make
./interpreter source_code_file
or you can just run ./interpreter and write source code as an input
- Definitions.hs - contains data types declarations for environments, stores and values
- Semantic.hs - describes what is the effect of particular statements and expressions
- Utils.hs - auxiliary functions used across the project, e.g. changing state, checking types
- Interpreter.hs - top-level module which starts and ends the execution of the program
- Main.hs - parsing code and running interpreter on abstract syntax tree
- AbsGrammar.hs - definition of abstract syntax tree (auto-generated)
- grammar.cf - formal description of the language grammar in EBNF notation
In order to execute a program interpreter processes instructions which changes current state of execution. State is in fact a pair (environment, store). First element tells which location in store is assigned to a particular identifier. Second represents a memory where values are stored.
Represents a single value which can be stored in store. Can be in one of 4 forms:
- I integer - integer value
- B bool - boolean value
- S string - string value
- V - void
Carries an information about identifiers bindings to store locations in the current scope. Especially useful in handling variables shadowing. Environment is a composition of three separated parts:
- variable environment (venv) - for storing actual variables names
- procedure environment (penv) - for storing actual procedure/function names
- cost environment (cenv) - for storing actually opened cost blocks
Used as a simple memory which stores values at specific locations. Is made of two parts:
- variable store (vst) - for storing variables values and special values (e.g. return values)
- cost store (cst) - for storing statistics needed by cost blocks
There is one conflict in the language grammar related to nested if-else constructs.
Consider code block if (a) if (b) f(); else g(); It can't be determined if else belongs to first or second if statement based on set of rules. Interpreter will parse it as
if (a) {
if (b) {
print "X";
} else {
print "Y";
}
}
Cost block turned out to be a little bit harder to implement than I originally thought. It works with some exceptions. Cost block must be defined inside a function. Calculating memory allocated used works only for integer and boolean types. Because string's length can change multiple times it requires more complicated approach to track the used memory.