commons-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Trejkaz (JIRA)" <>
Subject [jira] Commented: (IO-205) FileUtils.forceMkdir Javadoc is unclear
Date Fri, 15 May 2009 03:51:45 GMT


Trejkaz commented on IO-205:

That may be true also.  The Java docs say it's impossible for both to be false, but I think
that's also part of the same oversight as assuming they can't both be true.

> FileUtils.forceMkdir Javadoc is unclear
> ---------------------------------------
>                 Key: IO-205
>                 URL:
>             Project: Commons IO
>          Issue Type: Bug
>            Reporter: Trejkaz
> The Javadoc for FileUtils.forceMkdir doesn't say what is supposed to happen if a directory
already exists with the same path.
> It does (misleadingly) say that if a file already exists, it fails.  I interpreted this
to mean that a directory would also fail, as it said *file* and not *normal file*.  However
in reality, if the directory already exists, the method passes.
> For clarity it should say that it succeeds if the directory already exists, and should
probably also explicitly say "normal file" instead of just "file", as all directories are
> Also, internally, I notice it is checking for File.isFile() whereas our existing method
for implementing the same thing is checking !File.isDirectory().  On some filesystems, a file
can be both a directory and a normal file, and in this situation I would expect the method
to pass, whereas with Commons' implementation it would fail.  Maybe this is a trip-up because
Sun themselves documented "normal file" as meaning "something which is not a directory", which
is going to break the moment someone runs something on a more exotic filesystem. :-(

This message is automatically generated by JIRA.
You can reply to this email to add a comment to the issue online.

View raw message