NAME

regcomp, regcomplit, regcompnl, regexec, regsub, rregexec, rregsub, regerror − regular expression

SYNOPSIS

­#include <regexp.h>

Reprog∗regcomp(char ∗exp)

Reprog∗regcomplit(char ∗exp)

Reprog∗regcompnl(char ∗exp)

intregexec(Reprog ∗prog, char ∗string, Resub ∗match, int msize)

voidregsub(char ∗source, char ∗dest, Resub ∗match, int msize)

intrregexec(Reprog ∗prog, Rune ∗string, Resub ∗match, int msize)

voidrregsub(Rune ∗source, Rune ∗dest, Resub ∗match, int msize)

voidregerror(char ∗msg)

DESCRIPTION

­Regcomp compiles a regular expression and returns a pointer to the generated description.  The space is allocated by malloc(2) and may be released by ­free.  Regular expressions are exactly as in regexp(6).

­Regcomplit is like ­regcomp except that all characters are treated literally.  ­Regcompnl is like ­regcomp except that the ­.  metacharacter matches all characters, including newlines. 

­Regexec matches a null-terminated ­string against the compiled regular expression in ­prog.  If it matches, ­regexec returns ­1 and fills in the array ­match with character pointers to the substrings of ­string that correspond to the parenthesized subexpressions of exp: match[i].sp points to the beginning and match[i].ep points just beyond the end of the ith substring. (Subexpression ­i begins at the ith left parenthesis, counting from 1.) Pointers in ­match[0] pick out the substring that corresponds to the whole regular expression.  Unused elements of ­match are filled with zeros.  Matches involving ∗, +, and ­?  are extended as far as possible.  The number of array elements in ­match is given by ­msize.  The structure of elements of ­match is:

typedef struct {
union {
   char ∗sp;
   Rune ∗rsp;
};
union {
   char ∗ep;
   Rune ∗rep;
};
} Resub;

If ­match[0].sp is nonzero on entry, ­regexec starts matching at that point within string. If ­match[0].ep is nonzero on entry, the last character matched is the one preceding that point. 

­Regsub places in ­dest a substitution instance of ­source in the context of the last ­regexec performed using ­match.  Each instance of \n, where ­n is a digit, is replaced by the string delimited by match[n].sp and match[n].ep.  Each instance of ­& is replaced by the string delimited by ­match[0].sp and match[0].ep. 

­Regerror, called whenever an error is detected in ­regcomp, ­regexec, or ­regsub, writes the string ­msg on the standard error file and exits.  ­Regerror can be replaced to perform special error processing. 

­Rregexp and ­rregsub are variants of ­regexec and ­regsub that use strings of ­Runes instead of strings of chars.  With these routines, the ­rsp and ­rep fields of the ­match array elements should be used. 

SEE ALSO

grep(1)

DIAGNOSTICS

­Regcomp returns ­0 for an illegal expression or other failure.  ­Regexec returns 0 if ­string is not matched. 

Plan 9  —  January 03, 1993