The shell could have and option to call the Range.prefix() function. If you want a really precise prefix range you could look at its source code.
On Thu, Oct 29, 2015 at 10:52 AM, Brian Loss <[email protected]> wrote: > The -b and -e options specify the row portion of the key, and they are > inclusive. However, 1_abc > 1_ (and the same is true for all of your other > 1_ example keys), so none are returned. If you have multiple entries for > the same row (any of [column family, column qualifier, column visibility, > timestamp] differ but the row is the same), then specifying the same value > for -b and -e will give you all of those entries. In your case, you need to > do what you said last, and pick a character lexicographically higher than > the one you want. something like -b 1_ -e 1~ would work. > > > On Oct 29, 2015, at 10:41 AM, z11373 <[email protected]> wrote: > > > > Hi, > > Let say I have table T1 with following keys: > > 1_abc > > 1_def > > 1_xyz > > 2_xxx > > 2_yyy > > 3_a > > 3_ab > > ... > > > > I want to scan and only returns data with prefix '1_', so I run "scan -t > T1 > > -b 1_ -e 1_", but it returns no data. However, running "scan -t T1 -b 1_ > -e > > 2" will return what I want. I thought that -e should include that value > > (inclusive)? > > > > How can I scan fixed range from shell like in this example above? Do I > just > > need to pick character which is lexicographically sorted higher than the > one > > I want? > > > > Thanks, > > Z > > > > > > > > > > -- > > View this message in context: > http://apache-accumulo.1065345.n5.nabble.com/scan-fixed-range-from-shell-tp15427.html > > Sent from the Users mailing list archive at Nabble.com. > >
