Data Stewardship Wizard: Models
Models of Data Stewardship Wizard data, their transformations and tooling for scripts
Installation
pip install dsw-models
pip install 'dsw-models[rendering]' # Markdown helpers of the document context
What is inside
| Module | Purpose |
|---|---|
knowledge_model.flat |
Compiled knowledge model (entity maps and UUID lists) |
knowledge_model.events |
Knowledge model events |
knowledge_model.package |
.km bundles and their packages |
knowledge_model.tree |
Nested knowledge model for authoring (by hand or by tools) |
knowledge_model.graph, .visitor |
Read-only navigation and visitors |
knowledge_model.compiler, .bundle |
Events → knowledge model |
knowledge_model.diff, .squash |
Knowledge model → events, squashing events |
knowledge_model.builder, .validation |
Fluent building and semantic validation |
knowledge_model.migrations |
Upgrading bundles from older metamodel versions |
document_template.metadata, .migrations |
template.json (local and packaged) |
document_context.wire, .graph |
Document context and its object model for templates |
project.* |
Project events, replies, content, squashing and reports |
schemas |
JSON Schemas of the models |
Models use the JSON (camelCase) names of the Data Stewardship Wizard API. They reject unknown keys unless asked otherwise:
import json
from dsw.models.knowledge_model.package import KnowledgeModelBundle
from dsw.models.strictness import UnknownKeys, load
with open('dsw_root_2.8.1.km') as file:
bundle = load(KnowledgeModelBundle, json.load(file)) # strict
# load(KnowledgeModelBundle, data, unknown_keys=UnknownKeys.IGNORE) drops unknown keys
data = bundle.to_json_data() # JSON-compatible data as the server writes it
Knowledge models
Compile a bundle and navigate the result:
from dsw.models.knowledge_model import graph
from dsw.models.knowledge_model.bundle import compile_bundle
km = compile_bundle(bundle) # flat.KnowledgeModel of the bundle's package
km_graph = graph.KnowledgeModel(km)
for chapter in km_graph.chapters:
for question in chapter.questions:
print(question.title, [tag.name for tag in question.tags])
Every node knows its parent, children, ancestors and resolved references
(answer.follow_up_questions, question.required_phase, item_select.list_question …).
Reverse lookups answer questions such as "which questions use this integration"
(integration.questions).
Walk the knowledge model with a visitor; the most specific visit_<kind> method wins:
from dsw.models.knowledge_model.visitor import SKIP, KnowledgeModelVisitor, walk
class QuestionsWithoutReferences(KnowledgeModelVisitor):
def __init__(self):
self.found = []
def visit_question(self, question):
if not question.references:
self.found.append(question)
def visit_answer(self, answer):
if answer.label == 'No':
return SKIP # do not descend into follow-up questions
visitor = QuestionsWithoutReferences()
walk(km_graph, visitor)
Building and changing knowledge models
from dsw.models.knowledge_model.builder import KnowledgeModelBuilder
builder = KnowledgeModelBuilder()
phase = builder.phase('Before submitting the proposal')
findability = builder.metric('Findability', abbreviation='F')
chapter = builder.chapter('Data description')
question = chapter.options_question('Will you reuse existing data?', required_phase=phase)
question.answer('Yes', metric_measures=[(findability, 1.0, 1.0)]).value_question('Which data?')
question.answer('No')
km = builder.build()
bundle = builder.to_bundle(organization_id='example', km_id='data', version='1.0.0', name='Data')
To change an existing knowledge model, edit a copy of the flat model and let diff produce the
events (entities are matched by UUID; the result is verified by replaying it):
from dsw.models.knowledge_model.diff import decompile, diff
from dsw.models.knowledge_model.squash import squash
changed = km.model_copy(deep=True)
changed.entities.chapters[changed.chapter_uuids[0]].title = 'Data'
events = squash(diff(km, changed)) # or decompile(changed) for a new package
validation.validate(km) reports errors, warnings and infos with a location and a hint, e.g.
[warning] too-few-answers at Chapter 'Data' › OptionsQuestion 'Reuse?': ….
convert.flat_to_tree() and convert.tree_to_flat() convert to and from the nested model;
schemas.knowledge_model_schema('tree', compact=True) gives a smaller schema without titles,
defaults and discriminator mappings.
Older bundles
from dsw.models.knowledge_model.migrations import migrate_bundle
bundle = load(KnowledgeModelBundle, migrate_bundle(old_data), unknown_keys=UnknownKeys.IGNORE)
Projects
from dsw.models.project.content import compile_project_events
from dsw.models.project.report import generate_report
from dsw.models.project.walker import walk_replies
content = compile_project_events(events) # replies, labels, phase
report = generate_report(km, content.replies, content.phase_uuid)
for item in walk_replies(km_graph, content.replies): # questions as the questionnaire shows them
print(item.path, item.question.title, item.reply)
Document templates and contexts
from dsw.models.document_template.metadata import DocumentTemplateMetadata
from dsw.models.document_template.migrations import migrate_template_json
result = migrate_template_json(old_template_json) # best effort, with warnings
metadata = load(DocumentTemplateMetadata, result.data)
document_context.wire.DocumentContext validates the document context sent to the document
worker; document_context.graph.DocumentContext is the object model templates get from
ctx|to_context_obj.
JSON Schemas
from dsw.models import schemas
schema = schemas.knowledge_model_bundle_schema(schema_id='https://example.org/km-bundle.json')
Available: knowledge_model_bundle_schema, knowledge_model_schema, template_json_schema,
document_context_schema and project_events_schema (JSON Schema 2020-12).
Examples
Runnable scripts — validating a .km bundle, building one, diffing two knowledge models and
generating JSON Schemas — are in examples/.
License
This project is licensed under the Apache License v2.0 - see the LICENSE file for more details.
1 maintainer
Metadata
1 maintainer
Release files for dsw-models 4.35.0
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | |
|---|---|---|---|---|
| dsw_models-4.35.0.tar.gz | 2.3 MB | With provenance | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| dsw_models-4.35.0-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 2.4 MB
Release files / dsw_models-4.35.0.tar.gz
| Download URL | dsw_models-4.35.0.tar.gz |
|---|---|
| Size | 2.3 MB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
4c0458b63e981dcde318421ccfb2e5b15d72f803f7bb5b3fda37bc23e46f6882
|
|
BLAKE2b-256 checksum How to use checksums |
fe5412877cc79a5ac1f87ae641c246402818dbfb30cb9affd7af8f5028886096
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 6, 2026.
Transparency logRelease files / dsw_models-4.35.0-py3-none-any.whl
| Download URL | dsw_models-4.35.0-py3-none-any.whl |
|---|---|
| Size | 88.0 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
248cf1ba0f357c00e173fd2c1960c58d0d99e8911e004f5ae70c3dd33605398b
|
|
BLAKE2b-256 checksum How to use checksums |
2f6547a42ef1e471ef13c11758cd30b7febac175fe4c1202c7ecf79b8b0cd780
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
twine/7.0.0 CPython/3.13.14
|
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 6, 2026.
Transparency logSecurity for dsw-models 4.35.0
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
Full provenance
All 2 files in this release have been uploaded with provenance metadata.
Attestation for dsw_models-4.35.0.tar.gz
PyPI Publish Attestation
PyPI verified that this artifact, at this checksum, originated from the publisher listed below.
Signed by GitHub Actions, verified by PyPI on Oct 6, 2026.
Transparency logA note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.
