[jira] [Commented] (IO-483) getPrefixLength return -1 if unix file contains colon

["Bruno P. Kinoshita (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/IO-483?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15125914#comment-15125914 ] 

Bruno P. Kinoshita commented on IO-483:
---------------------------------------

Looks like FilenameUtils#getPrefixLength is finding the colon character in the second position, and then it gets the previous character (around line 660 and 666 r1727892) and checks the following:

{code}
ch0 = Character.toUpperCase(ch0);
if (ch0 >= 'A' && ch0 <= 'Z') {
    if (len == 2 || isSeparator(filename.charAt(2)) == false) {
        return 2;
    }
    return 3;
}
return NOT_FOUND;
{code}

The ch0 >= 'A' && ch0 <= 'Z' is evaluated to false, since ch0 is '/'.

Maybe this issue could be fixed by updating the Javadoc to state that one of the valid prefixes is a case insensitive character, from a to z, followed by a colon?


> getPrefixLength return -1 if unix file contains colon
> -----------------------------------------------------
>
>                 Key: IO-483
>                 URL: https://issues.apache.org/jira/browse/IO-483
>             Project: Commons IO
>          Issue Type: Bug
>          Components: Utilities
>    Affects Versions: 2.4
>            Reporter: Tristan Lins
>
> A simple
> {code:java}
> int prefixLength = FilenameUtils.getPrefixLength("/:foo/bar");
> {code}
> return {{-1}}, but {{0}} was expected. The path is absolutely valid for unix systems.



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