idb-iegap
idb-iegap copied to clipboard
fixes compond string calculation
The current algorithm used in iegap will not generate successive strings when the compound key values have varying lengths. This results in a DataError exception when trying to handle bound keys of the following kind:
IDBKeyRange.bound(['someValue'], ['someValue', 'someOtherValue'], false, false)
because the compounded lower bound value is "larger" than the upper bound value
This change would be a breaking change since all the existing keys would not be valid anymore. The implementation is taken straight out of ydn-db
I'm not sure if this could be fixed using the current algorithm without also breaking backwards compatibility.