Dear @magick93,

Thank you very much for your kind words and your interest in LEMMA. In fact, we would love to have a code generator for Typescript in LEMMA :-)

Unfortunately, the description of custom code generator development is not part of LEMMA's documentation, yet. Therefore, I have prepared an example code generator that illustrates the steps to realize custom code generators for LEMMA. You can find the example generator in LEMMA's master branch at https://github.com/SeelabFhdo/lemma/tree/master/examples/model-processing. I have also added a README.md to summarize the steps for custom code generator development with LEMMA. In addition, the code of the generator has extensive comments about what is going on. Nonetheless, please feel free to ask any questions in this issue, which may occur on your way to a Typescript generator for LEMMA :-)

Below, I give a few more background information on how you may process LEMMA models, e.g., for the sake of code generation.

Model Processing Using LEMMA's Own Model Processing Framework (aka "The Preferred Way")

To facilitate model processing with LEMMA, we provide our own model processing framework that is compatible with the Eclipse Modeling Framework. Actually, the example code generator at https://github.com/SeelabFhdo/lemma/tree/master/examples/model-processing is based on this framework. In one of the upcoming LEMMA releases we will even tune the framework to be more flexible in terms of provided and processed models. For now, however, the framework is sufficient to process LEMMA models based on their intermediate representations (cf. the README.md of the example code generator for more information on this topic).

Model Processing Using Xtext

LEMMA's modeling languages are based on Xtext. Consequently, you could alternatively use Xtext's standalone parsing mechanism to parse LEMMA models and derive code from the obtained AST. You can find the Xtext-based source code of the relevant LEMMA modeling languages on GitHub:

However, a benefit of LEMMA's model processing framework is that it hides the complexity of model parsing from the user.

Model Processing Using XML Metadata Interchange (XMI)

As mentioned above, LEMMA comes with intermediate specifications for its modeling languages (cf. the README.md of the example code generator for more information on this topic). These specifications aim to facilitate model processing as they incorporate preprocessed information depending on language semantics, e.g., applying default values or resolved inheritance hierarchies (please see the links to the intermediate specifications at the end of the README.md of the example code generator). LEMMA can produce intermediate representations for its models from the Eclipse IDE (see the description at the end of the README.md of the example code generator) in the XMI format. Since XMI is an open standard, you could theoretically parse the XMI files by yourself and generate code form them. While this approach might have the charm to not require a Java environment (as LEMMA's model processing framework or the Xtext approach do), I would consider it more complex given the necessity to resolve cross-model references and/or having to deal with XML.

Custom code generator #42

Description

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions