On Thursday, October 3, 2013 6:51:29 PM UTC-7, GregD wrote:

>
> On Oct 3, 2013, at 9:37 PM, Jeremy Evans <[email protected]<javascript:>> 
> wrote:
>
> On Thursday, October 3, 2013 6:07:02 PM UTC-7, GregD wrote:
>
>> Jeremy- 
>>
>> I tried: 
>>
>> 1) adding getting the meta data of column_type_name to the 
>> process_result_set method in the jdbc adapter: 
>> 2) changed the method process_result_set_convert to pass the 
>> column_type_name to the convert_type_proc 
>> 3) then in my sqlanywhere specific jdbc adapter I override the 
>> convert_type_proc and that will call super. 
>>
>> This worked for all the jdbc data types via the default super and allowed 
>> me to convert "smallint" to a boolean. 
>>
>> The gist is updated with this code:  
>> https://gist.github.com/gditrick/6240781 
>>
>> Do you see a problem with this approach?  I know this makes smallint a 
>> boolean no matter if the user wants a smallint.  But the native one will be 
>> doing that anyway.  Shame SqlAnywhere does not a boolean type.  In our java 
>> code and stored procs, we use either a smallint or a char of 'Y' or 'N'.  I 
>> could switch this to the char approach, but looking at the other adapters, 
>> I like using the smallint.  Thoughts? 
>>
>> With this change and the alter table changes:  The jdbc sqlanywhere 
>> integration test results are 539 tests, 0 failing, only 2 pending. 
>
>
> Personally, I'm fine with the jdbc/sqlanywhere adapter not emulating 
> booleans, and just marking the related boolean tests as pending.  As the 
> underlying jdbc driver doesn't automatically convert smallint to boolean, I 
> think at the very least we would want to make the behavior optional so that 
> the user can choose whether they want to do the conversion.  Otherwise, 
> Sequel would be basically unusable for users that have smallint columns 
> that aren't used to store booleans.
>
>
> Okay, I was trying to get the behavior of the jdbc 1 to be like how I set 
> up the native 1.  But, I can see your point and almost feel like I should 
> make the native one do that instead.  And looking at others, I thought this 
> was the approach. Thoughts?
>

I don't have strong feelings either way.  There's probably not going to be 
many users of it, so as long as the tests pass, I don't think we need to 
worry about now.  We can always make modifications later if necessary.
 

>
>
> It's great that you got the test suite passing cleanly.  I'd definitely 
> like to get this merged before 4.4.0.  Any change you could provide a 
> unified/git diff or a git repository I can pull from?
>
>
> I'm new to this.  So, what is your preference? If I do a git repo, do I 
> fork it first.  I did a clone of you repo, so I've been getting the updates 
> regularly with a pull.
>

If you have a GitHub account, just pushing your branch up to your fork on 
GitHub should be enough.  Alternative, hosting your git repo anywhere I can 
access it should be fine.  Worse cast scenario, providing the output of git 
diff against my master branch should be enough.

I still need to get a local VM with sqlanywhere setup.  I filled out the 
online form on Sybase's website, but they never sent a key for me to 
download the developer version.

Thanks,
Jeremy 

-- 
You received this message because you are subscribed to the Google Groups 
"sequel-talk" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To post to this group, send email to [email protected].
Visit this group at http://groups.google.com/group/sequel-talk.
For more options, visit https://groups.google.com/groups/opt_out.

Reply via email to