[EMAIL PROTECTED] wrote: > Hi. > > I have just merged the new-stm branch into the trunk. It contains the > implementation for the new stream architecture documented in > http://www.gnupdf.org/Lib:Architecture/Base_Layer/Stream_Module > > I have been marked the tasks for the re/writing of the filters. The > API that should be implemented by each single filter is documented in > the stream module architecture page (see above). > >
Great news :-) I'm having a brief look on the new API and (may caused by beer) I can't understand why the filter has to install its state in a hash map. I can understand that having parameters in a hash map provides an homogeneous and dynamic way of installing the filters that decreases coupling, but as the state should be private data, I can't see any advantage of having it stored in a hash map. Am I missing something? Thanks, JP PS: I will be a bit more busy this week but will try to get at least one or two filters ported.
