#ifndef TOKENIZER_CC
#define TOKENIZER_CC

#include "common.cc"
#include "source.cc"
#include "token.cc"
#include "utf8.cc"

struct Tokenizer {
  Buffer* buffer;
  Buffer_Stack* stack;

  Tokenizer(Buffer_Stack* stack) : buffer(nullptr), stack(stack) {}
};

static inline Buffer* tokenizer_get_buffer(Tokenizer* tokenizer) {
  assert_neq(tokenizer, nullptr);

  if (tokenizer->buffer != nullptr) return tokenizer->buffer;

  Buffer* buffer = nullptr;
  if (!buffer_stack_pop(tokenizer->stack, &buffer)) return nullptr;

  tokenizer->buffer = buffer;
  return buffer;
}

static inline char tokenizer_advance(const Tokenizer* tokenizer,
                                     usize* offset) {
  const String text = tokenizer->buffer->content;

  unsigned char c = *text[*offset];
  u8 nobytes = utf8_nobytes(c);
  if (nobytes > 1) panic("no support for multi-byte chars: %c:%d", c, nobytes);

  *offset += nobytes;
  return c;
}

static inline String tokenizer_make_lexeme(const Tokenizer* tokenizer,
                                           usize start, usize end) {
  assert_neq(tokenizer, nullptr);

  Buffer* buffer = tokenizer->buffer;
  return String(buffer->content[start], end - start);
}

bool tokenizer_next(Tokenizer* tokenizer, Token* out) {
  assert_neq(tokenizer, nullptr);
  assert_neq(out, nullptr);

again:
  Buffer* buffer = tokenizer_get_buffer(tokenizer);
  if (buffer == nullptr) return false;

  usize start = buffer->cursor;
  if (start == buffer->content.length) {
    tokenizer->buffer = nullptr;
    goto again;
  }

  usize offset = start;
  tokenizer_advance(tokenizer, &offset);

  String lexeme = tokenizer_make_lexeme(tokenizer, start, offset);
  *out = Token(Token_Kind_Eof, lexeme, Span(buffer->file, 0, 0));
  buffer->cursor = offset;
  return true;
}

#endif