Hello all,

 

Currently  SftpFileNameParser extends URLFileNameParser. When parsing file
name, later tries to extract the query string, stripping part of filename
starting with the first '?' character.

 

If file name contains non-ASCII characters (by any reason, i.e. the file was
created by FTP client with invalid control encoding setting), the results
will be quite unexpected. For instance, if file name is "???.txt", resulting
FileName will be equal to the parent' one, so that list of '/foo/bar' folder
children will contain '/foo/bar' entry, thus leading to infinite recursion.

 

The question is why URLFileNameParser is subclassed here. Any ideas?

 

Thanks,

 Kirill

Reply via email to