feat: Add a class diagram exporter

This is effectively the inverse operation of the existing json2capella,
i.e. capella2json.

Author: Wuestengecko

json2capella/datatypes.py

@@ -38,7 +38,9 @@ class Struct(_BaseModel):
 
     @p.field_validator("extends")
     @classmethod
-    def extends_is_valid_dotted_name(cls, value: str) -> str:
+    def extends_is_valid_dotted_name(cls, value: str | None) -> str | None:
+        if value is None:
+            return value
         numdots = sum(1 for char in value if char == ".")
         if numdots not in (0, 1):
             raise ValueError(

json2capella/export.py

@@ -0,0 +1,206 @@
+# Copyright DB InfraGO AG and contributors
+# SPDX-License-Identifier: Apache-2.0
+"""The json2capella exporter CLI, i.e. capella2json."""
+
+from __future__ import annotations
+
+import logging
+import typing as t
+
+import capellambse
+import click
+from capellambse import cli_helpers
+from capellambse.metamodel import information
+
+from . import datatypes
+
+logger = logging.getLogger(__name__)
+
+
+@click.command()
+@click.option(
+    "-m",
+    "--model",
+    required=True,
+    type=cli_helpers.ModelCLI(),
+    help="Path to the Capella model.",
+)
+@click.option(
+    "-p",
+    "--package",
+    "package_id",
+    help="ID or name of the DataPkg to export.",
+)
+@click.option(
+    "-l",
+    "--list",
+    "list_packages",
+    is_flag=True,
+    help="List all data packages found in the model.",
+)
+@click.option(
+    "-o",
+    "--output",
+    default="-",
+    type=click.File("w", atomic=True),
+)
+@click.option(
+    "--indent",
+    default=None,
+    type=click.IntRange(0),
+    help="Pretty-print the JSON output, indenting by INT spaces per level.",
+)
+def main(
+    *,
+    model: capellambse.MelodyModel,
+    package_id: str,
+    list_packages: bool,
+    output: t.IO[str],
+    indent: int | None,
+) -> None:
+    if list_packages:
+        print("The following data packages were found in the model:")
+        for i in model.search("DataPkg"):
+            print(" ", i._short_repr_())
+        return
+
+    if not package_id:
+        raise click.UsageError(
+            "--package is required (use --list to find a package)"
+        )
+
+    package_obj = _find_package(model, package_id)
+    package = _convert_package(package_obj)
+    dump = package.model_dump_json(indent=indent, exclude_defaults=True)
+    with output:
+        output.write(dump)
+
+
+def _find_package(
+    model: capellambse.MelodyModel, package_id: str, /
+) -> information.DataPkg:
+    try:
+        obj = model.by_uuid(package_id)
+    except KeyError:
+        pass
+    else:
+        if not isinstance(obj, information.DataPkg):
+            raise click.UsageError(
+                f"Expected a DataPkg at ID {package_id!r},"
+                f" but found a {type(obj).__name__} instead"
+            )
+        return obj
+
+    packages = model.search(information.DataPkg).by_name(
+        package_id, single=False
+    )
+    if len(packages) < 1:
+        raise click.UsageError(
+            f"Couldn't find a DataPkg with ID or name {package_id!r}"
+        )
+    if len(packages) > 1:
+        pkg_strings = []
+        for i in packages:
+            try:
+                p = i.parent
+            except ValueError:
+                pkg_strings.append(f"  {i.uuid!r} (orphaned)")
+            else:
+                pkg_strings.append(f"  {i.uuid!r} in {p._short_repr_()}")
+        raise click.UsageError(
+            f"Found more than one DataPkg with name {package_id!r}:\n"
+            + "\n".join(pkg_strings)
+        )
+    (obj,) = packages
+    assert isinstance(obj, information.DataPkg)
+    return obj
+
+
+def _convert_package(obj: information.DataPkg, /) -> datatypes.Package:
+    return datatypes.Package(
+        name=obj.name,
+        info=obj.description,
+        subPackages=[_convert_package(i) for i in obj.packages],  # type: ignore[call-arg]
+        structs=[_convert_class(i) for i in obj.classes],
+        enums=[_convert_enum(i) for i in obj.enumerations],
+        prefix=obj.uuid,
+    )
+
+
+def _convert_class(obj: information.Class, /) -> datatypes.Struct:
+    if obj.super is None:
+        extends = None
+    else:
+        extends = ""
+    return datatypes.Struct(
+        name=obj.name,
+        info=obj.description,
+        extends=extends,
+        attrs=[_convert_attr(i) for i in obj.properties],
+    )
+
+
+def _convert_attr(obj: information.Property, /) -> datatypes.StructAttrs:
+    min = 0
+    max = None
+    if (c := obj.max_card) and (v := c.value) and v != "*":
+        try:
+            max = int(v)
+        except ValueError:
+            logger.warning(
+                "Cannot convert max_card value %r of %s to int, ignoring",
+                v,
+                obj._short_repr_(),
+            )
+    if (c := obj.min_card) and (v := c.value):
+        try:
+            min = int(v)
+        except ValueError:
+            logger.warning(
+                "Cannot convert min_card value %r of %s to int, ignoring",
+                v,
+                obj._short_repr_(),
+            )
+
+    if (min, max) == (0, None) or min == max:
+        if max is None:
+            mult = "*"
+        else:
+            mult = str(max)
+    else:
+        mult = f"{min}..{max or '*'}"
+
+    if obj.type:
+        typename = obj.type.name
+    else:
+        logger.warning(
+            "No type set, falling back to 'string' for %s",
+            obj._short_repr_(),
+        )
+        typename = "string"
+
+    return datatypes.StructAttrs(
+        name=obj.name,
+        info=obj.description,
+        multiplicity=mult,
+        data_type=typename,
+    )
+
+
+def _convert_enum(obj: information.datatype.Enumeration, /) -> datatypes.Enum:
+    return datatypes.Enum(
+        name=obj.name,
+        info=obj.description,
+        enumLiterals=[  # type: ignore[call-arg]
+            datatypes.EnumLiteral(
+                intId=i,  # type: ignore[call-arg]
+                name=o.name,
+                info=o.description,
+            )
+            for i, o in enumerate(obj.literals)
+        ],
+    )
+
+
+if __name__ == "__main__":
+    main()

pyproject.toml

@@ -38,6 +38,10 @@ dependencies = [
 Homepage = "https://github.com/DSD-DBS/json2capella"
 Documentation = "https://dsd-dbs.github.io/json2capella"
 
+[project.scripts]
+json2capella = "json2capella.__main__:main"
+capella2json = "json2capella.export:main"
+
 [dependency-groups]
 dev = [
   "coverage>=7.9.1",