[
https://issues.apache.org/jira/browse/HBASE-18617?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16130241#comment-16130241
]
Chia-Ping Tsai commented on HBASE-18617:
----------------------------------------
{quote}
+ Pair<byte[], byte[]> p = new Pair<>();
+ // create a copy of pair bytes so that they are not modified by the
filter
+ p.setFirst(Arrays.copyOf(aFuzzyKeysData.getFirst(),
aFuzzyKeysData.getFirst().length));
+ p.setSecond(Arrays.copyOf(aFuzzyKeysData.getSecond(),
aFuzzyKeysData.getSecond().length));
+
if (p.getFirst().length != p.getSecond().length) {
Pair<String, String> readable =
- new Pair<>(Bytes.toStringBinary(p.getFirst()),
Bytes.toStringBinary(p
{quote}
Consider coping the pair after the length check. If the check fails, it is a
waste of time to copy input data.
HBASE-17890 covers this issue but it has not be reviewed. It clone the input
data in preprocessSearchKey/preprocessMask.
> FuzzyRowKeyFilter should not modify the filter pairs
> ----------------------------------------------------
>
> Key: HBASE-18617
> URL: https://issues.apache.org/jira/browse/HBASE-18617
> Project: HBase
> Issue Type: Bug
> Affects Versions: 1.1.2
> Reporter: vinisha
> Priority: Minor
> Attachments: FuzzyRowFilter_HBASE_18617.patch
>
>
> Current implementation of FuzzyRowKeyFilter modifies the provided filter
> pairs:
> https://github.com/apache/hbase/blob/master/hbase-client/src/main/java/org/apache/hadoop/hbase/filter/FuzzyRowFilter.java#L89-L90
> Filter should not change the filter pair.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
