This simple interpreter is written in C and uses a recursive-decent parser. It was made with an intention of learning more about compiler internals. This documentaion contains some of the internals of the language. This was also given as an exploitation challenge in InCTF-2019 and BSidesCTF 2019 where the aim was to execute code in this interpreter and escape it :)

The grammer for this language can be found in the grammer.md file. Here are some of the essentials anyway :)

• Math Expressions:

  In [1]: 1+2
  Out[1]: 3
  
  In [2]: 5*4-2/1*4
  Out[2]: 12

• Logical Operations:

  In [1]: 1 == 1
  Out[1]: 1
  
  In [2]: 100 > 1000
  Out[2]: 0

• Variable Assignment:

  In [1]: a1 = 0x10
  
  In [2]: a2 = 1
  
  In [3]: a1 + a2
  Out[3]: 17
  
  In [4]: str = "this is a string"
  
  In [5]: str
  Out[5]: "this is a string"
  
  In [6]: list = [1, 2+2, 0x100, "list with string", [12, str]]
  
  In [7]: list
  Out[7]: [1, 4, 256, "list with string", [12, "this is a string"]]

• Loop:

  In [1]: i=0
  
  In [2]: sum=0
  
  In [3]: while (i<10){sum=sum+i; i=i+1}
  
  In [4]: sum
  Out[4]: 45

• Conditional Statement:

  In [1]: a=1234
  
  In [2]: b=5678
  
  In [3]: c=0
  
  In [4]: if(a==b){c=1} else{c=100}
  
  In [5]: c
  Out[5]: 100

• Function: Not working as expected :(

• global.h: Contains global variables, tokens, vm instructions and all the structures used.
• tokens.h: The tonkenizer
• parser.h: Holds the code for the parser
• vm.h: Just a wrapper for calling vm_operations which are defined in vmop.h
• vmop.h: Contains the code for all the VM operations.
• interp.c: The entry point of the interpreter.

Essentially, Navarint parses the input code into bytecode which is then executed by it's VM.

In Navarint, there is only one datatype - val. It has the following structure:

typedef struct val
{
  long type;
  long data;
}val;

type is the underlying primitive type of this val. type can take the following values -

• 0 - None Type - Default type
• 1 - Integer type
• 2 - Strings type
• 3 - List type
• 4 - Pointer type
• 5 - Function type

The data pointer of the val structure points to the actual data of this val. If the type is Int, then the data field contains the integer itself. In all other cases, it points to the objects of their corresponding types. For more details, feel free to go through globals.h.

There are essentially 2 stacks - the VM's operation stack and the call stack.

The operations stack is used for all VM operations and operands are passed to operations via this stack. Each entry in the operations stack has the datatype val

The call stack is used to save the pc in case of a function call. It also keeps track of the current scope.

• sp: operations stack top pointer
• ax: auxilary register. has the type val
• cs_sp: call stack top pointer.
• cs_bp: call stack base pointer.

A brief description for each of the bytecodes.

• LEA: Not implemented

• IMM: Load the immediate value that follows into the ax register

• JMP: Jump (set the pc) to the address pointed to by pc

• CALL:

• JZ: Jump if ax=0

• JNZ: Jump if ax != 0

• RET: Return to the caller. Pop old scope and pc from call stack

• SET: Set ax to the content of the varible on stack top.

• GET: Put the content of a variable pointed to by ax into ax (dereference ax)

• PUSH: Push the contents of ax on to the stack

• GETELEM: Get the nth item from the string/list that is there on top of the stack. n, the index, is taken from ax

• SETELEM: Set the nth item of a string/list to value. n the index, is present on top of the stack. Second on stack top comes the list/string. The value is taken from ax

• ADDELEM: Add the element in ax to the list present on top of the stack.

• EXIT: Exit the VM

• Math Operations: All these take the first operand from stack top and second from ax. The result of the operation is also moved into ax: OR, XOR, AND, EQ, NE, LT, GT, LE, GE, SHL, SHR, ADD, SUB, MUL, DIV, MOD

• Not Implemented: OPEN, READ, CLOSE, PRTF, MALC, MSET, MCMP, LI, LC, SI, SC, ADJ, LEV, ENT

The initial idea for this interpreter and also many parts of the code come from this awesome tutorial on writting interpreters - [https://github.com/lotabout/write-a-C- interpreter/blob/master/tutorial/en/0-Preface.md](https://github.com/lotabout/write-a-C- interpreter/blob/master/tutorial/en/0-Preface.md)