-
Trie has great advantages in terms of average performances over hashtables or B-Trees giving on worst case O(L) insert, delete and search time complexity where L is the length of the key. This comes a…
-
So far language data are stored as dictionaries, i.e. key/value pairs of words and ground forms. Python dictionaries are quite efficient in terms of speed and rather efficient space-wise.
However t…
adbar updated
4 months ago
-
I'm using latest `go-graphite v0.15.5` and recently have both enabled `trie-index` and switched CPU to [GCP N2 family](https://cloud.google.com/blog/products/compute/expanding-virtual-machine-types-to…
-
Copied from: https://github.com/johann-petrak/gateplugin-StringAnnotation/issues/4
Consider: always store the actual case (if case-normalization is wanted, must be a preparation step for the list f…
-
Hello! I am a graduate student who is interested in audio confrontation samples. May I ask, where do I get the "models" file "models/session_dump" in the source code? Thanks!
-
possible to include the benchmark of space / mem used?
-
Rationale: `wheel`s use `zip` archives. Zip archives don't allow gaps. It is a practically needed thing to edit wheels metadata. It can be done 3 ways:
1. creating a new archive and copying files f…
-
Hey, I saw that IPv6 support was on your roadmap. I'm attempting to use your library to increase the speed of [logstash-filter-cidr](https://github.com/logstash-plugins/logstash-filter-cidr), which is…
-
## Background and Motivation
`dotnet trace collect` generates large files while tracing so it is more for taking snippets of activity rather than continuous monitoring.
## Proposed Feature
If…
-