last part of my quest.  hope not to annoy, but i get slightly
triggered by strlen(foo) == 0 since it need to traverse the whole
string just to assert that the first byte is not NUL.  After seeing it
a few times in gotwebd i decided to do a sweep, and that explains my
recent diffs.

Unlinke the previous, this is highly subjective and opinable.  It
avoids computing the string length just to see if it's empty, and
changes some for() loops too along the way.

I haven't checked, but won't be surprised if the compiler emits the
same code before and after the diff.  I'd still prefer without all
these strlen(), and I personally find it even easier to read, but I'm
not attached to it.  I can't play the optimization card since these
places are nowhere near being performance-critical.  since i wrote i
figured out i'd send it.  now back to real work ;)

diffstat refs/heads/main refs/heads/strlen
 M  got/got.c                 |  5+  5-
 M  gotwebd/config.c          |  1+  1-
 M  gotwebd/got_operations.c  |  2+  2-
 M  gotwebd/gotweb.c          |  4+  4-
 M  gotwebd/sockets.c         |  1+  1-
 M  tog/tog.c                 |  1+  1-

6 files changed, 14 insertions(+), 14 deletions(-)

diff refs/heads/main refs/heads/strlen
commit - f4a5cef1546205afab47f148edefabcf77c06d3b
commit + bedba0b194efdbd9ade4ce8005b575a8ad74b24c
blob - 94bf306447c1f76eadeb7357e88c4c7b6fce7790
blob + b713aa2e5f2e3920e0055fb2366825c0e1487eff
--- got/got.c
+++ got/got.c
@@ -878,7 +878,7 @@ cmd_import(int argc, char *argv[])
 	 * unveil(2) traverses exec(2); if an editor is used we have
 	 * to apply unveil after the log message has been written.
 	 */
-	if (logmsg == NULL || strlen(logmsg) == 0) {
+	if (logmsg == NULL || *logmsg == '\0') {
 		error = get_editor(&editor);
 		if (error)
 			goto done;
@@ -5969,7 +5969,7 @@ print_entry(struct got_tree_entry *te, const char *id,
 		err = got_tree_entry_get_symlink_target(&link_target, te, repo);
 		if (err)
 			return err;
-		for (i = 0; i < strlen(link_target); i++) {
+		for (i = 0; link_target[i] != '\0'; i++) {
 			if (!isprint((unsigned char)link_target[i]))
 				link_target[i] = '?';
 		}
@@ -6326,7 +6326,7 @@ cmd_status(int argc, char *argv[])
 			st.suppress = 1;
 			/* fallthrough */
 		case 's':
-			for (i = 0; i < strlen(optarg); i++) {
+			for (i = 0; optarg[i] != '\0'; i++) {
 				switch (optarg[i]) {
 				case GOT_STATUS_MODIFY:
 				case GOT_STATUS_ADD:
@@ -7989,7 +7989,7 @@ cmd_remove(int argc, char *argv[])
 			can_recurse = 1;
 			break;
 		case 's':
-			for (i = 0; i < strlen(optarg); i++) {
+			for (i = 0; optarg[i] != '\0'; i++) {
 				switch (optarg[i]) {
 				case GOT_STATUS_MODIFY:
 					delete_local_mods = 1;
@@ -8915,7 +8915,7 @@ collect_commit_logmsg(struct got_pathlist_head *commit
 	size_t len;
 
 	/* if a message was specified on the command line, just use it */
-	if (a->cmdline_log != NULL && strlen(a->cmdline_log) != 0) {
+	if (a->cmdline_log != NULL && *a->cmdline_log != '\0') {
 		len = strlen(a->cmdline_log) + 1;
 		*logmsg = malloc(len + 1);
 		if (*logmsg == NULL)
blob - b13f0b45ecb8c13c605728f061a869d69085a4e0
blob + 1c8057cb99a030e7a3b45871f35124b82101e918
--- gotwebd/config.c
+++ gotwebd/config.c
@@ -228,7 +228,7 @@ config_getsock(struct gotwebd *env, struct imsg *imsg)
 	    sock->conf.af_type == AF_UNIX ? "unix" :
 	    (sock->conf.af_type == AF_INET ? "inet" :
 	    (sock->conf.af_type == AF_INET6 ? "inet6" : "unknown")),
-	    strlen(sock->conf.unix_socket_name) ?
+	    *sock->conf.unix_socket_name != '\0' ?
 	    sock->conf.unix_socket_name : "none");
 
 	return 0;
blob - 61bee36696a4edd811768fba90d50e909418b91f
blob + a7525287e2c962f40938fd909cdb0b9142d1f605
--- gotwebd/got_operations.c
+++ gotwebd/got_operations.c
@@ -342,7 +342,7 @@ got_get_repo_commits(struct request *c, size_t limit)
 
 	TAILQ_INIT(&refs);
 
-	if (qs->file != NULL && strlen(qs->file) > 0)
+	if (qs->file != NULL && *qs->file != '\0')
 		if (asprintf(&file_path, "%s/%s", qs->folder ? qs->folder : "",
 		    qs->file) == -1)
 			return got_error_from_errno("asprintf");
@@ -376,7 +376,7 @@ got_get_repo_commits(struct request *c, size_t limit)
 	if (error)
 		goto done;
 
-	if (qs->file != NULL && strlen(qs->file) > 0) {
+	if (qs->file != NULL && *qs->file != '\0') {
 		error = got_commit_graph_open(&graph, file_path, 0);
 		if (error)
 			goto done;
blob - 983fb20185d96176ed732dfe115d8e8dd355b575
blob + 57acfb3983fd313a2a166e1c1296ed5262cb5b3d
--- gotwebd/gotweb.c
+++ gotwebd/gotweb.c
@@ -392,13 +392,13 @@ gotweb_get_server(uint8_t *server_name, uint8_t *subdo
 	struct server *srv = NULL;
 
 	/* check against the server name first */
-	if (strlen(server_name) > 0)
+	if (*server_name != '\0')
 		TAILQ_FOREACH(srv, &gotwebd_env->servers, entry)
 			if (strcmp(srv->name, server_name) == 0)
 				goto done;
 
 	/* check against subdomain second */
-	if (strlen(subdomain) > 0)
+	if (*subdomain != '\0')
 		TAILQ_FOREACH(srv, &gotwebd_env->servers, entry)
 			if (strcmp(srv->name, subdomain) == 0)
 				goto done;
@@ -614,7 +614,7 @@ qa_found:
 			}
 			break;
 		case INDEX_PAGE:
-			if (strlen(value) == 0)
+			if (*value == '\0')
 				break;
 			(*qs)->index_page = strtonum(value, INT64_MIN,
 			    INT64_MAX, &errstr);
@@ -635,7 +635,7 @@ qa_found:
 			}
 			break;
 		case PAGE:
-			if (strlen(value) == 0)
+			if (*value == '\0')
 				break;
 			(*qs)->page = strtonum(value, INT64_MIN,
 			    INT64_MAX, &errstr);
blob - bf53d6f9bff50c34f85de2733fe7ae29ff2401d7
blob + 3f932978bf05a203c0c4bcd43bf7390fb4d153c8
--- gotwebd/sockets.c
+++ gotwebd/sockets.c
@@ -231,7 +231,7 @@ sockets_conf_new_socket_fcgi(struct gotwebd *env, stru
 	memcpy(&acp->ss, &a->ss, sizeof(acp->ss));
 	acp->ipproto = a->ipproto;
 	acp->port = a->port;
-	if (strlen(a->ifname) != 0) {
+	if (*a->ifname != '\0') {
 		if (strlcpy(acp->ifname, a->ifname,
 		    sizeof(acp->ifname)) >= sizeof(acp->ifname)) {
 			fatalx("%s: interface name truncated",
blob - d694a6164057730509b8e3db4e24c4ae1ef16c64
blob + 50c38cd6d28d539dcfa273bc3233bbda7f02fe7a
--- tog/tog.c
+++ tog/tog.c
@@ -7304,7 +7304,7 @@ draw_tree_entries(struct tog_view *view, const char *p
 				free(id_str);
 				return err;
 			}
-			for (i = 0; i < strlen(link_target); i++) {
+			for (i = 0; i < link_target[i] != '\0'; i++) {
 				if (!isprint((unsigned char)link_target[i]))
 					link_target[i] = '?';
 			}