbd{nf} due to integer factorization difficulties when
 attempting to compute the field discriminant and maximal order.

 \misctitle{Caveat} \kbd{nfinit([T, listP])} allows to compute in polynomial
 time a conditional \var{nf} structure, which sets \kbd{nf.zk} to an order
 which is not guaranteed to be maximal at all primes. Always either use
 \kbd{nfcertify} first (which may not run in polynomial time) or make sure
 to input \kbd{nf.pol} instead of the conditional \var{nf}: \kbd{nffactor} is
 able to recover in polynomial time in this case, instead of potentially
 missing a factor.
