NDatabase - Fast data model creation - Powerful Async / Sync

Update master

Few fixes & build

Optimisations

fix build

Notes https://github.com/NivixX/NDatabase/releases

SQLite driver

Optimisations

· Fixed connection issue during concurrent access of SQLite · Improved SQLite performances with HikariCP

Fix build

Fix NEntity

Update SQLite

Fixed build

Fixed query when using NEntity that haven't indexed field

fixes in sqlite queries

Packaging changes

Update Libby for drivers

Sponge detection fixes

Jar shading detection improvements

Reduced jar file

Previous 1.5.2 jar was not correct

Update SQLITE driver to the latest (3.47.0.0)

MySQL Connector 8.0.30 update

Quite a big update: · MySQL / SQLite drivers are now provided by NDatabase and not your minecraft server software · These drivers are modern version and may improve performances · Driver are installed only if needed during runtime (if you use SQLite, the MySQL driver won't be downloaded) · Added platform Sponge (that doesn't have drivers by default, that's why we now use runtime driver installation · Since needed drivers are loaded on runtime, the size of NDatabase jar is smaller

Tracking if NDatabase is shaded

Added logging and enforcement on platform load

All DBMS are packaged from different modules for more flexibility

Validate if connection is valid with basic query on platform load

add coverage for NQuery

Added adapter at DAO level that will adapt the query depending on the MySQL version. SQLite adapter has also been implemented. In future update, more adapter for different MariaDB version will also be added. Fix the issue for #14

Some servers may have Google Guice issues, it should fix some of them (that cause null resolved instances)

You are now able to save external objects inside your NEntity such as Bukkit domain related objects. The serialization can be done using Jackson mappers Code (Text): @JsonSerialize(using = LocationSerializer.class) @JsonDeserialize(using = LocationDeserializer.class) private Location location;

Better api accesses and documentation

Changed the way how the single node path construct the full path for @ Indexed fields

SQLite throw exception earlier if the JDBC sqlite driver is not found.

The ON CONFLICT clause didn't worked in some specific sqlite driver versions.

SQL formatting issue was introduced with the last update (for the upsert method). Only SQLITE is concerned about that in the case were you didn't used the Indexed annotation on your NEntity

Only affect SQLITE (for more robust performences I recommend using MongoDB or MySQL). Initially, I was using GENERATED COLUMN with JSON operators to do the denormalization for the indexing. But it looks like most of minecraft servers software have a sqlite driver bellow 3.38.0 (2022-02-22). As the sqlite driver is loaded as a native library (.dll) it's not possible to load a new one. So I implemented the denormalization in another way. It's a bit less performant than the original solution.

see https://www.spigotmc.org/threads/tu...-framework-for-mine…

Don't affect other databases

Any schema modification related to indexing which fail will throw a DatabaseCreationException and details the issue.

· MariaDB is now supported ! see new config : https://github.com/NivixX/NDatabase/blob/1.1.0/ndatabase-… · fixed bug with denormalization (for indexing)

First NDatabase release ! You have now the possibility to index some field of your Key-value store and query it using NQuery read more in the wiki page https://github.com/NivixX/NDatabase/wiki API change from v0 to v1

· Exception handling fix: if your async callback throw an exception, nothing was printed to the console and it failed silently · MongoDB decoding didn't worked for embedded objects and some specific types.

Test the database at startup (fail faster at the plugin loading and not at the first interaction)

In this version, a new method is now available in order to generate a top K. Very useful if you want to display a ranking ! example : Code (Java): Comparator <PlayerEntity > comparator = (o1, o2 ) -> o2. getScore ( ) - o1. getScore ( ) ; repository. computeTopAsync ( 3, comparator ) . thenAsync ( (playerEntities ) -> { playerEntities. get ( 0 ) ; // TOP 1 playerEntities. get ( 1 ) ; // TOP 2 playerEntities. get ( 2 ) ; // TOP 3 } ) ;

Fix logging in debug mode and added logging for MongoDB

MongoDB is now supported ! Use the following config : https://github.com/NivixX/NDatabase/blob/0.4.0/ndatabase-…

This version add the support with SQLite, I also added the javadoc within the api, so feel free to update the api New config file, you can copy paste it if you want to use sqlite: https://github.com/NivixX/NDatabase/blob/0.3.0/ndatabase-…

The API has been documented with some clean-up, note that I probably won't break the API even at the 1.0.0 release. There is no breaking changes so you can pull the new version Code (Text): allprojects { repositories { ... maven { url 'https://jitpack.io' } } } dependencies { implementation 'com.github.NivixX.NDatabase:ndatabase-api:0.2.0' } Code (Text): <repositories> <repository> <id>jitpack.io</id> <url>https://jitpack.io</url> </repository> </repositories> <dependency> <groupId>com.github.NivixX.NDatabase</groupId> <artifactId>ndatabase-api</artifactId> <version>0.2.0</version> </dependency>

NDatabase use reflection to instantiate your data model, some Java version didn't worked.

This version fix plugin loading error for modern version of minecraft and java. As I intend to support Spigot 1.8 from 1.19 and java 8 from Java 18+ I had to shade the guava library inside the plugin because every spigot server use different guava versions. Note: tested with spigot 1.8 / 1.19 and java 8 / 18

You have 2 version of a promise's callback Code (Text): thenAsync(returnedEntity -> ) thenAsync((returnedEntity, throwable) -> ) If an exception occurred, throwable will not be null. Now, if you decide to not handle de exception and a exception actually occured, NDatabase will catch it and warn you where in your code you didn't handled the exception.

this pre-release introduce the usage of a thread pool for async database operations. Using thread pool will improve performances as there is no need to recreate thread on the fly as we can reuse one from the pool Configure the threadpool in the config : Code (Text): # The number of thread that will be used for your async database call # If you are using MySQL, I would advice to set the same number as "minimum idle connection" # In the case were you have more than 3 running operation, extra thread will be created temporarily idle-thread-pool-size: 3

As it's a key-value store the field name "key" make more sense

this is the first usable version