[
https://issues.apache.org/jira/browse/PHOENIX-1211?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14122226#comment-14122226
]
Kyle Buzsaki commented on PHOENIX-1211:
---------------------------------------
what about the case here:
{code}
// Increment key and backtrack until in range. We know at this
point that we'll be
// issuing a seek next hint.
seek = true;
Arrays.fill(position, i, position.length, 0);
int j = i - 1; // should this be "i - (slotSpan[i] + 1)"
instead? <-----------------------
// If we're positioned at a single key, no need to copy the
current key and get the next key .
// Instead, just increment to the next key and continue.
boolean incremented = false;
while (j >= 0 && slots.get(j).get(position[j]).isSingleKey() &&
(incremented=true) && (position[j] = (position[j] + 1) % slots.get(j).size())
== 0) {
j -= slotSpan[i] + 1;
incremented = false;
}
{code}
or here:
{code}
if (incremented) {
// Continue the loop after setting the start key, because
our start key maybe smaller than
// the current key, so we'll end up incrementing the start
key until it's bigger than the
// current key.
setStartKey();
schema.reposition(ptr, i, j, minOffset, maxOffset);
} else {
int currentLength = setStartKey(ptr, minOffset, j+1); //
should this be j + slotSpan[j] + 1 <-----------------------
// From here on, we use startKey as our buffer (resetting
minOffset and maxOffset)
// We've copied the part of the current key above that we
need into startKey
// Reinitialize the iterator to be positioned at previous
slot position
minOffset = 0;
maxOffset = startKeyLength;
schema.iterator(startKey, minOffset, maxOffset, ptr, j+1);
// should this be j + slotSpan[j] + 1 <-----------------------
// Do nextKey after setting the accessor b/c otherwise the
null byte may have
// been incremented causing us not to find it
ByteUtil.nextKey(startKey, currentLength);
}
{code}
> Use skip scan when row value constructor uses leading row key columns
> ----------------------------------------------------------------------
>
> Key: PHOENIX-1211
> URL: https://issues.apache.org/jira/browse/PHOENIX-1211
> Project: Phoenix
> Issue Type: Bug
> Reporter: James Taylor
> Assignee: Kyle Buzsaki
> Attachments: PHOENIX-1211-diff.patch
>
>
> We currently only use a skip scan for a row value constructor equality or in
> list expression is fully qualified. We can fairly easily use it when only
> some of the leading pk columns are used.
> For example:
> {code}
> WHERE (a,b) IN ((1,2),(3,4))
> {code}
> If the PK is (a,b,c), we can still use a skip scan. We need to pass through
> the slotSpan array in this case too, to the SkipScanFilter.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
