rule for common typos

[JuanPelota]

Hello

I am discovering rules and I'd like to know if there is a good existing rule for common typos according to a keyboard layout.

For example inversion of e and r due to proximity on the keyboard

Thank you

[atom]

No, but it's a nice idea. That would be many rules, not just one. The 's' rule should be able to do it. Because of the number rules allowed in a ruleser per line you need to use one rule per line. The disadvantage is that passwords with multiple typos in the same password would not be found.

[JuanPelota]

Ok thanks

[royce]

I did a rough pass a while back, but it has exactly the limitations that atom described.

https://gist.github.com/roycewilliams/9d...f0e8de8371

[Neorusi]

No, but it's a nice idea. That would be many rules, not just one. The 's' rule should be able to do it. Because of the number rules allowed in a ruleser per line you need to use one rule per line. The disadvantage is that passwords with multiple typos in the same password would not be found.

Hi, I am solving a similar problem right now, what about using multiple rules, like the way I am doing 
for example if the approximate  password I have is Bon-Aqua 

I would delete a letter with D0 and in another rule file in the same line I would insert an adjacent key with i0C
this way would Bon-Aqua -> Con-Aqua

for more than 1 typo I would duplicate the rules so the would be applied together:
 
1st line delete_1st_letter.rule D0
1st line insert_1st_letter.rule i0X
2nd line delete_2st_letter.rule D1
2nd line insert_2st_letter.rule i01

The disadvantage I see in this way is that you should exactly now the length of the approximate password
Am I right, or should I try another way ?