[jira] [Updated] (ZOOKEEPER-1716) jute/Utils.fromCSVBuffer cannot parse data returnd by toCSVBuffer

["Charlie Helin (JIRA)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/ZOOKEEPER-1716?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Charlie Helin updated ZOOKEEPER-1716:
-------------------------------------
    Attachment:     (was: ZOOKEEPER-1716.patch)

> jute/Utils.fromCSVBuffer cannot parse data returnd by toCSVBuffer 
> ------------------------------------------------------------------
>
>                 Key: ZOOKEEPER-1716
>                 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-1716
>             Project: ZooKeeper
>          Issue Type: Bug
>          Components: jute
>    Affects Versions: 3.5.0
>            Reporter: Robert Joseph Evans
>         Attachments: ZOOKEEPER-1716.patch, ZOOKEEPER-1716.patch
>
>
> I was trying to use org.apache.zookeeper.server.LogFormatter to analyze the access pattern of a particular application.  As part of this I wanted to get the size of the data that was being written into ZK.
> I ran into an issue where in some cases the hex data had an odd length.  I looked into it and found that the buffer is being written out using Integer.toHexString(barr[idx])
> Looking at the javadoce for toHexString it indicates that it does not pad the bits at all, and will output the twos compliment of the number if it is negative.  I then looked at how the data was being parsed and it assumed that every byte consisted of exactly two characters, which is not true.
> {code}
> Utils.toCSVBuffer(new byte[] {0xff}) returns "#ffffffff"
> Utils.toCSVBuffer(new byte[] {0x01}) returns "#1"
> If I combine those 
> Utils.fromCSVBuffer(Utils.toCSVBuffer(new byte[] {0xff, 0x01, 0xff})) will return {0xff, 0xff, 0xff, 0xff, 0x1f, 0xff, 0xff, 0xff}
> {code}
> I think what we want is something like
> {code}
> static final char[] NIBBLE_TO_HEX = {
>   '0', '1', '2', '3', '4', '5', '6', '7',
>   '8', '9', 'a', 'b', 'c', 'd', 'e', 'f'
> };
> static String toCSVBuffer(byte barr[]) {
>     if (barr == null || barr.length == 0) {
>         return "";
>     }
>     StringBuilder sb = new StringBuilder(barr.length + 1);
>     sb.append('#');
>     for(int idx = 0; idx < barr.length; idx++) {
>         byte b = barr[idx];
>         sb.append(NIBBLE_TO_HEX[b&0x0f]);
>         sb.append(NIBBLE_TO_HEX[(b&0xf0)>>4]);
>     }
>     return sb.toString();
> }
> {code}



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)