Message ID | 20180612120441.22325-1-dirk@gouders.net (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
On 06/12/2018 05:04 AM, Dirk Gouders wrote: > This patch adds boundary checks for nconf's i-search pattern. > > Further, the pattern buffer is always bzero'ed when '/' is > pressed, so the second line in the code below was not needed (and > otherwise wouldn't have worked as expected): > > state->pattern[strlen(state->pattern)] = c; > state->pattern[strlen(state->pattern)] = '\0'; > > Finally, the pattern length was reduced to a length that still > seems sufficient but will not fill more than the top line of the > screen, thus eliminating special treatment needs on resizes or normal > exit of i-search. > > Signed-off-by: Dirk Gouders S-O-B: line should also have your email address after your name. > --- > scripts/kconfig/nconf.c | 10 ++++++---- > 1 file changed, 6 insertions(+), 4 deletions(-) Your patch descsription makes sense. I tested nconf (make nconfig) both with and without this patch. I didn't like the /search results in either case. F1 (Help) says: ││ Start incremental, case-insensitive search for STRING in menu entries, ││ ││ no regex support, STRING is displayed in upper left corner ││ ││ </>STRING ││ ││ Remove last character <Backspace> ││ ││ Jump to next hit <Down> ││ ││ Jump to previous hit <Up> ││ ││ Exit menu search mode </> <Esc> but I couldn't get <Down> == "Jump to next hit" to work for me at all (with or without the patch). thanks,
Randy Dunlap <rdunlap@infradead.org> writes: > On 06/12/2018 05:04 AM, Dirk Gouders wrote: >> This patch adds boundary checks for nconf's i-search pattern. >> >> Further, the pattern buffer is always bzero'ed when '/' is >> pressed, so the second line in the code below was not needed (and >> otherwise wouldn't have worked as expected): >> >> state->pattern[strlen(state->pattern)] = c; >> state->pattern[strlen(state->pattern)] = '\0'; >> >> Finally, the pattern length was reduced to a length that still >> seems sufficient but will not fill more than the top line of the >> screen, thus eliminating special treatment needs on resizes or normal >> exit of i-search. >> >> Signed-off-by: Dirk Gouders > > S-O-B: line should also have your email address after your name. > >> --- >> scripts/kconfig/nconf.c | 10 ++++++---- >> 1 file changed, 6 insertions(+), 4 deletions(-) > > Your patch descsription makes sense. I tested nconf (make nconfig) > both with and without this patch. I didn't like the /search results > in either case. F1 (Help) says: > > ││ Start incremental, case-insensitive search for STRING in menu entries, ││ > ││ no regex support, STRING is displayed in upper left corner ││ > ││ </>STRING ││ > ││ Remove last character <Backspace> ││ > ││ Jump to next hit <Down> ││ > ││ Jump to previous hit <Up> ││ > ││ Exit menu search mode </> <Esc> > > but I couldn't get <Down> == "Jump to next hit" to work for me at all > (with or without the patch). Thank you for testing and commenting on the patch. I will correct the S-O-B line and also remove one "search" from the subject. The navigation through matches works here, did use the up/down arrow keys? Because of your comment, I also tested page-up/page-down and noticed that those keys add 'S', 'R' to the search pattern, respectively... Dirk -- To unsubscribe from this list: send the line "unsubscribe linux-kbuild" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/scripts/kconfig/nconf.c b/scripts/kconfig/nconf.c index 97b78445584b..0e224528aaf9 100644 --- a/scripts/kconfig/nconf.c +++ b/scripts/kconfig/nconf.c @@ -1007,11 +1007,12 @@ static void adj_match_dir(match_f *match_direction) /* else, do no change.. */ } +#define PATTERN_LENGTH 64 struct match_state { int in_search; match_f match_direction; - char pattern[256]; + char pattern[PATTERN_LENGTH]; }; /* Return 0 means I have handled the key. In such a case, ans should hold the @@ -1035,8 +1036,8 @@ static int do_match(int key, struct match_state *state, int *ans) return 1; if (isalnum(c) || isgraph(c) || c == ' ') { - state->pattern[strlen(state->pattern)] = c; - state->pattern[strlen(state->pattern)] = '\0'; + if (strlen(state->pattern) + 1 < PATTERN_LENGTH) + state->pattern[strlen(state->pattern)] = c; adj_match_dir(&state->match_direction); *ans = get_mext_match(state->pattern, state->match_direction); @@ -1049,7 +1050,8 @@ static int do_match(int key, struct match_state *state, int *ans) *ans = get_mext_match(state->pattern, state->match_direction); } else if (key == KEY_BACKSPACE || key == 127) { - state->pattern[strlen(state->pattern)-1] = '\0'; + if (state->pattern[0] != '\0') + state->pattern[strlen(state->pattern)-1] = '\0'; adj_match_dir(&state->match_direction); } else terminate_search = 1;
This patch adds boundary checks for nconf's i-search pattern. Further, the pattern buffer is always bzero'ed when '/' is pressed, so the second line in the code below was not needed (and otherwise wouldn't have worked as expected): state->pattern[strlen(state->pattern)] = c; state->pattern[strlen(state->pattern)] = '\0'; Finally, the pattern length was reduced to a length that still seems sufficient but will not fill more than the top line of the screen, thus eliminating special treatment needs on resizes or normal exit of i-search. Signed-off-by: Dirk Gouders --- scripts/kconfig/nconf.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-)