From: Paul Marquess [mailto:Paul.Marquess@btinternet.com] > > From: Michael G Schwern [mailto:schwern@pobox.com] > > > > > On Tue, Oct 28, 2003 at 10:34:29PM -0000, Paul Marquess wrote: > > > I'm a bit confused though - there are an identical set of tests > > in DB_File > > > (BerkeleyDB inherited a lot of tests from DB_File, including > > these). If this > > > is in 5.8.2-tobe, why hasn't DB_File fallen over? > > > > In the DB_File case, SubDB inherits from DB_File which inherits > > from Exporter. > > It then copies @DB_File::Exporter. This is correct. > > > > In the BerkelyDB case, SubDB inherits from BerkeleyDB::Hash (or Btree or > > Queue or Recno) which does not inherit from Exporter. It then copies > > @BerkeleyDB::Exporter. This is incorrect. It was only working because > > SubDB incidentally inherited import() from UNIVERSAL. > > > > Either SubDB should inherit from both BerkeleyDB and > BerkeleyDB::Salad or > > the BerkeleyDB::Salad should inherit from BerkeleyDB or SubDB > > should inherit > > from both BerkeleyDB and Exporter. I did the latter, since it had the > > smallest impact on your code, but I think since SubDB is exporting > > BerkeleyDB functions it should do one of the two former ones. #2 is > > probably most correct if you expect folks inheriting from > > BerkeleyDB::Salad > > to be able to export BerkeleyDB functions. > > > Thanks for the explanation. I'll sort it out & get a new BerkeleyDB > uploaded. Just built a snapshot of 5.8.2 (@21558) to see if I can reproduce the bug, but it refuses to fail. Has something changed since this bug was reported? PaulThread Previous | Thread Next