Relax

A relational algebra calculator

How to build

• Install yarn https://yarnpkg.com/
• Install node https://nodejs.org/en/
• Clone the repo
• Checkout the development branch
• Execute yarn install to install all dependencies
• Execute yarn serve to locally run the webapp on port 8088
• (Optional) open folder with https://code.visualstudio.com/ and install the workspace recommended extensions (includes tasks)

Alternatively, you can make use of the Visual Studio Code Dev Containers extension to develop inside a Docker container and take advantage of Visual Studio Code's full feature set. A development container is a running container with a well-defined tool/runtime stack and its prerequisites.

To get started, follow these steps:

• Install and configure Docker Desktop for your operating system;
• Install VS Code and the Dev Containers extension;
• Clone this repository and open it in VS Code;
• Use the Dev Containers: Reopen in Container command from the Command Palette;
• When VS Code restarts, all dependencies will be installed automatically in the dev container. After that you can open a browser window and access the webapp at http://localhost:8088/;
• Close the window to stop the container;
• Any change in the files within the container is kept in the existing repository source code on your filesystem so that you can easily push your contributions.

How to Release

• Clone the repo
• Checkout the development branch
• Execute yarn install to install all dependencies
• Use yarn build to create release build in the dist folder
• Copy the content of your dist folder somewhere outside of the repository
• Checkout the branch gh-pages
• Remove everything there and paste content of the former dist folder
• Push to github (the person doing this must be an administrator of the project)

Contributions

• Contributions are highly appreciated
• Please create a pull request for the development branch

Add a new Language

• Open src/locales/languages.csv as csv with e.g. libreoffice (use ',' as delimiter and '"' as string marker)
• Every column represents a language, every row a value
• Create a new column (use the country code as first row/header cell)
• Insert a string for every language key there is (pay attention to {{variables}} as they must stay untouched)
• When finished, execute the script writeLanguageFiles.py which generates the json files based on the csv
• In src/calc2/i18n.tsx add an import for this file as well as initializing it in the i18n init function (see other languages for examples)
• Add a new dropdown entry to src/calc2/calculator.entry.tsx and src/calc2/calculator.entry.tsx
• Test your language
• Create a pull request to the development branch

Features

Data Sources

There are two ways to provide a group of tables:

• static groups: the groups and tables are loaded from a text file stored on the server
• github gist: the groups and tables are loaded from a github gits via its unique id

In the group's description markdown code can be used but javascript code will be removed for security reasons.

Static Groups

Static groups are loaded from the file data/local_groups. In this file multiple groups can be declared in the following text format.

A group starts with the group header:

    group:sample group
    description:this is the <b>description</b>

Every keyword can be followed by either a colon and a single line of content terminated with a line break or the content is surrounded with '[[' and ']]' if the content spans multiple lines.

    group:sample group
    description[[this is the <b>description</b>
        that spans multiple lines]]

Headers can also be provided for different languages e.g. description@de:....

After the group header the tables are defined using the variable assignment syntax. The name of the variable is used for the tables name and the assigned RA-expression defines its content. Every valid RA-expression can be used if the only relations used are defined in that group. The qualifier of each column is its tables name and is automatically added at runtime.

    group:sample group
    description:this is the <b>description</b>
    
    A = {a, b
        1, 2
        3, 4
    }
    B = {a, c, d
        1, 'test', 1970-01-01
		3, 'test2', null
		42, 'test3', null
    }
    C = A x B

It is also possible to define example queries in both relational algebra and SQL. These will be automatically loaded into the editor when the group is loaded for the first time.

    group:sample group
    description:this is the <b>description</b>
    
    exampleSql - {
      Select * from A where a = 1
    }

    exampleRelAlg - {
      sigma R.a = 1 R
    }       

    A = {a, b
        1, 2
        3, 4
    }
    B = {a, c, d
        1, 'test', 1970-01-01
		3, 'test2', null
		42, 'test3', null
    }
    C = A x B

Github Gist

The program can load groups from a gist. This allows any user to use their own tables and share them with others.

The format to define a group in the gist ist the same format as for the static group definition.

The data can be loaded by adding the git's id to the query string. If you call the application with the query string data=gist:ac267b9cc810ac5f20e2, it tries to load additional groups from the gits with the id ac267b9cc810ac5f20e2.

So a user could publish a url like http://example.com/?data=gist:ac267b9cc810ac5f20e2 to share the group with other users.

Contributors

• Spanish translation was contributed by https://github.com/Nifled
• Korean translation was contributed by https://github.com/jeongwhanchoi