textX logo

textX is a meta-language (i.e. a language for language definition) for domain-specific language (DSL) specification in Python.

In a nutshell, textX will help you build your textual language in an easy way. You can invent your own language or build a support for an already existing textual language or file format.

From a single grammar description, textX automatically builds a meta-model (in the form of Python classes) and a parser for your language. The parser will parse expressions of your language and automatically build a graph of Python objects (i.e. the model) corresponding to the meta-model.

textX is inspired by Xtext - a Java based language workbench for building DSLs with full tooling support (editors, debuggers etc.) on the Eclipse platform. If you like Java and Eclipse check it out. It is a great tool.

A video tutorial for textX installation and implementation of a simple data modeling language is below.

For a not-so-basic video tutorial check out State Machine video tutorial.

For an introduction to DSLs in general here are some references:

For an in-depth coverage on the subject we recommend the following books:

Feature highlights

  • Meta-model/parser from a single description

    A single description is used to define both language concrete syntax and its meta-model (a.k.a. abstract syntax). See the description of grammar and metamodel.

  • Automatic model (AST) construction

    Parse tree will automatically be transformed to a graph of python objects (a.k.a. the model). See the model section.

    Python classes will be created by textX but, if needed, user supplied classes may be used. See custom classes.

  • Automatic linking

    You can have references to other objects in your language and the textual representation of the reference will be resolved to the proper python reference automatically.

  • Automatic parent-child relationships

    textX will maintain a parent-child relationships imposed by the grammar. See parent-child relationships.

  • Parser control

    Parser can be configured with regard to case handling, whitespace handling, keyword handling etc. See parser configuration.

  • Model/object post-processing

    A callbacks (so called processors) can be registered for models and individual classes. This enables model/object postprocessing (validation, additional changes etc.). See processors section.

  • Grammar modularization - imports

    Grammar can be split into multiple files and then files/grammars can be imported where needed. See Grammar modularization.

  • Scope Providers

    Scope Providers allow different types of scoping. See Scoping.

  • Multi-meta-model support

    Different meta-models can be combined. Typically some of these meta-models extend other meta-models (grammar modularization) and reference each other. Special scope providers support file-extension-based allocation of model files to meta models. See Multi meta-model support

  • Meta-model/model visualization

    Both meta-model and parsed models can be visulized using GraphViz software package. See visualization section.

Installation

$ pip install textX[cli]

Note

Previous command requires pip to be installed.

Also, notice the use of [cli] which means that we would like to use CLI textx command. If you just want to deploy your language most probably you won't need CLI support.

To verify that textX is properly installed run:

$ textx

You should get output like this:

Usage: textx [OPTIONS] COMMAND [ARGS]...

Options:
  --debug  Debug/trace output.
  --help   Show this message and exit.

Commands:
  check            Check/validate model given its file path.
  generate         Run code generator on a provided model(s).
  list-generators  List all registered generators
  list-languages   List all registered languages
  version          Print version info.

To install development (master branch) version:

$ pip install --upgrade https://github.com/textX/textX/archive/master.zip

Python versions

textX works with Python 3.6+. Other versions might work but are not tested.

Getting started

See textX Tutorials to get you started:

For specific information read various User Guide sections.

Also, you can check out examples.

A full example project that shows how multi-meta-modeling feature can be used is also available in a separate git repository.

To create the initial layout of your project quickly take a look at project scaffolding.

Discussion and help

For general questions and help please use StackOverflow. Just make sure to tag your question with the textx tag.

For issues, suggestions and feature request please use GitHub issue tracker.

Projects using textX

Here is a non-complete list of projects using textX.

  • Open-source

    • pyecore - ECore implementation in Python. Vincent Aranega is doing a great work on integrating textX with pyecore. The idea is that the integration eventually gets merged to the main textX repo. For now, you can follow his work on his fork of textX.
    • pyTabs - A Domain-Specific Language (DSL) for simplified music notation
    • applang - Textual DSL for generating mobile applications
    • pyFlies - A DSL for designing experiments in psychology
    • ppci - Pure python compiler infrastructure.
    • Expremigen - Expressive midi generation
    • fanalyse - Fortran code parser/analyser
    • Silvera - A DSL for microservice based software development
    • cutevariant - A standalone and free application to explore genetics variations from VCF file. Developed by labsquare - A community for genomics software
    • osxphotos - Python app to export pictures and associated metadata from Apple Photos on macOS.
    • StrictDoc - Software for technical documentation and requirements management.

  • Commercial

    • textX is used as a part of Typhoon-HIL's schematic editor for the description of power electronic and DSP schemes and components.

If you are using textX to build some cool stuff drop me a line at igor dot dejanovic at gmail. I would like to hear from you!

Editor/IDE support

Visual Studio Code support

There is currently an ongoing effort to build tooling support around Visual Studio Code. The idea is to auto-generate VCS plugin with syntax highlighting, outline, InteliSense, navigation, visualization. The input for the generator would be your language grammar and additional information specified using various DSLs.

Projects that are currently in progress are:

  • textX-LS - support for Language Server Protocol and VS Code for any textX based language.

  • textx-gen-coloring - a textX generator which generates syntax highlighting configuration for TextMate compatible editors (e.g. VSCode) from textX grammars.

  • textx-gen-vscode - a textX generator which generates VSCode extension from textX grammar.

  • viewX - creating visualizers for textX languages

Stay tuned ;)

Other editors

If you are a vim editor user check out support for vim.

For emacs there is textx-mode which is also available in MELPA.

You can also check out textX-ninja project. It is currently unmaintained.

Citing textX

If you are using textX in your research project we would be very grateful if you cite our paper:

Dejanović I., Vaderna R., Milosavljević G., Vuković Ž. (2017). TextX: A Python tool for Domain-Specific Languages implementation. Knowledge-Based Systems, 115, 1-4.

@article{Dejanovic2017,
    author = {Dejanovi\'{c}, I. and Vaderna, R. and Milosavljevi\'{c}, G. and Vukovi\'{c}, \v{Z}.},
    doi = {10.1016/j.knosys.2016.10.023},
    issn = {0950-7051},
    journal = {Knowledge-Based Systems},
    keywords = {Domain-Specific Language; Meta-model; Model; Model-Driven software development; Parser; Python },
    note = {},
    pages = {1--4},
    title = {{TextX: A Python tool for Domain-Specific Languages implementation}},
    url = {http://www.sciencedirect.com/science/article/pii/S0950705116304178},
    volume = {115},
    year = {2017}
}