regexp

Regular expression processor.

Available in:

Apps (win)	Apps (char)	Reportwriter	RPC	Standalone PL
X	X	X	X	X

Syntax

int regexp(option [| flags][,expr1[,expr2]])
int        option
string     expr1
string     expr2

Description

Applies a regular expression against a string and returns true if there is a match or false if there is no match. The options and flags are defined in trim.h and are:

regexp_end (option)	returns the ending offset of the match after regexp_exists returns true. Requires the regexp_offsets flag.
regexp_exist (option)	executes the cached regular expression tree against `expr1` and returns true if there is a match, otherwise false. If `expr2` is supplied, then the regexp processing uses it as the regular expression instead of the cached one. The cached regular expression is not affected.
regexp_free (option)	frees the cached regular expresssion.
regexp_init (option)	initializes and caches the regular expresssion `expr1`. Any existing regular expression is automatically replaced.
regexp_start (option)	returns the starting offset of the match after regexp_exists returns true. Requires the regexp_offsets flag.
regexp_icase (flag)	ignore case when checking for matches.
regexp_newline (flag)	"match any" in expr1 does not match newline.
regexp_offsets (flag)	calculate start/end offsets.

Notes

If the same regular expression is to be used multiple times, then it is much more efficient to use regexp_init, regexp_exist, regexp_free than regexp_exist with the regular expression.

Regexp_start and regexp_end are only valid when regexp_exists is used without the regular expression argument.

Example

The following demonstrates the performance difference between cached and uncached regular expressions.

#define iters 100000

{
int  i;
char exp[20] = "N*s";
char str[20] = "Niklas Back";

timestamp();

regexp(regexp_init,exp);

for (i=iters;i;i--) regexp(regexp_exist,str);

printf(regexp(regexp_exist,str) ^^ ", Seconds: " ^^ timestamp());

for (i=iters;i;i--) regexp(regexp_exist,str,exp);

printf(regexp(regexp_exist,str,exp) ^^ ", Seconds: " ^^ timestamp());
}

The following finds the beginning and ending offsets in a string.

{
char exp[20] = "N.*?s";
char str[20] = "Back, Niklas";

regexp(regexp_init|regexp_offsets,exp);
regexp(regexp_exist,str);
printf(regexp(regexp_start));
printf(regexp(regexp_end));
}

returns

6
12