Apache
/
hbase
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

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

This commit is contained in:
chenheng 2016-09-15 19:18:47 +08:00
parent 8ef6c76344
commit 1d6c90b496
3 changed files with 86 additions and 24 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

81
hbase-shell/src/main/ruby/hbase/table.rb
View File

@ -159,6 +159,62 @@ EOF
@table.put(p)
end
#----------------------------------------------------------------------------------------------
# Create a Delete mutation
def _createdelete_internal(row, column = nil,
timestamp = org.apache.hadoop.hbase.HConstants::LATEST_TIMESTAMP, args = {})
temptimestamp = timestamp
if temptimestamp.kind_of?(Hash)
timestamp = org.apache.hadoop.hbase.HConstants::LATEST_TIMESTAMP
end
d = org.apache.hadoop.hbase.client.Delete.new(row.to_s.to_java_bytes, timestamp)
if temptimestamp.kind_of?(Hash)
temptimestamp.each do |k, v|
if v.kind_of?(String)
set_cell_visibility(d, v) if v
end
end
end
if args.any?
visibility = args[VISIBILITY]
set_cell_visibility(d, visibility) if visibility
end
if column
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,
@ -175,27 +231,12 @@ EOF
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
if row.kind_of?(Hash)
_deleterows_internal(row, column, timestamp, args)
else
d = _createdelete_internal(row, column, timestamp, args)
@table.delete(d)
end
d = org.apache.hadoop.hbase.client.Delete.new(row.to_s.to_java_bytes, timestamp)
if temptimestamp.kind_of?(Hash)
temptimestamp.each do |k, v|
if v.kind_of?(String)
set_cell_visibility(d, v) if v
end
end
end
if args.any?
visibility = args[VISIBILITY]
set_cell_visibility(d, visibility) if visibility
end
if column
family, qualifier = parse_column_name(column)
d.addColumns(family, qualifier, timestamp)
end
@table.delete(d)
end
#----------------------------------------------------------------------------------------------

17
hbase-shell/src/main/ruby/shell/commands/deleteall.rb
View File

@ -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

12
hbase-shell/src/test/ruby/hbase/table_test.rb
View File

@ -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