(WIP) Fix optimize-kicks == 2

[michaelortmann]

Found by: michaelortmann
Patch by: michaelortmann
Fixes:

One-line summary:
Fix optimize-kicks == 2

Additional description (if needed):

• Fix snprintf() overlaps destination object for optimize-kicks == 2
  This problem was hidden because the code used egg_snprintf(), which gcc cant check and throw the following warning properly:

gcc -fPIC -g -O2 -pipe -Wall -I. -I../../.. -I../../.. -I../../../src/mod  -DHAVE_CONFIG_H -I/usr/include -g3 -DDEBUG -DDEBUG_ASSERT -DDEBUG_MEM -DDEBUG_DNS  -DMAKING_MODS -c .././server.mod/server.c && mv -f server.o ../
.././server.mod/server.c: In function ‘parse_q’:
[...]
  592 |             snprintf(newnicks, sizeof newnicks, "%s,%s", newnicks, newnick);
      |             ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.././server.mod/server.c:592:13: warning: ‘snprintf’ argument 4 overlaps destination object ‘newnicks’ [-Wrestrict]
.././server.mod/server.c:571:54: note: destination object referenced by ‘restrict’-qualified argument 1 was declared here
  571 |   char buf[SENDLINEMAX], *msg, *nicks, *nick, *chan, newnicks[SENDLINEMAX],
      |                                                      ^~~~~~~~

• egg_snprintf() -> snprintf()
• Fix a format truncation warning by properly adjusting size of newnicks[]
• The rest of the code compares the message in the queue with "KICK" (4 chars) and not "KICK " (5 chars), which is sufficient.
  lll
• Do snprintf(newmsg) only if we have to and no need to strlen() when we already got the len returned from snprintf()

Test cases demonstrating functionality (if applicable):
Is there a logic error in

eggdrop/src/mod/server.mod/server.c

Line 592 in 322bddb

egg_snprintf(newnicks, sizeof newnicks, ",%s", nick);

? No concatenation?
Still untested. Please can someone step in and test this PR?

[vanosg]

Please test and paste results, or close. Thanks!