Increase performance of parser

[J535D165]

As we approach the v1 release, I believe it's time for this API change. This will result in a savings of around 15-20% on a large dataset.

------------------------------------------------------------------------------------------------- benchmark: 2 tests -------------------------------------------------------------------------------------------------
Name (time in ms)                                Min                   Max                  Mean             StdDev                Median                IQR            Outliers     OPS            Rounds  Iterations
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
test_benchmark_real_data (0004_use_set)     800.9379 (1.0)        883.0898 (1.0)        838.6729 (1.0)      29.1617 (1.0)        836.2794 (1.0)      35.6849 (1.0)           4;0  1.1924 (1.0)          10           5
test_benchmark_real_data (0005_no_set_)     935.0793 (1.17)     1,097.2022 (1.24)     1,013.3880 (1.21)     59.4632 (2.04)     1,003.6009 (1.20)     88.8859 (2.49)          4;0  0.9868 (0.83)         10           5
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

[J535D165]

Tests fail on the README test. I suggest turning off the README.md testing for the time being to avoid a mismatch between the examples and the v1 changes.

[shapiromatron]

@J535D165 can you update the README.rst so it shows a set instead of a list? This would be a nice addition.