Revert "Make grep more memory-efficient"

This reverts commit d9a098ae6bc07188cbaefcd188e5911dec41815d.

Conflicts:
	grep.c

Hiltjo Posthuma <hiltjo@codemadness.org> said:

(Re)compiling the regex for each line doesn't make sense (imho) and
slows it down a lot.

A simple test:

time seq 1 100000000 | grep 'a'

sbase grep (before patch):
    0m22.21s real     0m0.00s user     0m0.00s system

sbase grep (after patch):
    2m16.28s real     0m0.00s user     0m0.00s system

coreutils grep:
    0m1.15s real     0m0.00s user     0m0.00s system

So the patch should be reverted I think.
This commit is contained in:
sin 2014-05-12 11:59:39 +01:00
parent 5b3a0a9382
commit 4896bdf6de

36
grep.c
View File

@ -11,7 +11,7 @@
enum { Match = 0, NoMatch = 1, Error = 2 }; enum { Match = 0, NoMatch = 1, Error = 2 };
static void addpattern(const char *); static void addpattern(const char *);
static bool grep(FILE *, const char *, int); static bool grep(FILE *, const char *);
static bool eflag = false; static bool eflag = false;
static bool vflag = false; static bool vflag = false;
@ -20,6 +20,7 @@ static char mode = 0;
static struct plist { static struct plist {
char *pattern; char *pattern;
regex_t preg;
struct plist *next; struct plist *next;
} *phead; } *phead;
@ -34,7 +35,7 @@ main(int argc, char *argv[])
{ {
bool match = false; bool match = false;
struct plist *pnode, *tmp; struct plist *pnode, *tmp;
int i, flags = REG_NOSUB; int i, n, flags = REG_NOSUB;
FILE *fp; FILE *fp;
ARGBEGIN { ARGBEGIN {
@ -61,9 +62,8 @@ main(int argc, char *argv[])
usage(); usage();
} ARGEND; } ARGEND;
/* no pattern */
if(argc == 0 && !eflag) if(argc == 0 && !eflag)
usage(); usage(); /* no pattern */
/* If -e is not specified treat it as if it were */ /* If -e is not specified treat it as if it were */
if(!eflag) { if(!eflag) {
@ -72,14 +72,23 @@ main(int argc, char *argv[])
argv++; argv++;
} }
/* Compile regex for all search patterns */
for(pnode = phead; pnode; pnode = pnode->next) {
if((n = regcomp(&pnode->preg, pnode->pattern, flags)) != 0) {
char buf[BUFSIZ];
regerror(n, &pnode->preg, buf, sizeof buf);
enprintf(Error, "invalid pattern: %s\n", buf);
}
}
many = (argc > 1); many = (argc > 1);
if(argc == 0) { if(argc == 0) {
match = grep(stdin, "<stdin>", flags); match = grep(stdin, "<stdin>");
} else { } else {
for(i = 0; i < argc; i++) { for(i = 0; i < argc; i++) {
if(!(fp = fopen(argv[i], "r"))) if(!(fp = fopen(argv[i], "r")))
enprintf(Error, "fopen %s:", argv[i]); enprintf(Error, "fopen %s:", argv[i]);
if(grep(fp, argv[i], flags)) if(grep(fp, argv[i]))
match = true; match = true;
fclose(fp); fclose(fp);
} }
@ -87,6 +96,7 @@ main(int argc, char *argv[])
pnode = phead; pnode = phead;
while(pnode) { while(pnode) {
tmp = pnode->next; tmp = pnode->next;
regfree(&pnode->preg);
free(pnode->pattern); free(pnode->pattern);
free(pnode); free(pnode);
pnode = tmp; pnode = tmp;
@ -110,29 +120,20 @@ addpattern(const char *pattern)
} }
bool bool
grep(FILE *fp, const char *str, int flags) grep(FILE *fp, const char *str)
{ {
char err[BUFSIZ];
char *buf = NULL; char *buf = NULL;
long n, c = 0; long n, c = 0;
int r;
static regex_t preg;
size_t size = 0, len; size_t size = 0, len;
struct plist *pnode; struct plist *pnode;
bool match = false; bool match = false;
for(n = 1; afgets(&buf, &size, fp); n++) { for(n = 1; afgets(&buf, &size, fp); n++) {
for(pnode = phead; pnode; pnode = pnode->next) { for(pnode = phead; pnode; pnode = pnode->next) {
if((r = regcomp(&preg, pnode->pattern, flags)) != 0) {
regerror(r, &preg, err, sizeof err);
enprintf(Error, "invalid pattern: %s\n", err);
}
if(buf[(len = strlen(buf))-1] == '\n') if(buf[(len = strlen(buf))-1] == '\n')
buf[--len] = '\0'; buf[--len] = '\0';
if(regexec(&preg, buf, 0, NULL, 0) ^ vflag){ if(regexec(&pnode->preg, buf, 0, NULL, 0) ^ vflag)
regfree(&preg);
continue; continue;
}
switch(mode) { switch(mode) {
case 'c': case 'c':
c++; c++;
@ -158,7 +159,6 @@ grep(FILE *fp, const char *str, int flags)
end: end:
if(ferror(fp)) if(ferror(fp))
enprintf(Error, "%s: read error:", str); enprintf(Error, "%s: read error:", str);
regfree(&preg);
free(buf); free(buf);
return match; return match;
} }