This repository contains a set of JSON objects that developers of MARCspec parser libraries can use to test their implementations.
Tests are divided into valid and invalid. Thus there are two directories valid/ and invalid/.
Each directory contains multiple .json files and each file contains one root JSON object with multiple test objects.
The filenames might give you a hint on which tests are covered. The files starting with wildCombination_ are (wild) combinations of the test data in the files starting with valid or invalid.
Each root JSON object contains an object with the description, a schema to validate against and some tests. Each test is an object itself with a description, the test data and a validation statement (true or false).
Here is an example of teh content of a .json file
{
"description": "field tags are strings and match pattern",
"schema": {
"type": "string",
"pattern": "^([.a-z0-9]{3,3}|[.A-Z0-9]{3,3})$"
},
"tests": [
{
"description": "all wildcards",
"data": "...",
"valid": true
},
{
"description": "two wildcards left with digit",
"data": "..0",
"valid": true
},
{
"description": "one wildcard left with two digits",
"data": ".00",
"valid": true
}
]
}Not every file contains complete MARCspec references. Only the files starting with wildCombination_ and the files *FieldTag.json does.
The other files:
- *ComparisonString.json
- *Indicators.json
- *PositionOrRange.json
- *SubfieldRange.json
- *SubfieldTag.json
- *SubSpec.json
contain either valid or invalid fragments of MARCspec references. This makes it easy to test distinct functionalities (e.g. a ComparisonString paser).