|
|
| [ < ] | [ > ] | [ << ] | [ Up ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |
Most user input is in the form of strings; you type something and then the game is supposed to act upon your answer. This calls for functions that can analyze your input and translate it to values you can use in your program. The syntactical analysis is very complicated, to say the least, and I'm going to leave that part for chapter three. However, let's look a bit at the value transformation bit. As stated input is in the form of strings, this makes it very interesting to convert strings to integers and floats, and vice versa.
Let's start with integers. Suppose you have received a string holding
a numerical value and you want to use it computations. In order to
convert it to the proper integer value you use the efun atoi(),
very simply. It takes a string as argument and converts it to the
corresponding integer. However, if the string contained non-numerical
characters apart from leading or trailing spaces, 0 will be returned.
The name atoi() is derived from 'a(scii) to i(nteger)', for those
of you who are interested to know.
int atoi(string str)
e.g.
int val;
val = atoi("23");
write("23 + 3 = " + (val + 3) + "\n");
|
Floats have a corresponding efun, atof(), which converts a string
to float. As you know by now, floats can't be converted to strings the
same way integers can by simply adding them to another string, but
require some other kind of treatment. The efun ftoa() will
convert a float to a string, and the reverse functoin atof() will
turn a string to a float, provided it contains a floating point
number. Again, if the string contains any non-numerical characters the
result will be 0.
For conversion between integer and float you have the efuns itof()
and ftoi(). Just keep in mind that when you convert a float to
integer, the decimal part will be cut off, not rounded.
There are many occasions when you would want to store a value as a string
and later convert it back to a value. For this purpose you have the two
efuns val2str() and str2val(). The output from
val2str() can be printed, but is not intended to. You can store
any kind of variable contents as a string using these efuns.
The most used data converter, however, is the efun sscanf(). With
sscanf() you can specify a pattern that should be scanned for
a certain value, extract that and put it into a variable. This makes
sscanf() a bit special since it handles variables given as
arguments, so it's impossible to get the function address of
sscanf(). I know this sounds pretty much like garbled Greek
to you at this moment, but trust me. I'll explain more in detail in
chapter 3. Anyway, otherwise sscanf() is fairly simple; you
provide a string to to search through, a pattern and the variables it
should store data in, and it returns the number of matches it actually
managed to make.
The string you give for pattern matching is interpreted literally apart from these control strings:
int sscanf(string str, string pattern, ...);
e.g.
int wide;
float weight;
string orgstr;
string wide_type, weight_type;
/*
* Assume the question "How wide and heavy do you think Fatty is?"
* has been posed and answered to. Furthermore, assume that the
* answer is given on the form '<amount> <sort> and <amount> <sort>',
* as for example '4 yards and 3.2 tons'. Assume the first value
* always is an integer and that the second is a float.
*
* Assume that this answer is given in the variable 'orgstr'
*
* The above is naturally only convenient assumptions to make the
* example easy to write. In reality you'd better be prepared for
* any kind of format being given as answer.
*/
if (sscanf(orgstr, "%d %s and %f %s", wide, wide_type,
weight, weight_type) != 4)
{
write("Please give a full answer!\n");
return;
}
write("Aha, you think Fatty is " + wide + " " + wide_type +
" wide and " + ftoa(weight) + " " + weight_type + " heavy.\n");
|
| [ < ] | [ > ] | [ << ] | [ Up ] | [ >> ] | [Top] | [Contents] | [Index] | [ ? ] |