Replace inspect.stack() with traceback.extract_stack() to solve performance bottleneck

[aploium]

Use traceback.extract_stack() instead of inspect.stack() to check upper stack method, they are at same functionality, but the traceback(5μs) is 400x faster than inspect.stack(2ms). This method affects many common operation, such as add() and from_timestamp(tz), 2ms delay typically cause performance issues.
This method was available since early py 2.x

EDIT: the performance diff is not 50x, but 400x

Performance

before

%%timeit
inspect.stack()[1][3]
# 2.11 ms ± 14.4 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)

%%timeit
pendulum.from_timestamp(1687658855, "Asia/Shanghai")
# 2.21 ms ± 63 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)

after

%%timeit
traceback.extract_stack(limit=2)[0].name
# 3.7 µs ± 73.5 ns per loop (mean ± std. dev. of 7 runs, 100,000 loops each)

%%timeit
pendulum.from_timestamp(1687658855, "Asia/Shanghai")
# 8.45 µs ± 142 ns per loop (mean ± std. dev. of 7 runs, 100,000 loops each)

Calling Result Demo

import inspect
import traceback

def call1(f):
    f()
def call2(f):
    call1(f)

def old():
    print("old return is->", inspect.stack()[1][3])
def new():
    print("new return is->", traceback.extract_stack(limit=2)[0].name)


call2(old)
call2(new)

The output is

old return is-> call1
new return is-> call1

Pull Request Check List

• [ x ] Added tests for changed code.
  No new tests required
• [ x ] Updated documentation for changed code.
  No new document required