From eb8bfa3eaa0846d685e4d12f9ee2e4273b85edcf Mon Sep 17 00:00:00 2001
From: Nathan Howell <nhowell@godaddy.com>
Date: Wed, 5 Aug 2015 22:16:56 +0800
Subject: [PATCH] [SPARK-9618] [SQL] Use the specified schema when reading
Parquet files
The user specified schema is currently ignored when loading Parquet files.
One workaround is to use the `format` and `load` methods instead of `parquet`, e.g.:
```
val schema = ???
// schema is ignored
sqlContext.read.schema(schema).parquet("hdfs:///test")
// schema is retained
sqlContext.read.schema(schema).format("parquet").load("hdfs:///test")
```
The fix is simple, but I wonder if the `parquet` method should instead be written in a similar fashion to `orc`:
```
def parquet(path: String): DataFrame = format("parquet").load(path)
```
Author: Nathan Howell <nhowell@godaddy.com>
Closes #7947 from NathanHowell/SPARK-9618 and squashes the following commits:
d1ea62c [Nathan Howell] [SPARK-9618] [SQL] Use the specified schema when reading Parquet files
---
.../src/main/scala/org/apache/spark/sql/DataFrameReader.scala | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/sql/core/src/main/scala/org/apache/spark/sql/DataFrameReader.scala b/sql/core/src/main/scala/org/apache/spark/sql/DataFrameReader.scala
index eb09807f9d..b90de8ef09 100644
--- a/sql/core/src/main/scala/org/apache/spark/sql/DataFrameReader.scala
+++ b/sql/core/src/main/scala/org/apache/spark/sql/DataFrameReader.scala
@@ -260,7 +260,7 @@ class DataFrameReader private[sql](sqlContext: SQLContext) extends Logging {
sqlContext.baseRelationToDataFrame(
new ParquetRelation(
- globbedPaths.map(_.toString), None, None, extraOptions.toMap)(sqlContext))
+ globbedPaths.map(_.toString), userSpecifiedSchema, None, extraOptions.toMap)(sqlContext))
}
}
--
GitLab