fgets() vs. gets() timing.

report_for_vle.txt

dima@dima-x1:~/github/dkorolev/current/examples/datafest_talk/tier2_rides_by_months@unstable $ cat step5_rides_by_months_unoptimized.c 
// To run: gcc -O3 step5_rides_by_months_unoptimized.c && time ./a.out | tee >(md5sum)
// 
// Count rides by month, print the counters in the lexicographically sorted order of keys.
// The "canonical C" implementation, with no extra checks, unsafe in a few ways, but still only 2x slower than `wc -l`.

#include <stdio.h>

int MM[12];

int main(int argc, char** argv) {
  char line[10000];  // Unsafe if we encounter longer lines. -- D.K.
  FILE* f = fopen(argc >= 2 ? argv[1] : "../cooked.csv", "r");
  while (fgets(line, sizeof(line), f)) {
    ++MM[(line[7] - '0') * 10 + (line[8] - '0') - 1];  // Unsafe if those two raw characters are not "01" .. "12" range.
  }
  for (int m = 0; m < 12; ++m) {
    printf("%d %02d/%d\n", MM[m], m + 1, 2016);
  }
  fclose(f);
}
dima@dima-x1:~/github/dkorolev/current/examples/datafest_talk/tier2_rides_by_months@unstable $ diff step5_rides_by_months_unoptimized.c step5a_rides_by_months_unoptimized.c 
1c1
< // To run: gcc -O3 step5_rides_by_months_unoptimized.c && time ./a.out | tee >(md5sum)
---
> // To run: gcc -O3 step5a_rides_by_months_unoptimized.c && time cat ../cooked.csv | ./a.out | tee >(md5sum)
10c10
< int main(int argc, char** argv) {
---
> int main() {
12,13c12
<   FILE* f = fopen(argc >= 2 ? argv[1] : "../cooked.csv", "r");
<   while (fgets(line, sizeof(line), f)) {
---
>   while (gets(line)) {
19d17
<   fclose(f);
dima@dima-x1:~/github/dkorolev/current/examples/datafest_talk/tier2_rides_by_months@unstable $ gcc -O3 step5_rides_by_months_unoptimized.c 
dima@dima-x1:~/github/dkorolev/current/examples/datafest_talk/tier2_rides_by_months@unstable $ time cat ../cooked.csv | ./a.out | md5sum
bd5b6d7f389e64aab50c56fed61eaded  -

real	0m1.002s
user	0m0.589s
sys	0m0.417s
dima@dima-x1:~/github/dkorolev/current/examples/datafest_talk/tier2_rides_by_months@unstable $ time cat ../cooked.csv | ./a.out | md5sum
bd5b6d7f389e64aab50c56fed61eaded  -

real	0m0.874s
user	0m0.487s
sys	0m0.389s
dima@dima-x1:~/github/dkorolev/current/examples/datafest_talk/tier2_rides_by_months@unstable $ time cat ../cooked.csv | ./a.out | md5sum
bd5b6d7f389e64aab50c56fed61eaded  -

real	0m0.877s
user	0m0.431s
sys	0m0.448s
dima@dima-x1:~/github/dkorolev/current/examples/datafest_talk/tier2_rides_by_months@unstable $ gcc -O3 step5a_rides_by_months_unoptimized.c 
step5a_rides_by_months_unoptimized.c: In function ‘main’:
step5a_rides_by_months_unoptimized.c:12:10: warning: implicit declaration of function ‘gets’; did you mean ‘fgets’? [-Wimplicit-function-declaration]
   while (gets(line)) {
          ^~~~
          fgets
/tmp/cc7WVDJk.o: In function `main':
step5a_rides_by_months_unoptimized.c:(.text.startup+0x51): warning: the `gets' function is dangerous and should not be used.
dima@dima-x1:~/github/dkorolev/current/examples/datafest_talk/tier2_rides_by_months@unstable $ time cat ../cooked.csv | ./a.out | md5sum
bd5b6d7f389e64aab50c56fed61eaded  -

real	0m1.430s
user	0m0.584s
sys	0m1.957s
dima@dima-x1:~/github/dkorolev/current/examples/datafest_talk/tier2_rides_by_months@unstable $ time cat ../cooked.csv | ./a.out | md5sum
bd5b6d7f389e64aab50c56fed61eaded  -

real	0m1.392s
user	0m0.614s
sys	0m1.853s
dima@dima-x1:~/github/dkorolev/current/examples/datafest_talk/tier2_rides_by_months@unstable $ time cat ../cooked.csv | ./a.out | md5sum
bd5b6d7f389e64aab50c56fed61eaded  -

real	0m1.502s
user	0m0.615s
sys	0m2.049s