commons-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Gary D. Gregory (Commented) (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (CODEC-130) Base64InputStream.skip skips underlying stream, not output
Date Fri, 16 Mar 2012 20:53:39 GMT

    [ https://issues.apache.org/jira/browse/CODEC-130?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13231614#comment-13231614
] 

Gary D. Gregory commented on CODEC-130:
---------------------------------------

Hm. I am almost sold on this. Nice job to all posters for clarifying the requirements and
implementations.

My question is: what do we want to be skipping? underlying bytes or decoded chars?

InflaterInputStream for example says:

{quote}
     * Skips specified number of bytes of uncompressed data.
     * @param n the number of bytes to skip
     * @return the actual number of bytes skipped.
{quote}

If we do that on a B64 and B32 stream, will skipping underlying bytes risk of messing up further
decoding? Can you prove that it will not?

:)

                
> Base64InputStream.skip skips underlying stream, not output
> ----------------------------------------------------------
>
>                 Key: CODEC-130
>                 URL: https://issues.apache.org/jira/browse/CODEC-130
>             Project: Commons Codec
>          Issue Type: Bug
>    Affects Versions: 1.5
>            Reporter: James Pickering
>            Priority: Minor
>         Attachments: base64snippet.java
>
>
> Base64InputStream.skip() skips within underlying stream, leading to unexpected behaviour.
> The following code will reproduce the issue:
> @Test
> public void testSkip() throws Throwable {
>     InputStream ins =
>             new ByteArrayInputStream("AAAA////".getBytes("ISO-8859-1"));//should decode
to {0, 0, 0, 255, 255, 255}
>     Base64InputStream instance = new Base64InputStream(ins);
>     assertEquals(3L, instance.skip(3L)); //should skip 3 decoded characters, or 4 encoded
characters
>     assertEquals(255, instance.read()); //Currently returns 3, as it is decoding "A/",
not "//" 
> }
> The following code, if added to Base64InputStream, or (BaseNCodecInputStream in the dev
build) would resolve the issue:
> @Override
> public long skip(long n) throws IOException {
>     //delegate to read()
>     long bytesRead = 0;
>     while ((bytesRead < n) && (read() != -1)) {
>         bytesRead++;
>     }
>     return bytesRead;
> }
> More efficient code may be possible.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira

        

Mime
View raw message