[
https://issues.apache.org/jira/browse/SPARK-18123?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15613088#comment-15613088
]
Dongjoon Hyun edited comment on SPARK-18123 at 10/27/16 8:20 PM:
-----------------------------------------------------------------
Hi, [~zwu....@gmail.com].
There are two related things about this. Actually, current implementation
supports the followings.
- Allow reserved keyword as a field name, e.g., 'order'.
- Allow mixed-case data frame like the following, e.g., `[a: int, A: int]`.
{code}
scala> val df = sql("select 1 a, 1 A")
df: org.apache.spark.sql.DataFrame = [a: int, A: int]
scala> val option = Map("url" -> "jdbc:postgresql:postgres", "dbtable" ->
"mixed", "user" -> "postgres", "password" -> "test")
scala> df.write.mode("overwrite").format("jdbc").options(option).save()
scala> df.write.mode("append").format("jdbc").options(option).save()
{code}
Your case should be supported without any regression the above cases. Your case
might be the following.
{code}
val df1 = sql("select 1 a")
val df2 = sql("select 1 A")
val option = Map("url" -> "jdbc:postgresql:postgres", "dbtable" -> "tx", "user"
-> "postgres", "password" -> "test")
df1.write.mode("overwrite").format("jdbc").options(option).save()
df2.write.mode("append").format("jdbc").options(option).save()
{code}
I'm testing more cases to solve all three cases including yours.
was (Author: dongjoon):
Hi, [~zwu....@gmail.com].
There is two related things about this. Actually, current implementation
supports the followings.
- Allow reserved keyword as a field name, e.g., 'order'.
- Allow mixed-case data frame like the following.
{code}
scala> val df = sql("select 1 a, 1 A")
df: org.apache.spark.sql.DataFrame = [a: int, A: int]
scala> val option = Map("url" -> "jdbc:postgresql:postgres", "dbtable" ->
"mixed", "user" -> "postgres", "password" -> "test")
scala> df.write.mode("overwrite").format("jdbc").options(option).save()
scala> df.write.mode("append").format("jdbc").options(option).save()
{code}
Your case should be supported without any regression the above cases. Your case
might be the following.
{code}
val df1 = sql("select 1 a")
val df2 = sql("select 1 A")
val option = Map("url" -> "jdbc:postgresql:postgres", "dbtable" -> "tx", "user"
-> "postgres", "password" -> "test")
df1.write.mode("overwrite").format("jdbc").options(option).save()
df2.write.mode("append").format("jdbc").options(option).save()
{code}
I'm testing more cases to solve all three cases including yours.
> org.apache.spark.sql.execution.datasources.jdbc.JdbcUtils.saveTable the case
> senstivity issue
> ----------------------------------------------------------------------------------------------
>
> Key: SPARK-18123
> URL: https://issues.apache.org/jira/browse/SPARK-18123
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Affects Versions: 2.0.1
> Reporter: Paul Wu
>
> Blindly quoting every field name for inserting is the issue (Line 110-119,
> https://github.com/apache/spark/blob/master/sql/core/src/main/scala/org/apache/spark/sql/execution/datasources/jdbc/JdbcUtils.scala).
> /**
> * Returns a PreparedStatement that inserts a row into table via conn.
> */
> def insertStatement(conn: Connection, table: String, rddSchema: StructType,
> dialect: JdbcDialect)
> : PreparedStatement = {
> val columns = rddSchema.fields.map(x =>
> dialect.quoteIdentifier(x.name)).mkString(",")
> val placeholders = rddSchema.fields.map(_ => "?").mkString(",")
> val sql = s"INSERT INTO $table ($columns) VALUES ($placeholders)"
> conn.prepareStatement(sql)
> }
> This code causes the following issue (it does not happen to 1.6.x):
> I have issue with the saveTable method in Spark 2.0/2.0.1. I tried to save a
> dataset to Oracle database, but the fields must be uppercase to succeed. This
> is not an expected behavior: If only the table names were quoted, this
> utility should concern the case sensitivity. The code below throws the
> exception: Caused by: java.sql.SQLSyntaxErrorException: ORA-00904:
> "DATETIME_gmt": invalid identifier.
> String detailSQL ="select CAST('2016-09-25 17:00:00' AS TIMESTAMP)
> DATETIME_gmt, '1' NODEB";
> hc.sql("set spark.sql.caseSensitive=false");
> Dataset<Row> ds = hc.sql(detailSQL);
> ds.show();
>
> org.apache.spark.sql.execution.datasources.jdbc.JdbcUtils.saveTable(ds, url,
> detailTable, p);
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
