Hi, I have created jira issue OPENJPA-360, and attached OPENJPA-360.patch. As soon as it passes regressions, I will commit the patch.
Catalina On 9/8/07, Kevin Sutter <[EMAIL PROTECTED]> wrote: > > Thanks, Catalina! > > On 9/7/07, catalina wei <[EMAIL PROTECTED]> wrote: > > > > Patrick & Kevin, > > Both of you are right. > > On the other hand, we do have a testcase to test FOR UPDATE OF string > > where > > we assert SQL string for "FOR UPDATE OF optimize for 1 row" under > > org.apache.openjpa.jdbc.TestSelectForUpdateOverride. > > > > I mislead you guys by only looking at the testcase without verifying it > > with > > right version of DB2. > > I doubt that the testcase ever ran successfully for DB2 type: > > * > > > > db2ISeriesV5R3OrEarlier or db2UDBV81OrEarlier > > * I will have someone who has the right version of DB2 to run the test, > > and > > will post the result shortly. > > Sorry for the confusion that I have caused. > > > > Catalina > > > > On 9/7/07, Kevin Sutter <[EMAIL PROTECTED]> wrote: > > > > > > Catalina, > > > > > > On 9/6/07, Patrick Linskey <[EMAIL PROTECTED]> wrote: > > > > > > > > > > > > FWIW, I was also seeing an exception complaining about the 'FOR > UPDATE > > > > OF' clause. I was surprised by the syntax, as I would have expected > it > > > > to say 'FOR UPDATE OF <something>'. > > > > > > > > > I agree with Patrick on this question. According to the db2 manual, > > the > > > "FOR UPDATE OF" clause is supposed to be followed by a list of > > column(s). > > > Otherwise, the clause should just be "FOR UPDATE". I looked at the > sql > > > generation code and I didn't see where we are supplying any column > > > identifiers. Am I missing something? > > > > > > For reference: > > > > > > > > > http://publib.boulder.ibm.com/infocenter/dzichelp/v2r2/topic/com.ibm.db29.doc.sqlref/xf6a19.htm#xf6a19 > > > > > > Thanks, > > > Kevin > > > > > > -Patrick > > > > > > > > On 9/6/07, catalina wei <[EMAIL PROTECTED]> wrote: > > > > > Hi Patrick, > > > > > That SQL is correct syntax if you are running DB2 UDB version 8.1or > > > > earlier > > > > > and the isolation level is set pessimistic. > > > > > > > > > > If you are not running the said DB2 version and still seeing "FOR > > > UPDATE > > > > OF" > > > > > string, then we have a problem in DB2Dictionary. > > > > > > > > > > Catalina > > > > > > > > > > > > > > > On 9/6/07, Patrick Linskey <[EMAIL PROTECTED]> wrote: > > > > > > > > > > > > Hi, > > > > > > > > > > > > I'm seeing SQL like so: > > > > > > > > > > > > SELECT t0.ID, t0.VERSN, t0.STRINGFIELD FROM > SIMPLEPERSISTENTCLASS > > > t0 > > > > > > FOR UPDATE OF > > > > > > > > > > > > Is this valid DB2 SQL? I'm using a DB2 database that returns > > > SQL08016 > > > > > > from a call to getDatabaseProductVersion(). I'm guessing that > the > > > > > > checks for the FOR UPDATE clauses are getting tripped up > > somewhere. > > > > > > Any suggestions about what it should be for this version of DB2? > > > > > > > > > > > > -Patrick > > > > > > > > > > > > -- > > > > > > Patrick Linskey > > > > > > 202 669 5907 > > > > > > > > > > > > > > > > > > > > > > > -- > > > > Patrick Linskey > > > > 202 669 5907 > > > > > > > > > >
