getfields, getmfields, setfields, tokenize - break a string into fields #include <u.h>

#include <libc.h>

int getfields(char *str, char **ptrs, int nptrs)

int getmfields(char *str, char **ptrs, int nptrs)

char* setfields(char *fielddelim)

int tokenize(char *str, char **args, int max)

Getfields breaks the null-terminated string str into at most nptrs null-terminated fields and places pointers to the start of these fields in the array ptrs . It returns the number of fields and terminates the list of pointers with a zero pointer. It overwrites some of the bytes in str . If there are nptr or more fields, the list will not end with zero and the last `field' will extend to the end of the input string and may contain delimiters.

A field is defined as a maximal sequence of characters not in a set of field delimiters. Adjacent fields are separated by exactly one delimiter. No field follows a delimiter at the end of string. Thus a string of just two delimiter characters contains two empty fields, and a nonempty string with no delimiters contains one field.

Getmfields is the same as getfields except that fields are separated by maximal strings of field delimiters rather than just one.

Setfields makes the field delimiters (space and tab by default) be the characters of the string fielddelim and returns a pointer to a string of the previous delimiters.

Tokenize breaks null-terminated string str into tokens by replacing every blank or newline with a null byte. Pointers to successive non-empty tokens are placed in args . Processing stops after max tokens are processed. Tokenize returns the number of tokens processed. Tokenize does not terminate args with a null pointer.

Of these routines, only tokenize is in Alef. /sys/src/libc/port/getfields.c

/sys/src/libc/port/tokenize.c

strtok in strcat (2)