from cspProblem import CSP

# implement nodes as CSP problems as nodes with cost functions

class CSP$\_$with$\_$Cost$($CSP$)$:

$"""$ cost$\_$functions maps a CSP var, here a meeting name, to a list of functions for the constraints that apply $"""$

def $\_\_$init$\_\_($self$,$ domains, constraints, cost$\_$functions$)$:

self.domains $=$ domains

self.variables $=$ self.domains.keys$()$

super$().\_\_$init$\_\_("$title of csp$",$ self.variables, constraints$)$

self.cost$\_$functions $=$ cost$\_$functions

self.cost $=$ self.calculate$\_$cost$()$

# specific to the visitor hosting csp

def calculate$\_$cost$($self$)$:

$"""$ this is really a function f $=$ path cost $+$ heuristic to be used by the constraint solver $"""$

cost $=0$

for var in self.variables:

best$\_$value$\_$cost $=$ float$(\text{'}$inf$\text{'})$ # Start with an infinite cost

# Loop through all possible values in the variable's domain

for val in self.domains$[$var$]$:

current$\_$value$\_$cost $=0$ # Cost of this particular value

# Apply all cost functions $($soft constraints$)$ for this variable

for cost$\_$function in self.cost$\_$functions$[$var$]$:

day, hour $=$ Day$\_$Time$().$hour$\_$day$\_$split$($val$)$

current$\_$value$\_$cost $+=$ cost$\_$function$($day$,$ hour$)$

# Track the minimum cost for the best value of the variable

best$\_$value$\_$cost $=$ min$($best$\_$value$\_$cost$,$ current$\_$value$\_$cost$)$

# Add the best value's cost to the total cost

cost $+=$ best$\_$value$\_$cost

return cost

def $\_\_$repr$\_\_($self$)$:

$"""$ string representation of an arc"""

return $"$CSP$\_$with$\_$Cost$("+$str$($list$($self$.$domains.keys$()))+\text{'}$:$\text{'}+$str$($self$.$cost$)+")"$