Optimize netloc extraction

[elliotwutingfeng]

SCHEME_RE can be replaced with an if-else equivalent for tangible speed improvement.

Benchmarks of this optimization together with improvements in #283

Python 3.10, Linux x64, Ryzen 7 5800X

import tldextract

%timeit tldextract.extract("")
%timeit tldextract.extract("com")
%timeit tldextract.extract("example\u3002com")
%timeit tldextract.extract("subdomain\uff0eexample\uff61com")
%timeit tldextract.extract("a\u3002very\uff0elong\uff61subdomain\u3002example\uff0ecom")
%timeit tldextract.extract("an\uff61even\u3002longer\uff0eand\uff61complex\u3002subdomain\uff0eexample\uff61com")
%timeit tldextract.extract("https://a\u3002b\uff0ec\uff61d\u3002e\uff0ef\uff61g\u3002h\uff0ei\uff61j\u3002k\uff0el\uff61m\u3002n\uff0eoo\uff61pp\u3002qqq\uff0errrr\uff61ssssss\u3002tttttttt\uff0euuuuuuuuuuu\uff61vvvvvvvvvvvvvvv\u3002wwwwwwwwwwwwwwwwwwwwww\uff0exxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx\uff61yyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy\u3002zzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzzz.tw")
%timeit tldextract.extract("\u3002\u3002\u3002\u3002\u3002\u3002\u3002\u3002\u3002\u3002\u3002\u3002\u3002\u3002\u3002\u3002\u3002\u3002\u3002\u3002\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff0e\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61\uff61")

1.72 µs ± 14.4 ns per loop (mean ± std. dev. of 7 runs, 1,000,000 loops each)
1.75 µs ± 15.7 ns per loop (mean ± std. dev. of 7 runs, 1,000,000 loops each)
2.34 µs ± 9.41 ns per loop (mean ± std. dev. of 7 runs, 100,000 loops each)
2.41 µs ± 4.97 ns per loop (mean ± std. dev. of 7 runs, 100,000 loops each)
2.6 µs ± 4.92 ns per loop (mean ± std. dev. of 7 runs, 100,000 loops each)
2.67 µs ± 18.9 ns per loop (mean ± std. dev. of 7 runs, 100,000 loops each)
3.76 µs ± 13.4 ns per loop (mean ± std. dev. of 7 runs, 100,000 loops each)
4.49 µs ± 21.2 ns per loop (mean ± std. dev. of 7 runs, 100,000 loops each)

Changes

• Replaced SCHEME_RE with faster if-else equivalent

[john-kurkowski]

Nice!

I took your code in a little different direction, 1 public function in the module calling 1 private utility function, vs. 2 public functions with 1 wrapping the other.

[john-kurkowski]

Thank you!