On 20-Aug-2013 08:06, Krzysztof Kosiński wrote:
I know that there is working EMF import on Linux, and you can paste things from e.g. ChemDraw. Is it possible to make this code also work on Linux, using e.g. libemf, or would it require writing a completely new input filter?
I am not quite sure what you are asking.
Branch lp988601 has working EMF (and WMF) import and export on Linux and I believe OS X, although I do not have a Mac to test that on. It uses libUEMF, which I wrote for this purpose.
http://sourceforge.net/projects/libuemf/
The pictures you will see on that page are screenshots of three test metafiles produced by libUEMF. These can be read into and then exported from Inkscape.
If you want this in trunk then lp988601 will need to be merged into trunk. Which it probably should be at this point. There is only one outstanding bug that I know of: in some EMF files written by Illustrator on OSX image objects render with the sign of y reversed, even though the rest of the diagram renders correctly. Also, apparently the version of the GPL used by libUEMF is somehow incompatible with that required by Inkscape. Since libUEMF is my code I can change the GPL version to whatever is needed.
EMF+ support is coming, at some point, I hope. The only program I have that emits EMF+ is PowerPoint, and it does so with some pretty dramatic limitations - for instance, it drops all text! Also while PowerPoint can "read" EMF+, it keeps it in a sort of container that it apparently passes to GDI+. I do not think any version of PowerPoint is capable of converting EMF+ records into native PowerPoint graphics objects. My hope is that EMF+ will handle import/export of gradients without having to emulate them through hundreds of little slices, but that may or may not pan out. (EMF, in theory supports gradients specifications too, but every program uses the emulation method instead of the native EMF command. So far I have been unable to get the EMF linear gradient record to work with any program, including Windows Preview, which may explain why none of them use it.)
Regards,
David Mathog mathog@...1176... Manager, Sequence Analysis Facility, Biology Division, Caltech