🇫🇷 Version française

[!IMPORTANT] Every fixture, rulebook, and calculator output in this lab is synthetic and non-binding. Nothing here represents an actual Desjardins product, rate, or policy, and no regulator or insurer has reviewed or endorsed this material.

Overview

Item	Value
Duration	20 minutes
Level	Beginner
Prerequisites	Lab 00

Learning Objectives

By the end of this lab, you will be able to:

• Describe the five synthetic fixtures in data/synthetic/fixtures/ and the calculation status each one produces
• Explain how data/synthetic/quote-contract.schema.json uses if/then/else to tie a calculation’s status to its amountCents and issues fields
• Explain how the schema ties a case’s workflow state to its approvedRevision field
• Validate every fixture against the schema yourself

Exercises

Exercise 1.1: Read the Fixture Set

Open each file in data/synthetic/fixtures/ and note its fixtureId, input, and expectedCalculation.status:

Fixture	Vehicle class	Plan	Expected status
case-syn-001.json	COMPACT	TRAINING_EXTENDED	READY (100000 cents)
case-syn-002-sedan.json	SEDAN	TRAINING_BASIC	READY
case-syn-003-unsupported.json	UNKNOWN	TRAINING_BASIC	UNSUPPORTED
case-syn-004-revision.json	SEDAN	TRAINING_EXTENDED	READY (draftRevision 2)
case-syn-005-missing-plan.json	COMPACT	null	INCOMPLETE

Every fixture carries "dataClass": "SYNTHETIC_ONLY" and the same RULEBOOK-SYN-ON rulebook, pinned at "authority": "WORKSHOP_AUTHORS_ONLY".

Exercise 1.2: Read the Schema’s Conditional Logic

Open data/synthetic/quote-contract.schema.json and find the calculation definition’s if/then/else block. When status is READY, the schema requires a populated amountCents, exactly three ruleIds, and zero issues. For any other status, it requires a null amountCents and at least one issue code. Find the equivalent rule for workflow: an APPROVED state requires a non-null approvedRevision, while every other state requires it to be null.

Exercise 1.3: Validate the Fixtures Against the Schema

python -c "
import json, pathlib
from jsonschema import validate

schema = json.loads(pathlib.Path('data/synthetic/quote-contract.schema.json').read_text())
for path in sorted(pathlib.Path('data/synthetic/fixtures').glob('*.json')):
    fixture = json.loads(path.read_text())
    validate(instance=fixture, schema=schema)
    print(f'{path.name}: OK')
"

Expected result: every fixture prints OK, with no ValidationError raised.

Exercise 1.4 (Hands-on): Explain the Unsupported and Revision Cases

Answer, in your own words, using only the fixture content and the schema:

1. Why does case-syn-003-unsupported.json produce UNSUPPORTED instead of INCOMPLETE, even though every input field is present?
2. case-syn-004-revision.json has draftRevision: 2. What does that tell you about this case’s history, and why would the schema reject draftRevision: 2 paired with approvedRevision: null if the workflow state were APPROVED?

Validation Checklist

• You can name all five fixtures and the calculation status each is expected to produce
• The Exercise 1.3 script prints OK for every fixture with no validation error
• You can explain why an unknown vehicle class produces UNSUPPORTED rather than a missing-field issue
• You can explain the schema’s APPROVED requires approvedRevision rule from Exercise 1.2

Knowledge Check

• Which fixture is designed to exercise revision invalidation, and which field number changed to signal it?
• What value does rulebook.authority carry on every fixture, and why does that matter for a learner reading this data?

Next Steps

Continue to Lab 02: The Deterministic Calculator.