Revise printing methods as they are not coherent

[albinahlback]

Printing fmpz_mat and fmpq_mat gives the following

**fmpz_mat_print_pretty**
[[0 0 0]
[0 0 0]
[0 0 0]
]

**fmpz_mat_print**
3 3  0 0 0 0 0 0 0 0 0

**fmpq_mat_print**
<3 x 3 matrix over Q>
[0, 0, 0]
[0, 0, 0]
[0, 0, 0]

Some questions:

• Should there be commas or not? I am fine with removing the commas in fmpq_mat.
• fmpq_mat_print is quite nicely printed. Should we call this fmpq_mat_print_pretty and remove <n x n matrix over Q>?
• Should an "ugly" version of printing for fmpq_mat be implemented, like the one for fmpz_mat?
• Should there be double square brackets or not, like seen in fmpz_mat_print_pretty?
• Should fmpz_mat_print_pretty print the last two brackets with a newline character in between?

Perhaps it goes for other types, I have not checked them out yet.

I suppose these are kind of breaking changes, but they need to be coherent at least (in my opinion).

[wbhart]

I agree. We need to check what all the other printing methods do though. Probably one of these methods is inconsistent with all the others.

[albinahlback]

Wrong documentation for fmpq_poly_fprint_pretty (also note FILT instead of FILE): https://flintlib.org/doc/fmpq_poly.html?highlight=print#c.fmpq_poly_fprint_pretty

[albinahlback]

I think the "ugly" versions are important and should be there to ease parsing. With that being said, I think fmpz_[mat/vec]_print does the correct thing, and thus think that fmpq_mat_print should print something very similar.

Thus, I think it remains to discuss the formatting of the output for the pretty printing. For matrix/vector-type objects, I am in favor of using one pair of brackets for each row (vectors can be interpretted as rows in order to not output so many newline-characters). I also think that commas are nice.

[wbhart]

I thought the issue was consistency. The only way to achieve that is to check what all the other functions are doing.

[albinahlback]

Yeah, we could either investigate how everything is printed and then do something about it. Or we could try to come up with a standardized format for everything based, and change it from there.