Add bindNoCopy methods to allow binding std::string with SQLITE_STATIC

[DouglasHeriot]

Should be safe, as long as you can guarantee the std::string exists while executing the query.

• Added an accessor to Column that returns a std::string, that can handle BLOB or TEXT values that contain null-bytes.
• Also more binding & Column cast support for uint32_t - fixes ambiguous overload errors when using unsigned-integer types.
  Note that I didn't use uint64_t, because unsigned 64-bit integers doesn't fit into SQLite (except for using int64_t and dealing with overflow with custom functions).
• Added a C++11 move constructor to Statement, to allow storing it inside STL containers (eg. vector).
• When binding, don’t pass int/double types as references – inefficient and unnecessary

[SRombauts]

Hi, thanks, I agree with what you wrote and will review the code.

In the mean time, please have a look on the build failures.

Cheers!

[coveralls]

Coverage increased (+16.3%) to 92.288% when pulling a3ea94c on DouglasHeriot:bindNoCopy into 50501a4 on SRombauts:master.

[DouglasHeriot]

Fixed the build errors and cpplint style issues.
Might have a look at writing some more tests, especially covering edge-cases like dealing with null-bytes inside strings.

[SRombauts]

I did a first review of the code, looks great!

If you can add some unit tests it would be great. Else I hope to look into it in a few days.

[SRombauts]

Hi @DouglasHeriot, any news on unit tests for this?

If you want to do some, try to rebase to my master to get the new unit tests, so that you can complete them.

If you don't think you will have the time I can merge this and complete afterward.

Cheers!

[DouglasHeriot]

Sorry, going to be a little while before I get back to this. Deadlines to meet over the next week, and then later I’ll try and come back and look at it.

[SRombauts]

Totally understandable!

I will merge part of your work, and add some unit test.

I will not take the unfinished Statement move (see comments in code: needs a move constructor on Statement::Ptr). It needs some more work, and that would require even more unit tests that I am not able to do for now.

Cheers!

[SRombauts]

Hi @DouglasHeriot, did you see that I managed to add unit tests for all your methods?

Also, I removed the Statement move constructor as I believe it was incorrect.
If you need it, please provide a use case, even better with a use case.

In fact, to better exploit C++11 features we should rewrite the use of Statement::Ptr to wrap it in a std:::shared_ptr (hence the name).

[SRombauts]

Also, thanks a lot again for your work, it's always inspiring to get some help like that!