On 2020/05/02 10:22:15, hahnjo wrote: > https://codereview.appspot.com/548030043/diff/559960055/lily/general-scheme.cc > File lily/general-scheme.cc (right): > > https://codereview.appspot.com/548030043/diff/559960055/lily/general-scheme.cc#newcode778 > lily/general-scheme.cc:778: free (a); > On 2020/05/02 10:15:40, hanwenn wrote: > > the code mixes setting up the GS instance (memory management etc) with > handling > > the file. Does it have to be this way? Can we have a > > > > class Ghostscript { > > process(string file, string device); > > close(); > > }; > > > > Ghostscript *get_gs(vector<string> args); > > > > instead? > > > > I think it should be possible to construct the API such that we always have > > ly:gs , and that it falls back to shelling out to GS if the API is not > > available. > > No, because there are two types of arguments when using the API: args and > device_args where the latter is added to command below. This uses a different > syntax and some properties are called differently (HWResolution vs -r for > example).
It may well be that calling them "HWResolution" instead of -r would work fine and -r is just provided as a courtesy abbreviation. I don't know whether this holds true, just bringing this up in case it might make for a different view of the situation. https://codereview.appspot.com/548030043/