Query-tree-modules

Automatic multi-languages (SQL, MongoDB, ...) query rewriter for polystore dataset.

Description

We develop a prototype to analyze a query and translate it into its algberaic tree form. Using other modules such as a mapping dictionary and translation rules, the tree is transformed into a more complex one showing data location, data transfers and data distribution.

Getting started

1. Clone repository

cd /project_folder
git clone https://gitlab.irit.fr/sig/these-lea-el-ahdab/query-tree-modules.git

After the previous step a popup window will ask for your git username and password.

Note : If you cannot succeed to connect with your password you can try used a gitlab personal access token instead

How Generate a gitlab token ?

2. Requirements

Note : All the tools listed below are indicated with the versions used to develop the project. To be sure you can run the project, use at least these versions or newer.

• Java 20 : https://www.oracle.com/fr/java/technologies/downloads/#jdk20-windows
• Select one :
  • IntelliJ IDE 2023 (contain maven): https://www.jetbrains.com/fr-fr/idea/
  • NetBeans 18 (contain maven): https://netbeans.apache.org/
  • Maven 3.9.2 https://maven.apache.org/download.cgi

3. Installation

Note 1 : Thanks to maven dependencies management you can find all dependencies in pom.xml

Note 2 : Program can produce algebraic tree in output but that require unicode characters to display it. Have a look to the following link to enable all unicode characters in windows : https://itsontreinamentos.com/2022/03/31/comment-changer-codage-des-caracteres-windows-11/

3.1 With IntelliJ IDEA

• Open /cloned_project_folder in IntelliJ
• Open the maven tool window (top right of the IDE) and in Lifecycle click on install

You can now run the application using IntelliJ run or use the polyglot.jar generated under /target folder with the following cmd :

java -jar .\target\polyglot.jar --help

3.2 With NetBeans

• Open query-tree-modules project under /cloned_project_folder in NetBeans
• Under /Project_Files right click on pom.xml and click on Run maven > Goals...
• In Goals field type 'install' and then OK

You can now run the application using NetBeans run or use the polyglot.jar generated under /target folder with the following cmd :

java -jar .\target\polyglot.jar --help

3.3 Without IDE

cd /cloned_project_folder
mvn install
java -jar .\target\polyglot.jar --help

Project Structure

Bac_A_Sable : Previous module 1 version

doc : Folder containing uml diagrams (see doc README)

Resources : Files containing database data (see resources README)

src : Java project (see next section)

Java project

Folder structure

You can choose between 2 main programs to run the app :

• java.fr.irit.PolyglotCli : is a CLI version of the app where you can put query in parameter
• java.fr.irit.App : is for development use, you have to put the query in the code and then run the app without its interface
• java.fr.sae.Application : is the java classes allowing object manipulation in the front-end application
• moduleInterface: is the java classes that build screens, controllers and run the front-end application

└──src
    ├── main
    |   ├── antlr4
    |   └── java (Back-End)
    |       ├── irit 
    |             ├── fr.irit.App (Main)
    |             ├── fr.irit.PolyglotCli (Main)
    |             ├── fr.irit.algebraictree (Algebraic tree classes)
    |             ├── fr.irit.commands (CLI commands classes)
    |             ├── fr.irit.module1 (module 1 classes)
    |             ├── fr.irit.module2 (module 2 classes)
    |             └── fr.irit.module3 (modules 3 classes)
    |       └── sae
    |             ├── fr.sae.algebraictree (Same algebraic tree classes than irit)
    |             ├── fr.sae.querybuilder
    |             ├── fr.sae.queryparser
    |             └── Application
    |
    |   └── moduleInterface (Front-End)
    |       ├── com.main.irit
    |             └── iritMainApplication (Main)
    |       └── fxgraph (Modules for building the tree)
    └── test

Front-End Application

Pre-requisites

• JavaFX SDK Gluon
• Unzip .zip file (without deleting the zip)
• (IntelliJ) Run >> Edit Configurations >> Modify options >> Add VM Options to iritMainApplication class

--module-path 
path/to/the/specific/directory/javaFxJDK/lib
--add-modules=javafx.controls,javafx.fxml

• (IntelliJ) Project Structure >> Modules >> (+) >> JAR or directories >> Add SDK.zip >> Apply changes

Java Class

• iritMainApplication
• Method

public static void main(String[] args) {
  launch(args);
} 

Tests

Unit tests are execute using the JUnit framework under the folder /src/test

They are automatically run when you compile the app with maven (mvn install or mvn compile).

/!\ There are not updated since June 2023

Command Line Interface (CLI)

Main file fr.irit.PolyglotCli allow launch the application as a command line interface.

For create the CLI we use a tool named Picocli

All CLI commands are defined under the folder fr.irit.commands

Usage: java -jar polyglot.jar [-h] [COMMAND]
Commands:
- run (run a sql or mongo query)
  ├── Usage: java -jar polyglot.jar run <query> [-gt] [-mt] [-tt] [--time] [--help]
  ├── Parameters:
  |   <query>               Database query (ex: SELECT * FROM Customers)
  └── Options:
      -gt, --globalTree     Print global tree
      -h, --help            display this help message
      -mt, --multistoreTree Print multi-store tree
      --target-db-name=<TARGET_DB_NAME>
                            Target database name (default : DB1)
      --target-db-type=<TARGET_DB_TYPE>
                            Target database type (default : RELATIONAL)
      --time                Show execution time
      -tt, --transferTree   Print transfer tree
 
- unified-view (Display polystore unified view)
  ├── Usage: java -jar polyglot.jar unified-view <dbType> [--help] [-pimpl]
  ├── Parameters:
  |   <dbType>              Database type (relational or document)
  └── Options:
      -h, --help   display this help message
      -pimpl, --physical-implementation
               show physical implementation in addition of unified view
    

ANTLR

ANTLR is a parser generator tool that we use for read and recognize DB Queries (sql for the moment) to translate it into a structured tree.

How to use it ?

Write a .g4 grammar file (lexer & parser rules) under /src/main/antlr4 and generate parser with following cmd :

mvn generate-sources

Copy newly generated antlr folder (here : ./target/generated-sources/antlr4) to antlr4 folder in java folder (replace : ./src/main/java/fr/irit/module1/antlr4)

Support

@IRIT-GD-SIG

Roadmap

GHES release version dates:

Authors and acknowledgment

@IRIT-GD-SIG

License

@IRIT-GD-SIG

Project status

Working