The ANTLR IP Issue (Or Which Parser To Use?)

In order to be able to parse models written in your language, Xtext needs to provide a special parser for it. The parser is generated from the self defined language grammar.

It is recommended to use the ANTLR -based parser. ANTLR is a very sophisticated parser generator framework which implements a so called LL(*) algorithm. It is fast, simple and at the same time has some very nice and sophisticated features. Especially its support for error recovery is much better than what other parser generators provide.

Xtext uses ANTLR 3 which comes in two parts: the runtime and the generator. Both are shipped under the BSD license and have a clean intellectual property history. However the ANTLR parser generator is unfortunately still implemented in an older version of itself (v 2.x), where it was not possible for the Eclipse Foundation to ensure where exactly every line of code originated. Therefore ANTLR 2 didn’t get the required IP approval. Eclipse has a strict IP policy, which makes sure that everything provided by Eclipse can be consumed under the terms of the Eclipse Public License. The details are described in this document. That is why we are not allowed to ship Xtext with the ANTLR generator but only with the IP approvied runtime components. We have to provide it separately and for your convenience you’ll be asked to download the ANTLR generator when you run your language generator for the first time. You can even download it directly or install an additional plug-in into Eclipse:

The workflow will not bother you with this issue on subsequent executions as the archive will be stored in your project’s root directory and can thereby be reused during the next run.

IMPORTANT : Although if you use the non-IP approved ANTLR generator, you can still ship any languages and the IDEs you’ve developed with Xtext without any worrying, because it is not needed at runtime