C

https://netradiant.gitlab.io/post/2017-07-10-pk3dir-dpkdir/
would have done an if chain myself , to avoid a for loop for... 4...? 3 elements?


char pkgsomething = 'pk3';
char pkgsomething2 = 'dvksoaap';
char pkgsomething3 = 'pk77';

if (bla == pkgsomething) {
stuff;
s = 1;
} else if (!s && bla == pkgsomething2) {
stuff;
s = 1;
} else if (!s && bla == pkgsomething3) {
stuff;
s = 1;
}


#The original code was
#if (bla == pkgsomething) {
# stuff;
# s = 1;
#}


No I do not think you should "risk" a for loop for 3 unchanging elements.

I know you trans fucking faggots and feminist cunts will have a problem with that;
But you faggots also use switch statements for non-sequential non-intergers.

#Anime style would also work fine:
if (bla == pkgsomething) {
stuff;
s = 1;
}

if (!s && bla == pkgsomething2) {
stuff;
s = 1;
}

if (!s && bla == pkgsomething3) {
stuff;
s = 1;
}

That's cleaner, as it doesn't suggest that s is unreachable
(though s was defined somewhere above this bloc level)

#BUT NOO U HAVE TO HAVE AN IF ELSIF OR YOU ARE YANDERECHAN!

s is set if a fake package is used,
and that code will work on all platforms and compilers.
While .. the else if might be optimized away and have unknown compiler specific implementations.

on some compilers and some optimization levels the s might not be seen as changed by the same-level bloc.

https://github.com/TTimo/GtkRadiant/pull/533/commits/ba300b17519e4865de348044c2068e7a83976abe

For loops used for unchanging array, of 3 elements+null

This:
const char* cur_ext = pak_ext;
if ( strcasecmp( ext, cur_ext ) == 0 ) {
is_pak = TRUE;
}
cur_ext = g_strconcat(cur_ext, pakdir_suf, NULL);
if ( strcasecmp( ext, cur_ext ) == 0 ) {
is_pak = TRUE;


Was changed to:
for ( int i = 0; pak_ext_list[i] != NULL ; i++ ) {
const char* cur_ext = pak_ext_list[i];
if ( strcasecmp( ext, cur_ext ) == 0 ) {
is_pak = TRUE;
}
cur_ext = g_strconcat(cur_ext, pakdir_suf, NULL);
if ( strcasecmp( ext, cur_ext ) == 0 ) {
is_pak = TRUE;
}

(note I see they are not following ISOC90)

const char* pak_ext = ".pk3";
const char* pak2_ext = ".dvks";
const char* pak3_ext = ".pk77";
I'd rather do something like this,
const char* cur_ext = pak_ext;
if ( strcasecmp( ext, cur_ext ) == 0 ) {
is_pak = TRUE;
}

cur_ext = g_strconcat(cur_ext, pakdir_suf, NULL);
if ( strcasecmp( ext, cur_ext ) == 0 ) {
is_pak = TRUE;
}
//Note: since cur_ext was changed even though it's a const,
//This sets the pattern

//Other additional package "formats" for no particular reason
if (!is_pak) {
cur_ext = pak2_ext;
if ( strcasecmp( ext, cur_ext ) == 0 ) {
is_pak = TRUE;
}

cur_ext = g_strconcat(cur_ext, pakdir_suf, NULL);
if ( strcasecmp( ext, cur_ext ) == 0 ) {
is_pak = TRUE;
}
}

//Other additional package "formats" for no particular reason
if (!is_pak) {
cur_ext = pak3_ext;
if ( strcasecmp( ext, cur_ext ) == 0 ) {
is_pak = TRUE;
}

cur_ext = g_strconcat(cur_ext, pakdir_suf, NULL);
if ( strcasecmp( ext, cur_ext ) == 0 ) {
is_pak = TRUE;
}
}

Yes it uses "more code" and "boilerplate"
But it doesn't run a for loop for almost no reason for just 2 more extension "types"
Which are all zip files anyway.

It's not like these are added frequently.
Nor is this a dynamic list of supported formats?
So why not set it in stone.

Why risk a for loop running away for no fucking reason?
If the counter bit is changed in ram you have a runaway

Since almost all will be pk3 and some pk3dir
add a check around the other pack code block totally to check if we allready are a pak
this will almost always be true, thus skipping the seldom used code.
Only will be run when it's one of the new weird gay paks

if (!is_pak) {
//all that stuff past the first 2

}

if it ain't lisp its crap

>>11
RMS wrote the gnu C compiler, retard.
So try again fucking idiot fake.

>>10-12
HAX MY ANUS!!!!!!!!!!!!!!!!!!!!!! u mad? u so mad mad fuck anus mad
u madder one multiple but more???
?????
fuck u and ur ded dof

>13
is the last numerical age a girl should be married off at.
It's too late by then anyway.
Fat fucking cunt bitches.

It's a bad deal by then anyway.