varaq-wasm-c/compiler.c

#include "compiler.h"

#include "common.h"
#include "tokenizer.h"

#include <math.h>

Code*
signedLEB128(size_t num)
{
  bool more = true;
  Code* buffer = (Code*)malloc(sizeof(Code));
  buffer->count = 0;
  int n = (int)num;

  while (more) {
    uint8_t byte = n & 0x7f;
    n >>= 7;
    if ((n == 0 && (byte & 0x40) == 0) || (n == -1 && (byte & 0x40) != 0)) {
      more = false;
    } else {
      byte |= 0x80;
    }

    size_t old_count = buffer->count;

    uint8_t* tmp =
      (uint8_t*)calloc((old_count + 1),
                       sizeof(uint8_t)); // really slow and bad refactor later
    memcpy(tmp, buffer->cells, buffer->count * sizeof(uint8_t));
    if (tmp) {
      buffer->cells = tmp;
    }
    buffer->cells[old_count] = byte;
    buffer->count += 1;
  }

  return buffer;
}

Code*
unsignedLEB128(size_t num)
{
  Code* buffer = (Code*)malloc(sizeof(Code));
  buffer->count = 0;
  int n = (int)num;

  do {
    uint8_t byte = n & 0x7f;
    n >>= 7;
    if (n != 0) {
      byte |= 0x80;
    }

    size_t old_count = buffer->count;

    uint8_t* tmp =
      (uint8_t*)calloc((old_count + 1),
                       sizeof(uint8_t)); // really slow and bad refactor later
    memcpy(tmp, buffer->cells, buffer->count * sizeof(uint8_t));
    if (tmp) {
      buffer->cells = tmp;
    }
    buffer->cells[old_count] = byte;
    buffer->count += 1;
  } while (n != 0);

  return buffer;
}

Code*
generic_append(Code* tape, void* data, size_t data_size)
{
  size_t old_count = tape->count;

  uint8_t* tmp = (uint8_t*)calloc((old_count + data_size), sizeof(uint8_t));
  if (old_count > 0) {
    memcpy(tmp, tape->cells, tape->count * sizeof(uint8_t));
  }
  if (tmp) {
    tape->cells = tmp;
  }

  memcpy((tape->cells + old_count), (unsigned char*)data, data_size);
  tape->count += data_size;
  return tape;
}

Code*
append_byte(Code* tape, uint8_t data)
{
  return generic_append(tape, &data, sizeof(uint8_t));
}

Code*
append_f64(Code* tape, double data)
{
  return generic_append(tape, &data, sizeof(double));
}

Code*
append(Code* tape, Code* data)
{
  size_t old_count = tape->count;

  uint8_t* tmp = (uint8_t*)calloc((old_count + data->count), sizeof(uint8_t));
  memcpy(tmp, tape->cells, tape->count * sizeof(uint8_t));
  if (tmp) {
    tape->cells = tmp;
  }

  memcpy((tape->cells + old_count), data->cells, data->count * sizeof(uint8_t));
  tape->count += data->count;
  return tape;
}

Code*
encodeString(char* string)
{
  Code* buffer = (Code*)malloc(sizeof(Code));
  buffer->cells = (uint8_t*)malloc(sizeof(uint8_t));
  buffer->cells[0] = strlen(string);
  buffer->count = 1;

  uint8_t* tmp = (uint8_t*)malloc((1 + strlen(string)) * sizeof(char));
  memcpy(tmp, buffer->cells, buffer->count * sizeof(uint8_t));
  if (tmp) {
    buffer->cells = tmp;
  }

  memcpy((buffer->cells + 1), string, strlen(string) * sizeof(char));
  buffer->count += strlen(string);
  return buffer;
}

Code*
encodeVector(Code* data)
{
  size_t count = data->override ? data->override_count : data->count;
  Code* buffer = unsignedLEB128(count);
  append(buffer, data);
  return buffer;
}

Code*
createSection(uint8_t section, Code* data)
{
  Code* buffer = (Code*)malloc(sizeof(Code));
  buffer->cells = (uint8_t*)malloc(sizeof(uint8_t));
  buffer->cells[0] = section;
  buffer->count = 1;
  return append(buffer, encodeVector(data));
}

Code*
compile(char* buffer)
{
  char number[100];

  Code* add_args_code = (Code*)calloc(1, sizeof(Code));
  Code* add_return_code = (Code*)calloc(1, sizeof(Code));
  append_byte(add_return_code, f64);

  Code* add_function_type = (Code*)calloc(1, sizeof(Code));
  append_byte(add_function_type, FUNCTION);

  append(add_function_type, encodeVector(add_args_code));
  append(add_function_type, encodeVector(add_return_code));
  add_function_type->override = true;
  add_function_type->override_count = 1;

  Code* type_section = createSection(TYPE, encodeVector(add_function_type));

  Code* return_type_code = (Code*)calloc(1, sizeof(Code));
  append_byte(return_type_code, 0x00);
  return_type_code->override = true;
  return_type_code->override_count = 1;

  Code* func_section = createSection(FUNC, encodeVector(return_type_code));

  Code* exp = encodeString("main");
  append_byte(exp, EXPORT_FUNC);
  append_byte(exp, 0x00);
  exp->override = true;
  exp->override_count = 1;
  Code* export_section = createSection(EXPORT, encodeVector(exp));

  // Code* code = (Code*)calloc(1, sizeof(Code));
  // append_byte(code, F64_CONST);
  // append_f64(code, 1.0);
  // append_byte(code, F64_CONST);
  // append_f64(code, 2.0);
  // append_byte(code, F64_ADD);

  Code* code = (Code*)calloc(1, sizeof(Code));
  initTokenizer(buffer);
  Token t = nextToken();
  do {
    // debug_printToken(t);
    switch (t.type) {
      case TOKEN_LEFT_PAREN:
        break;
      case TOKEN_RIGHT_PAREN:
        break;
      case TOKEN_LEFT_BRACE:
        break;
      case TOKEN_RIGHT_BRACE:
        break;
      case TOKEN_TILDE:
        break;
      case TOKEN_SLASH:
        break;
      case TOKEN_MINUS:
        break;
      case TOKEN_IDENTIFIER:
        break;
      case TOKEN_STRING:
        break;
      case TOKEN_FLOAT:
        strncpy(number, t.start, t.length);
        number[t.length + 1] = '\0';
        append_byte(code, F64_CONST);
        append_f64(code, strtod(number, NULL));
        break;
      case TOKEN_LIST:
        break;
      case TOKEN_ERROR:
        break;
      case TOKEN_FALSE:
        break;
      case TOKEN_TRUE:
        break;
      case TOKEN_PI:
        append_byte(code, F64_CONST);
        append_f64(code, M_PI);
        break;
      case TOKEN_E:
        append_byte(code, F64_CONST);
        append_f64(code, M_E);
        break;
      case TOKEN_EOF:
        break;
      case TOKEN_POP:
        append_byte(code, DROP);
        break;
      case TOKEN_DUP:
        break;
      case TOKEN_EXCH:
        break;
      case TOKEN_CLEAR:
        break;
      case TOKEN_REMEMBER:
        break;
      case TOKEN_FORGET:
        break;
      case TOKEN_DUMP:
        break;
      case TOKEN_NAME:
        break;
      case TOKEN_SET:
        break;
      case TOKEN_IFYES:
        break;
      case TOKEN_IFNO:
        break;
      case TOKEN_CHOOSE:
        break;
      case TOKEN_EVAL:
        break;
      case TOKEN_ESCAPE:
        break;
      case TOKEN_REPEAT:
        break;
      case TOKEN_SPLIT:
        break;
      case TOKEN_CONS:
        break;
      case TOKEN_SHATTER:
        break;
      case TOKEN_EMPTY:
        break;
      case TOKEN_COMPOSE:
        break;
      case TOKEN_STREQ:
        break;
      case TOKEN_STRCUT:
        break;
      case TOKEN_STRMEASURE:
        break;
      case TOKEN_STRTIE:
        break;
      case TOKEN_EXPLODE:
        break;
      case TOKEN_ADD:
        append_byte(code, F64_ADD);
        break;
      case TOKEN_SUB:
        append_byte(code, F64_SUB);
        break;
      case TOKEN_MUL:
        append_byte(code, F64_MUL);
        break;
      case TOKEN_DIV:
        append_byte(code, F64_DIV);
        break;
      case TOKEN_IDIV:
        break;
      case TOKEN_MOD:
        break;
      case TOKEN_POW:
        break;
      case TOKEN_SQRT:
        append_byte(code, F64_SQRT);
        break;
      case TOKEN_ADD1:
        append_byte(code, F64_CONST);
        append_f64(code, 1.0);
        append_byte(code, F64_ADD);
        break;
      case TOKEN_SUB1:
        append_byte(code, F64_CONST);
        append_f64(code, 1.0);
        append_byte(code, F64_SUB);
        break;
      case TOKEN_SIN:
        break;
      case TOKEN_COS:
        break;
      case TOKEN_TAN:
        break;
      case TOKEN_ATAN:
        break;
      case TOKEN_LN:
        break;
      case TOKEN_LOG:
        break;
      case TOKEN_LOG3:
        break;
      case TOKEN_CLIP:
        break;
      case TOKEN_SMOOTH:
        break;
      case TOKEN_HOWMUCH:
        break;
      case TOKEN_SETRAND:
        break;
      case TOKEN_RAND:
        break;
      case TOKEN_INT:
        break;
      case TOKEN_NUMBERIZE:
        break;
      case TOKEN_ISOLATE:
        break;
      case TOKEN_MIX:
        break;
      case TOKEN_CONTRADICT:
        break;
      case TOKEN_COMPL:
        break;
      case TOKEN_SHIFTRIGHT:
        break;
      case TOKEN_SHIFTLEFT:
        break;
      case TOKEN_GT:
        break;
      case TOKEN_LT:
        break;
      case TOKEN_EQ:
        break;
      case TOKEN_GE:
        break;
      case TOKEN_LE:
        break;
      case TOKEN_NE:
        break;
      case TOKEN_NULL:
        break;
      case TOKEN_NEGATIVE:
        break;
      case TOKEN_ISNULL:
        break;
      case TOKEN_ISINT:
        break;
      case TOKEN_ISNUMBER:
        break;
      case TOKEN_AND:
        break;
      case TOKEN_OR:
        break;
      case TOKEN_XOR:
        break;
      case TOKEN_DISP:
        break;
      case TOKEN_LISTEN:
        break;
      case TOKEN_COMPLAIN:
        break;
      case TOKEN_TIME:
        break;
      case TOKEN_GARBAGE_COLLECT:
        break;
    }
    t = nextToken();
  } while (t.type != TOKEN_EOF);

  Code* body = (Code*)calloc(1, sizeof(Code));
  append_byte(body, EMPTY_ARRAY);
  append(body, code);
  append_byte(body, END);

  Code* function_body = (Code*)calloc(1, sizeof(Code));
  append(function_body, encodeVector(body));
  function_body->override = true;
  function_body->override_count = 1;
  Code* code_section = createSection(CODE, encodeVector(function_body));

  Code* tape = (Code*)malloc(sizeof(Code));
  tape->cells = calloc(8, sizeof(uint8_t));
  tape->cells[0] = 0;
  tape->cells[1] = 'a';
  tape->cells[2] = 's';
  tape->cells[3] = 'm';
  tape->cells[4] = 1;
  tape->cells[5] = 0;
  tape->cells[6] = 0;
  tape->cells[7] = 0;
  tape->count = 8;

  append(tape, type_section);
  append(tape, func_section);
  append(tape, export_section);
  append(tape, code_section);
  return tape;
}