2011/7/20 Chris Ridd <chrisr...@mac.com>: > > On 20 Jul 2011, at 13:36, Graham Barr wrote: > >> >> On Jul 20, 2011, at 07:32 , Francis Swasey wrote: >> >>> >>> >>> On Jul 20, 2011, at 8:26, Chris Ridd <chrisr...@mac.com> wrote: >>> >>>> >>>>> >>>>> >>>>> >>>> >>>> Graham, should _escape be made public? It seems like it would be useful. >>>> Or is manipulating the data structure returned from new the better >>>> approach? >>> >>> Perhaps a flag on the new call that indicates there are no escapes in the >>> string so that the existing \, is not assumed to be a pre-existing escape >>> and the \ gets escaped? >> >> That would not work consistently. consider (attr=()) >> >> Some guessing would have to go one to know that the first ) needs to be >> escaped >> >> I think exporting _escape as escape_ldap_filter is the best approach and >> users should use that as they build their filters > > I'd call it escape_value - or at least something with "value" in the name - > to make it clearer that you shouldn't pass a complete filter string into it. >
The solution seems ok to me. Should I open an issue on the CPAN tracker? Clément.
