[dev] [st patch queue 07/12] Remove unnecessary break;s

[noname]

---
 st.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/st.c b/st.c
index 9de7f81..061dafd 100644
--- a/st.c
+++ b/st.c
@@ -1778,7 +1778,6 @@ tsetmode(bool priv, bool set, int *args, int narg) {
for(lim = args + narg; args  lim; ++args) {
if(priv) {
switch(*args) {
-   break;
case 1: /* DECCKM -- Cursor key */
MODBIT(term.mode, set, MODE_APPCURSOR);
break;
@@ -1897,7 +1896,6 @@ tsetmode(bool priv, bool set, int *args, int narg) {
fprintf(stderr,
erresc: unknown set/reset mode %d\n,
*args);
-   break;
}
}
}
-- 
1.8.4

Re: [dev] [st patch queue 07/12] Remove unnecessary break;s

[Markus Teich]

noname wrote:
 @@ -1897,7 +1896,6 @@ tsetmode(bool priv, bool set, int *args, int narg) {
   fprintf(stderr,
   erresc: unknown set/reset mode %d\n,
   *args);
 - break;
   }
   }
   }

Heyho,

although a break statement in the last case is unnecessary, I think it should be
left there. If someone appends a new case, he could forget to break this one and
get an unwanted fallthrough. The style in st.c is not consistent, there are
multiple places without the break in the last case and multiples with it. For a

switch(pid = fork())

it is pretty unlikely, that someone legitimately adds a new case other than -1,
0 or default, but it would not harm and help a more consistent style. If we can
agree on it, I'll write the patch to add break;s for the last switch-cases,
where they are missing.

--Markus

Re: [dev] [st patch queue 07/12] Remove unnecessary break;s

[Roberto E. Vargas Caballero]

 although a break statement in the last case is unnecessary, I think it should 
 be
 left there. If someone appends a new case, he could forget to break this one 
 and
 get an unwanted fallthrough. The style in st.c is not consistent, there are
 multiple places without the break in the last case and multiples with it. For 
 a

Yeah, usually it is a good idea put the break in the last case to avoid a
unwanted fallthrough. Maybe, if you put always default case in the end of
the switch, then it is also posible to remove the break in this case, but
due to the kind of switch you can find in st, it is impossible ensure
you are going to have always a default, so I vote for putting always a break
in the last.

 switch(pid = fork())
 
 it is pretty unlikely, that someone legitimately adds a new case other than 
 -1,
 0 or default, but it would not harm and help a more consistent style. If we 
 can
 agree on it, I'll write the patch to add break;s for the last switch-cases,
 where they are missing.

Yeah, I agree, and I am going to commit the patch of nonamed only with the
break in the beginnig of the switch.

Regards,

-- 
Roberto E. Vargas Caballero