HBASE-16381 Shell deleteall command should support row key prefixes (Yi Liang)

hbase-shell/src/main/ruby/hbase/table.rb

@@ -160,21 +160,9 @@ EOF
     end
 
     #----------------------------------------------------------------------------------------------
-    # Delete a cell
+    # Create a Delete mutation
-    def _delete_internal(row, column,
+    def _createdelete_internal(row, column = nil,
         timestamp = org.apache.hadoop.hbase.HConstants::LATEST_TIMESTAMP, args = {})
-      _deleteall_internal(row, column, timestamp, args)
-    end
-
-    #----------------------------------------------------------------------------------------------
-    # Delete a row
-    def _deleteall_internal(row, column = nil,
-    		timestamp = org.apache.hadoop.hbase.HConstants::LATEST_TIMESTAMP, args = {})
-      # delete operation doesn't need read permission. Retaining the read check for
-      # meta table as a part of HBASE-5837.
-      if is_meta_table?
-        raise ArgumentError, "Row Not Found" if _get_internal(row).nil?
-      end
       temptimestamp = timestamp
       if temptimestamp.kind_of?(Hash)
         timestamp = org.apache.hadoop.hbase.HConstants::LATEST_TIMESTAMP
@@ -195,8 +183,61 @@ EOF
         family, qualifier = parse_column_name(column)
         d.addColumns(family, qualifier, timestamp)
       end
+      return d
+    end
+
+    #----------------------------------------------------------------------------------------------
+    # Delete rows using prefix
+    def _deleterows_internal(row, column = nil,
+        timestamp = org.apache.hadoop.hbase.HConstants::LATEST_TIMESTAMP, args={})
+      cache = row["CACHE"] ? row["CACHE"] : 100
+      prefix = row["ROWPREFIXFILTER"]
+
+      # create scan to get table names using prefix
+      scan = org.apache.hadoop.hbase.client.Scan.new
+      scan.setRowPrefixFilter(prefix.to_java_bytes)
+      # Run the scanner to get all rowkeys
+      scanner = @table.getScanner(scan)
+      # Create a list to store all deletes
+      list = java.util.ArrayList.new
+      # Iterate results
+      iter = scanner.iterator
+      while iter.hasNext
+        row = iter.next
+        key = org.apache.hadoop.hbase.util.Bytes::toStringBinary(row.getRow)
+        d = _createdelete_internal(key, column, timestamp, args)
+        list.add(d)
+        if list.size >= cache
+          @table.delete(list)
+          list.clear
+        end
+      end
+      @table.delete(list)
+    end
+
+    #----------------------------------------------------------------------------------------------
+    # Delete a cell
+    def _delete_internal(row, column,
+    			timestamp = org.apache.hadoop.hbase.HConstants::LATEST_TIMESTAMP, args = {})
+      _deleteall_internal(row, column, timestamp, args)
+    end
+
+    #----------------------------------------------------------------------------------------------
+    # Delete a row
+    def _deleteall_internal(row, column = nil,
+    		timestamp = org.apache.hadoop.hbase.HConstants::LATEST_TIMESTAMP, args = {})
+      # delete operation doesn't need read permission. Retaining the read check for
+      # meta table as a part of HBASE-5837.
+      if is_meta_table?
+        raise ArgumentError, "Row Not Found" if _get_internal(row).nil?
+      end
+      if row.kind_of?(Hash)
+        _deleterows_internal(row, column, timestamp, args)
+      else
+        d = _createdelete_internal(row, column, timestamp, args)
         @table.delete(d)
       end
+    end
 
     #----------------------------------------------------------------------------------------------
     # Increment a counter atomically

hbase-shell/src/main/ruby/shell/commands/deleteall.rb

@@ -23,7 +23,8 @@ module Shell
       def help
         return <<-EOF
 Delete all cells in a given row; pass a table name, row, and optionally
-a column and timestamp. Examples:
+a column and timestamp. Deleteall also support deleting a row range using a
+row key prefix. Examples:
 
   hbase> deleteall 'ns1:t1', 'r1'
   hbase> deleteall 't1', 'r1'
@@ -31,13 +32,21 @@ a column and timestamp. Examples:
   hbase> deleteall 't1', 'r1', 'c1', ts1
   hbase> deleteall 't1', 'r1', 'c1', ts1, {VISIBILITY=>'PRIVATE|SECRET'}
 
+ROWPREFIXFILTER can be used to delete row ranges
+  hbase> deleteall 't1', {ROWPREFIXFILTER => 'prefix'}
+  hbase> deleteall 't1', {ROWPREFIXFILTER => 'prefix'}, 'c1'        //delete certain column family in the row ranges
+  hbase> deleteall 't1', {ROWPREFIXFILTER => 'prefix'}, 'c1', ts1
+  hbase> deleteall 't1', {ROWPREFIXFILTER => 'prefix'}, 'c1', ts1, {VISIBILITY=>'PRIVATE|SECRET'}
+
+CACHE can be used to specify how many deletes batched to be sent to server at one time, default is 100
+  hbase> deleteall 't1', {ROWPREFIXFILTER => 'prefix', CACHE => 100}
+
+
 The same commands also can be run on a table reference. Suppose you had a reference
 t to table 't1', the corresponding command would be:
 
-  hbase> t.deleteall 'r1'
-  hbase> t.deleteall 'r1', 'c1'
-  hbase> t.deleteall 'r1', 'c1', ts1
   hbase> t.deleteall 'r1', 'c1', ts1, {VISIBILITY=>'PRIVATE|SECRET'}
+  hbase> t.deleteall {ROWPREFIXFILTER => 'prefix', CACHE => 100}, 'c1', ts1, {VISIBILITY=>'PRIVATE|SECRET'}
 EOF
       end
 

hbase-shell/src/test/ruby/hbase/table_test.rb

@@ -120,6 +120,10 @@ module Hbase
       
       @test_table.put(105, "x:a", "3")
       @test_table.put(105, "x:a", "4")
+
+      @test_table.put("111", "x:a", "5")
+      @test_table.put("111", "x:b", "6")
+      @test_table.put("112", "x:a", "5")
     end
 
     def teardown
@@ -181,6 +185,14 @@ module Hbase
       assert_nil(res)
     end
 
+    define_test "deletall should work with row prefix" do
+      @test_table.deleteall({ROWPREFIXFILTER => '11'})
+      res1 = @test_table._get_internal('111')
+      assert_nil(res1)
+      res2 = @test_table._get_internal('112')
+      assert_nil(res2)
+    end
+
     define_test "append should work with value" do
       @test_table.append("123", 'x:cnt2', '123')
     end