[
https://issues.apache.org/jira/browse/HTTPCORE-333?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13573749#comment-13573749
]
Anton Khitrenovich commented on HTTPCORE-333:
---------------------------------------------
Oleg,
As you can see from the history of HTTPCORE-190 (the issue this one was cloned
from), it was created as HTTPCLIENT-829 with "affected version" set to "3.1
Final". Later it was moved to HttpCore project, which starts from version 4.0,
so the original version was lost. The bug was fixed for 4.x, but it still
exists in 3.x versions.
The patch is not mine, but was created for HttpClient 3.1 by Adam Bryzak in
2009 and attached to HTTPCLIENT-829 issue. Integrating the patch to 3.x branch
will not help unless the official version of HttpClient 3.1.x with this patch
inside will be available through standard Maven repositories. Once I have to
build my own patched version of HttpClient, it does really matter whether the
original source was pre-patched or the patch was applied later. And I'm not the
only one - people that use Axis2 encounter this bug [1,2,3] and patched
versions already appear in some strange places on the net [4].
Regards,
Anton.
[1] https://wso2.org/jira/browse/ESBJAVA-1764
[2]
http://stackoverflow.com/questions/8846833/old-version-of-commons-httpclient-gives-ioexception-in-the-axis2-web-service-cli
[3]
http://wso2-oxygen-tank.10903.n7.nabble.com/IOException-on-the-MTOM-web-service-client-side-td27873.html
[4]
http://help.retsiq.com/discussions/problems/7-search-response-parsing-failed-crlf-expected-at-end-of-chunk-4998
> ChunkedInputStream does not handle sockets with timeouts - Client 3.1
> ---------------------------------------------------------------------
>
> Key: HTTPCORE-333
> URL: https://issues.apache.org/jira/browse/HTTPCORE-333
> Project: HttpComponents HttpCore
> Issue Type: Bug
> Reporter: Anton Khitrenovich
> Priority: Minor
> Attachments: HttpClient_829.java, HTTPCLIENT-829.patch
>
>
> ChunkedInputStream does not handle socket timeouts when reading data. If a
> socket timeout is received after finishing reading a chunk (including it's
> CRLF combination) and before reading the next chunk's length, the next call
> to read will attempt to re-read the CRLF, causing an IOException with the
> message "CRLF expected at end of chunk" to be thrown. I have only tested this
> with 3.0 but it doesn't appear to be fixed in 3.1 or any 4.0 from looking at
> the code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
