Skip to content

Programmatically modify exiting Modelica files using Python!

License

Notifications You must be signed in to change notification settings

urbanopt/modelica-builder

Repository files navigation

Modelica Builder

The Modelica Builder project aims to make in-place modifications to Modelica language files easier. The principal use case is to load, modify using higher level abstracted methods, and then save the resulting file. The user has access to the entire Abstract Syntax Tree of the entire Modelica grammar.

The Modelica Builder project does not:

  • Compile nor check for syntax validity

Install

pip install modelica-builder

Usage

ModBuild provides specific methods for reading and modifying files.

from modelica_builder.model import Model

# parse the model file
source_file = 'DCMotor.mo'
model = Model(source_file)

# do read and modify the model
# refer to modelica_builder.model.Model class methods to see what's available
name = model.get_name()
model.set_name('New' + name)
model.add_connect('some.component.port_a', 'another.component.port_b')
model.insert_component('MyComponentType', 'myInst',
                        modifications={'arg1': '1234'}, string_comment='my comment',
                        annotations=['my annotation'], insert_index=0)

# save the result
model.save_as('NewDCMotor.mo')

You can also define your own classes for editing the file

from modelica_builder.edit import Edit
from modelica_builder.selector import Selector
from modelica_builder.transformation import SimpleTransformation

class MySelector(Selector):
    # implement class for selecting AST nodes

# define the edit to make to the node's text and combine it with the selector
edit = Edit.make_replace('FOOBAR')
selector = MySelector()
transformation = SimpleTransformation(selector, edit)

model = Model('my_modelica_file.mo')
# add your custom transformation
model.add(transformation)
model.save_as('result.mo')

Transformations specify what nodes to change and how to change them. This is done by combining Selectors and Edits. Selectors specify how to select nodes in the AST, and edits are modifications (insert, replace, delete) to the text of selected nodes.

A Transformer is a collection of Transformations, which can then be applied to a file.

See the tests for more examples and information.

Development

For developers, dependency management is through Poetry. Installation is accomplished by running pip install poetry.

    pip install poetry

    # install after cloning repo
    poetry install

If you change the source grammar file you need to regenerate the parser and lexer.

With docker installed, run these commands from this the repo's root directory

    # build Antlr container
    docker build -t antlr4:latest -f antlr/Dockerfile .

    # run parser generator for python
    docker run -v "$(pwd)/modelica_builder/modelica_parser":/var/antlrResult \
        antlr4:latest \
        -Dlanguage=Python3 /var/antlrResult/modelica.g4

    # commit results along with grammar file

If not using Docker, install antlr4 following these instructions.

    # in modelica_building/modelica_parser
    antlr4 -Dlanguage=Python3 modelica.g4

    # commit results along with grammar file

Managed Tasks

Updating Licenses

To apply the copyright/license to all the files, run the following managed task. To update, then update the script, and then rerun the managed task.

    ./setup.py update_licenses

Testing

To run the tests, simply run the following:

    poetry run pytest

Known Issues

  • The transformations occur on strings which are immutable. Need to investigate using byte arrays. This does not cause errors, but can be slow when parsing really large modelica files.

Release Instructions

  1. Bump version to <NEW_VERSION> in pyproject.toml (use semantic versioning, e.g., 0.5.0, 0.5.1.rc1).
  2. Run poetry run pre-commit run --all-files
  3. In a prep-release branch, push the changes to GitHub and draft a release against the latest branch.
    • Run 'auto-generate changelog' and copy the contents to the CHANGELOG.md. Cull any items that are repeated.
    • Discard the draft release (you will create an official one off of the main branch later)
  4. Merge the prep-release branch into develop.
  5. From command line, merge develop into main with: git checkout main; git pull; git merge --ff-only origin develop; git push
  6. In GitHub, tag the release against main. Copy and paste the changelog entry into the notes. Verify the release is posted to PyPI.