Symbol init cleanup 2 #691
Conversation
rocky
force-pushed
the
Symbol-init-cleanup_2
branch
5 times, most recently
from
ae421de to
e3c545c
Compare
PredefinedSymbol acts more like a numeric constant.
Update builtin/atomic/symbols.py according to current standards.
* Some long lines split * some linting * correct a wrong symbol
We should be using Symbols insead of string where possible. builtins_precedence is really used in eval/makeboxes so it is defined there. It is just initialized in `mathics.buitins`.
and try to clearify what is up here.
rocky
force-pushed
the
Symbol-init-cleanup_2
branch
from
2fde486 to
632b83a
Compare
|
@mmatera I have rebased this with the current master and made small changes to get this working. Please look over all of the boxing changes here and advise if this is relevant, correct, and should go in. |
rocky
force-pushed
the
Symbol-init-cleanup_2
branch
from
632b83a to
c4c1e6d
Compare
Looking over myself, I am happy to put in a new PR with the boxing stuff removed until after release. I don't fully understand this, and the focus of this PR of this wasn't about boxing which somehow slipped in here as a bug fix for the Symbol work. |
|
@rocky, as you prefer. The idea was to move out of mathics.builtin.makeboxes the implementation of |
Thanks @mmatera to getting this to the point where I can see the light at the end of the tunnel.