Negative VOTE Missing methods

[Craig L Russell]

Javadogs,

Here's the spec update for the missing makeTransient methods.

Please reply if there are any issues.

Craig

<proposed>
void makeTransient (Object pc, boolean useFetchPlan);
void makeTransientAll (Object[] pcs, boolean useFetchPlan);
void makeTransientAll (Collection pcs, boolean useFetchPlan);

</proposed>

On Feb 15, 2006, at 12:12 PM, Craig L Russell wrote:

If the useFetchPlan parameter is false, these methods behave exactly as the corresponding makeTransient methods. If the useFetchPlan parameter is true, the current FetchPlan, including MaxFetchDepth, DETACH_LOAD_FIELDS, and DETACH_UNLOAD_FIELDS, is applied to the pc or pcs parameter instance (s) to load fields and instances from the datastore. The DetachmentRoots is not affected. After the fetch plan is used to load instances, the entire graph of instances reachable via loaded fields of the parameter instances is made transient. Transient fields are not modified by the method. If the parameter instance or instances are detached, then JDOUserException is thrown.

Hi Andy,

Yes, you are right. I accidentally omitted this from the draft.

I'll add all three methods:
void makeTransient(Object pc, boolean useFetchPlan);
void makeTransient(Collection pcs, boolean useFetchPlan);
void makeTransient(Object[] pcs, boolean useFetchPlan);



Thanks,

Craig

On Feb 15, 2006, at 8:05 AM, Andy Jefferson wrote:

I haven't reviewed your patch but I assume you only added the first method. The others are needed for consistency.

I thought a new makeTransient() method was to be added. This is Issue 38 on the JDO2Issues. A negative VOTE was called on 07 Feb and nobody showed up to

decry it. Therefore it should be in the spec.

Section 12.6.7 needs the following (or similar adding)
<spec>
void makeTransient (Object pc, boolean useFetchPlan);

This method would behave like retrieve(rootObject, true) [FGOnly] in
terms of loading instances according to the FetchPlan, followed by

</spec>


--
Andy

recursive makeTransient on all instances reachable via the fetch plan.

Craig Russell
Architect, Sun Java Enterprise System http://java.sun.com/products/jdo
408 276-5638 mailto:[EMAIL PROTECTED]
P.S. A good JDO? O, Gasp!

Craig Russell
Architect, Sun Java Enterprise System http://java.sun.com/products/jdo
408 276-5638 mailto:[EMAIL PROTECTED]
P.S. A good JDO? O, Gasp!

smime.p7s
Description: S/MIME cryptographic signature