[
https://issues.apache.org/jira/browse/PHOENIX-1863?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14499133#comment-14499133
]
Dumindu Buddhika commented on PHOENIX-1863:
-------------------------------------------
[~jamestaylor]
bq. For example test the sort order for a CHAR(5) that's DESC and ASC wrt to
byte[] (you can use Bytes.BYTE_COMPARATOR) for the following kinds of values:
I do not clearly understand the above. Do you mean a following kind of a test?
:)
{code}
PDataType dataType = PChar.INSTANCE;
String str = "foo";
byte[] result = new byte[]{-103, -112, -112, -33, -33, -33, -33, -33};
LiteralExpression expr = LiteralExpression.newConstant(str, dataType, 5, null,
SortOrder.DESC, Determinism.ALWAYS);
ImmutableBytesPtr ptr = new ImmutableBytesPtr(expr.getBytes());
dataType.pad(ptr, 8, SortOrder.DESC);
String resultValue = (String) dataType.toObject(ptr, dataType, SortOrder.DESC);
assertTrue(Arrays.equals(result, ptr.get()));
assertEquals(str, resultValue);
{code}
I also found that, StringUtil.padChar also does not care about the SortOrder.
StringUtil.padChar is used in LiteralExpression.newConstant method. So the
above test case does not pass actually.
> Create unit test for inverting the pad character
> ------------------------------------------------
>
> Key: PHOENIX-1863
> URL: https://issues.apache.org/jira/browse/PHOENIX-1863
> Project: Phoenix
> Issue Type: Sub-task
> Reporter: James Taylor
> Assignee: Dumindu Buddhika
> Fix For: 5.0.0, 4.4.0
>
> Attachments: PHOENIX-1863.patch
>
>
> It'd be good to have unit tests that test the necessity of inverting the pad
> character (the parent JIRA).
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
