fz-lang/src/compiler/fzl_reader.c

#include <compiler/fzl_reader.h>
#include <compiler/fzl_token.h>
#include <compiler/internals/reader.h>
#include <compiler/utils/fzl_linked.h>
#include <stdbool.h>
#include <stddef.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>

// Last error from reader
char last_err[256];

fzl_reader_t *fzl_mkreader(fzl_linked_t *linked)
{
  fzl_reader_ll_t *first_node = malloc(sizeof(fzl_reader_ll_t));
  first_node->token           = NULL;
  first_node->next            = NULL;

  fzl_reader_t *reader = malloc(sizeof(fzl_reader_t));
  reader->first        = first_node;
  reader->last         = first_node;
  reader->linked       = linked;
  reader->source       = linked->source;
  reader->size         = linked->total_sz;
  reader->position     = 0;

  memset(last_err, 0, sizeof(last_err));

  return reader;
}

int fzl_read(fzl_reader_t *reader)
{
  if (reader->position >= reader->size)
    return 0;

  bool is_format_str = false;

  do
  {
    size_t p = reader->position;
    fzl_token_t *token;
    char curr = reader->source[p];

    if (is_format_str)
    {
      token = rd_fmtstr(reader);
      if (token && token->type == FZL_TOKTYPE_STR_FORMAT_END)
        is_format_str = false;
    }
    else if (curr == '\n' || curr == ' ' || curr == '\r')
    {
      move(reader, 1);
      continue;
    }
    else if (curr == '\'' || curr == '"')
      token = rd_string_literal(reader);
    else if (curr == '/' && reader->source[p + 1] == '/')
    {
      rd_comment(reader);
      continue;
    }
    else if (curr == '=' && reader->source[p + 1] == '=')
      token = rd_double(reader, FZL_TOKTYPE_EQEQ);
    else if (curr == '>' && reader->source[p + 1] == '=')
      token = rd_double(reader, FZL_TOKTYPE_REQ);
    else if (curr == '<' && reader->source[p + 1] == '=')
      token = rd_double(reader, FZL_TOKTYPE_LEQ);
    else if (curr == '!' && reader->source[p + 1] == '=')
      token = rd_double(reader, FZL_TOKTYPE_NEQ);
    else if (curr == '|' && reader->source[p + 1] == '|')
      token = rd_double(reader, FZL_TOKTYPE_DVBAR);
    else if (curr == '&' && reader->source[p + 1] == '&')
      token = rd_double(reader, FZL_TOKTYPE_DAMPERSAND);
    else if (curr == '>')
      token = rd_single(reader, FZL_TOKTYPE_RANGLE);
    else if (curr == '<')
      token = rd_single(reader, FZL_TOKTYPE_LANGLE);
    else if (curr == '!')
      token = rd_single(reader, FZL_TOKTYPE_NOT);
    else if (curr == ';')
      token = rd_single(reader, FZL_TOKTYPE_SEMICOLON);
    else if (curr == '{')
      token = rd_single(reader, FZL_TOKTYPE_LBRACE);
    else if (curr == '}')
      token = rd_single(reader, FZL_TOKTYPE_RBRACE);
    else if (curr == '(')
      token = rd_single(reader, FZL_TOKTYPE_LPAREN);
    else if (curr == ')')
      token = rd_single(reader, FZL_TOKTYPE_RPAREN);
    else if (curr == ',')
      token = rd_single(reader, FZL_TOKTYPE_COMMA);
    else if (curr == '.')
      token = rd_single(reader, FZL_TOKTYPE_DOT);
    else if (curr == '*')
      token = rd_single(reader, FZL_TOKTYPE_STAR);
    else if (curr == '/')
      token = rd_single(reader, FZL_TOKTYPE_SLASH);
    else if (curr == '+')
      token = rd_single(reader, FZL_TOKTYPE_PLUS);
    else if (curr == '-')
      token = rd_single(reader, FZL_TOKTYPE_MINUS);
    else if (curr == '^')
      token = rd_single(reader, FZL_TOKTYPE_CARET);
    else if (curr == '%')
      token = rd_single(reader, FZL_TOKTYPE_PERCENT);
    else if (curr == '=')
      token = rd_single(reader, FZL_TOKTYPE_EQ);
    else if (curr == '$' && reader->source[p + 1] == '"')
    {
      move(reader, 2);
      is_format_str = true;
      token = fzl_mktoken(FZL_TOKTYPE_STR_FORMAT_START, 0, 0, NULL, 0, "");
    }
    else if ((curr >= 'A' && curr <= 'Z') || (curr >= 'a' && curr <= 'z') ||
             curr == '_')
      token = rd_name(reader);
    else if (curr >= '0' && curr <= '9')
      token = rd_number(reader);
    else
    {
      fzl_file_loc l;
      bool ok = fzl_linked_getpos(reader->linked, p, &l);
      if (!ok)
      {
        snprintf(last_err, sizeof(last_err), "failed to locate line:col");
        return -1;
      }

      sprintf(last_err, "'%s': undefined token '%c' at %ld:%ld", l.filename,
              curr, l.line, l.col);
      return -1;
    }

    if (!token)
    {
      if (!last_err[0])
      {
        fzl_file_loc l;
        bool ok = fzl_linked_getpos(reader->linked, p, &l);
        if (!ok)
        {
          snprintf(last_err, sizeof(last_err), "failed to locate line:col");
          return -1;
        }

        sprintf(last_err, "'%s': cannot parse token at %ld:%ld", l.filename,
                l.line, l.col);
      }

      return -1;
    }

    fzl_reader_ll_t *tok_node = malloc(sizeof(fzl_reader_ll_t));
    tok_node->next            = NULL;
    tok_node->token           = token;

    reader->last->next = tok_node;
    reader->last       = tok_node;
  } while (reader->position < reader->size);

  return 0;
}

char *fzl_reader_geterror() { return last_err; }

void free_node_and_next(fzl_reader_ll_t *node)
{
  if (!node)
    return;
  if (node->next)
    free_node_and_next(node->next);

  if (node->token)
    fzl_freetoken(node->token);

  free(node);
}

void fzl_freereader(fzl_reader_t *reader)
{
  if (!reader)
    return;
  if (reader->first)
    free_node_and_next(reader->first);

  free(reader);
}