Increase default dictionary size, use more efficient dictionary algorithm

[peterkovacs]

I increased the size of the default dictionary to the top 10,000 passwords. I swapped in a more efficient dictionary comparison algorithm (based on regexps). A few miscellaneous cleanups.

[bdmac]

Hi Peter,

Thanks for the PR! I'm a bit worried that some of the changes you made seem to have altered the calculation of the entropy of a given password. Can you explain if this impact is truly significant or not?

[peterkovacs]

Hi Brian --

I believe that the entropy changes are due to two factors:

1. I'm including the entire 10k common password list since the comparisons are processed more efficiently.
2. I'm looking for longest matches first, so when we look at something like password12345678 we're now finding password as the initial substring (for example, we might actually be finding something like password1).

I believe that accounts for the entropy changes in the tests.

[bdmac]

Thanks!