A general purpose number parser.
The parser is not intended to parse a string itself, but rather to receive the input from an external parser.
It starts interpreting the number as a signed integer and switches to floating-point if either the number is too long to represent it as an integer, the radix point is set or the an exponent digit is added.
After initialization with number_parser_init(), the parser functions can be called in any order. number_parser_end() terminates the parser and calculates the final number.
Numbers exceeding the value range of a double will have the value INFINITY or -INFINITY if positive or negative, respectively.
// Define parser.
number_parser parser;
// Initialize parser with base 10.
number_parser_init(&parser, 10);
// Parse the number "-12.3e4".
// Notice that the call order does not have to match the input, except for the digits.
number_parser_add_exp_digit(&parser, 4); // add exponent digit "4"
number_parser_add_digit(&parser, 1); // add integer digit "1"
number_parser_add_digit(&parser, 2); // add integer digit "2"
number_parser_set_rad_point(&parser); // set radix point at current offset
number_parser_add_digit(&parser, 3); // add integer digit "3"
number_parser_set_neg(&parser, 1); // set number negative
// Terminate parser.
int is_float = number_parser_end(&parser);
if (is_float) {
printf("float: %lf\n", parser.fval);
}
else {
printf("int: %lld\n", parser.ival);
}