Different implementations of bufferpool.
Designed around prior experiences with “larger than memory” problems. Data analysis, append-only databases, etc. All things I’ve done professionally or personally.
If you happen upon this project and think that, yes, you’d like to help, or if you want to use it, I welcome help (but I’ll be quite stubborn about the license).
This is a prototype, used for testing ideas and experiementing.
The Python library is intentionally “pure” stdlib Python, not using any third party packages. This simplifies testing, deployments, etc.
Go: half done.
This will be used in my own projects.
This will be the gold standard implementation.
The top-level concept will be to load a block of data into pool and then mutate it via get/write from there.
From a pure refchecking perspective, the conventional references are difficult to address with a lifetime.
Dataframes are therefore Box<T> for now.
Sophisticated projects.
(C) Affero GPL v 3.0 - https://www.gnu.org/licenses/agpl-3.0.en.html
This means, in short, you MUST use share all source code this is linked or associated with.
Contact me if you want to use it commercially.
paul@nathan.house
https://15445.courses.cs.cmu.edu/fall2020/notes/05-bufferpool.pdf