java - How do I create a row key prefix scan using asynchbase -

i'm trying row-key prefix scan using java asynchbase library. e.g. want find first 10 records row key has prefix "abc":

byte[] prefix = bytes.tobytes("abc"); scanner scanner = hbaseclient.newscanner(tblname); scanner.setfamily(bytes.tobytes("cf_test"));  scanner.setfilter(new rowfilter(comparefilter.compareop.equal, new binaryprefixcomparator(prefix)));  result = scanner.nextrows(10).join(); // ... 

i use rowfilter job, doesn't work. seems trigger full-table scan.

i don't find filter class rowprefixfilter, neither #setrowkeyprefix() method in scanner class.

how create scan match row key prefix?

update:
read code of hbase-client project , found 1 solution. in org.apache.hadoop.hbase.client.scan class, #setrowprefixfilter() method implemented below:

public scan setrowprefixfilter(byte[] rowprefix) {   if (rowprefix == null) {     setstartrow(hconstants.empty_start_row);     setstoprow(hconstants.empty_end_row);   } else {     this.setstartrow(rowprefix);     this.setstoprow(calculatetheclosestnextrowkeyforprefix(rowprefix));   }   return this; } 

it implemented via setting start row , stop row.
change code to:

byte[] prefix = bytes.tobytes("abc"); scanner scanner = hbaseclient.newscanner(tblname); scanner.setfamily(bytes.tobytes("cf_test"));  scanner.setstartkey(prefix); scanner.setstopkey(calculatetheclosestnextrowkeyforprefix(prefix));  result = scanner.nextrows(10).join(); 

the implementation of #calculatetheclosestnextrowkeyforprefix() same 1 in org.apache.hadoop.hbase.client.scan:

private byte[] calculatetheclosestnextrowkeyforprefix(byte[] rowkeyprefix) {     // treating 'unsigned very long' ,     // doing +1 manually.     // search place trailing 0xffs start     int offset = rowkeyprefix.length;     while (offset > 0) {         if (rowkeyprefix[offset - 1] != (byte) 0xff) {             break;         }         offset--;     }      if (offset == 0) {         // got 0xffff... (only ffs) stoprow value         // last possible prefix before end of table.         // set stop @ 'end of table'         return hconstants.empty_end_row;     }      // copy right length of original     byte[] newstoprow = arrays.copyofrange(rowkeyprefix, 0, offset);     // , increment last 1     newstoprow[newstoprow.length - 1]++;     return newstoprow; }