[
https://issues.apache.org/jira/browse/COMPRESS-239?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Stefan Bodewig updated COMPRESS-239:
------------------------------------
Fix Version/s: 1.6
> ArchiveStreamFactory cannot create an ArchiveInputStream from any input
> stream that is blocking
> -----------------------------------------------------------------------------------------------
>
> Key: COMPRESS-239
> URL: https://issues.apache.org/jira/browse/COMPRESS-239
> Project: Commons Compress
> Issue Type: Bug
> Components: Archivers
> Affects Versions: 1.5
> Environment: Windows Server 2008r1/r2, Ubuntu 12.10
> Reporter: Jon Neel
> Labels: easyfix
> Fix For: 1.6
>
>
> Encountered while streaming zip data over a network:
> In the createArchiveInputStream method of ArchiveStreamFactory, when the
> provided input stream is read, and it blocks before 12 bytes are available
> for reading, due to the contract of the java.io.InputStream class, the
> archive signature will not be completely read, and an ArchiveException will
> be thrown ("No Archiver found for the stream signature").
> In ZipArchiveInputStream, you have implemented a readFully method, which
> should solve this issue, but since you are checking the length of the
> signature read against the expected length, you are never getting to do that.
> When you try to read the signature, you should be using readFully.
> For reference, here is important part of the contract of
> java.io.InputStream.read():
> This method blocks until [ANY] input data is available, end of file is
> detected, or an exception is thrown.
> If len is zero, then no bytes are read and 0 is returned; otherwise, there is
> an attempt to read _at least one byte_. If no byte is available because the
> stream is at end of file, the value -1 is returned; otherwise, at least one
> byte is read and stored into b.
--
This message was sent by Atlassian JIRA
(v6.1#6144)
