Go to file
Clément Renault 58c07e7f8c
Merge pull request #23 from meilisearch/update-builder-thread-pool
Allow library users to specify the rayon ThreadPool for UpdateBuilder
2020-11-02 19:11:50 +01:00
.github/workflows Add a Github actions workflows 2020-11-01 14:47:44 +01:00
benches Fix the benchmarks 2020-10-31 22:18:29 +01:00
public Switch to a JSON protocol for the front page 2020-10-21 18:26:29 +02:00
src Allow library users to specify the rayon ThreadPool for UpdateBuilder 2020-11-02 19:11:22 +01:00
templates Switch to a JSON protocol for the front page 2020-10-21 18:26:29 +02:00
.gitignore Introduce the words-docids command for the infos binary 2020-09-07 22:36:35 +02:00
bors.toml Introduce bors to the project 2020-11-01 14:49:07 +01:00
build.rs Introduce a simple FST based chinese word segmenter 2020-10-04 17:04:33 +02:00
Cargo.lock Bump heed to be able to reference a RoTxn from multiple threads 2020-11-02 12:49:23 +01:00
Cargo.toml Bump heed to be able to reference a RoTxn from multiple threads 2020-11-02 12:49:23 +01:00
chinese-words.txt Introduce a simple FST based chinese word segmenter 2020-10-04 17:04:33 +02:00
LICENSE Initial commit 2020-05-31 14:21:56 +02:00
qc_loop.sh Initial commit 2020-05-31 14:22:06 +02:00
README.md Update the README to be up to date with the recent updates 2020-11-02 18:07:24 +01:00

the milli logo

a concurrent indexer combined with fast and relevant search algorithms

Introduction

This engine is a prototype, do not use it in production. This is one of the most advanced search engine I have worked on. It currently only supports the proximity criterion.

Compile and Run the server

You can specify the number of threads to use to index documents and many other settings too.

cargo run --release -- serve --db my-database.mdb -vvv --indexing-jobs 8

Index your documents

It can index a massive amount of documents in not much time, I already achieved to index:

  • 115m songs (song and artist name) in ~1h and take 107GB on disk.
  • 12m cities (name, timezone and country ID) in 15min and take 10GB on disk.

All of that on a 39$/month machine with 4cores.

You can feed the engine with your CSV (comma-seperated, yes) data like this:

cat "name,age\nhello,32\nkiki,24\n" | http POST 127.0.0.1:9700/documents content-type:text/csv

Here ids will be automatically generated as UUID v4 if they doesn't exist in some or every documents.

Note that it also support JSON and JSON streaming, you can send them to the engine by using the content-type:application/json and content-type:application/x-ndjson headers respectively.

Querying the engine via the website

You can query the engine by going to the HTML page itself.