Trustin Lee
|
e8665ea867
|
Added a convenience constructor to LocalAddress
|
2009-02-13 06:53:22 +00:00 |
|
Trustin Lee
|
395fcbbd29
|
Simplified LocalAddress implementation - no more global reference map that causes memory leak
|
2009-02-13 06:52:04 +00:00 |
|
Trustin Lee
|
1b28c126c5
|
Fixed a problem where DefaultNioSocketChannelConfig allowed incorrect configuration of write buffer watermark values
|
2009-02-13 06:36:30 +00:00 |
|
Trustin Lee
|
65048c930f
|
Made DirectChannelBufferFactory's preallocatedBufferCapacity configurable
|
2009-02-13 06:25:29 +00:00 |
|
Trustin Lee
|
91d7a329d4
|
* Renamed Concurrent*Weak*HashMap to Concurrent*WeakKey*HashMap to avoid confusion
* Added some FIXMEs related with potential memory leak
|
2009-02-13 06:19:10 +00:00 |
|
Trustin Lee
|
8aff5914d9
|
Added IdleState-aware ChannelHandler implementations
|
2009-02-13 06:05:21 +00:00 |
|
Trustin Lee
|
7b721ef1d4
|
Prettier LocalAddress.toString()
|
2009-02-13 04:51:36 +00:00 |
|
Trustin Lee
|
fa8890580e
|
Indentation
|
2009-02-13 04:41:01 +00:00 |
|
Trustin Lee
|
bf6a7db90a
|
Added more convenience constructors to HashedWheelTimer
|
2009-02-13 04:29:56 +00:00 |
|
Trustin Lee
|
789bfd3221
|
Changed the type of second unit timeout from long to int
|
2009-02-13 04:24:29 +00:00 |
|
Trustin Lee
|
11478edb5b
|
Fixed NPE
|
2009-02-13 04:17:36 +00:00 |
|
Trustin Lee
|
b49f8d8353
|
* Renamed the callback methods for Read/WriteTimeoutHandler
* More convenient exception handling
|
2009-02-12 15:41:58 +00:00 |
|
Trustin Lee
|
b06e33f5ee
|
Changed the default time unit to SECONDS
|
2009-02-12 15:33:28 +00:00 |
|
Trustin Lee
|
7f16fa352a
|
Removed unnecessary variable in IdleStateHandler
|
2009-02-12 15:31:39 +00:00 |
|
Trustin Lee
|
2c8f669030
|
Added IdleState.ALL_IDLE - was able to simplify even more
|
2009-02-12 15:29:35 +00:00 |
|
Trustin Lee
|
cbd2dec0fd
|
* Reverted back recent changes
* Renamed IdlenessHandler to IdleStateHandler
* Renamed IdlenessEvent to IdleStateEvent
* Added IdleState enum
|
2009-02-12 15:21:25 +00:00 |
|
Trustin Lee
|
2c3e7565fa
|
Implemented both idleness event roughly (needs some fix)
|
2009-02-12 11:03:10 +00:00 |
|
Trustin Lee
|
ed69d8bc15
|
Added another missing class
|
2009-02-12 10:51:38 +00:00 |
|
Trustin Lee
|
eab8623f4a
|
Added the missing class
|
2009-02-12 10:51:06 +00:00 |
|
Trustin Lee
|
0e5f798889
|
* Removed the getter methods in IdlenedssEvent and replaced them with subtypes
** Added ReaderIdlenessEvent and WriterIdlenessEvent
* Added protected methods which enables a user to override what to do on idleness / timeout
|
2009-02-12 10:50:43 +00:00 |
|
Trustin Lee
|
59b67420bf
|
Prettier output
|
2009-02-12 07:44:51 +00:00 |
|
Trustin Lee
|
2ebce4c987
|
Better keep-alive behavior
|
2009-02-12 07:39:20 +00:00 |
|
Trustin Lee
|
431151b8e5
|
Fixed issue: NETTY-107 - HttpMessageDecoder can not handle the content with no 'Content-Length' header.
|
2009-02-12 07:32:53 +00:00 |
|
Trustin Lee
|
2d682dc2a3
|
Fixed NETTY-107 HttpMessageDecoder can not handle the content with no 'Content-Length' header.
* Added HttpMessage.getContentLength(int defaultValue)
* Renamed state enum values
|
2009-02-12 07:17:29 +00:00 |
|
Trustin Lee
|
5faa7e9ea8
|
Removed unnecessary boxing / unboxing
|
2009-02-12 07:09:51 +00:00 |
|
Trustin Lee
|
e787378024
|
US-ASCII -> ASCII
|
2009-02-12 07:03:35 +00:00 |
|
Trustin Lee
|
912c66b931
|
Forgot to encode the last chunk correctly
|
2009-02-12 07:01:26 +00:00 |
|
Trustin Lee
|
cf1b915340
|
Improved the HttpServer example to demonstrate chunked encoding support
|
2009-02-12 06:44:52 +00:00 |
|
Trustin Lee
|
2a555082bb
|
Implemented chunked encoding support for the HTTP encoder
|
2009-02-12 06:37:22 +00:00 |
|
Trustin Lee
|
602c99c532
|
Fixed a problem where the last chunk is decoded too early
|
2009-02-12 06:27:17 +00:00 |
|
Trustin Lee
|
ebe77cef21
|
Fixed a 'shouldn't reach here' error
|
2009-02-12 06:26:05 +00:00 |
|
Trustin Lee
|
5ee6da3892
|
Fixed a bug where HttpRequest message is decoded twice
|
2009-02-12 06:25:13 +00:00 |
|
Trustin Lee
|
2ac8d11a34
|
Fixed a problem where DefaultHttpChunk.isLast returns a wrong value
|
2009-02-12 06:22:06 +00:00 |
|
Trustin Lee
|
44d56d2c1a
|
Added HttpChunk message for users who want to receive the content chunk by chunk
|
2009-02-12 06:09:29 +00:00 |
|
Trustin Lee
|
b6c992de8b
|
More optimistic initial line buffer size
|
2009-02-12 05:49:19 +00:00 |
|
Trustin Lee
|
a4fe52559b
|
More robust CR/LF handling in chunked encoding
|
2009-02-12 05:48:25 +00:00 |
|
Trustin Lee
|
9125ff5616
|
Fixed more issues related with chunked encoding
|
2009-02-12 05:41:22 +00:00 |
|
Trustin Lee
|
fb74832dc8
|
* Removed the usage of unnecessary state variable
* Renamed READ_CRLF to READ_END_OF_CHUNK
|
2009-02-12 05:28:55 +00:00 |
|
Trustin Lee
|
66423e0c1d
|
More robust chunk size parser
|
2009-02-12 05:23:39 +00:00 |
|
Trustin Lee
|
44498e067f
|
* Improved the detection of chunked encoding
* Fixed broken chunk length parser
|
2009-02-12 05:16:38 +00:00 |
|
Trustin Lee
|
01b0beab50
|
Typo
|
2009-02-12 05:14:16 +00:00 |
|
Trustin Lee
|
1243baa05b
|
More robust HTTP message decoding
|
2009-02-12 05:10:25 +00:00 |
|
Trustin Lee
|
9cdc4a959e
|
Better white space handling in HTTP
|
2009-02-12 05:02:22 +00:00 |
|
Trustin Lee
|
0198da7b6c
|
Related issue: NETTY-116 HTTP issues including chunked request / response patch
* Applied Dave's patch for QueryStringDecoder
|
2009-02-12 04:39:17 +00:00 |
|
Trustin Lee
|
32738273e0
|
Related issue: NETTY-116 HTTP issues including chunked request / response patch
* Applied Dave's patch for HttpMessageDecoder
|
2009-02-12 04:37:48 +00:00 |
|
Trustin Lee
|
334ea6f6e6
|
Now that there are two MessageEvent implementation, there's no point of having Channels.messageEvent() - a user could just create a new instance by calling the respective implementation's constructor
|
2009-02-11 09:56:28 +00:00 |
|
Trustin Lee
|
2b5a00e8b8
|
Removed unused code block
|
2009-02-11 09:53:14 +00:00 |
|
Trustin Lee
|
73cb446310
|
Removed the usage of Channels.messageEvent() which is obsolete
|
2009-02-11 09:52:36 +00:00 |
|
Trustin Lee
|
2030411759
|
Reverted r825 which doesn't make much sense
|
2009-02-11 09:47:53 +00:00 |
|
Trustin Lee
|
84aa72a70f
|
Made sure all timeout and idle time can be disabled with the non-positive constructor parameters
|
2009-02-11 09:38:24 +00:00 |
|