Accept --sleep <ms> in addition to --sleep=<ms>

[ChrisJr404]

Closes #357.

--sleep is the only common rate-limiting flag in sslscan and the manual arg parser only matched the strncmp("--sleep=", ...) form. The space-separated form --sleep <ms> silently fell through to the next else-if (and was eventually treated as an unrecognized argument or as a hostname), so users who tried --sleep 100 saw it appear to have no effect — that's exactly the confusion the issue describes.

Fix

Add a parallel else if (strcmp("--sleep", argv[argLoop]) == 0) branch that advances argLoop to consume the value, atoi-parses it, and sets options->sleep using the same >=0 guard as the existing branch. If the user passes --sleep with no following argument, print a friendly error pointing at both working forms:

$ ./sslscan --sleep
--sleep requires a value in milliseconds (e.g. --sleep 100 or --sleep=100)

Verification

End-to-end timing against example.com confirms both forms now sleep identically:

$ python3 -c "..." # measure wall-clock for each
--- baseline (no --sleep) ---  0.26s
--- --sleep=200 (existing) ---  2.49s
--- --sleep 200 (was broken)--- 2.51s

The build is clean against system OpenSSL on Linux:

$ make
Building against system OpenSSL...
$ ls -la sslscan
-rwxrwxr-x ... sslscan

I scoped this fix to --sleep only (the issue) rather than reworking the whole arg parser to handle both --option value and --option=value for every flag — happy to extend the pattern if you'd prefer it consistent across all options.

[rbsec]

I'm not against the idea of accepting both --sleep=200 (as per the help text and manpage) and --sleep 200 as equivalent - but making that change for just --sleep would make it inconsistent with the other arguments that require parameters, which is likely to cause far more problems than not accepting --sleep 200.

The error condition in this PR also don't work properly, because it only checks for --sleep as the final argument. If you try that then you get an error:

$ ./sslscan example.org --sleep   
--sleep requires a value in milliseconds (e.g. --sleep 100 or --sleep=100)

But that command would never have been valid the first place, because options need to come before the hostname. So if you try and follow the recommendation and run ./sslscan example.org --sleep 100 then it still doesn't work. And if you have an empty sleep in the middle of the command like $ ./sslscan --sleep --no-heartbleed example.org then it still doesn't error (which is the current behaviour).