Christian Tu created OPENJPA-2848:
-------------------------------------
Summary: Duplicate DROP SEQUENCE statements generated by
MappingTool/SchemaTool
Key: OPENJPA-2848
URL: https://issues.apache.org/jira/browse/OPENJPA-2848
Project: OpenJPA
Issue Type: Bug
Components: jdbc
Affects Versions: 3.1.2
Reporter: Christian Tu
Attachments: openjpa-mappingtool-test.zip
Using the MappingTool with the action dropSchema to generate DDL statements
from Java classes, results in duplicate DROP SEQUENCE statements, if a SEQUENCE
of the same name already exists in the database.
I have added a small Maven example that demonstrates this. It uses the
openjpa-maven-plugin with the sql goal and build/dropSchema action on a Friend
class. The sql generated from the build action is used to initialize an h2 db
with a sequence. The subsequent dropSchema action then creates a file
target/drop.sql where the duplicate DROP SEQUENCE statement can be observed.
Using Version 2.4.3 of openjpa-maven-plugin results in a single DROP SEQUENCE
statement.
Not initializing the database with the sequence first, also results in a single
DROP SEQUENCE statement.
I'm not familiar with OpenJPA an its code base. But it seems reasonable to me
that the offending code is in SchemaTool, where this if should be an else if to
prevent dropSequence from being called twice:
{code:java}
diff --git
a/openjpa-jdbc/src/main/java/org/apache/openjpa/jdbc/schema/SchemaTool.java
b/openjpa-jdbc/src/main/java/org/apache/openjpa/jdbc/schema/SchemaTool.java
index 1bdbf6416..32a803ce5 100644
--- a/openjpa-jdbc/src/main/java/org/apache/openjpa/jdbc/schema/SchemaTool.java
+++ b/openjpa-jdbc/src/main/java/org/apache/openjpa/jdbc/schema/SchemaTool.java
@@ -918,9 +918,7 @@ public class SchemaTool {
dbSeq.getSchema().removeSequence(dbSeq);
else
_log.warn(_loc.get("drop-seq", seqs[j]));
- }
-
- if (_writer != null) {
+ } else if (_writer != null) {
dropSequence(seqs[j]);
}
}
{code}
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
