[RFC] Netx: clean up path sanitization code
Andrew John Hughes
ahughes at redhat.com
Tue Jun 29 09:58:09 PDT 2010
On 29 June 2010 00:47, Omair Majid <omajid at redhat.com> wrote:
> On 06/24/2010 07:37 PM, Andrew John Hughes wrote:
>> On 24 June 2010 17:04, Omair Majid<omajid at redhat.com> wrote:
>>> The attached patch does a little bit of refactoring, moving path
>>> manipulation functions to a common location.
>>> Any comments?
>> Looks good. It's more than just cleanup though, as the behaviour of
>> sanitizeFileName has changed. It now only replaces ILLEGAL_CHARS and
>> not all characters that aren't in the legal set.
> Yes, it is more than cleanup. There were two conflicting file sanitization
> schemes. One (which sanitized path names) used a blacklisting approach and
> only disallowed certain characters. The other (which sanitized file names)
> used a whitelisting approach and disallowed everything except alphanumeric
> characters. I wrote this code. For the sake of consistency, we want both
> sanitization schemes to return the same file name. There is a problem with
> the whitelisting approach: it can not deal with international file names.
> Hence I am making both the functions use the black listing approach. Please
> let me know if there are any issues I may have missed.
Yes, I understood the need for and agree with the change. But you do
need to make sure it's mentioned clearly in the ChangeLog... :-)
>> I'd also make the two arrays private and final.
Free Java Software Engineer
Red Hat, Inc. (http://www.redhat.com)
Support Free Java!
Contribute to GNU Classpath and the OpenJDK
PGP Key: 94EFD9D8 (http://subkeys.pgp.net)
Fingerprint: F8EF F1EA 401E 2E60 15FA 7927 142C 2591 94EF D9D8
More information about the distro-pkg-dev