method submatrix of matrix_mod2_dense needs default arguments

[cnassau]

The submatrix methods for dense matrices mod 2^e have a different signature than the other submatrix methods; the latter treat their last two arguments as optional. This leads to errors like this (from 6.8.beta5):

sage: d0=matrix(GF(2),[[1, 1], [1, 1], [1, 1], [1, 1]])
sage: d0._echelon_form_PID()---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
...
TypeError: submatrix() takes exactly 4 positional arguments (2 given)

or this

sage: d0=matrix(GF(2),[[1, 1], [1, 1], [1, 1], [1, 1]])
sage: d1=matrix(GF(2),[[1, 1, 0, 0],
 [1, 1, 0, 0],
 [1, 1, 1, 1],
 [1, 1, 1, 1],
 [0, 0, 1, 1],
 [0, 0, 1, 1]]
)
sage: C=ChainComplex(data=(d0,d1))
sage: C.homology(1,generators=True)
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
...
TypeError: submatrix() takes exactly 4 positional arguments (2 given)

[cnassau]

New commits:

​3549edf

Ticket #18761 - change the signature of the submatrix method for dense

[cnassau]

I have given the submatrix methods the same signature as in file "matrix1.pyx". This includes a name change of the arguments, because in parts of Sage (eg, ChainComplex._homology_generators_snf) the method is called with named arguments.

I also made this change:

-        if self._ncols == 0 or self._nrows == 0:
+        if ncols == 0 or nrows == 0:

This fixed a segfault in my tests, and also seems to make more sense that the original: one returns an uninitialized matrix if *it* is empty, not if *self* is empty.

Caveat: I have no detailed understanding of m4rie, someone knowledgeable might want to have a look at my changes.

[malb]

Looks good to me.

[vbraun]

Reviewer name is missing

[cnassau]

I have added the name of the reviewer, and taken the liberty to revert the status to positive-review again myself.