hawkinchina / zmdcstl Goto Github PK

zmdcstl is a data structure  and algorithm library that mimic c++ stl. 

Bugs, support requests, and feature requests are welcome.

Below are comparisons  results between c++ stl and zmdcstl:
currently without fully optimised, in average zmdcstl slower 
by 10% than c++stl in some tests cases. in some other test cases, 
zmdcstl can be faster than c++stl.

benchmark_vector:
zmdcstlvec:0.000255ms, stdvec:0ms, ratio: inf
PASS benchmark_vector_ctor:  (1 ticks, 0.001 sec)

zmdcstlvec:681.317ms, stdvec:964.43ms, ratio: 0.706445
PASS benchmark_vector_ctor_n:  (2935 ticks, 2.935 sec)

vec_ctor_n_v_non_pod:1008.4ms, stdvec_n_v_non_pod:994.384ms, ratio: 1.0141
vec_ctor_n_v_pod:57.88ms, stdvec_n_v_pod:52.2712ms, ratio: 1.1073
vec_ctor_n_v_ctype:4.77279ms, stdvec_n_v_ctype:6.27634ms, ratio: 0.760441
PASS benchmark_vector_ctor_n_v:  (3524 ticks, 3.524 sec)

zmdcstlvec:1227.39ms, stdvec:1196.01ms, ratio: 1.02624
PASS benchmark_vector_ctor_range:  (7499 ticks, 7.499 sec)

zmdcstlvec:1256.16ms, stdvec:1211.05ms, ratio: 1.03725
PASS benchmark_vector_ctor_range_n:  (7812 ticks, 7.812 sec)

cstl_vector_ctor_vector:1225.48ms, std_vector_ctor_vector:1229.65ms, ratio: 0.996611
PASS benchmark_vector_ctor_vector:  (7857 ticks, 7.857 sec)

zmdcstlvec:20.4933ms, stdvec:540.694ms, ratio: 0.0379018
PASS benchmark_vector_erase:  (3640 ticks, 3.640 sec)

cstlvec_swap_ctype:0.000511ms, stdvec_swap_ctype:0ms, ratio: inf
PASS benchmark_vector_swap:  (32 ticks, 0.032 sec)

cstlvec_reserve_non_pod:634.302ms, stdvec_reserve_non_pod:1963.87ms, ratio: 0.322986
cstlvec_reserve_pod:92.4203ms, stdvec_reserve_pod:83.7403ms, ratio: 1.10365
cstlvec_reserve_ctype:8.57114ms, stdvec_reserve_ctype:8.10692ms, ratio: 1.05726
PASS benchmark_vector_reserve:  (6465 ticks, 6.465 sec)


9 tests - 9 passed, 0 failed, 0 skipped (39770 ticks, 39.770 sec)

Total: 11 tests (39771 ticks, 39.771 sec), 0 assertions
Pass: 11, fail: 0, skip: 0.

Enjoy coding, cheers !