Mailing-List: contact commons-dev-help@jakarta.apache.org; run by ezmlm
Precedence: bulk
Reply-To: "Jakarta Commons Developers List" <commons-dev@jakarta.apache.org>
Date: 13 Sep 2002 08:18:49 -0000
Message-ID: <20020913081849.2586.qmail@nagoya.betaversion.org>
From: bugzilla@apache.org
To: commons-dev@jakarta.apache.org
Cc: 
Subject: DO NOT REPLY [Bug 12607] New:  -
    ChunkedInputStream broken (2 bugs + fixes, 1 suggestion)

DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG 
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://nagoya.apache.org/bugzilla/show_bug.cgi?id=12607>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND 
INSERTED IN THE BUG DATABASE.

http://nagoya.apache.org/bugzilla/show_bug.cgi?id=12607

ChunkedInputStream broken (2 bugs + fixes, 1 suggestion)

           Summary: ChunkedInputStream broken (2 bugs + fixes, 1 suggestion)
           Product: Commons
           Version: Nightly Builds
          Platform: PC
        OS/Version: Windows NT/2K
            Status: NEW
          Severity: Major
          Priority: Other
         Component: HttpClient
        AssignedTo: commons-dev@jakarta.apache.org
        ReportedBy: erik.van-der-goot@jrc.it


Bug 1.

In the     

read(byte[] b, int off, int len)  method

of ChunkedInputStream, the number of bytes to read from the underlying 
InputStream is calculated wrongly. In the code this is done by

len = Math.min(len, chunkSize);

This could (and will) cause the server (also Apache) to indeed serve that 
number of bytes (let's say chunkSize), but it may be that we already had a 
number of bytes on the first read. The result is that the input is now NOT 
positioned on the end of a chunk and the rest of the reader fails because it 
cannot find CRLF or a valid chunksize.

Proposed fix (works, tested)
len = Math.min(len, chunkSize-pos);

Bug 2.

In the calculation of the chunkSize (method getChunkSizeFromInputStream) the 
conversion to int is done by calling             

result = Integer.parseInt(dataString, 16);

This is not robust and causes the occasional crash. The fix is simple and in 
fact implements what is done when the chunkSize is commented (see lines in code 
above)

result = Integer.parseInt(dataString.trim(), 16);

Tested and works.


Suggestion:
Same routine, input state machine. Perhaps just being pedantic..change while 
loop to:

        while (state != 2) {
            int b = in.read();
            if (b == -1) throw new IOException("chunked stream ended 
unexpectedly");
            switch (state) {
                case 0:
                    if (b == '\r')
                      state = 1;
                    else
                      baos.write(b);
                    break;
                case 1:
                    if (b == '\n')
                      state = 2;
                    else{
                     // this was not CRLF, so now write '\r' + this char
                      baos.write('\r');
                      baos.write(b);
                      state = 0;
                    }
                    break;
                default: throw new RuntimeException("assertion failed");
            }
        }

--
To unsubscribe, e-mail:   <mailto:commons-dev-unsubscribe@jakarta.apache.org>
For additional commands, e-mail: <mailto:commons-dev-help@jakarta.apache.org>