big-data/hadoop
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

HADOOP-10555. Add offset support to MurmurHash. Contributed by Sergey Shelukhin.

Browse Source
This commit is contained in:
Haohui Mai 2015-11-22 18:00:29 -08:00
parent 077fa529e1
commit aac260faa1
2 changed files with 10 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
hadoop-common-project/hadoop-common/CHANGES.txt
View File

@ -975,6 +975,9 @@ Release 2.8.0 - UNRELEASED
HADOOP-10035. Cleanup TestFilterFileSystem. (Suresh Srinivas via wheat9)
HADOOP-10555. Add offset support to MurmurHash.
(Sergey Shelukhin via wheat9)
OPTIMIZATIONS
HADOOP-11785. Reduce the number of listStatus operation in distcp

9
hadoop-common-project/hadoop-common/src/main/java/org/apache/hadoop/util/hash/MurmurHash.java
View File

@ -36,9 +36,13 @@ public class MurmurHash extends Hash {
public static Hash getInstance() {
return _instance;
}
@Override
public int hash(byte[] data, int length, int seed) {
return hash(data, 0, length, seed);
}
public int hash(byte[] data, int offset, int length, int seed) {
int m = 0x5bd1e995;
int r = 24;
@ -47,7 +51,7 @@ public int hash(byte[] data, int length, int seed) {
int len_4 = length >> 2;
for (int i = 0; i < len_4; i++) {
int i_4 = i << 2;
int i_4 = offset + (i << 2);
int k = data[i_4 + 3];
k = k << 8;
k = k | (data[i_4 + 2] & 0xff);
@ -67,6 +71,7 @@ public int hash(byte[] data, int length, int seed) {
int left = length - len_m;
if (left != 0) {
length += offset;
if (left >= 3) {
h ^= (int) data[length - 3] << 16;
}