Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.

Bug 336414

Summary:

[DB] Make the used CDOIDObject subtype and storage pluggable

Product:

[Modeling] EMF

Reporter:

Eike Stepper <stepper>

Component:

cdo.db

Assignee:

Eike Stepper <stepper>

Status:

CLOSED FIXED

QA Contact:

Eike Stepper <stepper>

Severity:

enhancement

Priority:

CC:

stefan

Version:

4.0

Keywords:

noteworthy

Target Milestone:

---

Hardware:

All

OS:

All

Whiteboard:

Attachments:

Description	Flags
Patch v1	none
Patch v2	none
Patch v3 - comile error free	none
Patch v4 - most tests pass	none
Patch v5	none
Patch v6	none
Patch v7	none
Patch v8	none
Patch v9 - for future reference	none
Patch v10 - full patch with StringIDHandler	none
supplemental-patch-v1	none

Description Eike Stepper

2011-02-05 04:45:35 EST

The DBStore hardcodes long-valued CDOIDs everywhere. Make this configurable. This would also be needed for UUIDs.

Comment 1 Eike Stepper

2011-02-05 04:48:42 EST

Created attachment 188385 [details]
Patch v1

I renamed the typical helper methods that create, parse or deal otherwise with long value IDs. Then I fixed most of the resulting compile errors by delegating to the new IIDHandler which is maintained by IDBStore. Now I'm down to 18 errors.

I roughly identified some additional challenges:

ITypeMapping
ITypeMapping.Factory
initTable() methods

IMetaDataManager

ObjectTypeTable.idField

IExternalReferenceManager
CDODBSchema.EXTERNAL_ID!!!

FeatureMap tags???

repairAfterCrash()

Comment 2 Eike Stepper

2011-02-06 04:53:56 EST

Created attachment 188402 [details]
Patch v2

Comment 3 Eike Stepper

2011-02-06 06:04:44 EST

Created attachment 188403 [details]
Patch v3 - comile error free

All issues addressed, untested.

Comment 4 Eike Stepper

2011-02-06 07:44:37 EST

Created attachment 188405 [details]
Patch v4 - most tests pass

Comment 5 Eike Stepper

2011-02-06 13:25:56 EST

Created attachment 188409 [details]
Patch v5

Comment 6 Eike Stepper

2011-02-06 13:33:01 EST

Created attachment 188410 [details]
Patch v6

Comment 7 Eike Stepper

2011-02-06 15:48:57 EST

Created attachment 188412 [details]
Patch v7

Comment 8 Eike Stepper

2011-02-06 16:45:46 EST

Created attachment 188414 [details]
Patch v8

Comment 9 Eike Stepper

2011-02-07 02:18:40 EST

Created attachment 188420 [details]
Patch v9 - for future reference

Comment 10 Eike Stepper

2011-02-07 06:44:09 EST

Created attachment 188428 [details]
Patch v10 - full patch with StringIDHandler

feature maps fail with StringIDHandler

Comment 11 Eike Stepper

2011-02-07 06:48:53 EST

Committed revision 7028 (patch v10):

- trunk/org.eclipse.emf.cdo.server
- trunk/org.eclipse.emf.cdo.server.db
- trunk/org.eclipse.emf.cdo.server.hibernate
- trunk/org.eclipse.emf.cdo.server.objectivity

Comment 12 Stefan Winkler

2011-02-07 11:01:17 EST

Created attachment 188446 [details]
supplemental-patch-v1

Patch-v10 misses a handler-dependent DBField (which actually should have been BIGINT, not INTEGER, BTW). This patch fixes the field.
All non-Workspace tests pass now.

Comment 13 Eike Stepper

2011-02-08 08:13:17 EST

Committed revision 7029:
- trunk/plugins/org.eclipse.emf.cdo.server.db
- trunk/plugins/org.eclipse.emf.cdo.tests

Comment 14 Eike Stepper

2011-02-08 08:13:37 EST

Good catch, Stefan!

Comment 15 Eike Stepper

2011-06-23 03:41:07 EDT

Available in R20110608-1407