Implement Data Validation Error Format

Author: nichtich

README.md

@@ -182,6 +182,8 @@ If the data directory contains a file `bartoc.json` with an array of JSKOS recor
 
 There is a minimal HTML interface at root path (**GET /**) to try out the API. This is more useful than an interface generated automatically, for instance with Swagger. The API is not meant to be publically available (there is no authentification), so there is no need for an [OpenAPI](https://swagger.io/specification/) document anyway.
 
+A HTTP 400 error with response body in [Data Validation Error Format](https://gbv.github.io/validation-error-format/) is returned if collection metadata or mapping source metadata does not conform to its corresponding JSON Schema.
+
 ### General endpoints
 
 #### GET /status.json

app.py

@@ -41,16 +41,18 @@ def init(**config):
     mappings = MappingRegistry(**app.config)
 
 
-@app.errorhandler(ValidationError)
-def handle_validation_error(e):
-    return jsonify(error="ValidationError", message=str(e), code=400), 400
-
-
 @app.errorhandler(ApiError)
 def handle_error(e):
     return jsonify(e.to_dict()), type(e).code
 
 
+@app.errorhandler(ValidationError)
+def handle_error(e):
+    e = e.to_dict()
+    e["code"] = 400
+    return jsonify(e), 400
+
+
 def route(method, path, fn):
     fn.__name__ = f'{method}-{path}'
     app.add_url_rule(path, methods=[method], view_func=fn)

lib/__init__.py

@@ -1,13 +1,14 @@
 from .collections import CollectionRegistry
 from .terminologies import TerminologyRegistry
 from .mappings import MappingRegistry
-from .errors import ApiError, NotFound, NotAllowed, ValidationError, ServerError
+from .errors import ApiError, NotFound, NotAllowed, ServerError
 from .utils import read_json, write_json
 from .rdf import TripleStore, triple_iterator
 from .rdffilter import RDFFilter
+from .validate import validateJSON, ValidationError
 
 
 __all__ = [CollectionRegistry, TerminologyRegistry, triple_iterator,
            MappingRegistry, ApiError, NotFound, NotAllowed, read_json,
            write_json, ValidationError, ServerError, TripleStore,
-           RDFFilter]
+           RDFFilter, validateJSON]

lib/errors.py

@@ -1,4 +1,4 @@
-from jsonschema import ValidationError  # noqa
+from .validate import ValidationError  # noqa
 
 
 class ApiError(Exception):

lib/registry.py

@@ -2,7 +2,7 @@
 from datetime import datetime
 from shutil import copy, copyfileobj, rmtree
 import urllib
-from jsonschema import validate
+from .validate import validateJSON
 from .rdf import jsonld2nt, TripleStore
 from .rdffilter import RDFFilter
 from .log import Log
@@ -60,7 +60,7 @@ def validate(self, item, id=None):
         item["uri"] = self.prefix + id
         item["partOf"] = [self.graph]
         if self.schema:
-            validate(instance=item, schema=self.schema)
+            validateJSON(item, schema=self.schema)
         return item
 
     def list(self):

lib/validate.py

@@ -0,0 +1,29 @@
+import jsonschema
+
+# Data Validation Error Format
+
+
+class ValidationError(Exception):
+    def __init__(self, message, position=None):
+        super().__init__(message)
+        self.position = position
+
+    def to_dict(self):
+        e = {"message": str(self)}
+        if self.position:
+            e["position"] = self.position
+        return e
+
+
+def validateJSON(data, schema):
+    try:
+        jsonschema.validate(data, schema)
+    except jsonschema.ValidationError as err:
+        pos = ""
+        for elem in err.absolute_path:
+            if isinstance(elem, int):
+                pos += "/" + str(elem)
+            else:
+                pos += "/" + elem.replace("~", "~0").replace("/", "~1")
+        pos = {"jsonpointer": pos}
+        raise ValidationError(err.message, pos)

tests/test_validate.py

@@ -0,0 +1,32 @@
+import pytest
+
+from lib import ValidationError, validateJSON
+
+
+def test_jsonschema():
+    data = {  # Example from JSON Pointer RFC
+        "foo": ["bar"],
+        "": 0,
+        "a/b": 1,
+        "c%d": 2,
+        "e^\nf": 3,
+        " ": 7,
+        "m~n": 8
+    }
+
+    def fail(prop, pos, check={"type": "string"}):
+        try:
+            schema = {"type": "object", "properties": {}}
+            schema["properties"][prop] = check
+            validateJSON(data, schema)
+            assert prop == "ValidationError should have been thrown!"
+        except ValidationError as e:
+            assert e.position == {"jsonpointer": pos}
+
+    fail("foo", "/foo/0", {"type": "array", "items": {"type": "number"}})
+    fail("", "/")
+    fail("a/b", "/a~1b")
+    fail("c%d", "/c%d")
+    fail("e^\nf", "/e^\nf")
+    fail(" ", "/ ")
+    fail("m~n", "/m~0n")