Next: Traversable Collections, Previous: Imperative Language Syntax, Up: Imperative Language Interpreter [Index]
Interpreted programs may access some of the state of the application via variables.
Variables are introduced to the imperative language interpreter through variables collections (and not through a more flexible context and lookup method specification, like for the expression evaluator).
Variables collections are application wise typeless. They are declared as
pointers (to void presumably) and initialized with NULL
. A single
collection is allowed for program parsing, and it is linked into the parsing
context (the struct x1f4_c1_type
one) via variable_set.context
field of the parsing context.
See struct x1f4_c1_type.
Adding variables to variables collections is convenienced by the
x1f4_push_bill
, x1f4_push_mode
, x1f4_push_real
,
x1f4_push_text
and x1f4_push_variable
functions.
See x1f4_push_bill.
See x1f4_push_mode.
See x1f4_push_real.
See x1f4_push_text.
See x1f4_push_variable.
Variables collections may be examined via the x1f4_list_state
function.
See x1f4_list_state.
Once variables collections are no longer of use they are to be disposed via the
x1f4_air_state
function. Executable program representations do refer
variables in the variables collection in use during the construction of such
representations, hence disposing of variables collections should not occur
before the disposal of the executable program representations referring them.
See x1f4_air_state.
The variable collections are regular libaime named objects collections,
and can be operated via specific methods. The one exception is their initial
NULL
state - in this state they are not named objects collections.