lp:~fpstovall/nrtb/bk2thefuture

Branch merges

Propose for merging

No branches dependent on this one.

Merged into lp:nrtb at revision 15

Rick Stovall: Approve (code merge) on 2013-08-11

Diff: 7286 lines (+3314/-2762)

42 files modified

cpp/common/Makefile (+7/-8)
cpp/common/abs_queue/Makefile (+33/-0)
cpp/common/abs_queue/abs_queue.h (+187/-0)
cpp/common/abs_queue/abs_queue_test.cpp (+120/-0)
cpp/common/circular_queue/Makefile (+4/-2)
cpp/common/circular_queue/circular_queue.h (+23/-128)
cpp/common/circular_queue/circular_queue_test.cpp (+47/-81)
cpp/common/common_rl/Makefile (+6/-3)
cpp/common/common_rl/common.cpp (+168/-192)
cpp/common/common_rl/common.h (+29/-33)
cpp/common/common_rl/common_rl_test.cpp (+18/-0)
cpp/common/confreader/Makefile (+5/-3)
cpp/common/confreader/confreader.cpp (+140/-148)
cpp/common/confreader/confreader.h (+234/-233)
cpp/common/confreader/conftest.cpp (+54/-48)
cpp/common/linear_queue/Makefile (+4/-2)
cpp/common/linear_queue/linear_queue.h (+6/-117)
cpp/common/linear_queue/linear_queue_test.cpp (+52/-85)
cpp/common/logger/Makefile (+20/-22)
cpp/common/logger/logger.cpp (+141/-0)
cpp/common/logger/logger.h (+76/-0)
cpp/common/logger/logger_test.cpp (+105/-0)
cpp/common/point/Makefile (+4/-5)
cpp/common/point/common_test.cpp (+65/-65)
cpp/common/point/triad.h (+176/-176)
cpp/common/serializer/Makefile (+2/-2)
cpp/common/serializer/serializer.cpp (+0/-2)
cpp/common/serializer/serializer.h (+3/-3)
cpp/common/singleton/Makefile (+3/-2)
cpp/common/singleton/singleton.h (+64/-60)
cpp/common/singleton/singleton_test.cpp (+8/-7)
cpp/common/sockets/Makefile (+5/-3)
cpp/common/sockets/base_socket.cpp (+565/-598)
cpp/common/sockets/base_socket.h (+472/-528)
cpp/common/sockets/socket_test.cpp (+53/-87)
cpp/common/timer/Makefile (+38/-0)
cpp/common/timer/hires_timer.cpp (+174/-0)
cpp/common/timer/hires_timer.h (+131/-0)
cpp/common/timer/timer_test.cpp (+72/-0)
obsolete/Cpp/common/logger/log_setup.cpp (+0/-43)
obsolete/Cpp/common/logger/log_setup.h (+0/-30)
obsolete/Cpp/common/logger/log_test.cpp (+0/-46)

api

Related source package recipes

No recipes using this branch. (?)

Create packaging recipe

Related snap packages

Related bugs

Related blueprints

53. By Rick Stovall <fpstovall> on 2013-08-09

Added timer back into the common lib build process. No changes were needed. Unit tested and ready to go.

52. By Rick Stovall <fpstovall> on 2013-08-09

Reinstated the global_conf_reader singleton after removing all internal logging from the conf_reader class. Unit tested and ready to use.

51. By Rick Stovall <fpstovall> on 2013-08-09

Completed conf_reader cleanup and revised unit test. Removed the singleton version as problematic, altered logging to use the new logger class. Unit tested and ready to use.

50. By Rick Stovall <fpstovall> on 2013-08-09

Added the following convienence methods to log_recorder just so I could be lazy later:

  void critical(std::string msg);
  void severe(std::string msg);
  void warning(std::string msg);
  void info(std::string msg);
  void trace(std::string msg);

49. By Rick Stovall <fpstovall> on 2013-08-09

Added logger to the common top level Makefile.

48. By Rick Stovall <fpstovall> on 2013-08-09

Logger component for the alpha phase is complete and unit tested. Ready to use.

47. By Rick Stovall <fpstovall> on 2013-08-08

Checkpoint for logger development. This verision of the logger lib compiles. Unit test still needs to be completed.

46. By Rick Stovall <fpstovall> on 2013-08-07

Checkpoint save of the revised logger components. This does not compile clean yet.

45. By Rick Stovall <fpstovall> on 2013-07-18

Starting a complete redesign and rewrite of the logger.

44. By Rick Stovall <fpstovall> on 2013-07-14

Altered the tcp_server_socket_factory to store and return tcp_socket_p (shared pointers to tcp_sockets) instead of just storing the socket internally. This was driven by several considerations:

* It ensures that any tcp_sockets discarded due to queue overflow are automatically closed and cleaned up.
* It greatly simplies the passing of tcp_socket_p to the final consumer
* It allows us to keep the "close on destruct" syntax for the tcp_socket.
* It allows us to to avoid the cumbersome move syntax while still ensuring the socket is cleaned up when we are done with it.

Generally, a shared pointer is a less than optimal solution, but the balance in this case seems to favor it.

Unit tested and ready to go.