github-actions[bot] commented on code in PR #65346:
URL: https://github.com/apache/doris/pull/65346#discussion_r3548660612
##########
fe/fe-core/src/main/java/org/apache/doris/nereids/trees/expressions/SlotReference.java:
##########
@@ -223,11 +223,7 @@ public String toString() {
@Override
public String shapeInfo() {
- if (qualifier.isEmpty()) {
- return name.get();
- } else {
- return qualifier.get(qualifier.size() - 1) + "." + name.get();
- }
+ return name.get();
Review Comment:
This leaves the checked shape-plan output stale. The live diff now only
changes `SlotReference.shapeInfo()` to return the bare `name`, and shape-plan
join text is built by `PhysicalHashJoin` / `PhysicalNestedLoopJoin` through
`Expression.shapeInfo()` and `BinaryOperator.shapeInfo()`. Existing regression
outputs still expect qualified slots, e.g. `test_hint.out` still has
`hashCondition=((t1.c1 = t2.c2))`, and the PR body's cited `multi_leading.out`
line still expects `avg(t1.c11)`. With this change those direct slot references
render as `c1`/`c2`, and aggregate/function arguments still go through
`toSql()`, so the expected files need to be regenerated or the implementation
needs to keep the intended qualified display behavior. Otherwise CI can keep
comparing against stale shape-plan text.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
