TODO

TODO
====

Bson
----
+ implement deprecated types (were left out)
+ convert from/to json
+ tie regex type to type in regex library
+ Read instance for Documents that can read its Show representation

MongoDB
-------
+ When one connection in a pool fails, close all other since they will likely fail too
+ on insert/update: reject keys that start with "$" or "."
+ dereference dbref
+ functions for every command, eg.
	- findAndModify
	- reIndex (http://www.mongodb.org/display/DOCS/Indexes#Indexes-ReIndex)
	- createIndex attributes: background, min, max
	- createIndex Order [Asc, Dec, Geo2d]
	- getIndexInfo
	- logout
	- collectionsInfo
	- databasesInfo
	- getProfileInfo
+ Query attribute: timeout
+ Update If Current (http://www.mongodb.org/display/DOCS/Atomic+Operations)
+ Upon client exit, send killCursors for all open cursors, otherwise server will keep them open for 10 minutes and keep NoCursorTimeout cursors open for hours.
+ Upon cursor finalize (garbage collect) send killCursor even if you have to create a new connection, because server keeps cursors open for 10 minutes (or more).
+ Query option Exhaust
+ Reconnect on replica set primary stepdown, so no exception raised to user
+ Reconnect on query ioerror re-query, so no exception raised to user. Can't be done for writes because it is not safe to re-execute a write.
+ tailable cursor support
  - only close cursor when cursorID is 0
  - have to create loop that sleeps and retries
  - lazy list support
+ GridFS

Tests - none currently

Misc
----
+ javascript DSL
+ update tutorial to match new python one
+ custom types (see python examples)

Questions:
- In Mongo shell, db.foo.totalSize fetches storageSize of each index but does not use it

Notes:
- Remember that in the new version of MongoDB (>= 1.6), "ok" field can be a number (0 or 1) or boolean (False or True). Use 'true1' function defined in Database.MongoDB.Util
- A cursor will die on the server if not accessed (by any connection) within past 10 minutes (unless NoCursorTimeout option set). Accessing a dead (or non-existent) cursor raises a CursorNotFoundFailure.
- Unsafe to shrink pool and close connections because map/reduce temp tables that were created on the connection will get deleted. Note, other connections can access a map/reduce temp table as long as the original connection is still alive. Also, other connections can access cursors created on other connections, even if those die. Cursors will be deleted on server only if idle for more than 10 minutes. Accessing a deleted cursor returns an error.