MeiliSearch/README.md
2019-10-16 18:04:45 +02:00

79 lines
5.9 KiB
Markdown

# MeiliDB
[![Build Status](https://dev.azure.com/thomas0884/thomas/_apis/build/status/meilisearch.MeiliDB?branchName=master)](https://dev.azure.com/thomas0884/thomas/_build/latest?definitionId=1&branchName=master)
[![dependency status](https://deps.rs/repo/github/meilisearch/MeiliDB/status.svg)](https://deps.rs/repo/github/meilisearch/MeiliDB)
[![License](https://img.shields.io/badge/license-commons%20clause-lightgrey)](https://commonsclause.com/)
A _full-text search database_ based on the fast [LMDB key-value store](https://en.wikipedia.org/wiki/Lightning_Memory-Mapped_Database).
## Features
- Provides [6 default ranking criteria](https://github.com/meilisearch/MeiliDB/blob/dc5c42821e1340e96cb90a3da472264624a26326/meilidb-core/src/criterion/mod.rs#L107-L113) used to [bucket sort](https://en.wikipedia.org/wiki/Bucket_sort) documents
- Accepts [custom criteria](https://github.com/meilisearch/MeiliDB/blob/dc5c42821e1340e96cb90a3da472264624a26326/meilidb-core/src/criterion/mod.rs#L24-L33) and can apply them in any custom order
- Support [ranged queries](https://github.com/meilisearch/MeiliDB/blob/dc5c42821e1340e96cb90a3da472264624a26326/meilidb-core/src/query_builder.rs#L283), useful for paginating results
- Can [distinct](https://github.com/meilisearch/MeiliDB/blob/dc5c42821e1340e96cb90a3da472264624a26326/meilidb-core/src/query_builder.rs#L265-L270) and [filter](https://github.com/meilisearch/MeiliDB/blob/dc5c42821e1340e96cb90a3da472264624a26326/meilidb-core/src/query_builder.rs#L246-L259) returned documents based on context defined rules
- Can store complete documents or only [user schema specified fields](https://github.com/meilisearch/MeiliDB/blob/dc5c42821e1340e96cb90a3da472264624a26326/meilidb-schema/src/lib.rs#L265-L279)
- The [default tokenizer](https://github.com/meilisearch/MeiliDB/blob/dc5c42821e1340e96cb90a3da472264624a26326/meilidb-tokenizer/src/lib.rs) can index latin and kanji based languages
- Returns [the matching text areas](https://github.com/meilisearch/MeiliDB/blob/dc5c42821e1340e96cb90a3da472264624a26326/meilidb-core/src/lib.rs#L66-L88), useful to highlight matched words in results
- Accepts query time search config like the [searchable attributes](https://github.com/meilisearch/MeiliDB/blob/dc5c42821e1340e96cb90a3da472264624a26326/meilidb-core/src/query_builder.rs#L272-L275)
- Supports [runtime incremental indexing](https://github.com/meilisearch/MeiliDB/blob/dc5c42821e1340e96cb90a3da472264624a26326/meilidb-core/src/store/mod.rs#L143-L173)
It uses [LMDB](https://en.wikipedia.org/wiki/Lightning_Memory-Mapped_Database) as the internal key-value store. The key-value store allows us to handle updates and queries with small memory and CPU overheads. The whole ranking system is [data oriented](https://github.com/meilisearch/MeiliDB/issues/82) and provides great performances.
You can [read the deep dive](deep-dive.md) if you want more information on the engine, it describes the whole process of generating updates and handling queries or you can take a look at the [typos and ranking rules](typos-ranking-rules.md) if you want to know the default rules used to sort the documents.
We will be proud if you submit issues and pull requests. You can help to grow this project and start contributing by checking [issues tagged "good-first-issue"](https://github.com/meilisearch/MeiliDB/issues?q=is%3Aissue+is%3Aopen+label%3A%22good+first+issue%22). It is a good start!
The project is only a library yet. It means that there is no binary provided yet. To get started, you can check the examples wich are made to work with the data located in the `datasets/` folder.
MeiliDB will be a binary in a near future so you will be able to use it as a database out-of-the-box. We should be able to query it using HTTP. This is our current goal, [see the milestones](https://github.com/meilisearch/MeiliDB/milestones). In the end, the binary will be a bunch of network protocols and wrappers around the library - which will also be published on [crates.io](https://crates.io). Both the binary and the library will follow the same update cycle.
## Performances
With a database composed of _100 353_ documents with _352_ attributes each and _3_ of them indexed.
So more than _300 000_ fields indexed for _35 million_ stored we can handle more than _2.8k req/sec_ with an average response time of _9 ms_ on an Intel i7-7700 (8) @ 4.2GHz.
Requests are made using [wrk](https://github.com/wg/wrk) and scripted to simulate real users queries.
```
Running 10s test @ http://localhost:2230
2 threads and 25 connections
Thread Stats Avg Stdev Max +/- Stdev
Latency 9.52ms 7.61ms 99.25ms 84.58%
Req/Sec 1.41k 119.11 1.78k 64.50%
28080 requests in 10.01s, 7.42MB read
Requests/sec: 2806.46
Transfer/sec: 759.17KB
```
### Notes
With Rust 1.32 the allocator has been [changed to use the system allocator](https://blog.rust-lang.org/2019/01/17/Rust-1.32.0.html#jemalloc-is-removed-by-default).
We have seen much better performances when [using jemalloc as the global allocator](https://github.com/alexcrichton/jemallocator#documentation).
## Usage and examples
Currently MeiliDB do not provide an http server but you can run the example binary.
The _index_ subcommand has been made to create an index and inject documents into it. Using the command line below, the index will be named _movies_ and the _19 700_ movies of the `datasets/` will be injected in MeiliDB.
```bash
cargo run --release --example from_file -- \
index example.mdb datasets/movies/data.csv \
--schema datasets/movies/schema.toml
```
Once the first command is done, you can query the freshly created _movies_ index using the _search_ subcomand. In this example we filtered the dataset to only show _non-adult_ movies using the non-definitive `!adult` syntax filter.
```bash
cargo run --release --example from_file -- \
search example.mdb
--number 4 \
--filter '!adult' \
id popularity adult original_title
```