Query.bind implementation && behavior about sql2o HOT 6 CLOSED

I haven't used the bind method, but it sounds to me like the behavior is incorrect. If that's the case, I don't think we need a new method, we should just fix the existing one.

Also, what's the benefit of the marker interface in this case over a plain Object?

from sql2o.

Also, what's the benefit of the marker interface in this case over a plain Object?

if (obj instanceof ParamObject){
  // analyze fields and put them as param values
} else {
  // analyze getXXXX methods (bean-style) and put them as param values
}

from sql2o.

The inheritance issue should be fixed. It is also described in issue #55.

When binding objects, we should only bind expected parameters. So if we have 2 parameters in the sql-statement, but 3 properties on the binded object, it should only bind the 2 parameters and ignore the last property. It seems like this behavior has been changed recently from the original implementation of bind(). as @dimzon points out, it will throw exception if this happens today! This should be fixed.

Given the above, I don't think the changes suggested will break backwards-compatibility?

If we add analysis of fields (in addion to getters), we don't need the ParamObject interface. Then it should work with:

executeQuery("select @name, @age", new Object(){
    int age = 31;
    String name = "dimzon!"
})

as well as with:

executeQuery("select @name, @age", myPojo)

from sql2o.

who will do this (I can but if anyone else want feel free to stop me)

from sql2o.

@dimzon - Your on a killing spree :)
Go ahead. I'm really glad that sql2o gets the love it deserves.

BTW, you deserve your name under developers in the pom.xml file (If you would like it there, of cause - no pressure :) You have been a really valuable contributor! Without you and @aldenquimby, sql2o would never have been able to become such an awesome project!

from sql2o.

btw - When fixing an issue like this. I would recommend creating a new branch just needed for the fix.

git branch fix-issue-80
git checkout fix-issue-80

Do your stuff, and check it in with a commentary that mensions the issue number, like this: "ref #80 - comment here".
when you are done:

git push -u origin fix-issue-80    // -u sets upstream to origin/fix-issue-80, so that the next time you push on the same branch, you can just write "git push"

On github, create a pull request from the new branch.

This way you can start on working on other things, and not have to wait for me to merge in your work.

from sql2o.