[GitHub] [spark] amaliujia commented on a diff in pull request #38218: [SPARK-40587][CONNECT][FOLLOW-UP] Make sure python client support select *

[GitBox <gi...@apache.org>]

amaliujia commented on code in PR #38218:
URL: https://github.com/apache/spark/pull/38218#discussion_r995379653


##########
python/pyspark/sql/connect/plan.py:
##########
@@ -143,23 +143,29 @@ def __init__(self, child: Optional["LogicalPlan"], *columns: "ExpressionOrString
         self._verify_expressions()
 
     def _verify_expressions(self) -> None:
-        """Ensures that all input arguments are instances of Expression."""
+        """Ensures that all input arguments are instances of Expression or String."""
         for c in self._raw_columns:
-            if not isinstance(c, Expression):
-                raise InputValidationError(f"Only Expressions can be used for projections: '{c}'.")
+            if not isinstance(c, (Expression, str)):
+                raise InputValidationError(
+                    f"Only Expressions or String can be used for projections: '{c}'."
+                )
 
     def withAlias(self, alias: str) -> LogicalPlan:
         self.alias = alias
         return self
 
     def plan(self, session: Optional["RemoteSparkSession"]) -> proto.Relation:
         assert self._child is not None
-        proj_exprs = [
-            c.to_plan(session)
-            if isinstance(c, Expression)
-            else self.unresolved_attr(*(c.split(".")))
-            for c in self._raw_columns
-        ]
+        proj_exprs = []
+        for c in self._raw_columns:
+            if isinstance(c, Expression):
+                proj_exprs.append(c.to_plan(session))
+            elif c == "*":
+                exp = proto.Expression()
+                exp.unresolved_star.SetInParent()
+                proj_exprs.append(exp)
+            else:
+                proj_exprs.append(self.unresolved_attr(*(c.split("."))))

Review Comment:
   First of all this belongs to 
   
   In fact you asking reminds me a very tricky problem that I didn't think of: in the past PySpark does not care because it can use Py4j to reuse Scala implementation, however now the world has changed....
   
   For this specific problem, it is not about the separator. What we really need to do is to patch the scala implementation for `CatalystSqlParser.parseMultipartIdentifier(s)`. The identifier could contain multiple parts. 
   



-- 
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: reviews-unsubscribe@spark.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: reviews-unsubscribe@spark.apache.org
For additional commands, e-mail: reviews-help@spark.apache.org