#21 Wrong relation for the 1..1 → 0..1 associations

[Tomas]

The iUML-B Classdiagam generates a partial surjection for 1..1 → 0..1 associations, which is wrong as a partial surjection represents 1..n → 0..1 associations. However, until the Feature Request #349 "Partial bijection" is solved only a workaround is possible as it is very important, that the generated Event-B model matches the iUML-B representation not only in the mathematical, but also in the logical (semantic) sense.

[Colin Snook]

Fixed (subject to Feature Request #349) for next release
Added an additional invariant:
partialBijection∼ ∈ R⇸D
in generated output.