Translator – References

When processing operands, the Translator needs to distinguish between the values for and references to variables and arrays. This will be accomplished with a flag in Token to indicate whether it contains a reference or not. When a variable or array token is added to the output list, it's reference flag will be set. When the Encoder processes tokens from the output list to generate the internal code, it will generate a push reference instruction if the reference flag is set and a push value instruction if the reference flag is not set.

In the Translator, when a non-assignment operator is processed and it pops it's operands off of the done stack, it will clear the reference flag in the token of the operand since it needs values and not references. For an assignment operator, the reference flag of only the second operand will be cleared; however, the reference flag of the first operand needs to be checked to make sure that it is set, so that expressions like “5 = A” or “A+B = 4” will cause a “non-reference cannot be assigned” error.

As previously mentioned, the Translator does not know the difference between a variable and function with no arguments or an array and a function with arguments. Therefore, it could be setting the reference flag for a token that refers to a function and not an variable or array. This will not cause a problem however. Only tokens on the left side of an assignment operator will have the reference flag set and there should be no functions on the left side (except in this case of setting the return value of a function with it's own function name – therefore setting the reference flag makes sense, but functions are much later).