Ah, what timing.
I have just started thinking about updating the C++ drawing code (which I
agree is extremely basic) to improve the quality and make it more generally
useful. I agree that the abstract base class idea makes a lot of sense.
This is, in essence, what the Python code does already.
There's definitely interest from my side. I think it would help a lot.
If there's a way to do so, I would happy to help out as you proceed. I
certainly would be more than willing to add "refinements" like dealing with
chiral centers.
-greg
On Wednesday, July 2, 2014, David Cosgrove <davidacosgrov...@gmail.com>
wrote:
>
> Hi All,
>
> I am currently writing some code for which I needed a Qt widget with a 2D
> RDKit rendering that I could interact with. This needed
> $RDBASE/GraphMol/MolDrawing/MolDrawing.h, but I found it very difficult to
> find the drawing coordinates for all atoms in the RDKit drawing model and
> convert them to screen coordinates so I could add annotation to the
> drawing, find which atom users had picked, etc. So I have created new
> drawing classes using a design borrowed from the original OpenEye OEDepict
> toolkit written, I believe, by Roger Sayle. This uses an abstract base
> class that handles all the drawing in terms of pure virtual functions for
> drawing lines, writing text etc. The user derives a concrete class that
> implements these functions for his/her drawing library of choice.
>
> Currently I have something that's adequate for my purposes, but if there's
> enough interest from people on the list, I can tidy it up, provide some
> example code and submit it for inclusion. In the short term it would be an
> addition to the existing system so a non-breaking change, but in the longer
> term it could possibly be a route to unifying the python and C++ 2D drawing
> systems which seem at present to be quite separate.
>
> I mention this at this stage because it would be a non-trivial amount of
> work on my part to do this. To start with, I don't need to worry about
> drawing chiral centres in my current project, so I haven't done anything
> about that (MolDrawing.h doesn't either, for that matter!). The example
> codes would have to be written to do something useful that showed all the
> features, there might be a need for documentation, etc. Not insignificant
> is that I would have to bash this square peg through the round hole which
> is AstraZeneca's publication approval system that won't have seen anything
> like it before. There don't seem to be many of us C++ programmers posting
> to the list, so I'm really looking to gauge interest. I'm happy to put the
> work in to make this a useful submission to the project, but only if other
> people are likely to use it.
>
> Please let me know, either to this list or privately to
> david.cosgr...@astrazeneca.com.
>
> Thanks,
> Dave
>
>
------------------------------------------------------------------------------
Open source business process management suite built on Java and Eclipse
Turn processes into business applications with Bonita BPM Community Edition
Quickly connect people, data, and systems into organized workflows
Winner of BOSSIE, CODIE, OW2 and Gartner awards
http://p.sf.net/sfu/Bonitasoft
_______________________________________________
Rdkit-discuss mailing list
Rdkit-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
