Reestablish warning on duplicate column in query. #1825
Conversation
| Object rsv = JdbcUtils.getResultSetValue(resultSet, i + 1); | ||
| String columnName = md.getColumnLabel(i + 1); | ||
| document.put(columnName, rsv instanceof Array a ? a.getArray() : rsv); | ||
| Object old = document.put(columnName, rsv instanceof Array a ? a.getArray() : rsv); |
There was a problem hiding this comment.
I'm wondering why we're not using putIfAbsent semantics. Obtaining a value from the ResultSet typically uses the first occurrence of a column while we should be agnostic to the fact whether a column occurs multiple times in the result.
For reference, none of Spring JDBC's mappers care about duplicate columns, they just accept the fact that ResultSet either returns different values when iterating over the column count or use the value by name lookup.
Furthermore, when an external query declares the same column names, why is it us to validate this concern?
There was a problem hiding this comment.
FWIW, the JDBC spec says when you have duplicate column names, the first one should be returned.
There was a problem hiding this comment.
why is it us to validate this concern?
Because we are nice people and we notice something is weird.
| @@ -66,9 +71,13 @@ static RowDocument toRowDocument(ResultSet resultSet) throws SQLException { | |||
| RowDocument document = new RowDocument(columnCount); | |||
|
|
|||
| for (int i = 0; i < columnCount; i++) { | |||
|
|
|||
| Object rsv = JdbcUtils.getResultSetValue(resultSet, i + 1); | |||
| String columnName = md.getColumnLabel(i + 1); | |||
There was a problem hiding this comment.
Nit: We should use JdbcUtils.lookupColumnName(…) here.
|
Made the changes requested by the review and merged. |
Closes #1680