You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@commons.apache.org by "Mike Graham (Jira)" <ji...@apache.org> on 2020/02/14 15:32:00 UTC

[jira] [Commented] (VFS-257) FTP: Add support for MDTM to get more accurate last modified times

    [ https://issues.apache.org/jira/browse/VFS-257?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17037041#comment-17037041 ] 

Mike Graham commented on VFS-257:
---------------------------------

It's about time some support for this got added. :)

I had a requirement whereby all files timestamps needed to be obtained via MDTM command for timezone reasons (timestamps augmented from LIST cmd can not be relied upon – see above desc as to why).

The patch above is quite old, so I started fresh. 

Pull request: [https://github.com/apache/commons-vfs/pull/82]

> FTP: Add support for MDTM to get more accurate last modified times
> ------------------------------------------------------------------
>
>                 Key: VFS-257
>                 URL: https://issues.apache.org/jira/browse/VFS-257
>             Project: Commons VFS
>          Issue Type: Improvement
>    Affects Versions: 1.0
>            Reporter: Andrew Franklin
>            Priority: Major
>         Attachments: VFS-257.patch
>
>          Time Spent: 10m
>  Remaining Estimate: 0h
>
> The current FTPFileObject implementation delegates to the commons-net FtpFile to get the last modified date. Unfortunately this value is parsed from the result of the LIST command, which suffers from a few issues:
> * NET-212 - LMDs aren't persistable beyond a year as commons-net has to guess the year in many cases, based on the current year.
> * Most FTP parser formats are only accurate to the minute, and sometimes only the nearest day
> * Sometimes the FTP server will change the time format for a single file over time, changing the resulting lmd.
> By using the MDTM command you're able to get a last modified date that's accurate to the second (and sometimes even millisecond) and doesn't suffer from the above issues. As I understand it, the command is implemented in commons-net 2.1?



--
This message was sent by Atlassian Jira
(v8.3.4#803005)