https://bugs.documentfoundation.org/show_bug.cgi?id=150102
Armin Le Grand <[email protected]> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |ASSIGNED Assignee|[email protected] |[email protected] |desktop.org | --- Comment #10 from Armin Le Grand <[email protected]> --- Took a look. Due to https://bugs.documentfoundation.org/show_bug.cgi?id=126319 and that originally integer-based version GraphicExporter::GetGraphic and it's hairry adaption to do the right stuff in int-based metafiles it is *very* dangerous to scope around with it. If so, it is *urgently necessary* to keep the tests from https://bugs.documentfoundation.org/show_bug.cgi?id=126319 working, especially the file and what to do from https://bugs.documentfoundation.org/show_bug.cgi?id=126319#c19. I debugged what that black BG only does - it looks luckily it uses the !bSingleGraphic path in GraphicExporter::GetGraphic while the conversions from the previous task use the single-obj Metafile creation. So I experimented with that and it seems that adding aOnePixelInMtf by (a) removing from origin, see aOutMap.SetOrigin and (b) adding to Size, see SetPrefSize(aBoundSize) is too much. It looks like in aBound is the PageSize-1 (here locally I get 0/0/27999/15749), so to get the correct Size correcting by adding '1' should be enough, no need to expand by half a pixel-size. That is theoretically also good, and on the safe side, but seems to be too much at places where the Metafile gets interpreted, in this case SlideShow. Unfortunately this is the UNO API Metafile-export, so a whole lot of consumers might interpret in a whole lot of ways..sigh. What is true is that the correct integer size (here: 0/0/28000/15750) of the Page (and thus the fill of it) is more correct, so it's worth a try to correct just that. I tested the exports (selection) of the example above, it gets not changed. So I do some more checks and might then dare to change that... -- You are receiving this mail because: You are the assignee for the bug.
