/sqlite3cc

To get this branch, use:
bzr branch http://bzr.ed.am/sqlite3cc
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
- add columns() to row that returns a boost::tuple of various types so multple columns can be fetched at once
	- look in to using BOOST_PP_ITERATE macro
- make basic_statement and database keep a shared pointer to the database handle so the classes can be made copyable. Also:
	- the wrappers around the handle can clean them up after use
	- the actual wrappers wround the handles can be made in sqlite::detail
	- this will also make the implementation od rows (to get round the forced non-dependency of rows on querys) a little easier to swallow.
- committing a transaction during a query (i.e., when sqlite3_step() has returned SQLITE_ROW) causes an error. To counter this:
	- need to check it's not fixed in latest sqlite; write test program
	- calling query.reset() before the commit fixes the issue
	- will need to keep a list of querys that need resetting in the database  :o(
- turn on extended errcodes in open() and handle them in sqlite_error
- use sqlite3_db_mutex() to provide extended error information during sqlite_error construction - see sqlite::query::step() for example
- expand sqlite_error - perhaps use boost::system_error (see boost/asio/error.hpp for an example of extending system_error)
- see if we can #include "sqlite.h" in to a namespace.
	Pros:
		we better encapsulate the library
		we can reuse "sqlite3" as a namespace
	Cons:
		makes access to real sqlite stuff awkward to sqlite3cc users, but does this matter? they can't access database._handle anyway!
		potential incompatibility when linking to libraries that also link against sqlite
- fix step() inconsistency - query::step() returns a row, whereas basic_statement::step() and command::step() return an int return code
- query::prepare() isn't being called during construction (form
	basic_statement's constructor)