Failure to read Netlib MPS file

[jajhall]

In Netlib, the problem blend has no name for its RHS entries, so JuMP rejects the file with the comment "Malformed RHS line: expected three or five fields." There are several problems in the Netlib set that have no RHS name.

More generally, the CUTEst test set of optimization problems use SIF format that has no RHS name, so maybe it would be good if JuMP were to ignore the lack of a RHS name

[odow]

I hate these file formats.

At least https://jump.dev/MathOptFormat/ has a schema