#include <unistd.h>
while (opt = getopt(argc, argv, ":if:"i) != -1)){// -1 means no more options to process
// if no recoginized option is given, ? is returned
//f: means option f requires an option. if no option is given, getopt returns ?
// if starting with :, getopt will return : instead of ? when no value is given for f.
switch (opt){
case 'i':
do_stuff();
case 'f':
printf("input args: %s", optarg);
break;
case '?':
break;
}
}- opt is option
- optarg is the argument, **A global variable**
- optind is number of arguments.
- Misc
- **If you see errors, check if you have break in case.**
- std::getenv
#include <cstdlib>
const char* env_p = std::getenv("PATH");
- signal
- the right hand side is the handler.
- signal(SIGINT, somefunc) //this will hijack ctrl-c
- signal(SIGPIPE, SIG_IGN) //ignore broken pipe? (the process reading a pipe or socket. SIG_IGN means to ignore it.)
- signal(SIGTERM, finish) //SIGTERM is an administrative signal sent by another process to terminate. Kernel will never send this.
- pseudo-random: to generate [0,n)
- rand() 是[0,RAND_MAX]，两个在中define. 如果你要generate [0,n）的话，那么你会不会选择rand%(n)呢？ 其实不要这么做。
- 当n 比较小时，比方说n=2）,那么最后很有可能是0,1 alternate 的。
- 还有，比如RAND_MAX为32768，假如n = 20000， 那么[0,12768]的数可以被两个rand 值产生。 而剩下[12768,20000)只有一个数相对应。
所以，你可以:
int r, bin_size; bin_size = RAND_MAX/n; do r = rand()/bin_size; while(r>=n); (discard 那些最靠近RAND_MAX的数)
-
rand() may not generate evenly distributed values, instead in C++11:
#include <random>
#include <iostream>
int main()
{
std::random_device rd; //Will be used to obtain a seed for the random number engine
std::mt19937 gen(rd()); //Standard mersenne_twister_engine seeded with rd()
std::uniform_int_distribution<> distrib(1, 6);
for (int n=0; n<10; ++n)
//Use `distrib` to transform the random unsigned int generated by gen into an int in [1, 6]
std::cout << distrib(gen) << ' ';
std::cout << '\n';
}