(Sorry for the broken formatting. I am forced to use a bad email client on this account.)
> 1. It has failures on builds without OpenSSL. Hi, I'm not sure, does that cover your patch as well, please? It doesn't seem to, but i might overlook something. It is included, with an #ifdef PODOFO_HAVE_OPENSSL in EncryptTest.cpp. I'd prefer to have the verbosity either kept enabled or being able to enable it in runtime (or better disable it, aka have the default to have it enabled). The output is important when the test fails. Having It boils down to preference, which I’m sure for both of us is based on experience. My primary goal is to quickly see if something fails, and if and when it does, to be able to find the relevant information quickly. The level of verbosity in the test outputs makes both points impossible. Personally, I’ve always found it an improvement (for myself and others working on the code) to selectively opt into more verbose output. I’ll look into making this choice a runtime flag, when I find time for it. Cheers, Christopher The MathWorks GmbH | Friedlandstr.18 | 52064 Aachen | District Court Aachen | HRB 8082 | Managing Directors: Bertrand Dissler, Steven D. Barbo, Jeanne O’Keefe From: zyx <z...@gmx.us> Sent: Sunday, July 18, 2021 10:54 To: podofo-users@lists.sourceforge.net Subject: Re: [Podofo-users] Patch proposal: working & mostly silent unittests On Fri, 2021-05-21 at 08:40 +0200, Christopher Creutzig wrote: > test/unit/podofo-test currently has at least two properties I regard > as problems: > > 1. It has failures on builds without OpenSSL. Hi, I'm not sure, does that cover your patch as well, please? It doesn't seem to, but i might overlook something. > 2. It is very verbose. I’ve counted over 12,500 lines of output. > > Additionally, some test points do not use the CPPUNIT_ASSERT macros, > but rather fprintf(stderr,…) and exceptions. That makes the test > summary lack information. Good catch. > The attached patch (against r2033) changes these issues. The > verbosity is still in there and can be activated when needed by > uncommenting a #define. I'd prefer to have the verbosity either kept enabled or being able to enable it in runtime (or better disable it, aka have the default to have it enabled). The output is important when the test fails. Having the tests as installed test you do not have a chance to get verbose output for easier debugging. It also doesn't make much sense to me to use debug output, but not logging. By the way, at the end of the patch, there's "logging logging" text. You might use the debugging instead of the printf() calls (like the PoDoFo::PdfError::DebugMessage()). Bye, zyx P.S.: it would be nice to have the --selftest enabled by default too, I find unfortunate to not forget to add it, though others might have other opinion. _______________________________________________ Podofo-users mailing list Podofo-users@lists.sourceforge.net<mailto:Podofo-users@lists.sourceforge.net> https://lists.sourceforge.net/lists/listinfo/podofo-users<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
