I'm the author of Beancount v2 & v3 core.
Since the plan was laid out, beangulp (the ingest/import framework) and beanquery have been successfully out to distinct projects taken over by Daniele Nicolodi, here:
Daniele did lots of improvements to both of these projects and I'm using them.
beanprice has also taken a life of its own, Kirill Goncharov and others have done some maintenance work on it:
As for the core rewrite to C++, I've been sitting on its development for a long while, it's more or less been on ice. Work got busy and with the appearance of children I have less free cycles to devote to it, and have been building other projects for more pressing needs (e.g.,
https://github.com/beancount/johnny). The state of affairs is that
- the new c++ parser is 90% done (missing the ability to store comments though)
- some of the c++ core is done (maybe 20% of it)
- I implemented a working prototype for the proto / C++ / python / bazel integration, but recent changes upstream need to be integrated (changes that make this easier, open source changes from Google)
- the build is currently broken due to changes upstream.
I was actually looking at upgrading the build today to account for the many improvements in dependent libraries, but I couldn't get an hour of undisturbed time.
I've decided over recent vacations to allocate a specific weekly time slot from now on to restart making at least regular progress on it. I'm not so interested in help, as I have a clear and uncompromising picture of what I'd like for it to look like, my problem is time more than anything else. I'm not so sure how to accept contributions at this stage, I was hoping to have at least a good chunk of the C++ core picture laid out in code before having other people get involved. The core has to be ported from Python to C++, keeping the Python tests all the while, but I want to redesign the data structures as well and present a clean Python API to them as the primary interface.
Some help I could use is not very sexy work:
- straightforwardly porting the Python core to the C++ core equivalents (with tests)
- upgrading the bazel build to 5.3 and all dependencies (removing hacks I had to use if now possible, they're improved bazel quite a bit)