Mailing-List: contact dev-help@openjpa.apache.org; run by ezmlm
Precedence: bulk
Reply-To: dev@openjpa.apache.org
Message-ID: <21432738.1183996806181.JavaMail.jira@brutus>
Date: Mon, 9 Jul 2007 09:00:06 -0700 (PDT)
From: "Daniel Migowski (JIRA)" <jira@apache.org>
To: dev@openjpa.apache.org
Subject: [jira] Created: (OPENJPA-282) Postgresql does not support unique
 constraints
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit

Postgresql does not support unique constraints
----------------------------------------------

                 Key: OPENJPA-282
                 URL: https://issues.apache.org/jira/browse/OPENJPA-282
             Project: OpenJPA
          Issue Type: Bug
          Components: jdbc
         Environment: PostgreSQL
            Reporter: Daniel Migowski
            Priority: Minor


PostgreSQL does not support deferred unique constraints. Since the dictionary states deferred constraints capabilities, this results in an error on automatic database generation for unique fields.

Please use the attached patch to fix this. (Or try something else... ;). Maybe one could better define a variable to describe this capability.

With best regards,
Daniel Migowski

--------------------- PATCH BEGINS BELOW THIS LINE --------------------------
Index: C:/IKOfficeRoot/Projekte/OpenJPA/openjpa/openjpa-jdbc/src/main/java/org/apache/openjpa/jdbc/sql/PostgresDictionary.java
===================================================================
--- C:/IKOfficeRoot/Projekte/OpenJPA/openjpa/openjpa-jdbc/src/main/java/org/apache/openjpa/jdbc/sql/PostgresDictionary.java	(revision 554663)
+++ C:/IKOfficeRoot/Projekte/OpenJPA/openjpa/openjpa-jdbc/src/main/java/org/apache/openjpa/jdbc/sql/PostgresDictionary.java	(working copy)
@@ -32,6 +32,7 @@
 import org.apache.openjpa.jdbc.schema.Column;
 import org.apache.openjpa.jdbc.schema.Sequence;
 import org.apache.openjpa.jdbc.schema.Table;
+import org.apache.openjpa.jdbc.schema.Unique;
 import org.apache.openjpa.lib.jdbc.DelegatingConnection;
 import org.apache.openjpa.lib.jdbc.DelegatingPreparedStatement;
 import org.apache.openjpa.lib.util.Localizer;
@@ -36,6 +37,8 @@
 import org.apache.openjpa.lib.jdbc.DelegatingPreparedStatement;
 import org.apache.openjpa.lib.util.Localizer;
 
+import serp.util.Strings;
+
 /**
  * Dictionary for Postgres.
  */
@@ -93,10 +96,10 @@
         // PostgreSQL requires double-escape for strings
         searchStringEscape = "\\\\";
 
-        maxTableNameLength = 31;
-        maxColumnNameLength = 31;
-        maxIndexNameLength = 31;
-        maxConstraintNameLength = 31;
+        maxTableNameLength = 63;
+        maxColumnNameLength = 63;
+        maxIndexNameLength = 63;
+        maxConstraintNameLength = 63;
         schemaCase = SCHEMA_CASE_LOWER;
         rangePosition = RANGE_POST_LOCK;
         requiresAliasForSubselect = true;
@@ -301,6 +304,31 @@
         throws SQLException {
         return new PostgresConnection(super.decorate(conn), this);
     }
+    
+	/**
+     * Return the declaration SQL for the given unique constraint. This
+     * method is used from within {@link #getCreateTableSQL}.
+     * Returns	<code>CONSTRAINT &lt;name&gt; UNIQUE (&lt;col list&gt;)</code>
+     * by default. Only foreign key constraints can be deferred in PostgreSQL
+     * so we have to override the function in DbDictionary.
+     */
+    protected String getUniqueConstraintSQL(Unique unq) {
+        if (unq.isDeferred()) return null;
+
+        StringBuffer buf = new StringBuffer();
+        if (unq.getName() != null
+            && CONS_NAME_BEFORE.equals(constraintNameMode))
+            buf.append("CONSTRAINT ").append(unq.getName()).append(" ");
+        buf.append("UNIQUE ");
+        if (unq.getName() != null && CONS_NAME_MID.equals(constraintNameMode))
+            buf.append(unq.getName()).append(" ");
+        buf.append("(").append(Strings.join(unq.getColumns(), ", ")).
+            append(")");
+        if (unq.getName() != null
+            && CONS_NAME_AFTER.equals(constraintNameMode))
+            buf.append(" CONSTRAINT ").append(unq.getName());
+        return buf.toString();
+    }    
 
     /**
      * Connection wrapper to work around the postgres empty result set bug.


-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.