-
Notifications
You must be signed in to change notification settings - Fork 17
New issue
Have a question about this project? # for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “#”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? # to your account
filterx: native dict/list types #385
base: main
Are you sure you want to change the base?
Conversation
a049f79
to
83c7160
Compare
rebased against main. |
Signed-off-by: László Várady <laszlo.varady@anno.io>
Signed-off-by: László Várady <laszlo.varady@anno.io>
Signed-off-by: László Várady <laszlo.varady@anno.io>
Signed-off-by: László Várady <laszlo.varady@anno.io>
Signed-off-by: László Várady <laszlo.varady@anno.io>
Signed-off-by: László Várady <laszlo.varady@anno.io>
Signed-off-by: László Várady <laszlo.varady@anno.io>
Signed-off-by: László Várady <laszlo.varady@anno.io>
I rebased this to current main and followed up the change of prototype of built-in simple functions. |
This branch helps a lot in malloc() efficiency and json cloning, but the lookups take 2x as much than the JSON lookups. With the JSON based build, get_subscript() takes 7.5% of the same run where with this branch (and the changing of the configuration from json() to dict()), it takes 14.4% I was running a lot of tests, so this might not be apples-to-apples completely, but that would be the reason for not improving performance that much. So I would definitely merge this, it is so much better on all other aspects than the JSON based build, than I am inclined to ditch that (and protobuf) completely and use just one internal/optimized representation and convert on input/output. |
No description provided.