next up previous contents index
Next: Translating assignment expressions Up: Basic translation Previous: Translating global index expression   Contents   Index

Translating variable accesses

We only need to consider specially the case where the variable is a multiarray. The general scheme is illustrated in Figure A.21. This applies to the case of a variable that is a simple identifier--a local variable, method or constructor argument, or instance variable of the current object. The case of a field reference with an object prefix will be covered in section A.4.14.

The names $a'_{\mbox{\small dat}}$, $a'_{\mbox{\small bas}}$, and $a'_0, \ldots, a'_{R-1}$ are the names introduced by the translator when translating the corresponding field, local variable declaration or formal parameter in the source program (see sections A.4.2 and A.4.3).

Figure A.22: Translation of a multiarray assignment.
\begin{figure}\textbf{SOURCE:}
\begin{eqnarray*}
e & \equiv & e_t \verb$ = $ e_s...
...x{$R$ is the rank of both arrays.} \\
\end{array}\end{displaymath}\end{figure}



Bryan Carpenter 2003-04-15