schema-markdown is a schema definition and validation library.
Schemas are defined using the Schema Markdown language, which is parsed by the parse_schema_markdown function. For example:
from schema_markdown import parse_schema_markdown
model_types = parse_schema_markdown('''\
# An aggregate numerical operation
struct Aggregation
# The aggregation function - default is "Sum"
optional AggregationFunction aggregation
# The numbers to aggregate on
int[len > 0] numbers
# An aggregation function
enum AggregationFunction
Average
Sum
''')
To validate an object using the schema, use the validate_type function. For example:
from schema_markdown import validate_type
validate_type(model_types, 'Aggregation', {'numbers': [1, 2, '3', 4]})
{'numbers': [1, 2, 3, 4]}
Notice that the numerical input '3' above is type-massaged to the integer 3 by validation.
Validation fails if the object does not match the schema:
from schema_markdown import ValidationError
try:
validate_type(model_types, 'Aggregation', {'numbers': [1, 2, 'asdf', 4]})
except ValidationError as exc:
str(exc)
"Invalid value 'asdf' (type 'str') for member 'numbers.2', expected type 'int'"
Validation also fails if a member constraint is violated:
try:
validate_type(model_types, 'Aggregation', {'numbers': []})
except ValidationError as exc:
str(exc)
"Invalid value [] (type 'list') for member 'numbers', expected type 'array' [len > 0]"
To document the schema, download the documentation application stub and save the type model as JSON:
curl -O https://craigahobbs.github.io/schema-markdown-doc/extra/index.html
python3 \
-c 'from model import model_types; import json; print(json.dumps(model_types))' \
> model.json
To host locally, start a local static web server:
python3 -m http.server
This package is developed using python-build. It was started using python-template as follows:
template-specialize python-template/template/ schema-markdown/ -k package schema-markdown -k name 'Craig A. Hobbs' -k email '[email protected]' -k github 'craigahobbs' -k nomain 1