pywraplp.Solver deep copy

[SanPen]

What language and solver does this apply to?
Python

Describe the problem you are trying to solve.

I want to copy a Solver object and in the copy, add a variable to each constraint.

These two features apply:

• Deep copy of the Solver object.
• Access to the variables used by a constraint and modify a constraint

from ortools.linear_solver import pywraplp

def create_constraints(model):
    x = model.NumVar(0, 1, 'x')
    y = model.NumVar(0, 1, 'y')
    constraint = model.Constraint(-pywraplp.Solver.infinity(), 1)
    constraint.SetCoefficient(x, 1)
    constraint.SetCoefficient(y, 1)
    return constraint

# Create the first solver and add constraints
solver1 = pywraplp.Solver.CreateSolver('SCIP')
constraint1 = create_constraints(solver1)

# Create the second solver
new_solver = pywraplp.Solver.CreateSolver('SCIP')

# copy vars... kinda ok
for var in solver.variables():
    new_solver.NumVar(lb=var.Lb(), ub=var.Ub(), name=var.name())

# I want to copy the constraints, but adding a new var to each one
slacks = list()
for i, cst in enumerate(solver.constraints()):
    sl = new_solver.NumVar(0, 1e20, 'Slack{}'.format(i))
    new_solver.Add(cst + sl)  # this line fails
    slacks.append(sl)

new_solver.Minimize(new_solver.Sum(slacks))

Describe the solution you'd like

I'd like something like solver.copy() to copy the solver.

I'd like an easy way to add variables to an existing constraint (if this does not exist already, which I haven't been able to accomplish)

Describe alternatives you've considered

Not sure I have alternatives for this

Additional context

I want to add MIP model debugging by the process described in the code

asked on Stack Overflow: https://stackoverflow.com/questions/77348474/ortools-mip-solver-copy-pywraplp-solver-copy

[lperron]

Just answered on stack overflow.