FWD » Core Development » Database

04/01/2023

08:28 AM Bug #7035: Extract query parameters from statically converted FQL

There is another peculiarity at least with PostgreSQL - @select '1' = 1;@ will show 'true'. The same for @select 'tr... Constantin Asofiei

04:57 AM Bug #7035: Extract query parameters from statically converted FQL

The root cause seems to be an FQL like @tt1.f1 <> 0@ where @f1@ is an integer field. In this case, @0@ is extracted ... Constantin Asofiei

08:58 PM Bug #7108: Simulate upper/rtrim directly in H2 using case specific columns

Dănuț Filimon wrote:
> Ovidiu, is there any case where a rtrim is used but should not be removed when using @MariaDb... Ovidiu Maxiniuc

03/31/2023

07:28 PM Bug #6767: Possible name clashes in DMO classes

Eric Faulhaber wrote:
> Vladimir Tsichevski wrote:
>
> > What am I expected to do to make this patch merged to th... Vladimir Tsichevski

03:06 PM Feature #6371: implement SAVE CACHE statement

Merged into trunk as revision 14524. Boris Schegolev

02:18 PM Feature #6371 (Review): implement SAVE CACHE statement

Code review 6371b/14529: the changes look good. Eric Faulhaber

01:53 PM Feature #6371: implement SAVE CACHE statement

I pushed cleanup and headers update for modified files as revision 14529. Also, normalized extent support is replaced... Boris Schegolev

12:08 PM Feature #6371: implement SAVE CACHE statement

I've looked at 6371b and the changes look safe, they are limited to SAVE CACHE package in FWD. Constantin Asofiei

11:25 AM Bug #7035: Extract query parameters from statically converted FQL

7035a still fails, I think is related to some date literal. I need to find the actual FQL and SQL, will get back to ... Constantin Asofiei

07:13 AM Bug #7035: Extract query parameters from statically converted FQL

Committed it as 7035a/rev. 14525. I will also redo some of my regression tests for this. Alexandru Lungu

04:51 AM Bug #7035: Extract query parameters from statically converted FQL

Dănuț Filimon wrote:
> I attached a patch that should solve the issue. When handling strings in @FQLPreprocessor.cre... Constantin Asofiei

04:08 AM Bug #7035: Extract query parameters from statically converted FQL

I attached a patch that should solve the issue. When handling strings in @FQLPreprocessor.createParameterDiff@, they ... Dănuț Filimon

02:03 AM Feature #6444: dataset improvements

Ovidiu Maxiniuc wrote:
> Yes, That is correct. The time-stamps remain, but they need to be grouped together at the e... Igor Skornyakov

03/30/2023

07:29 PM Feature #6444: dataset improvements

Yes, That is correct. The time-stamps remain, but they need to be grouped together at the end of history list, under ... Ovidiu Maxiniuc

04:32 PM Feature #6444: dataset improvements

Igor Skornyakov wrote:
> Ovidiu Maxiniuc wrote:
> > Actually, not what I meant by updating the history entries.
> ... Igor Skornyakov

03:15 PM Feature #6444: dataset improvements

Constantin Asofiei wrote:
> The 'scale' computed in @int colDecs = Math.max(1, column.getDecimals());@ results in '0... Igor Skornyakov

03:04 PM Feature #6444: dataset improvements

The 'scale' computed in @int colDecs = Math.max(1, column.getDecimals());@ results in '0' so colDecs becomes 1, in Ov... Constantin Asofiei

02:28 PM Feature #6444: dataset improvements

Ovidiu Maxiniuc wrote:
> Here is the isolated testcase:
> [...]
> And the results are as follows:
> |_. Reference... Igor Skornyakov

02:22 PM Feature #6444: dataset improvements

Constantin Asofiei wrote:
> We need Ovidiu's test first. Also, try the test in #6444-201 with and without DECIMAL... Igor Skornyakov

02:20 PM Feature #6444: dataset improvements

Here is the isolated testcase:... Ovidiu Maxiniuc

01:13 PM Feature #6444: dataset improvements

Igor Skornyakov wrote:
> Constantin Asofiei wrote:
> > If DECIMALS is not specified, it defaults to 10 in OE. In F... Constantin Asofiei

01:12 PM Feature #6444: dataset improvements

Constantin Asofiei wrote:
> If DECIMALS is not specified, it defaults to 10 in OE. In FWD, there may be a problem a... Igor Skornyakov

01:08 PM Feature #6444: dataset improvements

If DECIMALS is not specified, it defaults to 10 in OE. In FWD, there may be a problem as we assume '0' in @persist.o... Constantin Asofiei

12:54 PM Feature #6444: dataset improvements

Well, in fact I've used the attached test for WRITE-JSON. Igor Skornyakov

12:40 PM Feature #6444: dataset improvements

For example, this test:... Constantin Asofiei

12:39 PM Feature #6444: dataset improvements

Constantin Asofiei wrote:
> Do you have a test which just uses @WRITE-XML@ for a temp-table with some records?
I me... Constantin Asofiei

12:39 PM Feature #6444: dataset improvements

This test uses appserver calls - the data may be affected by the transfer over the appserver. Do you have a test whi... Constantin Asofiei

12:27 PM Feature #6444: dataset improvements

Constantin Asofiei wrote:
> Igor, please specify a testcase I can use for the decimal serialization - where you saw ... Igor Skornyakov

12:23 PM Feature #6444: dataset improvements

Igor, please specify a testcase I can use for the decimal serialization - where you saw that the rounding is necessary. Constantin Asofiei

12:22 PM Feature #6444: dataset improvements

Branch 6444a was rebased to the trunk rev. 14523.
Pushed up to revision 14548. Igor Skornyakov

01:35 AM Feature #6444: dataset improvements

Ovidiu Maxiniuc wrote:
> Actually, not what I meant by updating the history entries.
> Your entry must be the last ... Igor Skornyakov

12:15 AM Feature #6444: dataset improvements

Igor Skornyakov wrote:
> Ovidiu Maxiniuc wrote:
> > Please update the H entries first. I will do some tests regardi... Ovidiu Maxiniuc

02:51 PM Feature #6371: implement SAVE CACHE statement

Boris Schegolev wrote:
> I rebased the branch (well, merged trunk into 6371b) without major issues.
>
> > As not... Eric Faulhaber

02:44 PM Feature #6371: implement SAVE CACHE statement

Awesome, I will apply headers update and continue cleanup and testing. Thank you! Boris Schegolev

02:42 PM Feature #6371: implement SAVE CACHE statement

Eric Faulhaber wrote:
> Boris Schegolev wrote:
> > I rebased the branch (well, merged trunk into 6371b) without maj... Eric Faulhaber

01:37 PM Feature #6371: implement SAVE CACHE statement

Boris Schegolev wrote:
> I rebased the branch (well, merged trunk into 6371b) without major issues.
This is not ... Eric Faulhaber

01:17 PM Feature #6371: implement SAVE CACHE statement

I rebased the branch (well, merged trunk into 6371b) without major issues.
> As noted in #6371-100, #6371-101, an... Boris Schegolev

11:23 AM Bug #7035: Extract query parameters from statically converted FQL

There are lots of these:... Constantin Asofiei

11:21 AM Bug #7035: Extract query parameters from statically converted FQL

Alexandru Lungu wrote:
> Rebased 7035a with latest trunk/rev. 14523.
ETF testing fails badly. Constantin Asofiei

09:53 AM Bug #7035: Extract query parameters from statically converted FQL

Rebased 7035a with latest trunk/rev. 14523. Alexandru Lungu

09:34 AM Bug #7035: Extract query parameters from statically converted FQL

Alexandru Lungu wrote:
> Tested it against two large customer applications. Also profiled against a customer POC. Th... Constantin Asofiei

06:37 AM Bug #7035: Extract query parameters from statically converted FQL

Tested it against two large customer applications. Also profiled against a customer POC. There were no regressions.
... Alexandru Lungu

06:31 AM Bug #7035: Extract query parameters from statically converted FQL

Alexandru, what's the state of testing for 7035a? Constantin Asofiei

10:08 AM Bug #7108: Simulate upper/rtrim directly in H2 using case specific columns

*Committed 7108a_h2/rev.18*. Fix for cached @CompareMode@ where the same instance is used but a different one is requ... Dănuț Filimon

05:39 AM Bug #7108: Simulate upper/rtrim directly in H2 using case specific columns

*Committed 7108a/rev.14519*. Skip adding upper and rtrim in the @SortCriterion.toString@ represensation when necessar... Dănuț Filimon

09:08 AM Feature #7066: Implement multi-table indexed query in H2

Alexandru Lungu wrote:
> * @s.setType(INDEX);@ means that you set the @index_name@ AST type to INDEX. This way, in a... Radu Apetrii

08:27 AM Feature #7066: Implement multi-table indexed query in H2

*Review for 7066a/rev. 14532*
The changes are reasonable considering the need of enforcing USE-INDEX in some scenari... Alexandru Lungu

06:26 AM Feature #7066: Implement multi-table indexed query in H2

I've committed to 7066a, rev. 14532 and 7066a_h2, rev. 14 some changes that target the use of @use index@.
As Alex p... Radu Apetrii

05:36 AM Feature #7194: Avoid generating an ORDER-BY clause if not required

Alexandru Lungu wrote:
> Eric Faulhaber wrote:
> > If you are intending to remove it at the FQL level, please note ... Radu Apetrii

03:05 AM Bug #6767: Possible name clashes in DMO classes

Vladimir Tsichevski wrote:
> What am I expected to do to make this patch merged to the trunk?
Create a task bra... Eric Faulhaber

03/29/2023

05:10 PM Bug #6767: Possible name clashes in DMO classes

Eric Faulhaber wrote:
> Code review 6767-full.diff:
>
> The changes look good.
>
> Just some minor issues with... Vladimir Tsichevski

12:58 PM Bug #6767: Possible name clashes in DMO classes

Code review 6767-full.diff:
The changes look good.
Just some minor issues with the file headers:
* The dates... Eric Faulhaber

04:09 PM Feature #6444: dataset improvements

Ovidiu, please post when you have some test results, so we can figure out where this falls in the merge order. Thanks. Eric Faulhaber

12:23 AM Feature #6444: dataset improvements

Ovidiu Maxiniuc wrote:
> Please update the H entries first. I will do some tests regarding the synchronize, fill and... Igor Skornyakov

03:30 PM Feature #6371: implement SAVE CACHE statement

Code review 6371b/14485-14488:
Overall, it seems good functionally; nice work. I do have some considerations and c... Eric Faulhaber

11:40 AM Bug #7185: H2 in-memory lazy hydration

Committed 7185a/rev. 14532 and 7185a_h2/rev. 16: handle cases when result-set is composed out of multiple row structu... Alexandru Lungu

11:09 AM Bug #7185: H2 in-memory lazy hydration

*Done some tests with latest 7185a/rev. 14531 and latest 7185a_h2:*
* @FWDDirectAccess.retrieveResponse@ you creat... Alexandru Lungu

09:02 AM Bug #7185: H2 in-memory lazy hydration

Based on #7185-14 I committed *7185a/rev.14531*. I confirmed that @TempRecord.checkIncomplete@ is redundant and remov... Dănuț Filimon

07:15 AM Bug #7185: H2 in-memory lazy hydration

Eric Faulhaber wrote:
> The idea when I first implemented @BaseRecord@ was to make the @data@ array directly accessi... Alexandru Lungu

06:28 AM Bug #7185: H2 in-memory lazy hydration

Alexandru Lungu wrote:
> *Eric, Constantin, please advice*. If we attempt to implement lazy-hydration in any form ... Eric Faulhaber

06:03 AM Bug #7185: H2 in-memory lazy hydration

*Review for 7185a/rev. 14529 and 14530*
This is close to a conceptually correct solution. However, there are still... Alexandru Lungu

05:19 AM Bug #7185: H2 in-memory lazy hydration

*Committed 7185a/rev.14530*. Fixed a NPE in @TempRecord.getDatum()@ where fallback was assigned but readFields was null. Dănuț Filimon

10:16 AM Bug #7108: Simulate upper/rtrim directly in H2 using case specific columns

Danut, please mark #7108-31 as high priority for now. Alexandru Lungu

07:47 AM Bug #7235: Avoid selecting all fields for an CAN-FIND sub-select

4GL code:... Alexandru Lungu

07:39 AM Bug #7235 (Closed): Avoid selecting all fields for an CAN-FIND sub-select

Alexandru Lungu

07:17 AM Feature #7194: Avoid generating an ORDER-BY clause if not required

Eric Faulhaber wrote:
> If you are intending to remove it at the FQL level, please note that we do a reverse lookup ... Alexandru Lungu

06:37 AM Feature #7194: Avoid generating an ORDER-BY clause if not required

Is the idea to just remove the ORDER BY at the SQL level (as opposed to the FQL level)?
If you are intending to re... Eric Faulhaber

05:25 AM Feature #7066: Implement multi-table indexed query in H2

Created 7066a from trunk rev. 14519 and rebased with 7026b changes. Also created 7066a_h2 from FWD-H2 trunk rev. 13. Alexandru Lungu

03/28/2023

07:37 PM Feature #6444: dataset improvements

Igor Skornyakov wrote:
> > *** according to https://docs.progress.com/bundle/openedge-prodatasets-117/page/Using-the... Ovidiu Maxiniuc

07:11 AM Feature #6444: dataset improvements

Ovidiu Maxiniuc wrote:
> Review of 6444a/14538:
> * @AbstractQuery.java@, @CompoundComponent.java@, Dialect.java, e... Igor Skornyakov

02:30 AM Feature #6444: dataset improvements

Branch 6444a was rebase to the trunk rev. 14518.
Pushed up to revision 14541. Igor Skornyakov

02:12 AM Feature #6444 (WIP): dataset improvements

Igor Skornyakov

11:34 PM Feature #6444: dataset improvements

Review of 6444a/14538:
* @AbstractQuery.java@, @CompoundComponent.java@, Dialect.java, etc: GCD coding standard: ple... Ovidiu Maxiniuc

07:05 PM Feature #6371: implement SAVE CACHE statement

I added support for comments / meta info for extent fields as revision 6371b/14488. Documentation is also updated - s... Boris Schegolev

10:58 AM Bug #7108: Simulate upper/rtrim directly in H2 using case specific columns

h2. Reviews
*Review for 7108a*
* @SortCriterion@ instances are still generated with @upper@ and @rtrim@! Please c... Alexandru Lungu

08:25 AM Bug #7108: Simulate upper/rtrim directly in H2 using case specific columns

Rebased 7108a from trunk rev. 14515. 7108a is now at rev. 14518. Alexandru Lungu

09:50 AM Bug #7185: H2 in-memory lazy hydration

Based on #7185-10 I committed *7185a/rev.14529* and *7185a_h2/rev.15.*.
* Fixed @TempRecord@.
* Made @BaseRecord.ja... Dănuț Filimon

05:27 AM Bug #7185: H2 in-memory lazy hydration

*Review for 7185a/rev. 14528*
* In @TempRecord@:
** @readFields.set(i);@ is redundant as @readProperty@ already s... Alexandru Lungu

02:57 AM Bug #7185: H2 in-memory lazy hydration

*Committed 7185a_h2/rev.14* and *7185a/rev.14528* as part of the initial implementation for lazy hydration.
I've m... Dănuț Filimon

09:23 AM Feature #7194: Avoid generating an ORDER-BY clause if not required

Radu, please address #7194 once you get a stable solution for #7066. AFAIK, you already have some implementation rega... Alexandru Lungu

07:16 AM Feature #7060 (Review): Consider using wildcard selection to reduce SQL size

Alexandru Lungu

04:27 AM Bug #7138 (WIP): Short circuit fake-update (update with a new value eqaul to the old value) in H2

Alexandru Lungu

03/27/2023

10:29 AM Bug #7220: Record not created when property is set to empty string

I remember that Danut also tested the example with a temporary table. In that case, the record was found because @isA... Alexandru Lungu

03/24/2023

06:56 PM Feature #6371: implement SAVE CACHE statement

I puhsed revision 6371b/14487. It provides support for DTTZ columns, fixes for MariaDB in @SchemaCheck@ and support f... Boris Schegolev

03/23/2023

04:27 AM Feature #7061 (Test): Enable the use of lazy result sets in H2

Alexandru Lungu

04:25 AM Bug #7165 (Test): Close multiplex scope using prepared statement

Alexandru Lungu

10:49 PM Bug #7220: Record not created when property is set to empty string

Interesting finding.
The record is not forced to flush to DB when it was created. The 'touch' marks the field as d... Ovidiu Maxiniuc

03/22/2023

07:18 AM Bug #7220: Record not created when property is set to empty string

Found an issue related to persistent tables where a record is not created when the property is set to an empty string... Dănuț Filimon

07:10 AM Bug #7220 (New): Record not created when property is set to empty string

Dănuț Filimon

06:34 AM Bug #7108: Simulate upper/rtrim directly in H2 using case specific columns

When I added rtrimming to H2, I set it so that it will always rtrim when comparing strings. Then I noticed an issue d... Dănuț Filimon

03/20/2023

01:30 PM Bug #7211: Several H2 connections for one single session

-Adding a server log that contains around 100 of such connections. Note that is just during server startup without th... Thanasis Giannakopoulos

09:54 AM Feature #7062: Direct access to H2 internals to solve basic queries

I also committed an optimization for #7062-27 (7062b_h2/rev. 14 and 7026b/rev. 14512), including a mean of lazily com... Alexandru Lungu

03/17/2023

08:48 AM Feature #7062: Direct access to H2 internals to solve basic queries

Created 7062b_h2 for further work on H2.
Committed 7062b_h2/rev. 13 - generating a dummy jdbc statement for each r... Alexandru Lungu

06:30 AM Feature #6444: dataset improvements

Please don't. They will get forgotten. Marian has already confirmed he will handle the merge. Greg Shah

06:15 AM Feature #6444: dataset improvements

Greg Shah wrote:
> Marian/Constantin: Can you recommend a better place for these tests than @6453-6444/@?
>
> Igo... Igor Skornyakov

06:11 AM Feature #6444: dataset improvements

Marian/Constantin: Can you recommend a better place for these tests than @6453-6444/@?
Igor: We don't want to stor... Greg Shah

05:49 AM Feature #6444: dataset improvements

Tests for #6453 and #6444 has been added to @6453-6444@ subfolder of the @xfer.goldencode.com/opt/testcases/@ rev. 13... Igor Skornyakov

04:14 AM Feature #6444: dataset improvements

Marian Edu wrote:
> Igor, I think the easiest way to do that is to simply use a 'pipe through' routine - I think we ... Igor Skornyakov

03:16 AM Feature #6444: dataset improvements

Igor Skornyakov wrote:
> This is for testing TEMP-TABLEs DATA-SETs as arguments in the remote calls.
Igor, I th... Marian Edu

03:12 AM Feature #6444: dataset improvements

Greg Shah wrote:
> Marian: Do you want Igor to commit his tests into the current [[p2j:Testcases]] or do you want so... Marian Edu

03:09 AM Feature #6444: dataset improvements

Greg Shah wrote:
> Marian: Has your team had to do this with the existing tests?
We've followed a similar approac... Marian Edu

05:51 AM Bug #5306: AdaptiveQuery sets a found record into the buffer twice in dynamic mode

Radu Apetrii wrote:
> Alexandru Lungu wrote:
> > Can you do a run on the latest rev. 14507?
>
> Since the tests ... Alexandru Lungu

05:07 AM Bug #5306: AdaptiveQuery sets a found record into the buffer twice in dynamic mode

Alexandru Lungu wrote:
> Can you do a run on the latest rev. 14507?
Since the tests were already set up, I ran th... Radu Apetrii

03/16/2023

01:08 PM Feature #6444: dataset improvements

Greg Shah wrote:
> We may want this to be configurable.
>
> Why does it generate output on the server side? That... Igor Skornyakov

12:55 PM Feature #6444: dataset improvements

We may want this to be configurable.
Why does it generate output on the server side? That doesn't sound expected. Greg Shah

12:53 PM Feature #6444: dataset improvements

Please note also that some tests generate files not only in the client home but also in the app. server home. This ma... Igor Skornyakov

12:46 PM Feature #6444: dataset improvements

Greg Shah wrote:
> I agree. It is good to be able to look deeply at any issue. I just want the comparison/detectio... Igor Skornyakov

12:40 PM Feature #6444: dataset improvements

I agree. It is good to be able to look deeply at any issue. I just want the comparison/detection of a problem to be... Greg Shah

12:25 PM Feature #6444: dataset improvements

Please note that my approach for testing is based on visual comparision of multiple output files generated by the tes... Igor Skornyakov

12:08 PM Feature #6444: dataset improvements

Marian: Do you want Igor to commit his tests into the current [[p2j:Testcases]] or do you want some other process to ... Greg Shah

12:06 PM Feature #6444: dataset improvements

Greg Shah wrote:
> Are all of your testcases integrated into the dataset tests in [[p2j:Testcases]]? If not, they s... Igor Skornyakov

12:02 PM Feature #6444: dataset improvements

Are all of your testcases integrated into the dataset tests in [[p2j:Testcases]]? If not, they should be.
Marian'... Greg Shah

10:31 AM Feature #6444 (Review): dataset improvements

Radu Apetrii wrote:
> Igor, I've attached a patch that should solve the problem.
>
> There was a parameter in @Ad... Igor Skornyakov

10:06 AM Feature #6444: dataset improvements

Radu Apetrii wrote:
> Igor, I've attached a patch that should solve the problem.
>
> There was a parameter in @Ad... Igor Skornyakov

09:54 AM Feature #6444: dataset improvements

Igor, I've attached a patch that should solve the problem.
There was a parameter in @AdaptiveQuery.executeScroll@ ... Radu Apetrii

05:41 AM Feature #6444: dataset improvements

Radu Apetrii wrote:
> In @AdaptiveQuery.executeQuery@, due to the introduction of @lazyMode@, all the @ProgressiveRe... Radu Apetrii

05:12 AM Feature #6444: dataset improvements

Eric Faulhaber wrote:
> Radu, have you been able to recreate the problem with branch #6444a?
Yes, I have.
In @Ad... Radu Apetrii

02:14 AM Feature #6444: dataset improvements

Eric Faulhaber wrote:
> Rather than just commenting out the lazy mode code in @AdaptiveQuery@, as in #6444-147, I st... Igor Skornyakov

11:48 PM Feature #6444: dataset improvements

Igor Skornyakov wrote:
> Apart from READ-JSON support and #6444-149 all is done and 6444a is ready for the code re... Eric Faulhaber

11:53 AM Bug #6887: improve performance of dataset/table[-handle] parameters

This is the entire JMX result for the large application:... Constantin Asofiei

02:59 AM Bug #6887: improve performance of dataset/table[-handle] parameters

Boris Schegolev wrote:
> I found a few points where we spend most CPU time:
>
> 1. Reflexive invocation @java.lan... Eric Faulhaber

02:31 AM Bug #6887: improve performance of dataset/table[-handle] parameters

Constantin Asofiei wrote:
> In a large application, there is this JMX reported times (count/millis/nanos):
> ... Eric Faulhaber

11:50 AM Feature #7062: Direct access to H2 internals to solve basic queries

Go with option 2 - on @getStatement()@ call, lazily create the @java.sql.Statement@ instance which has all the entire... Constantin Asofiei

11:31 AM Feature #7062: Direct access to H2 internals to solve basic queries

Response to #7026-97
Constantin Asofiei wrote:
> Alexandru, I'm not sure where to post this. The changes for @Di... Alexandru Lungu

11:16 AM Bug #7211: Several H2 connections for one single session

This is an issue as there are several messages like the following in the server log:... Alexandru Lungu

11:04 AM Bug #7211 (WIP): Several H2 connections for one single session

Alexandru Lungu

10:58 AM Bug #5306: AdaptiveQuery sets a found record into the buffer twice in dynamic mode

Radu, I think you have both 7026b (at least rev. 14505) and the @adaptive_scrolling@ tests (scrolling and non-scrolli... Alexandru Lungu

10:28 AM Feature #7060 (WIP): Consider using wildcard selection to reduce SQL size

I committed 7026b/rev. 14506 for wildcard selection only for P2JH2 dialect which already supports @INVISIBLE@ keyword... Alexandru Lungu

09:34 AM Bug #7185: H2 in-memory lazy hydration

> override checkIncomplete or create checkFetched, to automatically fetch the data at the specified index from the st... Alexandru Lungu

08:38 AM Bug #7174: Resolve simple CAN-FIND statements faster

Alexandru Lungu wrote:
> I think your <10% rates are error thresholds, considering that the first two optimizations ... Dănuț Filimon

07:23 AM Bug #7035 (Review): Extract query parameters from statically converted FQL

Committed 7035a/rev.14507. I applied the patch from #7035-11 which refactors @ParameterIndices@ into @ParameterDiff@ ... Dănuț Filimon

05:07 AM Bug #7035: Extract query parameters from statically converted FQL

I profiled #7035 fix again and I noticed a slight performance improvement this time. I think the optimization is havi... Alexandru Lungu

06:17 AM Bug #7167: Associating records from opened buffers to new sessions is slow

After a bit of investigation, the database sessions @com.goldencode.p2j.persist.orm.Session@ are closed quite eagerly... Alexandru Lungu

05:52 AM Bug #7165: Close multiplex scope using prepared statement

Alexandru Lungu wrote:
> * In @TemporaryBuffer.java@, add a proper default capacity for @additionalArgs@ and @ooArgs... Radu Apetrii

03/15/2023

01:18 PM Bug #6887: improve performance of dataset/table[-handle] parameters

I found a few points where we spend most CPU time:
# Reflexive invocation @java.lang.reflect.Method.invoke()@ - th... Boris Schegolev

12:45 PM Feature #6444: dataset improvements

Fixed the the NESTED/FOREIGN-KEY-HIDDEN support for WRITE-JSON.
6444a/14525.
The current status of the NESTED/FORE... Igor Skornyakov

11:13 AM Feature #6830 (Test): find and fix all SQL SELECT statements with inlined literal arguments, and ...

This reached trunk with 6129b. It can be closed.
#7165 was opened in the meantime targeting a "sub-issue" from here. Alexandru Lungu

11:10 AM Feature #6995 (Test): possibly support NO-UNDO temporary tables directly in H2

This reached trunk with 7026a. It can be closed. Alexandru Lungu

07:00 AM Bug #5306 (Review): AdaptiveQuery sets a found record into the buffer twice in dynamic mode

Committed fix for repetitive fetching in 7026b/rev. 14505. I also attached the changes here.
The fix is the same a... Alexandru Lungu

06:38 AM Bug #7108: Simulate upper/rtrim directly in H2 using case specific columns

I noticed a problem related to unique index/primary keys where the value that needs to be inserted is rtrimmed, but a... Dănuț Filimon

05:26 AM Feature #7061: Enable the use of lazy result sets in H2

Alexandru Lungu wrote:
> Radu, please commit #7061-28 patch to 7026b if there are no obvious problems in your tests.... Radu Apetrii

05:05 AM Feature #7061: Enable the use of lazy result sets in H2

Radu, please commit #7061-28 patch to 7026b if there are no obvious problems in your tests. I am looking forward to s... Alexandru Lungu

05:25 AM Bug #7174 (Review): Resolve simple CAN-FIND statements faster

> There is an improvement when using ANY instead of ONE for persistent tables when a findByUniqueIndex is detected. T... Alexandru Lungu

05:19 AM Bug #7165: Close multiplex scope using prepared statement

*Review of 7026b/rev. 14501*
* In @TemporaryBuffer.java@, add a proper default capacity for @additionalArgs@ and @... Alexandru Lungu

03/14/2023

02:11 PM Feature #6444: dataset improvements

Added FOREIGN-KEY-HIDDEN support for WRITE-XML,(READ|WRITE)-XMLSCHEMA.
Committed to 6444a/14524
The remaining thi... Igor Skornyakov

04:59 AM Feature #6444: dataset improvements

Rebased 6444a to the trunk/14503.
Pushed up to revision 14523.
Igor Skornyakov

04:51 AM Feature #6444: dataset improvements

Fixed @READ-XML(SCHEMA)?@ support for DATA-SETs with NESTED DATA-RELATIONs and multiple namespaces.
It works correct... Igor Skornyakov

12:30 PM Bug #7174: Resolve simple CAN-FIND statements faster

This semantic is why we have the @has{Any|One}@ nomenclature in @FindQuery@.
* @CAN-FIND({FIRST|LAST} ...)@ is con... Eric Faulhaber

09:35 AM Bug #7174: Resolve simple CAN-FIND statements faster

These are the expected results, where ONE is @CAN-FIND(tt)@, ANY is @CAN-FIND(FIRST tt)@.
|_. Number of records |_... Alexandru Lungu

09:23 AM Bug #7174: Resolve simple CAN-FIND statements faster

In OE, will a CAN-FIND ONE result return true if there is more than one match? In other words, is CAN-FIND ONE meant... Greg Shah

05:29 AM Bug #7174: Resolve simple CAN-FIND statements faster

I created tests that use simple queries, those tests consist of executing CAN-FIND for each record in a temporary/per... Dănuț Filimon

10:50 AM Bug #7138: Short circuit fake-update (update with a new value eqaul to the old value) in H2

Ovidiu Maxiniuc wrote:
> Alexandru Lungu wrote:
> > Right now, I am still dragged back by @String.hashCode@:
> > [... Alexandru Lungu

10:32 AM Bug #7138: Short circuit fake-update (update with a new value eqaul to the old value) in H2

Alexandru Lungu wrote:
> Right now, I am still dragged back by @String.hashCode@:
> [...]
Why is that? It seems li... Ovidiu Maxiniuc

09:42 AM Bug #7138: Short circuit fake-update (update with a new value eqaul to the old value) in H2

Ovidiu Maxiniuc wrote:
> Apparently, the hash short-circuit is not used. The @length@ of the strings are compared, b... Alexandru Lungu

08:46 AM Bug #7138: Short circuit fake-update (update with a new value eqaul to the old value) in H2

H2 database engine is built in Java. My first reaction to your hash suggestion was that it unneeded since Java has th... Ovidiu Maxiniuc

05:10 AM Bug #7138: Short circuit fake-update (update with a new value eqaul to the old value) in H2

... Alexandru Lungu

10:02 AM Feature #7194: Avoid generating an ORDER-BY clause if not required

This task targets the following scenario:... Alexandru Lungu

09:49 AM Feature #7194 (Closed): Avoid generating an ORDER-BY clause if not required

Alexandru Lungu

08:03 AM Feature #7062: Direct access to H2 internals to solve basic queries

Moved the @isFindByRowid@ short-circuit from #7062-30 to @RAQ.execute@. This is because @isSimpleQuery@ was always re... Alexandru Lungu

07:33 AM Bug #7185: H2 in-memory lazy hydration

Danut, please focus on implementing a demo for point 5 - we should have a fast mock to test its performance. Therefor... Alexandru Lungu

07:15 AM Bug #7185 (WIP): H2 in-memory lazy hydration

Alexandru Lungu

04:47 AM Bug #7165: Close multiplex scope using prepared statement

Alexandru Lungu wrote:
> Radu, please move 7061a, rev. 14487 to 7026b.
> I will review, test and profile it once th... Radu Apetrii

04:03 AM Bug #7165 (Review): Close multiplex scope using prepared statement

Radu, please move 7061a, rev. 14487 to 7026b.
I will review, test and profile it once there. Alexandru Lungu

03/13/2023

06:47 PM Bug #6887 (WIP): improve performance of dataset/table[-handle] parameters

I did more testing on provided code. The INPUT-OUTPUT DATASET-HANDLE is slightly slower in FWD, but it's definitely n... Boris Schegolev

03:44 PM Feature #6444: dataset improvements

Ovidiu Maxiniuc wrote:
> The main difference between @ds.xsd@ and the deserialied @hds.xsd@ is that the namespace, t... Igor Skornyakov

03:21 PM Feature #6444: dataset improvements

Igor Skornyakov wrote:
> I've noticed a strange thing. Consider the attached test @nested-test.p@.
> It defines a D... Ovidiu Maxiniuc

09:55 AM Feature #6444: dataset improvements

Radu Apetrii wrote:
> Igor Skornyakov wrote:
> > The problem is with 'lazyMode'. After applying the attached path t... Igor Skornyakov

09:32 AM Feature #6444: dataset improvements

Igor Skornyakov wrote:
> The problem is with 'lazyMode'. After applying the attached path the regression has gone.
... Radu Apetrii

05:28 AM Feature #6444: dataset improvements

Fixed minor issues with @WRITE-XML(SCHEMA)?@ for tables.
Committed to 6444a/14519. Igor Skornyakov

03:59 AM Feature #6444: dataset improvements

Eric Faulhaber wrote:
> Please post the test case, explain the regression to Radu, and let him fix it.
I'm workin... Radu Apetrii

03:09 AM Feature #6444: dataset improvements

Sorry, I've attached a wrong file. Igor Skornyakov

02:51 AM Feature #6444: dataset improvements

Eric Faulhaber wrote:
> Igor Skornyakov wrote:
> > Igor Skornyakov wrote:
> > > After rebasing I see a regression ... Igor Skornyakov

02:14 AM Feature #6444: dataset improvements

Igor Skornyakov wrote:
> Igor Skornyakov wrote:
> > After rebasing I see a regression in the FILL test for temporar... Eric Faulhaber

10:32 AM Bug #5306 (WIP): AdaptiveQuery sets a found record into the buffer twice in dynamic mode

Alexandru Lungu

07:37 AM Bug #7185: H2 in-memory lazy hydration

There is already some effort into lazy hydration as a general topic:
* supporting FIELDS/EXCEPT clause, so we can fe... Alexandru Lungu

03/12/2023

12:40 PM Feature #6444: dataset improvements

Igor Skornyakov wrote:
> After rebasing I see a regression in the FILL test for temporary source tables and recursiv... Igor Skornyakov

10:28 AM Feature #6444: dataset improvements

After rebasing I see a regression in the FILL test for temporary source tables and recursive DATA-RELATION. Igor Skornyakov

10:13 AM Feature #6444: dataset improvements

Rebased 6444a to the trunk/14500.
Pushed up to revision 14517. Igor Skornyakov

09:26 AM Feature #6444: dataset improvements

Fixed WRITE-XML(SCHEMA) for multiple namespaces and NESTED relations.
Committed to 6444a/14508. Igor Skornyakov

03/10/2023

02:09 PM Feature #6444: dataset improvements

Igor Skornyakov wrote:
> [...]
> This looks incorrect. As far as I understand which tables are missed depends on t... Eric Faulhaber

04:22 AM Feature #7061: Enable the use of lazy result sets in H2

Alexandru Lungu wrote:
> Radu, please review and test with your adaptive test cases (scrolling and non-scrolling).
... Radu Apetrii

03/09/2023

03:03 PM Bug #6919: fix ErrorManager.displayError in TempTableBuilder

My understanding is that we can close it. Boris Schegolev

01:20 PM Feature #6444: dataset improvements

More about the situation described in #6444-141.
Although in the Progress Developer Studio debugger @hds@ after im... Igor Skornyakov

08:39 AM Feature #6444: dataset improvements

I've noticed a strange thing. Consider the attached test @nested-test.p@.
It defines a DATA-SET with 5 tables (tt1... Igor Skornyakov

09:03 AM Bug #7185 (Test): H2 in-memory lazy hydration

Alexandru Lungu

07:54 AM Feature #7061: Enable the use of lazy result sets in H2

Constantin Asofiei wrote:
> Alexandru, is this patch meant to solve regressions? Because we are getting close to fin... Alexandru Lungu

07:48 AM Feature #7061: Enable the use of lazy result sets in H2

Alexandru, is this patch meant to solve regressions? Because we are getting close to finish testing 7026a, and unless... Constantin Asofiei

07:46 AM Feature #7061: Enable the use of lazy result sets in H2

After a self-review of the patch, I see that we still register the change broker for @AdaptiveQuery.open@ and there i... Alexandru Lungu

05:52 AM Feature #7061: Enable the use of lazy result sets in H2

Radu, this is the patch I am planning to integrate into 7026a to wrap-up the effort from #7061 and have a complete im... Alexandru Lungu

05:38 AM Feature #7061 (Review): Enable the use of lazy result sets in H2

Rebased 7061a to trunk (rev. 14497). Planning to merge your latest changes to 7026a as they target a temporary fix fr... Alexandru Lungu

05:00 AM Feature #7061: Enable the use of lazy result sets in H2

Alexandru Lungu wrote:
> Radu, is the multi-table @AdaptiveQuery@ considered for your implementation? I understood t... Radu Apetrii

07:39 AM Bug #7108: Simulate upper/rtrim directly in H2 using case specific columns

Alexandru Lungu wrote:
> * Note that you apply @s.replace(CASE_TYPE_TOK, "");@ to any type. You can do a check befor... Dănuț Filimon

03/08/2023

08:16 AM Bug #7108: Simulate upper/rtrim directly in H2 using case specific columns

Alexandru Lungu wrote:
> * testFWD shows regressions. I can't tell if they are false positives, but you should look ... Alexandru Lungu

07:43 AM Bug #7108 (Review): Simulate upper/rtrim directly in H2 using case specific columns

First thing to notice: the application always requires reconversion due to the new DDL for the H2 meta database (even... Alexandru Lungu

03/07/2023

09:43 AM Feature #7066 (WIP): Implement multi-table indexed query in H2

The purpose of this task too wide, especially after the implementation of the @lazy@ keyword. We should address some ... Alexandru Lungu

09:28 AM Feature #7062: Direct access to H2 internals to solve basic queries

The results were obtained using 7026a/rev. 14518. I don't have some results for 7026a/rev. 14503 for this specific ap... Alexandru Lungu

09:10 AM Feature #7062: Direct access to H2 internals to solve basic queries

Do these results differ from 7026a rev 14503? Greg Shah

08:44 AM Feature #7062: Direct access to H2 internals to solve basic queries

I've done the performance test on another large customer application. I've added more details to cover temp vs persis... Alexandru Lungu

08:59 AM Feature #7061: Enable the use of lazy result sets in H2

Radu, is the multi-table @AdaptiveQuery@ considered for your implementation? I understood that @lazy@ keyword works o... Alexandru Lungu

07:23 AM Feature #6444: dataset improvements

I've created a more advanced test with five tables, multiple namespaces and multiple nested data-reltions.
It appear... Igor Skornyakov

05:08 AM Bug #7174: Resolve simple CAN-FIND statements faster

Please note that we have a task regarding CAN-FIND (#7113) waiting for the code review.
Maybe it makes sense to wait... Igor Skornyakov

04:59 AM Bug #7174 (WIP): Resolve simple CAN-FIND statements faster

There are some opportunities to short-circuit the CAN-FIND query in FWD. A CAN-FIND statement is relaxed, in the sens... Alexandru Lungu

04:40 AM Bug #7174 (Closed): Resolve simple CAN-FIND statements faster

Alexandru Lungu

03:47 AM Bug #7165: Close multiplex scope using prepared statement

Alexandru Lungu wrote:
> @TemporaryBuffer.removeRecords@ is widely used when closing a multiplex scope, but not only... Radu Apetrii

03/06/2023

05:10 PM Bug #6812 (Review): fix FastFindCache to use a shared instance per persistent database

I prepared a change for the last missing bit (@SavepointManager@) - see revision 6812a/14487. It needs more testing, ... Boris Schegolev

06:52 AM Bug #7108: Simulate upper/rtrim directly in H2 using case specific columns

Alexandru Lungu wrote:
> We should integrate the first two for now to benefit from the "free meal". Danut, please co... Dănuț Filimon