Github user gatorsmile commented on a diff in the pull request:
https://github.com/apache/spark/pull/12081#discussion_r59652599
--- Diff:
sql/catalyst/src/main/scala/org/apache/spark/sql/catalyst/catalog/SessionCatalog.scala
---
@@ -121,8 +123,13 @@ class SessionCatalog(
currentDb = db
}
- def getDefaultDBPath(db: String): String = {
- System.getProperty("java.io.tmpdir") + File.separator + db + ".db"
+ def getDatabasePath(dbName: String, path: Option[String]): String = {
+ val dbPath =
+ path.map(new Path(_))
+ .getOrElse(
+ new Path(new Path(System.getProperty("java.io.tmpdir")),
dbName.toLowerCase() + ".db"))
+ val dbLocation = dbPath.toString
+ if (dbLocation.endsWith(File.separator)) dbLocation.dropRight(1) else
dbLocation
--- End diff --
So far, the code is not used. Basically, this implementation is a dead
code, but it still has two issues.
First, on some platform, the temporary directory `java.io.tmpdir` does not
include a trailing slash. Thus, using `new Path` will handle all the issues we
might miss. It is consistent with what we did in `HiveSessionCatalog`.
Second, we need to call `toLowerCase()` when generating the path. This is
to make it consistent with what we did in `HiveSessionCatalog`.
So far, we are not creating any physical directory when we creating a
database in `SQLContext`. We just store the information in the memory.
Basically, this database is a logical concept, which corresponds to a set of
tables. We can set it to an empty string, it should be still OK. Let me know if
you want me to set it to an empty string. Thanks!
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
