Hi, I can reproduce that "critical" message only when having built the test after backing out the patch. A static analysis (done "by hand") also shows that with the patch applied GetPageNode() can't get called when there are no pages recorded in the pages tree except when deleting pages from the pages tree, which isn't done in helloworld-base14, only in PagesTreeTest. I don't have libcppunit installed right now so can't test the latter but I expect the message to not show up also when running that (it doesn't test deleting from empty AFAICS).
So could you please recheck if you applied the patch before building PoDoFo and the test (with the freshly-built PoDoFo)? I have tested with all except PdfPagesTree and PdfPagesTreeCache mocked out (with an old PoDoFo version) because of the security policy for the (lent) computer I'm using, sorry if that isn't enough. I haven't yet had time for PdfFontFlags due to this issue, I'll work on that after this is solved (if you still want it soon). I'm willing to amend the patch for this issue with the "critical" message if it really isn't enough, please refrain from reverting (to be frank, I was terrified you seemed to have considered that, if I read your post right). Also Nenad Novak had posted the message before in this discussion thread that his issue (also with that message, but on Windows) is solved by my patch (URL http://sourceforge.net/p/podofo/mailman/message/34529262/ in case thread order was mangled) and apologised (in a post quoted by him). Best regards, mabri ----- Original Message ----- From: zyx <z...@litepdf.cz> To: podofo-users@lists.sourceforge.net CC: Sent: 20:19 Tuesday, 13 October 2015 Subject: Re: [Podofo-users] unreachable-code and robustness fixes in PdfPagesTree::GetPageNode() On Sun, 2015-10-04 at 15:53 +0000, Matthew Brincke wrote: > You don't rely on outcome of a method call which has invalid > parameters, do you? Hi, maybe I'm missing something here, but I rather ask before committing your patch: If I apply it on top of revision 1686, then the example helloworld-base14 still claims the critical about accessing page 0 when there are 0 pages. Reverting back to revision 1682 avoids the critical on the console, but it might not be the right "fix". The resulting PDF files from the example look the same. That is, if the helloworld-base14 does anything wrong with pages (as the warning suggests it does), then it should be fixed as well. Thus I'd like to ask you to correct it, either by updating the previous patch or by sending a new one. I'll left it up to you. Bonus points if you find more similar issues in the podofo examples/code/.... Or correct me, if I'm wrong. Thanks and bye, zyx -- http://www.litePDF.cz i...@litepdf.cz ------------------------------------------------------------------------------ _______________________________________________ Podofo-users mailing list Podofo-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/podofo-users ------------------------------------------------------------------------------ _______________________________________________ Podofo-users mailing list Podofo-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/podofo-users
