Re: [Quantlib-users] including problem!

[Luigi B. <lui...@gm...>]

Hello Donghee,
    hard to say without seeing exactly how you modified the library files.
My suggestion is to use the QuantLib sources unchanged (because they're
known to build correctly) and to either qualify types with the QuantLib
namespace in your application or use "using namespace QuantLib".  This will
also help you when you upgrade the version of QuantLib you're using, since
you won't have to modify the namespace all over again.

Hope this helps,
    Luigi


On Tue, May 16, 2023 at 1:57 AM Donghee Shim <dh7...@gm...> wrote:

> Dear all of experts!
>
> My previous colleague had developed some programs for evaluating
> non-standard financial products (not included in QuantLib) using QuantLib
> with a test-suite approach! He named these code as analyticresolver. I’m
> trying to include a simple fixedratebond code to this analyticresolver for
> my training exercise. When I built with Cmake , errors (Days, Months, Years
> are not declared) occurred at both instruments/bond.cpp and
> time/daycounter/actualactual.cpp. When I included time/all.hpp at
> corresponding instruments/bond.hpp and time/daycounter/actualactual.hpp,
> the same errors occurred. (I changed the namespace QuantLib as
> anlyticsolver for every code files.)
>
>
>
> To eliminate the error stating that the variables were not declared, I
> resorted to a workaround by modifying Days, Months, and Years to 2, 4, and
> 8 respectively, and completed the library build.
>
> My application fixedratebond.cpp including the corresponding header files
> bond.hpp and actualactual.hpp files, also using the builted library at link
> phase, runs successfully.
>
>
>
> But these ridiculous modification are not normal, How can I fix this
> problem?
>
>
>
> Thanks!!!
>
>
> _______________________________________________
> QuantLib-users mailing list
> Qua...@li...
> https://lists.sourceforge.net/lists/listinfo/quantlib-users
>