Memoize expensive polynomial searches

[mhostetter]

Use LRU cache to speed up subsequent calls to galois.irreducible_poly() and galois.primtiive_poly().

Before

In [2]: %time galois.primitive_poly(7, 4)
CPU times: user 677 ms, sys: 13.2 ms, total: 691 ms
Wall time: 745 ms
Out[2]: Poly(x^4 + x^2 + 3x + 5, GF(7))

In [3]: %time galois.primitive_poly(7, 4)
CPU times: user 338 ms, sys: 10.6 ms, total: 348 ms
Wall time: 337 ms
Out[3]: Poly(x^4 + x^2 + 3x + 5, GF(7))

In [4]: %time galois.primitive_poly(7, 4)
CPU times: user 331 ms, sys: 17 ms, total: 348 ms
Wall time: 336 ms
Out[4]: Poly(x^4 + x^2 + 3x + 5, GF(7))

In [5]: %time galois.primitive_poly(7, 4)
CPU times: user 355 ms, sys: 12.8 ms, total: 368 ms
Wall time: 348 ms
Out[5]: Poly(x^4 + x^2 + 3x + 5, GF(7))

After

In [1]: import galois

In [2]: %time galois.primitive_poly(7, 4)
CPU times: user 691 ms, sys: 12.7 ms, total: 703 ms
Wall time: 756 ms
Out[2]: Poly(x^4 + x^2 + 3x + 5, GF(7))

In [3]: %time galois.primitive_poly(7, 4)
CPU times: user 20 µs, sys: 13 µs, total: 33 µs
Wall time: 35 µs
Out[3]: Poly(x^4 + x^2 + 3x + 5, GF(7))

In [4]: %time galois.primitive_poly(7, 4)
CPU times: user 20 µs, sys: 13 µs, total: 33 µs
Wall time: 34.8 µs
Out[4]: Poly(x^4 + x^2 + 3x + 5, GF(7))

In [5]: %time galois.primitive_poly(7, 4)
CPU times: user 33 µs, sys: 0 ns, total: 33 µs
Wall time: 34.6 µs
Out[5]: Poly(x^4 + x^2 + 3x + 5, GF(7))