Libraries
Background
tollosUtil.c – useful macros and functions for Tollos
The macros are defined in tollosUtil.h, and are:
abs(x) – return the absolute value of x
deg2rad(d) – degrees to radians
iswap(x,y) – swap the values of the ints x and y using XOR
max(a,b) – return the larger of x and y
min(a,b) – return the smaller of x and y
rad2deg(r) – radians to degrees
signum(x) – return the sign of x (−1, 0, or 1)
The functions have various purposes.
| Functions | |
|---|---|
| byte2hex | convert a byte to two hex chars |
| fastrand | return a random number in the range 0 to n−1 |
| hex2byte | convert two hexadecimal characters to a byte |
| isqrtg | return the square root of an int, with guess |
| i2str | layout an integer as a string |
| nop | do nothing with an int |
| streq | caseless string compare |
| strf | a reentrant mini printf to send a formatted string |
byte2hex – convert a byte to two hex chars
void byte2hex(byte b, char *hex, flag upper);
b – byte to convert
hex – points to where ASCII bytes should go
upper – 1 to use ABCDEF, 0 for abcdef
This writes only the two characters; no terminating zero char is added. No error is possible.
fastrand – return a random number in the range 0 to n−1
int fastrand(uint n, uint *seed);
n – defines the random number range
seed – static or persistent seed
returns the pseudo-random number
n should be <30,000,000; the seed is provided by the caller initially but should not then be altered by the caller.
This uses the conventional X(n+1)=mod((A*X(n) + C),M) as used in the Rexx function package, but without some refinements.
The values for A and M are from the table in D.E.Knuth’s book, “Seminumerical Algorithms”, Addison-Wesley, 1981, page 102 (section 3.3.2). These particular values are attributed to Lavaux and Janssens.
- C is chosen as 1
- M is 2**32
- A is 1664525 (RNDFACT)
hex2byte – convert two hexadecimal characters to a byte
int hex2byte(char *hex, byte *b);
hex – points to two ASCII bytes
b – points to byte which will receive the converted hexadecimal; set to 0x00 if error
returns 0 if OK, −1 if error
This accepts both uppercase and lowercase A->F and a->f.
isqrtg – return the square root of an int, with guess
int isqrtg(int num, int guess);
num – the number to square root
guess – is the number to converge from (e.g., last similar result); if < 1, isqrtg will make the first guess
returns the integer square root of num, such that result*result < num; if num is less than zero then 0 is returned
Note: the maximum input value is 2**31−1 (2147483647) hence the maximum possible returned value is 46340.
i2str – layout an integer as a string
int i2str(char *string, int num, int len);
string – where to layout the integer
num – the integer to lay out (not 0x80000000)
len – width of the field (string) to lay out into
returns 0 if OK, 1 if overflow (does not fit)
This writes num right-aligned and signed in the string field. No terminating zero char is written.
nop – do nothing with an int
int nop(int i);
i – an integer
returns i
This is used for placating compilers when parameters are intentionally or temporarily unused.
streq – caseless string compare
int streq(const char *char1, const char *char2);
char1 – first string to compare
char2 – second string
returns 1 if the strings caseless-compare equal, 0 otherwise
If both strings are empty they are equal.
strf – a reentrant mini printf to send a formatted string
int strf(putter putC, const char *format, va_list args);
putC – the function to which to send formatted characters
format – format and data string
args – list of other arguments as required
returns the number (count) of characters sent
The following format indicators are supported:
%% – ‘%’
%c – single char
%d, i – signed int
%m – minidec decimal number [iff MINIDEC is defined]
%s – string
%u – unsigned int
%x, X – hexadecimal
%\n – LF only (otherwise \n becomes CRLF)
The five printf format flags (+ - 0 # and space) are supported for all formats as in C99 (e.g., # only affects ‘x’, here).
The ‘h’, ‘l’, and ‘L’ modifiers are not supported.
Field width (e.g., %10i, %08x) is supported for all formats.
Field precision (e.g., %5.7i) is not supported; for decimals, use Quantize on a number to set its ‘precision’.
A \n in the string is expanded to \r\n unless preceded by %.
Almost no error checking is carried out, so, for example, the flags and field width can be used even if the format character is not one of those listed; the format character then appears in the output.