xie-dongping / modparc

modparc is a Modelica parser in Python based on parser combinator

modparc

modparc is a Modelica parser in Python based on parser combinator.

• Free software: GNU General Public License v3
• Source code: https://github.com/xie-dongping/modparc.
• Documentation: https://modparc.readthedocs.io.

Quickstart

Install the package from PyPI:

$ pip install modparc

To parse a Modelica source file `"your_modelica_file.mo"`:

import modparc
model_definition = modparc.parse_file("your_modelica_file.mo")

To list all the equations in the model_definition instance:

all_equations = model_definition.search('Equation')
for equation in all_equations:
    print(equation.code())  # The code of the equation as string

To get the name of the model loaded:

print(model_definition.name())  # get the name of the stored class
print(model_definition.class_type())  # get the type of the class

Features

• Experimentally parses Modelica Standard Library 3.2.1
• Search element of a certain class

Known Issues

• Handling tokenization of Q-IDENT and comments, which comes first?
• Assertion syntax not defined in Modelica specification
• Default recursion depth is not enough for long vector literals
• Cyclic import is neccessary for the Modelica syntax definition

Credits

This package was created with Cookiecutter and the audreyr/cookiecutter-pypackage project template.

The test cases used code from the ModelicaByExample library (MIT License by Michael Tiller).

modparc's Issues

List Connections [from François]

model_definition.connects() would give the list of connects (with lhs(), and rhs() for example)

Query with GraphQL?

http://graphql.org/learn/

List Model Instances [from François]

model_definition.model_instances() would give a list of model objects (and each model object could be called with id(), type(), redeclare()

List Model Parameters [by François]

model_definition.parameters() would give the list of parameters (with type(), id()...)

Make short_class_specifier traversable

The part after = sign is returned as the value of the short_class_specifier, but this value syntax element could itself have additional layers which could be traversed.

Related to #48

Implement Name Look-up Rules

Load a Directory of Libraries

Load the package.mo or modelica directory as a Library class which inihibits similar behaviour as StoredDefintions class

Make Name an Attribute for Class Defintion

Instead of name() we should use name for the class?

The tests should be changed too.

List Annotations

List Declared Classes [By François]

model_definition.model_declarations() would give all classes declared in the model

Make SyntaxElement.search an generator

Sometimes we just use the first element in the search function, and therefore it could be made an generator expression.

Support Element Query with XPath

Pretty Printing Syntax Elements

Printing syntax elements according to the best practices of Modelica

TOX Test for Python 3.6

Editting a Syntax Element

Editting a syntax element might changes the position for all the tokens, so this relates with issue #3

Improving Code Coverage

It would be nice to have a higher coverage in the unit test (although it works with MSL)

Create/Extends a Model

Modify Model Components with Modification

Get Type and IDENT for ClassDefinition

To know the name of the class and its type (if it is a package, model or a block)

Printing StoredDefinition According to the Original Token Position

Roundtripping the StoredDefinition and tokenizer.

License

Currently, there is a LICENSE file and a COPYING file, that seems to be redundant; the COPYING file can probably be deleted (github is looking for a LICENSE file first).

Additionally, the license text seems to be slightly changed, otherwise github would recognize what license is used and display it in the overview (next to the number of contributors):
https://github.com/blog/2252-license-now-displayed-on-repository-overview
https://github.com/blog/2335-open-source-license-descriptions-and-metadata
To fix this, I would recommend deleting both files LICENSE and COPYING and then recreate the LICENSE file using githubs webinterface Create new file. When you name the new files LICENSE, a dropdown menu appears on the right.

Support Unicode Characters in Tokenizer

Listing the Names of the Models in a Package Directory

Clarify the last clause of class specifier

class_specifier :
IDENT string_comment composition end IDENT
| IDENT "=" base_prefix name [ array_subscripts ]
[ class_modification ] comment
| IDENT "=" enumeration "(" ( [enum_list] | ":" ) ")" comment
| IDENT "=" der "(" name "," IDENT { "," IDENT } ")" comment
| extends IDENT [ class_modification ] string_comment composition
end IDENT