Transform¶

The transform command copies or edits a model within the database, from one schema to another. This is crucial for working with submodels, version control and multilinguality.

Basic Syntax¶

crunch_uml transform -ttp <type> -sch_to <target_schema> [options]

Transformation Types¶

Type	Option `-ttp`	Description
Copy	`copy`	Deep copy of a package hierarchy to another schema
Plugin	`plugin`	Custom transformation via a dynamically loaded plugin

Options¶

Option	Description
`-sch_from`	Source schema (default: `default`)
`-sch_to`	Target schema (required)
`-sch_to_cln`	Clean the target schema first
`-ttp, --transformationtype`	Type: `copy` or `plugin`
`-rt_pkg, --root_package`	ID of the root package to be copied
`-m_gen, --materialize_generalizations`	Copy attributes from superclasses to subclasses
`-plug_mod, --plugin_file_name`	Path to plugin Python file
`-plug_cl, --plugin_class_name`	Class name of the plugin

Copy Transformer¶

The copy transformer makes a deep copy of a package hierarchy (including all classes, attributes, associations, enumerations and generalizations) from one schema to another.

graph LR
    A["Schema: default<br/>━━━━━━━━━━━━<br/>Complete model"] -->|"copy"| B["Schema: my_model<br/>━━━━━━━━━━━━<br/>Selected<br/>submodel only"]

Example: Copy submodel¶

# Copy only the package with ID EAPK_12345 to schema "submodel"
crunch_uml transform -ttp copy -sch_to submodel -rt_pkg EAPK_12345

Example: Flatten inheritance¶

With --materialize_generalizations attributes from superclasses are copied to subclasses. This is useful when you need a "flat" model, for example for generating database tables.

crunch_uml transform -ttp copy -sch_to flat_model -rt_pkg EAPK_12345 \
    -m_gen True

Plugin Transformer¶

For custom transformations you can write your own Python plugin:

crunch_uml transform -ttp plugin \
    -plug_mod /path/to/my_plugin.py \
    -plug_cl MyTransformation \
    -sch_to result

The plugin class must extend crunch_uml.transformers.plugin.Plugin:

from crunch_uml.transformers.plugin import Plugin

class MyTransformation(Plugin):
    def transform(self, args, schema_from, schema_to):
        # Read data from schema_from
        # Edit and write to schema_to
        ...

Typical Workflows¶

Prepare version comparison¶

# 1. Import current version
crunch_uml import -f model_v2.xmi -t eaxmi -db_create

# 2. Copy to a named schema
crunch_uml transform -ttp copy -sch_to current_version -rt_pkg ROOT_ID

# 3. Import previous version in another schema
crunch_uml -sch previous_raw import -f model_v1.xmi -t eaxmi

# 4. Copy that as well to a named schema
crunch_uml transform -ttp copy -sch_from previous_raw -sch_to previous_version -rt_pkg ROOT_ID

# 5. Now you can generate a diff (see Export)

Build multilingual model¶

# 1. Import and copy the base model
crunch_uml import -f model.xmi -t eaxmi -db_create
crunch_uml transform -ttp copy -sch_to translation_en -rt_pkg ROOT_ID

# 2. Import English translations into that schema
crunch_uml -sch translation_en import -f translations_en.json -t i18n --language en