Log of #mupdf at irc.freenode.net.

Search:
 <<<Back 1 day (to 2020/03/01)Fwd 1 day (to 2020/03/03)>>>20200302 
mendel_munkis thanks01:36.18 
avih ator: i can make the utf8 api thingy a header-only thing with one additional file (the header) which is fully self contained and consists of strict wrappers using only public API, which the users can include (say mujsutf8.h) after mujs.h. this will be a service to clients and will have zero effect on mujs binary. it would be slightly less efficient because it can't use mujs internals, but OTOH it's in a way external to mujs09:57.48 
  though with mujs creating this problem in the first place, i think it should solve it officially too10:00.04 
  ator: why does js_trystring do pop on error? js_tostring doesn't push anything which needs to be popped... does it?13:35.23 
ator it pops the error13:57.57 
avih gotcha, thx14:02.24 
  ator: could there be any difference at all (under all circumstances) between the names returned by js_nextiterator and those returned by Object.keys(x) (for own properties) and for(key in x) for non-own?14:40.42 
  if not, why don't they share code? for instance, O_getOwnPropertyNames could use an iterator14:41.32 
  (or the other way around)14:42.21 
  (by "all circumstances i think mostly mean if the object is modified between js_nextiterator calls)14:44.04 
  (though i'm mainly interested in the if question, and not the code sharing question)14:49.03 
ator the iterator also returns keys from objects in the prototype chain14:59.12 
  Object.keys only returns an objects own keys14:59.31 
avih iterator has "own" input. i'm asking if when it's 1 it's the same as Object.keys(x), and if when it's 0 it's the same as for(key in x)15:04.16 
  (js_pushiterator(J, idx, own) )15:04.34 
ator you'd have to read the code, I can't remember all these details :)15:06.40 
avih k15:08.51 
 <<<Back 1 day (to 2020/03/01)Forward 1 day (to 2020/03/03)>>> 
ghostscript.com #ghostscript
Search: