Peter M. Groen / redox

subscribe/psubscribe/unsubscribe/punsubscribe methods that keep track of
subscribed topics to make sure we don't ask hiredis for bad things. It
appears all crashes are eliminated, though no stress testing has been
done.

subscribe/unsubscribe/publish with proper and working callbacks. Guard
added that throws an exception if a non-pubsub command is issued after a
subscribe.

Also removed all std:: prefixes from redox.cpp. Just made the decision
for "using namespace std" there for readability. Not a header, of
course.

Also added another async watcher for breaking the loop, since ev_break
doesn't actually do anything when called outside of an ev_run callback.

utils/logger.[ch]pp gives a logger implementation based on ostringstream
that is pretty nice. I took this from Stack Overflow and added some
things to make syntax nicer. Changed all cout and cerr statements in
Redox to be of the form logger.{debug/info/warning/error/fatal}() <<
stuff. Arguments for non-printed statements are still evaluated. Gives a
performance hit if we actually add statements on every callback and run
speed tests. For most use cases, doesn't matter. For now, not including
such low level output anyway. Would be nice to have a macro to leave
out/include the low-level logs at compile-time.

Implemented limited but useful binary data support without breaking the
API. If the last character of the command is a ", look for the first ",
and everything in between the quotes treat as binary data. Takes care of
setting a key with a binary value. Not useful if there need to be
multiple binary entries in one command, or for binary keys. Also need to
be careful if the last character of the value actually needs to be a
quote, in which case we need to quote the value.

Useful based on what kind of data you need.

Added GTest under test/. Great for unit testing commands. Added some
nice helper functions to make writing tests easy.

Added options to CMakeLists to build the library, tests, and examples
separately, configurable through ccmake. Also default the build type
to RelWithDebInfo, but don't override user settings.

Consolidated user callbacks into one, improved some logic.

Created an atomic_int connect_state that keeps track of not yet
connected, connected, disconnected, and errors. Used to implement good
error behavior when the server is down and you start() a Redox instance,
or when the server goes down in the middle of running commands. Now,
nothing should hang, but should return errors (or throw exceptions) when
the server goes down.

Also added hooks for a user connect/disconnect callback in the
constructor, which is helpful for client programs.

Added an example with three Redox clients in one thread.

Finally getting rid of the busy loop in the event thread. Now, we use
an ev_async watcher that processes commands in the queue, which we
can signal in a thread-safe manner using ev_async_send.

Instead of busy looping event thread, use a condition variable that
is notified every time a command is sent or received, or every 100
milliseconds anyway.

Better than using a laptop, because it can be reproduced by others
much more easily

submit_command_callback

Also plan to add in set<string> and unordered_set<string>. Maybe
queue<string>. Looking at likely removing char* as an option, if it
shows to be about the same speed as string.

Thoroughly cleaned up redox.hpp to expose a minimal public API, and
commented those well. Generally cleaned up a bunch of stuff.

Added wrapper function DEL, which could be useful.

Easy few lines for get and set commands that are easy to use.

Implemented a condition variable that blocks run() until the event
loop is running and Redis is connected. Also added a check to
command() that will throw an exception if you try to add a command
before calling run().

Fixed a couple of bugs found with a test case of 100 parallel
asynchronous clients. As of now, there are no known memory
leaks, segfaults, or deadlocks in Redox.

Ran into a nasty segfault when using 100 parallel asynchronous command
loops, where a command's address would match up with a previous
command's
address. To circumvent the whole issue, Redox keeps a counter of how
many Commands it has created and assigns each Command a unique id in
the order they are created.

Changed all data structures and methods to pass around Command IDs, so
nothing uses void* anymore. I consider this a very good thing, but it
did seem to slow down performance 20-30% because of the extra lookups.
Will see if there's a solution later.

Take that, hours of debugging memory leaks! Did lots of structural
tweaks to more smartly keep track of commands and make sure everything
is evenutally freed from the heap.

Created a pointer Command.rdx (Redox*) to allow commands
to increment Redox.cmd_count, and simplify a couple of
other things.

Did a little bit of moving and renaming context objects,
so the context is always .ctx and commands can always
be called c.

Added Command<nullptr_t> which successfully returns NIL
replies from Redis, with REDOX_OK. Also renamed constants
from REDISX_ -> REDOX_.

Moved a bunch of code from redox.cpp into command.cpp, where
it logically makes more sense.

Move as much as possible into the .cpp file, and clean things up.

Found out that ev_timer watchers have a void* data that allows
currying information to the libev callbacks! That means I could
stick the Command* in there and completely get rid of the global
timer_callbacks map that was needed before.

Additionally, we can move ev_timer into the Command object instead
of separately allocating it on the heap. Simplifies management.

Implemented a synchronous loop speed test based on the latest
and greatest command_blocking(). Happy to see it is around
30k/s, which is 100% network delays.

Several hours of multithreaded debugging to find a nasty extra
ampersand. command_blocking() now returns a Command object with
the value and status accessible through methods.

Added the option to free memory or not for Command objects when
command() is called. If free_memory = true, then the Command
object is freed after the callback or error callback returns.
If free_memory = false, then the user must call cmd->free().

To implement this, we have to pass in a blank function to the
freeObject entry in the hiredis redisReader, because otherwise
it automatically frees memory. This was not cool for the blocking
case.

Implemented error callbacks for command(). Now, either the success
or error callbacks are guaranteed to return for any command. Added
constants for error types.

Many improvements and fixes for memory management. Command objects
keep track of pending replies and delete themselves when completed.

Changed ReplyT value to remove the const and reference requirements.
Now, you do command<ReplyT> and the callback should take a
const ReplyT&, whatever the type is. Makes things more uniform.

Renamed CommandAsync to Command to make things simpler, and
refactored into its own file.

command() now returns a pointer to the created Command object,
which the client can pass into cancel() to stop any delayed
or repeating calls. This is very important for an asynchronous
client.

Uses libev timer watchers to queue commands on a regular
interval. Very fast. Still need to handle deactivation of
timers from the user side.

num_commands_processed(), trarcks using cmd_count and guarded
with queue_guard.

Used to block the main thread until the event loop is stopped by
calling the .stop() method. Implemented using a condition_variable,
and very useful for clients.

Improved examples, added a threaded getter/setter example
and cleaned up the other. Generally refactored code in the
core files and cleaned up the details to make the slickest
possible API. Implemented a to_exit variable as a stop
condition for the event thread.

The event loop (now libev) now runs in a separate detached thread,
which is abstracted away from the user, who only calls a nonblocking
.start() method. This thread loops continously, alternating telling
Redis about asynchronous commands ready to send, and running one
iteration of the event loop, where all pending events are taken
care of. This greatly simplifies the user code.

Additionally, some clever tricker is implemented now to handle
memory management well with the templated command types. The
difficulty comes from the fact that we pass redis a void pointer
only, and must retreive everything from that (so, we can't use
shared_ptr for the whole thing). Thus, we use maps of void pointers
to their templated data structure pointers, where the memory address
of the pointer serves as the key.