Hi Abel, Thank you for that contribution.
I don't think we should introduce so many new types, just for the ignore flag. This is a minor feature, it shouldn't have such a big impact. However, you were right in thinking that this also affects the non-fluent API. But I'd prefer to have a getter/setter on the org.jooq.InsertQuery type, rather than duplicating the whole type. This will keep things DRY and thus maintainable: http://en.wikipedia.org/wiki/DRY In the same way, the fluent API and its implementation shouldn't be duplicated. So here are the expected added artefacts: - Several overloaded Factory.insertIgnoreInto() methods - Getter / Setter on InsertQuery - Boolean flag in InsertQueryImpl Note, should you want to try again, your files seem to be a bit outdated. Before sending them, can you please merge SVN trunk into your changes? Do you want to try again? Or should I implement it? Please let me know Cheers Lukas 2012/4/13 Abel Birya <[email protected]>: > Hey Lukas, > > Thank you for your email. I was however unable to upload the files to > sourceforge. The page seems to be having an endless redirect loop. Find > attached the files that I created. Below is a list of the changes that I > made to already existing files. > > Factory Operations - I added insertIgnoreInto() and insertIgnoreQuery() > methods. > Factory - I implemented the above methods. > > Kindly let me know whether everything is satisfactory or whether I may need > to make adjustments to any of the code. > > Cheers. > > Abel > > ________________________________ > From: "Lukas Eder" <[email protected]> > To: [email protected] > Sent: Thursday, April 12, 2012 3:51:39 PM > Subject: Re: JOOQ INSERT IGNORE > > > Hi Abel > >> Apologies for the late reply... The screen to my laptop has been having >> problems the last 2 days. > > No worries! > >> I had however started on the implementation of the insert ignore >> functionalities. I had planned on doing it in 2 parts ie "insertignore" >> and >> "insertignoreselect". > > I don't think the latter is needed. These syntaxes will already be possible: > > new Factory(...).insertIgnoreInto(...).values(...) > new Factory(...).insertIgnoreInto(...).set(...).set(...) > new Factory(...).insertIgnoreInto(...).select(...) > >> I have been able to implement what I think is the "insertignore" >> functionality although with my laptop having issues I am unable to do any >> testing. Kindly let me send you the files with a list of the changes I >> have >> made so that you can start testing when you have time and maybe give me a >> few pointers on where you feel I need to make adjustments and/or changes >> (if >> any). > > You can send those files to the user group if the change is not too > big. Otherwise, you could attach them to this ticket with your > sourceforge account: > https://sourceforge.net/apps/trac/jooq/ticket/1295 > > Cheers > Lukas > >> >> Regards >> >> >> On Tuesday, April 10, 2012 2:28:47 PM UTC+3, Lukas Eder wrote: >>> >>> Hello Abel, >>> >>> > [...] if I had the right guide I can try to >>> > do an implementation of this functionality. I will use your suggestions >>> > as a >>> > guide and will get back to you as I progress along. >>> >>> OK, very nice. So the new method would probably be called >>> Factory.insertIgnore(...), and it would set a new "ignore" flag in >>> InsertQueryImpl. If you could try to make this work for MySQL, it >>> would be great. I can then take over and simulate this for other >>> dialects. >>> >>> Cheers >>> Lukas >>> >>> 2012/4/10 name256 <[email protected]>: >>> > Hi. >>> > >>> > Many thanks for your quick reply. I am currently working on a project >>> > that I >>> > am sure will benefit from the many features that JOOQ has to offer. >>> > Challenge is the "INSERT IGNORE" implementation is integral to the >>> > success >>> > of this project (I am thinking of changing the implementation from raw >>> > jdbc >>> > to JOOQ) and that is why I felt that if I had the right guide I can try >>> > to >>> > do an implementation of this functionality. I will use your suggestions >>> > as a >>> > guide and will get back to you as I progress along. >>> > >>> > I have also been looking at the implementation of >>> > onDuplicateKeyUpdate() to >>> > see whether I can make some headway on its implementation (no luck so >>> > far). >>> > I shall let you know how I am getting along in the coming days. >>> > >>> > Regards >>> > >>> > Abel >>> > >>> > On Tuesday, April 10, 2012 1:13:51 PM UTC+3, Lukas Eder wrote: >>> >> >>> >> Hello Abel, >>> >> >>> >> Good idea, and thanks for the offer of implementing this. I'll track >>> >> this idea as feature request #1295: >>> >> https://sourceforge.net/apps/trac/jooq/ticket/1295 >>> >> >>> >> For better maintainability, I would prefer if such DSL features >>> >> wouldn't be implemented in dialect-specific factories, but in the >>> >> global Factory with an appropriate org.jooq.Support annotation to >>> >> indicate the dialects supporting this clause. Specifically, MySQL's >>> >> INSERT IGNORE clause could be simulated using the SQL:2003 standard >>> >> MERGE statement in DB2, HSQLDB, Oracle, SQL Server, and Sybase SQL >>> >> Anywhere. >>> >> >>> >> On the other hand, there seems to be no difference between executing >>> >> INSERT IGNORE and INSERT .. ON DUPLICATE KEY UPDATE SET pk = pk, where >>> >> the UPDATE part is effectively optimised away in a simple test by this >>> >> Stack Overflow user: >>> >> >>> >> http://stackoverflow.com/a/7853634/521799 >>> >> >>> >> So, INSERT IGNORE could be rendered "as is" for MySQL, and simulated >>> >> for CUBRID (Which has the ON DUPLICATE KEY UPDATE clause), DB2, >>> >> HSQLDB, Oracle, SQL Server, Sybase SQL Anywhere (which have the MERGE >>> >> statement) >>> >> >>> >> Any other opinions welcome! >>> >> >>> >> Cheers >>> >> Lukas >>> >> >>> >> 2012/4/10 name256 <[email protected]>: >>> >> > Hi, >>> >> > >>> >> > I am a newbie to JOOQ. I was wondering whether someone has been able >>> >> > to >>> >> > implement an implementation of insert ignore for the MYSQLFactory. >>> >> > Kindly >>> >> > let me know. I would't mind a guide to doing this myself as well as >>> >> > it >>> >> > will >>> >> > give me a better insight into JOOQ. >>> >> > >>> >> > Regards. >>> >> > >>> >> > Abel >> >> >> On Tuesday, April 10, 2012 2:28:47 PM UTC+3, Lukas Eder wrote: >>> >>> Hello Abel, >>> >>> > [...] if I had the right guide I can try to >>> > do an implementation of this functionality. I will use your suggestions >>> > as a >>> > guide and will get back to you as I progress along. >>> >>> OK, very nice. So the new method would probably be called >>> Factory.insertIgnore(...), and it would set a new "ignore" flag in >>> InsertQueryImpl. If you could try to make this work for MySQL, it >>> would be great. I can then take over and simulate this for other >>> dialects. >>> >>> Cheers >>> Lukas >>> >>> 2012/4/10 name256 <[email protected]>: >>> > Hi. >>> > >>> > Many thanks for your quick reply. I am currently working on a project >>> > that I >>> > am sure will benefit from the many features that JOOQ has to offer. >>> > Challenge is the "INSERT IGNORE" implementation is integral to the >>> > success >>> > of this project (I am thinking of changing the implementation from raw >>> > jdbc >>> > to JOOQ) and that is why I felt that if I had the right guide I can try >>> > to >>> > do an implementation of this functionality. I will use your suggestions >>> > as a >>> > guide and will get back to you as I progress along. >>> > >>> > I have also been looking at the implementation of >>> > onDuplicateKeyUpdate() to >>> > see whether I can make some headway on its implementation (no luck so >>> > far). >>> > I shall let you know how I am getting along in the coming days. >>> > >>> > Regards >>> > >>> > Abel >>> > >>> > On Tuesday, April 10, 2012 1:13:51 PM UTC+3, Lukas Eder wrote: >>> >> >>> >> Hello Abel, >>> >> >>> >> Good idea, and thanks for the offer of implementing this. I'll track >>> >> this idea as feature request #1295: >>> >> https://sourceforge.net/apps/trac/jooq/ticket/1295 >>> >> >>> >> For better maintainability, I would prefer if such DSL features >>> >> wouldn't be implemented in dialect-specific factories, but in the >>> >> global Factory with an appropriate org.jooq.Support annotation to >>> >> indicate the dialects supporting this clause. Specifically, MySQL's >>> >> INSERT IGNORE clause could be simulated using the SQL:2003 standard >>> >> MERGE statement in DB2, HSQLDB, Oracle, SQL Server, and Sybase SQL >>> >> Anywhere. >>> >> >>> >> On the other hand, there seems to be no difference between executing >>> >> INSERT IGNORE and INSERT .. ON DUPLICATE KEY UPDATE SET pk = pk, where >>> >> the UPDATE part is effectively optimised away in a simple test by this >>> >> Stack Overflow user: >>> >> >>> >> http://stackoverflow.com/a/7853634/521799 >>> >> >>> >> So, INSERT IGNORE could be rendered "as is" for MySQL, and simulated >>> >> for CUBRID (Which has the ON DUPLICATE KEY UPDATE clause), DB2, >>> >> HSQLDB, Oracle, SQL Server, Sybase SQL Anywhere (which have the MERGE >>> >> statement) >>> >> >>> >> Any other opinions welcome! >>> >> >>> >> Cheers >>> >> Lukas >>> >> >>> >> 2012/4/10 name256 <[email protected]>: >>> >> > Hi, >>> >> > >>> >> > I am a newbie to JOOQ. I was wondering whether someone has been able >>> >> > to >>> >> > implement an implementation of insert ignore for the MYSQLFactory. >>> >> > Kindly >>> >> > let me know. I would't mind a guide to doing this myself as well as >>> >> > it >>> >> > will >>> >> > give me a better insight into JOOQ. >>> >> > >>> >> > Regards. >>> >> > >>> >> > Abel >> >> >> On Tuesday, April 10, 2012 2:28:47 PM UTC+3, Lukas Eder wrote: >>> >>> Hello Abel, >>> >>> > [...] if I had the right guide I can try to >>> > do an implementation of this functionality. I will use your suggestions >>> > as a >>> > guide and will get back to you as I progress along. >>> >>> OK, very nice. So the new method would probably be called >>> Factory.insertIgnore(...), and it would set a new "ignore" flag in >>> InsertQueryImpl. If you could try to make this work for MySQL, it >>> would be great. I can then take over and simulate this for other >>> dialects. >>> >>> Cheers >>> Lukas >>> >>> 2012/4/10 name256 <[email protected]>: >>> > Hi. >>> > >>> > Many thanks for your quick reply. I am currently working on a project >>> > that I >>> > am sure will benefit from the many features that JOOQ has to offer. >>> > Challenge is the "INSERT IGNORE" implementation is integral to the >>> > success >>> > of this project (I am thinking of changing the implementation from raw >>> > jdbc >>> > to JOOQ) and that is why I felt that if I had the right guide I can try >>> > to >>> > do an implementation of this functionality. I will use your suggestions >>> > as a >>> > guide and will get back to you as I progress along. >>> > >>> > I have also been looking at the implementation of >>> > onDuplicateKeyUpdate() to >>> > see whether I can make some headway on its implementation (no luck so >>> > far). >>> > I shall let you know how I am getting along in the coming days. >>> > >>> > Regards >>> > >>> > Abel >>> > >>> > On Tuesday, April 10, 2012 1:13:51 PM UTC+3, Lukas Eder wrote: >>> >> >>> >> Hello Abel, >>> >> >>> >> Good idea, and thanks for the offer of implementing this. I'll track >>> >> this idea as feature request #1295: >>> >> https://sourceforge.net/apps/trac/jooq/ticket/1295 >>> >> >>> >> For better maintainability, I would prefer if such DSL features >>> >> wouldn't be implemented in dialect-specific factories, but in the >>> >> global Factory with an appropriate org.jooq.Support annotation to >>> >> indicate the dialects supporting this clause. Specifically, MySQL's >>> >> INSERT IGNORE clause could be simulated using the SQL:2003 standard >>> >> MERGE statement in DB2, HSQLDB, Oracle, SQL Server, and Sybase SQL >>> >> Anywhere. >>> >> >>> >> On the other hand, there seems to be no difference between executing >>> >> INSERT IGNORE and INSERT .. ON DUPLICATE KEY UPDATE SET pk = pk, where >>> >> the UPDATE part is effectively optimised away in a simple test by this >>> >> Stack Overflow user: >>> >> >>> >> http://stackoverflow.com/a/7853634/521799 >>> >> >>> >> So, INSERT IGNORE could be rendered "as is" for MySQL, and simulated >>> >> for CUBRID (Which has the ON DUPLICATE KEY UPDATE clause), DB2, >>> >> HSQLDB, Oracle, SQL Server, Sybase SQL Anywhere (which have the MERGE >>> >> statement) >>> >> >>> >> Any other opinions welcome! >>> >> >>> >> Cheers >>> >> Lukas >>> >> >>> >> 2012/4/10 name256 <[email protected]>: >>> >> > Hi, >>> >> > >>> >> > I am a newbie to JOOQ. I was wondering whether someone has been able >>> >> > to >>> >> > implement an implementation of insert ignore for the MYSQLFactory. >>> >> > Kindly >>> >> > let me know. I would't mind a guide to doing this myself as well as >>> >> > it >>> >> > will >>> >> > give me a better insight into JOOQ. >>> >> > >>> >> > Regards. >>> >> > >>> >> > Abel >> >> >> On Tuesday, April 10, 2012 2:28:47 PM UTC+3, Lukas Eder wrote: >>> >>> Hello Abel, >>> >>> > [...] if I had the right guide I can try to >>> > do an implementation of this functionality. I will use your suggestions >>> > as a >>> > guide and will get back to you as I progress along. >>> >>> OK, very nice. So the new method would probably be called >>> Factory.insertIgnore(...), and it would set a new "ignore" flag in >>> InsertQueryImpl. If you could try to make this work for MySQL, it >>> would be great. I can then take over and simulate this for other >>> dialects. >>> >>> Cheers >>> Lukas >>> >>> 2012/4/10 name256 <[email protected]>: >>> > Hi. >>> > >>> > Many thanks for your quick reply. I am currently working on a project >>> > that I >>> > am sure will benefit from the many features that JOOQ has to offer. >>> > Challenge is the "INSERT IGNORE" implementation is integral to the >>> > success >>> > of this project (I am thinking of changing the implementation from raw >>> > jdbc >>> > to JOOQ) and that is why I felt that if I had the right guide I can try >>> > to >>> > do an implementation of this functionality. I will use your suggestions >>> > as a >>> > guide and will get back to you as I progress along. >>> > >>> > I have also been looking at the implementation of >>> > onDuplicateKeyUpdate() to >>> > see whether I can make some headway on its implementation (no luck so >>> > far). >>> > I shall let you know how I am getting along in the coming days. >>> > >>> > Regards >>> > >>> > Abel >>> > >>> > On Tuesday, April 10, 2012 1:13:51 PM UTC+3, Lukas Eder wrote: >>> >> >>> >> Hello Abel, >>> >> >>> >> Good idea, and thanks for the offer of implementing this. I'll track >>> >> this idea as feature request #1295: >>> >> https://sourceforge.net/apps/trac/jooq/ticket/1295 >>> >> >>> >> For better maintainability, I would prefer if such DSL features >>> >> wouldn't be implemented in dialect-specific factories, but in the >>> >> global Factory with an appropriate org.jooq.Support annotation to >>> >> indicate the dialects supporting this clause. Specifically, MySQL's >>> >> INSERT IGNORE clause could be simulated using the SQL:2003 standard >>> >> MERGE statement in DB2, HSQLDB, Oracle, SQL Server, and Sybase SQL >>> >> Anywhere. >>> >> >>> >> On the other hand, there seems to be no difference between executing >>> >> INSERT IGNORE and INSERT .. ON DUPLICATE KEY UPDATE SET pk = pk, where >>> >> the UPDATE part is effectively optimised away in a simple test by this >>> >> Stack Overflow user: >>> >> >>> >> http://stackoverflow.com/a/7853634/521799 >>> >> >>> >> So, INSERT IGNORE could be rendered "as is" for MySQL, and simulated >>> >> for CUBRID (Which has the ON DUPLICATE KEY UPDATE clause), DB2, >>> >> HSQLDB, Oracle, SQL Server, Sybase SQL Anywhere (which have the MERGE >>> >> statement) >>> >> >>> >> Any other opinions welcome! >>> >> >>> >> Cheers >>> >> Lukas >>> >> >>> >> 2012/4/10 name256 <[email protected]>: >>> >> > Hi, >>> >> > >>> >> > I am a newbie to JOOQ. I was wondering whether someone has been able >>> >> > to >>> >> > implement an implementation of insert ignore for the MYSQLFactory. >>> >> > Kindly >>> >> > let me know. I would't mind a guide to doing this myself as well as >>> >> > it >>> >> > will >>> >> > give me a better insight into JOOQ. >>> >> > >>> >> > Regards. >>> >> > >>> >> > Abel >
