H2 does not appear to support the index/FK syntax generated by hibernate for the MySQL5 dialects when using ;MODE=MYSQL and hibernate with a dialect of org.hibernate.dialect.MySQL5Dialect.
My goal here is to have one set of SQL scripts and use hibernate for the ORM parts. Everything works fine in MySQL 5.5 but when I try to use H2 for things like unit tests and starting up a demo version of my app I get hundreds of failures from hibernate generated alter table statements as shown below. Unfortunately, I have not been able to find a way to get hibernate to change the way the statements are generated but that might be an option as well. I tried using org.hibernate.dialect.H2Dialect but that produces more severe errors so I don't think that will work.
alter table SAM_PUBLISHEDSECUREDIP_T
add index FK1EDEA25B9482C945 (ASSESSMENTID),
add constraint FK1EDEA25B9482C945 foreign key (ASSESSMENTID)
references SAM_PUBLISHEDASSESSMENT_T (ID)
Results in an error like this in H2:
org.h2.jdbc.JdbcSQLException: Syntax error in SQL statement "ALTER
TABLE SAM_PUBLISHEDSECUREDIP_T ADD INDEX FK1EDEA25B9482C945
(ASSESSMENTID),[*] ADD CONSTRAINT FK1EDEA25B9482C945 FOREIGN KEY
(ASSESSMENTID) REFERENCES SAM_PUBLISHEDASSESSMENT_T (ID) "; SQL
statement: alter table SAM_PUBLISHEDSECUREDIP_T add index
FK1EDEA25B9482C945 (ASSESSMENTID), add constraint FK1EDEA25B9482C945
foreign key (ASSESSMENTID) references SAM_PUBLISHEDASSESSMENT_T (ID)
[42000-172]
NOTE: I am open to writing and providing a patch for H2 but I could use some tips on where to look in that codebase.
H2 is not 100% compatible with MySQL, even when using the MySQL mode. It seems some of your SQL statements are not supported by H2.
Creating an index is not done using alter table (does MySQL really use this strange syntax? Every DBMS I know uses CREATE INDEX to create an index).
You have to split this up in two statements:
CREATE INDEX fk_assessment_id_index
on SAM_PUBLISHEDSECUREDIP_T (ASSESSMENTID);
alter table SAM_PUBLISHEDSECUREDIP_T
add constraint FK1EDEA25B9482C945 foreign key (ASSESSMENTID)
references SAM_PUBLISHEDASSESSMENT_T (ID);
Those two statements should also work in MySQL if I'm not mistaken.
Related
I have an existing table in Sqlite. How do I add a unique constraint to it?
You can't add a constraint to existing table in SQLite,(In SQL there is a option for do that). You need to re-create the table with necessary contraints.
There is only a few option available for alter table command in sqlite. Please check the image:
Also check Sqlite org reference.
EDIT
However you can add unique index for your table to achieve the same effect.
So you can use the following query to achieve that:
CREATE UNIQUE INDEX your_unique_index ON your_table(column_name);
In most cases, UNIQUE and PRIMARY KEY constraints are implemented by
creating a unique index in the database.
Reference : SQLite Constraints
You can add constraint by creating unique index:
CREATE UNIQUE INDEX ux_friend_name ON friend(name);
where ux_friend_name is unique index name, friend(name) - table and its columns that will be affected by this constraint.
You can read more here.
I am assuming you are using SQLiteManager from FireFox, please create the table again, normally it doesn't allow to change the constraints when you have already created a table.
NOTE - It can be done using code too, this is another way of doing it..
Edited
See the image below
When I have SQL Server Management Studio generate a table creation script for me, the foreign key constraints are a bit different than how I would write them.
Here is one:
ALTER TABLE [dbo].[GeoBytesCountries]
WITH CHECK
ADD CONSTRAINT [FK_GeoBytesCountries_MapReferenceId]
FOREIGN KEY ([MapReferenceId])
REFERENCES [dbo].[GeoBytesMapReferences] ([MapReferenceId])
GO
ALTER TABLE [dbo].[GeoBytesCountries]
CHECK CONSTRAINT [FK_GeoBytesCountries_MapReferenceId]
GO
I would write this foreign key constraint without "WITH CHECK" and the 2nd "CHECK CONSTRAINT" statement and expect to get the same functionality.
Can someone explain to me the value of the using "WITH CHECK" and a separate "CHECK CONSTRAINT" statement when you are writing a foreign key constraint for a table?
Or is the code below completely / functionally equivalent to the code above?
ALTER TABLE [dbo].[GeoBytesCountries]
ADD CONSTRAINT [FK_GeoBytesCountries_MapReferenceId]
FOREIGN KEY ([MapReferenceId])
REFERENCES [dbo].[GeoBytesMapReferences] ([MapReferenceId])
GO
The way I see it, the two step approach allows you to at least keep more "bad" data from getting in assuming the with check part fails. That is, your constraint will exist and apply to DML from that point forward, but you may have to do some cleanup on your existing data to make it a trusted constraint.
I have inherited an old database which wasn't designed very well. It is a Sql Server 2008 database which is missing quite a lot of Foreign Key relationships. Below shows two of the tables, and I am trying to manually create a FK relationship between dbo.app_status.status_id and dbo.app_additional_info.application_id
I am using SQL Server Management Studio when trying to create the relationship using the query below
USE myDatabase;
GO ALTER TABLE dbo.app_additional_info
ADD CONSTRAINT FK_AddInfo_AppStatus FOREIGN KEY (application_id)
REFERENCES dbo.app_status (status_id)
ON DELETE CASCADE
ON UPDATE CASCADE ;
GO
However, I receive this error when I run the query
The ALTER TABLE statement conflicted with the FOREIGN KEY constraint
"FK_AddInfo_AppStatus". The conflict occurred in database
"myDatabase", table "dbo.app_status", column 'status_id'.
I am wondering if the query is failing because each table already contains approximately 130,000 records?
Please help.
Thanks.
The error is occuring because there is a value in dbo.app_additional_info.application_ID that is not in dbo.app_Status.Status_ID. Unless the naming convention is seriously messed up you are trying to add a relationship to unrelated columns, why would application_ID reference status_ID?
I expect that dbo.App_Additional_Info.Application_ID should be referencing dbo.Application.Application_ID (Guessing at the table and column names slightly) so you would want this:
USE MyDatabase
GO
ALTER TABLE dbo.App_Additional_Info
ADD CONSTRAINT FK_App_Additional_Info_Application_ID (Application_ID)
REFERENCES dbo.Application (Application_ID)
ON DELETE CASCADE
ON UPDATE CASCADE;
I'm having a problem with an SQL statement.
I want to activate a "ON UPDATE CASCADE" behavior for a foreign key in a table with this statement :
ALTER TABLE "DB"."RECORD" ADD CONSTRAINT "RECORD_PT_OUTIL_FK1" FOREIGN KEY ("CDE_PO")
REFERENCES "DB"."PT_OUTIL" ("CDE_PO") ON UPDATE CASCADE ENABLE;
But when i run the statement in Oracle Developer, i just get this error message : "ORA-00905 : missing keyword"
I can't find what could be this missing keyword, i tried several changes but always the same error occurs.
I reuse a code generated by Oracle Developer it self and just modify it with what i want.
This is the generated code :
ALTER TABLE "DB"."RECORD" ADD CONSTRAINT "RECORD_PT_OUTIL_FK1" FOREIGN KEY ("CDE_PO")
REFERENCES "DB"."PT_OUTIL" ("CDE_PO") ON DELETE CASCADE DISABLE;
See, i just change the end of it.
So what's the matter here ? Am i missing something ? (please don't bash if it's something obvious :) )
Thx !
Oracle does not support the ON UPDATE clause for foreign keys.
See the description of the REFERENCES clause in the manual:
http://download.oracle.com/docs/cd/B19306_01/server.102/b14200/clauses002.htm#CJAIHHGC
Try by removing the DISABLE/ ENABLE at the end of your command
As per the ADD CONSTRAINT reference, there doesnt seem to be any "Enable / Disable" as part of the command.
I think it is something that your Oracle Developer is adding at the end (it being part of Oracle Syntax) and it might be causing the problem!!
How do i delete a foreign key constraint programmatically in Microsoft Access, preferable using SQL. For starters i don't know how to find the name of the foreign key.
I connect to Access from a Java application using the JDBC-ODBC bridge. I want to execute the SQL from my Java application.
I can see the relationship in Access, in the RelationShip view, but there seems to be no way of finding out the name. If i could find out the name i expect i could drop it with an ALTER TABLE statement.
Determine the relationship using
SELECT szRelationship FROM
Msysrelationships WHERE szObject =
'childtablename' and
szReferencedObject = 'parenttablename'
THEN
Use the ALTER TABLE command. Something along the line of this
ALTER TABLE Table2 DROP CONSTRAINT Relation1
I've tried accessing the foreign key name via JDBC's DataBaseMetadata object, but the JDBC-ODBC bridge does not implement the required functions. So i've resorted to droping and recreating the entire table with the foreign key.