Thanks for code. Maybe you interested in proposed improvements: #10
Comments
|
Thanks igor, would you be able to submit a pull request? |
|
Hi ! For me it was learning task.
You free use the code as you want (MIT licence) Please see attachment Regards, On Tue, May 28, 2013 at 12:49 AM, Eric Butler notifications@git.luolix.topwrote:
|
|
Hi @igor--, your attachment appears to be missing? |
|
Opps, In attachment was sources in zip file ! Seems the mail have some filters Or maybe really submit it as clone in github ? Igor On Tue, May 28, 2013 at 8:09 PM, Eric Butler notifications@git.luolix.topwrote:
|
|
a pull request is definitely easiest for me, cheers. On Tue, May 28, 2013 at 10:57 AM, igor-- notifications@github.com wrote:
|
|
Hi Eric ! I create new project and commit the sources, please see: https://github.com/igor--/tinywebsocket Regards, On Tue, May 28, 2013 at 9:01 PM, Eric Butler notifications@git.luolix.topwrote:
|
|
Thanks! I'll look into integrating your work. I had considered removing the Android-specific code, I like the name "tinywebsockets". |
|
OK On Tue, May 28, 2013 at 9:54 PM, Eric Butler notifications@git.luolix.topwrote:
|
no need HappyDataInputStream extends DataInputStream,
if you use DataInputStream readFully method: stream.readFully(...) to read from input stream exactly some number
of bytes (not less as it can be if you use read() )
you use double arithmetic in the code (seems it's java script way But in Java we can work with bits fields.
So instead:
frame[1] = (byte) (masked | 127);
frame[2] = (byte) (((int) Math.floor(length / Math.pow(2, 56))) & BYTE);
frame[3] = (byte) (((int) Math.floor(length / Math.pow(2, 48))) & BYTE);
frame[4] = (byte) (((int) Math.floor(length / Math.pow(2, 40))) & BYTE);
frame[5] = (byte) (((int) Math.floor(length / Math.pow(2, 32))) & BYTE);
frame[6] = (byte) (((int) Math.floor(length / Math.pow(2, 24))) & BYTE);
frame[7] = (byte) (((int) Math.floor(length / Math.pow(2, 16))) & BYTE);
frame[8] = (byte) (((int) Math.floor(length / Math.pow(2, 8))) & BYTE);
frame[9] = (byte) (length & BYTE);
can be used:
frame[1] = (byte) (masked | 127);
//frame[2] = (byte) (0);
//frame[3] = (byte) (0);
//frame[4] = (byte) (0);
//frame[5] = (byte) (0);
frame[6] = (byte) ((length >> 24) & 0xFF);
frame[7] = (byte) ((length >> 16) & 0xFF);
frame[8] = (byte) ((length >> 8) & 0xFF);
frame[9] = (byte) (length & 0xFF);
variable 'length' has int type, so contains only 4 bytes, so
frame[2]..frame[5] will be always 0.
(your code ported from java script and wrote as length can contains 8 bytes)
you can use DataInputStream standard function to read 2 and 8 bytes integer, likes:
length = stream.readUnsignedShort(); // read 2 bytes length
long length8 = stream.readLong(); // read 8 bytes length
if( length8 > Integer.MAX_VALUE )
throw new IOException("too big frame length");
length = (int)length8;
instead of your method: byteArrayToLong()
The text was updated successfully, but these errors were encountered: