umeca74
umeca74
I found a memory leak too! Where the filter is getting document properties (author/title) with `GetPropertyTemp`, the conversion from UTF8 is done using `ToWStr`, whereas I think you should have...
I am talking about the PdfFilter subproject in sumatra source, where newlines are added, then replaced with \r\n what about the memory leaks? nothing huge I admit, just needs to...
another minor glitch: for identification (`IPersist::GetClassID`) all filter objects return the handler e,g, `kEpubFilterHandler`; instead they should return the created object clsid vis `kEpubFilterClsid`
all my comments apply to the PDF filter too, which is included in the distro