[
https://issues.apache.org/jira/browse/PHOENIX-852?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14110890#comment-14110890
]
James Taylor commented on PHOENIX-852:
--------------------------------------
And instead of the code below, it'd be better if you called into WhereOptimizer
to get this information. There are cases beyond a simple column reference that
we'd be able to optimize (like a SUBSTR(rowKeyCol, 1, 3), or even a row value
constructor construct):
{code}
+ private Pair<Expression, Expression>
extractKeyRangeExpressions(StatementContext context, PTable table, JoinType
type, final List<Expression> joinExpressions, final List<Expression>
hashExpressions) {
+ if (type != JoinType.Inner)
+ return new Pair<Expression, Expression>(null, null);
+
+ List<Integer> rowkeyColumnIndexes = Lists.newArrayList();
+ for (int i = 0; i < joinExpressions.size(); i++) {
+ Expression joinExpression = joinExpressions.get(i);
+ if (joinExpression instanceof RowKeyColumnExpression) {
+ rowkeyColumnIndexes.add(i);
+ }
+ }
+ Collections.sort(rowkeyColumnIndexes, new Comparator<Integer>() {
+ @Override
+ public int compare(Integer l, Integer r) {
+ return ((RowKeyColumnExpression)
joinExpressions.get(l)).getPosition() - ((RowKeyColumnExpression)
joinExpressions.get(r)).getPosition();
+ }
+ });
+ int positionOffset = (table.getBucketNum() ==null ? 0 : 1) +
(context.getConnection().getTenantId() != null && table.isMultiTenant() ? 1 :
0) + (table.getViewIndexId() == null ? 0 : 1);
+ int position = 0;
+ for (Integer index : rowkeyColumnIndexes) {
+ RowKeyColumnExpression exp = (RowKeyColumnExpression)
joinExpressions.get(index);
+ if (exp.getPosition() != position + positionOffset) {
+ break;
+ }
+ position++;
+ }
+
+ if (position == 0)
+ return new Pair<Expression, Expression>(null, null);
+
+ if (position == 1)
+ return new Pair<Expression,
Expression>(joinExpressions.get(rowkeyColumnIndexes.get(0)),
hashExpressions.get(rowkeyColumnIndexes.get(0)));
+
+ List<Expression> lChildren = Lists.newArrayList();
+ List<Expression> rChildren = Lists.newArrayList();
+ for (int i = 0; i < position; i++) {
+ Integer index = rowkeyColumnIndexes.get(i);
+ lChildren.add(joinExpressions.get(index));
+ rChildren.add(hashExpressions.get(index));
+ }
+
+ return new Pair<Expression, Expression>(new
RowValueConstructorExpression(lChildren, false), new
RowValueConstructorExpression(rChildren, false));
+ }
+
{code}
> Optimize child/parent foreign key joins
> ---------------------------------------
>
> Key: PHOENIX-852
> URL: https://issues.apache.org/jira/browse/PHOENIX-852
> Project: Phoenix
> Issue Type: Improvement
> Reporter: James Taylor
> Assignee: Maryann Xue
> Attachments: 852-2.patch, 852.patch, PHOENIX-852.patch
>
>
> Often times a join will occur from a child to a parent. Our current algorithm
> would do a full scan of one side or the other. We can do much better than
> that if the HashCache contains the PK (or even part of the PK) from the table
> being joined to. In these cases, we should drive the second scan through a
> skip scan on the server side.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
