[jetty-discuss] [ jetty-Bugs-1421265 ] Jetty6 TRUNK Off-by-one error in HttpConnection$OutputWriter

[SourceForge.net <no...@so...>]

Bugs item #1421265, was opened at 2006-02-01 08:43
Message generated for change (Settings changed) made by gregwilkins
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=107322&aid=1421265&group_id=7322

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: None
Group: None
>Status: Closed
Resolution: Fixed
Priority: 5
Submitted By: Christopher Taylor (c_s_taylor)
Assigned to: Nobody/Anonymous (nobody)
Summary: Jetty6 TRUNK Off-by-one error in HttpConnection$OutputWriter

Initial Comment:
There's an off-by-one error in both the
org.mortbay.jetty.HttpConnection$OutputWriter.write
(String...) and OutputWriter.write (char[]...) methods.

Consider a single JSP file that contains a large amount
of data that must be converted all at once.  The call
to _converter.write(s,i0,i-i0) will not include the
final character of that group because the third
parameter is a length and not a final index.

Assume _writeChunk = 6, i = 0, end > 6:

int i0 = i++; // i0 = 0, i = 1
i += _writeChunk / 2 // i0 = 0, i = 4
// ignore i>end since end is larger than i for this example
_converter.write (s,i0,i-i0); // We pass (s,0,4)
_converter.flush(); // not relevant for the bug

So, after this block is done, i (the index variable) is
4, _but_ as we come back to the top of the loop, we
increment i to 5, which means s[4] never got written to
the stream, hence we're "off-by-one".

I've attached a JSP file in UTF-8 format that quickly
illustrates the bug.  Just have Jetty load that JSP
file and note how the last row of full-width characters
doesn't match up with the above rows.

The fix is easy: after flush, decrement i

----------------------------------------------------------------------

Comment By: Greg Wilkins (gregwilkins)
Date: 2006-02-01 10:01

Message:
Logged In: YES 
user_id=44062

thanks- another faulty optimizatons!

I've refactored this code in SVN.

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=107322&aid=1421265&group_id=7322