Go error checking and chaining - go

Question

package main
import (
"errors"
"fmt"
)
type FieldElement struct {
Num int
Prime int
}
func (el *FieldElement) Add(o FieldElement) (FieldElement, error) {
if el.Prime != o.Prime {
return FieldElement{}, errors.New("cannot operate on two numbers in different Fields")
}
return FieldElement{(el.Num + o.Num) % el.Prime, el.Prime}, nil
}
// Sub, Mul, Div and Pow also members of FieldElement returning (FieldElement, error)
func main() {
a := FieldElement{13, 19}
b := FieldElement{9, 19}
fmt.Println(a.Add(b))
}
type Point struct {
X FieldElement
Y FieldElement
A FieldElement
B FieldElement
}
Now upon Point creation I want to check if a point is on the curve defined as y^2 = x^3 + ax + b
Each operation of FieldElement returns 2 values to satisfy the error checking.
So I can not do
x.Pow(3).Add(a.Mul(x)).Add(b)
and end up with
pow3, e1 := x.Pow(three)
mul, e2 := a.Mul(x)
add, e3 := pow3.Add(mul)
rval, e4 := add.Add(b)
// check that y^2 == rval
This makes the code hard to follow, what is the golang way to properly do this?
playground

Answer 1

Make your methods return only a single value, so they can be chained, and store the error value in that returned value, to check after. Example:
type FieldElement struct {
Num int
Prime int
Error error
}
func (el *FieldElement) Add(o FieldElement) *FieldElement {
if el.Error != nil {
// An error already happened, so do nothing
return el
}
if el.Prime != o.Prime {
el.Error = errors.New("cannot operate on two numbers in different Fields")
} else {
el.Num = (el.Num + o.Num) & el.Prime
}
return el
}
Then use it like this:
x.Add(...).Mul(...).Sub(...) // etc
if x.Error != nil {
// handle error here
}