On Tue, Dec 23, 2003 at 03:16:09PM +0800, Autrijus Tang wrote: } Seeing that the eval-based Bleach was causing troubles, and support } for B::Bytecode is not satisfactory (especially for 5.6.x users), } I hacked up yet another filter today: "PAR::Filter::Obfuscate". } } Ironically, it uses the B::Deobfuscate CPAN module, to mangle the } names of all variables in the program, as well as stripping away } POD sections and comments.
That's interesting and probably not the best use of B::Deobfuscate. I'd just finished writing B::Source and was going to use it to include Pod, comments and the original 'obfuscated' source code along-side the deparse output. I had a B::Obfuscate once which produced Stunnix-like output but never published it. I suppose I could publish it anyway. It'd make a nicer backend for this purpose than something intended for deobfuscation since that's likely to get even friendlier over time. } Apologies to Joshua, author of B::Deobfuscate (Cc'ed), for abusing } his work in the totally opposite direction. :-) It makes a nice early present. I grinned, thanks! Josh
