Am 19.06.2014 um 21:36 schrieb Esteban A. Maringolo <[email protected]>:
> Cool! > > It is a nice companion to my 8 byte "pretty much unique" generator > [1], and could be integrated in my friendly URLs :). > cool > However, I couldn't prove the case insensitivity of Base36, the > #toNumber: ignores characters not present in the Characters array. I > fixed Base36 toNumber: but broke Base62 tests :) > > self assert: (Base36 toNumber: (Base36 fromNumber: 16) asUppercase) > Good catch. It is fixed now. I've uploaded a new version Norbert > will fail. > > Regards, > > -- > Esteban. > > [1] https://gist.github.com/eMaringolo/be973dcf03b0783711f1 > > > > > > > Esteban A. Maringolo > > > 2014-06-12 8:16 GMT-03:00 Norbert Hartl <[email protected]>: >> Hmm *cough* >> >> http://smalltalkhub.com/#!/~NorbertHartl/Base62 >> >> >> Norbert >> >> Am 12.06.2014 um 13:12 schrieb Norbert Hartl <[email protected]>: >> >> Just for the record. I’ve uploaded a package to smalltalkhub that contains >> util classes to en-/decode values into/from Base62/Base36. >> >> Base62/Base36 encode numbers in short strings. These are used e.g. by url >> shortener services from google, bit.ly etc. >> >> Base36 uses lowercase letters and numbers to encode, Base62 uses all >> characters from Base36 plus all uppercase letters. Base62 produces smaller >> strings while Base36 produces case insensitive ones. >> >> Example: >> >> Base62 fromNumber: 100000 -> 'q0U‘ >> >> Base36 fromNumber: 100000 -> ‚255s' >> >> FYI, >> >> Norbert >> >> >> >
