Core: Support case-insensitive field lookups in SchemaUpdate #14734
Conversation
Adds support for case-insensitive field name lookups when adding and moving columns in SchemaUpdate. This ensures that move operations work correctly in case-insensitive mode, fixing a bug where newly added fields could not be referenced with different casing.
|
@singhpk234: What do you think about the change? Shall we invite someone else to review this too? |
|
|
||
| // update tracking for moves | ||
| addedNameToId.put(fullName, newId); | ||
| addedNameToId.put(caseSensitivityAwareName(fullName), newId); |
There was a problem hiding this comment.
initially I was thinking that we should not store the name in a case insensitive way but only keep the lookup case insensitive (similar to how we do it when we lookup fields in Schema. However, Schema keeps two maps, one which is case sensitive and the other which is case insensitive. I was also thinking whether this distinction would make sense for SchemaUpdate but couldn't find a good enough argument to justify this, so I think the changes you have here make sense to me
|
Merged to main. |
|
Thanks for the reviews @pvary @singhpk234 @nastra! Thanks for merging @pvary! |
4 participants
Adds support for case-insensitive field name lookups when adding and moving columns in SchemaUpdate. This ensures that move operations work correctly in case-insensitive mode, fixing a bug where newly added fields could not be referenced with different casing.