Activity
From 02/27/2021 to 03/28/2021
03/29/2021
03/28/2021
- 11:33 AM Feature #1587: implement full support for word indexes
- For the performance testing (@words/words-perf.p@) I use the following data:
The test table definition is:... - 04:56 AM Feature #1587: implement full support for word indexes
- Added documentation for the new @CONTAINS@ support to the 'Internals' section of Wiki. See [[p2j:Word Index and CONTA...
03/26/2021
- 05:04 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> For the "how it works" documentation you just posted, you could add a page to the @Internals... - 05:02 PM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> BTW: maybe it makes sense to add documentation to the Wiki? If so, then to what section?
... - 04:36 PM Feature #1587: implement full support for word indexes
- h1. Word tables and related database objects
For the table *{T}* and field *{f}* with word index on it the word ta... - 02:13 PM Feature #1587: implement full support for word indexes
- Merged wors tables tests' data to the @fwd@ test database. Performed cleanup of the tests.
Committed to @xfer.gold... - 01:33 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> What is the problem and do you have an idea about the level of effort required to fix it?
... - 01:22 PM Feature #1587: implement full support for word indexes
- What is the problem and do you have an idea about the level of effort required to fix it?
- 08:14 AM Feature #1587: implement full support for word indexes
- Found a regression in the denormalized extents' support.
Working on this. - 07:23 AM Feature #1587: implement full support for word indexes
- I've just found a stupid bug in the @schema_word_tables_<db>_<dialect>.sql@ script generation. It is not a problem fo...
03/23/2021
- 12:17 PM Bug #4163 (Closed): SHARED non-NEW buffer definitions must act as the equivalent of FIND or CREATE
03/22/2021
- 10:55 AM Support #5183: database collation
- I've committed a newer version of the collation data collector program to 3821c/12160.
The @locale/cp_coll/@ direc... - 05:26 AM Feature #1587: implement full support for word indexes
- Re-worked ordering looks a little slower than before. However, it is still much faster than with UDF.
| expr ...
03/21/2021
- 05:16 PM Feature #1587: implement full support for word indexes
- Implicit sorting for @_temp@ and 'dirty' databases is implemented. Will be committed tomorrow after additional testin...
- 09:04 AM Feature #1587: implement full support for word indexes
- Re-worked implicit ordering for word tables. See #1587-462.
Committed to 3821c/12157.
03/19/2021
- 01:18 PM Feature #1587: implement full support for word indexes
- I've realized that the implicit ordering described in #1587-407 is not correct and can result in the repeated records...
03/18/2021
- 11:11 AM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Igor Skornyakov wrote:
>
> > What about ordering for @_temp@ and 'dirty'?
>
> If you c... - 11:09 AM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> What about ordering for @_temp@ and 'dirty'?
If you can get this done today or tomorro... - 07:54 AM Feature #1587: implement full support for word indexes
- A refactored version of the word tables' support and creation of word tables for the @_temp@ database (currently disa...
- 07:42 AM Feature #1587: implement full support for word indexes
- The results of the performance test ("warmed" mode). Last 4 columns are times in ms:
| expr | records | 4... - 05:33 AM Feature #1587: implement full support for word indexes
- After adding "warming up" to my @words/words-perf.p@ performance test I've noticed the following. The screen remains ...
- 04:40 AM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> In 3821c/12143, I have added the 1688 conversion-time warning and code to remove the subscri... - 04:36 AM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> I'm sure we can adapt the @orm.Persister@ class to allow for some trigger-like feature which... - 12:01 AM Feature #1587: implement full support for word indexes
- In 3821c/12143, I have added the 1688 conversion-time warning and code to remove the subscript from an extent field l...
03/17/2021
- 06:23 PM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> Without triggers, we have to add the word tables population/update on every insert/update... - 02:04 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> The current implementation of triggers in H2 seems to be a dealbreaker, in that the trigger ... - 01:53 PM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> It seems that for H2 temporary tables it is impossible to use triggers. H2 trigger is invok... - 09:47 AM Feature #1587: implement full support for word indexes
- It seems that for H2 temporary tables it is impossible to use triggers. H2 trigger is invoked with a separate new con...
- 08:31 AM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> I've encountered a strange problem with trigger for @_temp@:
> [...]
>
> The @OnInsertW... - 07:30 AM Feature #1587: implement full support for word indexes
- I've encountered a strange problem with trigger for @_temp@:...
- 03:52 AM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> No runtime changes are required. By "will just not work" I mean that the corresponding piec... - 03:30 AM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Until now, I think yes. I think the other related tables (normalized extent field tables), i... - 03:27 AM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Sorry, to clarify "will just not work", do you mean that removing the subscript at conversio... - 04:36 PM Feature #3574 (Closed): finish implementation of temp-table XML support
- 04:26 PM Feature #3574: finish implementation of temp-table XML support
- Yes, please do. I am not aware of any remaining issue.
- 02:30 PM Feature #3574: finish implementation of temp-table XML support
- Can we close this?
- 02:29 PM Bug #5176 (Closed): Buffer field attribute leak
03/16/2021
- 06:41 PM Feature #1587: implement full support for word indexes
- Ovidiu Maxiniuc wrote:
> The only case that comes into my mind is that there is a single statement and the delimiter... - 06:35 PM Feature #1587: implement full support for word indexes
- I think the missing of @dialect.getDelimiter()@ is just an omission. I am sorry for that. Please add it in @TempTable...
- 06:34 PM Feature #1587: implement full support for word indexes
- Actually, we do append the delimiter to the end of the DROP INDEX statements, which are added to the JDBC batch first...
- 06:26 PM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> I think it is needed. Without it, the statements are not properly terminated and cannot b... - 06:22 PM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> Actually, I remove the subscript at runtime (now silently) at a very early stage of the p... - 06:19 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> > Is there any reason why we do not append @dialect.getDelimiter()@ to the @drop table@ stat... - 06:12 PM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> Is there any reason why we do not append @dialect.getDelimiter()@ to the @drop table@ sta... - 05:40 PM Feature #1587: implement full support for word indexes
- Is there any reason why we do not append @dialect.getDelimiter()@ to the @drop table@ statement in the TempTableHelpe...
- 04:18 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> I can make the conversion changes (including adding the 1688 warning), but I wanted to un... - 03:57 PM Feature #1587: implement full support for word indexes
- > This is a little bit of a grey area, because the converted output we are talking about is a string containing a con...
- 03:52 PM Feature #1587: implement full support for word indexes
- Greg Shah wrote:
> In other words, it is OK in the input 4GL code but we should only ever generate correct/clean Jav... - 03:39 PM Feature #1587: implement full support for word indexes
- > Actually, I do not think that leaving the subscript in the converted code is misleading since the definition of th...
- 03:05 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> If we fix the conversion to drop the extent field dereference, will the current runtime be a... - 02:14 PM Feature #1587: implement full support for word indexes
- If we fix the conversion to drop the extent field dereference, will the current runtime be able to handle the FQL? If...
- 02:00 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Please refresh my memory...for the 1688 warning, is adding the conversion-time warning the o... - 01:06 PM Feature #1587: implement full support for word indexes
- Please refresh my memory...for the 1688 warning, is adding the conversion-time warning the only thing left to be done...
- 04:01 AM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Igor, what is left for this task, besides adding the 1688 error to conversion?
Eric,
I u... - 03:56 AM Feature #1587: implement full support for word indexes
- Igor, what is left for this task, besides adding the 1688 error to conversion?
03/15/2021
- 03:00 PM Feature #1587: implement full support for word indexes
- * Re-worked implicit ordering to be compatible with one for @CONTAINS@ with UDF described in #1587-423
* Refactored ... - 03:29 AM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> One question: for a given record, is there any efficiency/performance difference evaluating ... - 01:35 PM Bug #5176: Buffer field attribute leak
- Eric Faulhaber wrote:
> Code review 3821c/12108:
> Generally, the changes look good to me, though the change was so...
03/14/2021
- 05:27 PM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> Hopefully is not not just a 'late night idea' ))
Well, it seems like an interesting on...
03/12/2021
- 05:36 PM Feature #1587: implement full support for word indexes
- BTW. I've just realized that it is possible to add a 'natural' implicit ordering to the queries using the @CONTAINS@ ...
- 02:27 PM Feature #1587: implement full support for word indexes
- I'm ok with using GUIDs.
- 02:25 PM Feature #1587: implement full support for word indexes
- Greg Shah wrote:
> Do dynamic temp-tables have names?
Yes, they are referenced in subsequent dynamic queries by n... - 02:20 PM Feature #1587: implement full support for word indexes
- Greg Shah wrote:
> Do dynamic temp-tables have names?
If they are in the H2 database they should have at least so... - 02:16 PM Feature #1587: implement full support for word indexes
- Do dynamic temp-tables have names?
- 02:10 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Igor Skornyakov wrote:
> When you say "ensure that data objects created for word tables are... - 02:09 PM Feature #1587: implement full support for word indexes
- Some google searches suggest '@@@' is a valid character in an SQL table name, but not in a 4GL table name. This will ...
- 01:59 PM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> Another question. I need to ensure that data objects created for word tables are unique, ev... - 01:45 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Yes, the dirty database should work properly with extent fields, if we go with the UDF appro... - 01:42 PM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> So I have to add a check for @isDirty@ and do not re-write the query for the 'dirty' data... - 01:32 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Igor Skornyakov wrote:
> > What is 'dirty' database and should I consider such databases fo... - 01:22 PM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> What is 'dirty' database and should I consider such databases for the word tables support?
... - 05:28 AM Feature #1587: implement full support for word indexes
- Another question. I need to ensure that data objects created for word tables are unique, even in the scope of a sessi...
- 05:19 AM Feature #1587: implement full support for word indexes
- Sorry for my ignorance. What is 'dirty' database and should I consider such databases for the word tables support?
T... - 04:17 PM Bug #5056 (Closed): Temp-tables and unique indexes
- 04:00 PM Bug #5056: Temp-tables and unique indexes
- The validation works correctly for temp-tables now. It does not really make sense why OE shows error 132 multiple tim...
- 11:39 AM Bug #5176: Buffer field attribute leak
- Code review 3821c/12108:
Generally, the changes look good to me, though the change was so dense in @TableMapper@, ...
03/11/2021
- 02:41 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Just saw this in the H2 mailing list, FWIW. Noel is one of the main contributors to the proj... - 02:30 PM Feature #1587: implement full support for word indexes
- Just saw this in the H2 mailing list, FWIW. Noel is one of the main contributors to the project. The OP was reporting...
- 07:49 AM Bug #5121: Converted temp table should be using the legacy name instead of the converted name in ...
- Eric Faulhaber wrote:
> Ovidiu, can you think of an efficient way to give access to @SharedVariableManager.lookupTem... - 09:22 PM Bug #5121: Converted temp table should be using the legacy name instead of the converted name in ...
- Roger Borrello wrote:
> Greg Shah wrote:
> > I think the legacy name can be found in @expected@. It is the DMO we ...
03/10/2021
- 04:27 PM Bug #5176 (WIP): Buffer field attribute leak
- I committed the fix in 3821c as revision 12108. The update is a aggregate, it contains fixes for other smaller issues...
- 04:08 PM Bug #5121: Converted temp table should be using the legacy name instead of the converted name in ...
- Eric?
- 03:51 PM Bug #5121: Converted temp table should be using the legacy name instead of the converted name in ...
- Greg Shah wrote:
> I think the legacy name can be found in @expected@. It is the DMO we are searching for, which in... - 01:17 PM Bug #5121: Converted temp table should be using the legacy name instead of the converted name in ...
- I think the legacy name can be found in @expected@. It is the DMO we are searching for, which inherently maps 1-to-1...
- 12:02 PM Bug #5121: Converted temp table should be using the legacy name instead of the converted name in ...
- At this point in processing I don't think we know the legacy name (since we haven't found it).
From @SharedVariabl... - 01:44 PM Bug #4173 (Closed): BUFFER-COPY/BUFFER-COMPARE compares table fields by name rather than data typ...
- 01:29 PM Bug #4173 (WIP): BUFFER-COPY/BUFFER-COMPARE compares table fields by name rather than data type a...
- This can be closed; the premises in #4173 are not accurate as there is a difference in implementation between @bufer-...
- 11:01 AM Bug #4173: BUFFER-COPY/BUFFER-COMPARE compares table fields by name rather than data type and order
- Is this still an open issue?
- 01:42 PM Feature #3958 (Closed): BUFFER-COPY optimization
- 01:22 PM Feature #3958: BUFFER-COPY optimization
- This can be closed; the BUFFER-COPY implementation reached all the optimizations proposed here.
- 11:00 AM Feature #3958: BUFFER-COPY optimization
- What is left to do on this task?
- 01:26 PM Feature #4011: database/persistence layer performance improvements
- We have the idea in #4011-774 implemented as @DmoSignature@.
- 11:01 AM Feature #4011: database/persistence layer performance improvements
- Is #4011-774 complete?
- 01:19 PM Feature #4055: optimize temp-table output parameter copying
- #4055-63 keeps track of the things left to do. The opened ones are still some possible directions, which were explore...
- 10:59 AM Feature #4055: optimize temp-table output parameter copying
- What is left to do on this task?
- 11:21 AM Feature #1587: implement full support for word indexes
- The final logic for implicit sorting appeared to be more tricky than I've described initially. In particular, a situa...
03/09/2021
- 05:04 PM Feature #1587: implement full support for word indexes
- Ovidiu Maxiniuc wrote:
> These two classes are really specialised:
> * the @DDLGeneratorWorker@ is used by static c... - 05:02 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Not as such. The documentation on FWD internals is the least developed of the various books.... - 04:25 PM Feature #1587: implement full support for word indexes
- These two classes are really specialised:
* the @DDLGeneratorWorker@ is used by static conversion to get DDLs for pe... - 03:39 PM Feature #1587: implement full support for word indexes
- Not as such. The documentation on FWD internals is the least developed of the various books.
The code which create... - 03:15 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> OK, please post any questions you may have about the temp-table creation, DDL, etc. Ovidiu o... - 03:07 PM Feature #1587: implement full support for word indexes
- OK, please post any questions you may have about the temp-table creation, DDL, etc. Ovidiu or I will be able to get y...
- 01:57 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Igor Skornyakov wrote:
> Given that support for H2 already is implemented, what is the incr... - 01:02 PM Feature #1587: implement full support for word indexes
- Igor Skornyakov wrote:
> According to #1587-327, you wanted also to implement word tables for @_temp@ db.
Given... - 11:58 AM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Igor, other than finishing the sort implementation (and gauging its performance) and the 168... - 11:50 AM Feature #1587: implement full support for word indexes
- Igor, other than finishing the sort implementation (and gauging its performance) and the 1688 conversion warning, wha...
- 02:48 AM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> W.r.t. the command line options, we don't want to use system properties long term. For any o... - 02:44 PM Bug #5176: Buffer field attribute leak
- I have really thought of this issue during last days.
The @DmoMeta@ carries immutable data, that is: information ... - 11:53 AM Bug #5176: Buffer field attribute leak
- Ovidiu, if every buffer has its own writable version of a certain set of attributes, does it make sense for each buff...
- 11:47 AM Support #5183: database collation
- Some additional research to consider:
*1.* Locales
A "locale":https://en.wikipedia.org/wiki/Locale_(computer_so...
03/08/2021
- 05:54 PM Feature #1587: implement full support for word indexes
- Code review 3821c, rev 12079, 12089:
The logic changes look good to me. I have not tested them.
W.r.t. the comm... - 03:35 AM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Another question/concern: it seems like this approach requires the CTE. It follows that it w... - 03:31 AM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> In the absence of a BY clause, I don't know of any guaranteed ordering of the results of a q... - 11:03 PM Feature #1587: implement full support for word indexes
- Another question/concern: it seems like this approach requires the CTE. It follows that it would disrupt your more pe...
- 10:46 PM Feature #1587: implement full support for word indexes
- This is a very interesting proposal, and I appreciate the creativity in coming up with it.
In the absence of a BY ...
03/07/2021
- 10:57 AM Feature #1587: implement full support for word indexes
- I doubt that it is possible to implement an implicit ordering imposed by @CONTAINS@ which is 100% compatible with 4GL...
- 10:25 AM Feature #1587: implement full support for word indexes
- I've created the following test to analyze the implicit sorting imposed by @CONTAINS@ (see @testcases/words/words-ord...
03/06/2021
- 02:41 PM Feature #1587: implement full support for word indexes
- The problem with ...
- 08:21 AM Feature #1587: implement full support for word indexes
- Added CNF ordering (see #1576-386).
Committed to 3821c/12089 - 05:10 AM Feature #1587: implement full support for word indexes
- I understand that the order of terms in the @CONTAINS@@ expression is important for implicit sorting. So I'm going to...
- 11:24 AM Support #5183: database collation
- In the "Internationalize ABL Application" docs there are good details about codepages and all the character configura...
03/05/2021
- 05:17 PM Support #5183: database collation
- The attached program (@cp_coll.p@) was written as a starting point to review the 4GL's collation, inspired by the exi...
- 05:05 PM Support #5183: database collation
- A review of how FWD collates its database data has raised some questions as to whether we are doing it correctly for ...
- 05:00 PM Support #5183 (WIP): database collation
03/04/2021
- 08:02 AM Bug #5176: Buffer field attribute leak
- My understanding is that in the 4GL every temp-table *instance* is a completely separate table. Multiple table insta...
- 08:10 PM Bug #5176: Buffer field attribute leak
- I will try to summarize an ugly issue I am investigating lately.
I was working on a issue related to #3574-11 and ... - 07:07 PM Bug #5176 (Closed): Buffer field attribute leak
- 06:29 AM Feature #1587: implement full support for word indexes
- I've noticed that queries with @CONTAINS@ run faster with CTE when the expression is 'simple' (its CNF contains only ...
03/03/2021
- 03:46 PM Feature #1587: implement full support for word indexes
- Added SQL re-writing using common table expressions (WITH clause).
Committed to 3821c/12077.
For PostgreSQL, it i...
03/02/2021
- 02:41 AM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> Well, it is extra work to add it in @HQLPreprocessor@ and I imagine there probably is some w...
03/01/2021
- 05:59 PM Feature #1587: implement full support for word indexes
- Well, it is extra work to add it in @HQLPreprocessor@ and I imagine there probably is some work to ignore it when gen...
- 05:19 PM Feature #1587: implement full support for word indexes
- Eric Faulhaber wrote:
> The RTRIM is injected for text property references by the @HQLPreprocessor@. It is needed fo... - 05:09 PM Feature #1587: implement full support for word indexes
- The RTRIM is injected for text property references by the @HQLPreprocessor@. It is needed for normal text property eq...
- 04:44 PM Feature #1587: implement full support for word indexes
- I've noticed some strange thing.
Consider the following 4GL code:...
Also available in: Atom