JSON schema for data.

1 files changed, 75 insertions, 0 deletions

diff --git a/src/schema.py b/src/schema.py
new file mode 100644
index 0000000..37da5aa
--- /dev/null
+++ b/src/schema.py
@@ -0,0 +1,75 @@
+import json
+
+import jsonschema
+
+schema = {
+    "$schema": "http://json-schema.org/schema#",
+    "type": "object",
+    "properties": {
+        "document_version": {
+            "type": "integer"
+        },
+        "ip": {
+            "type": "string"
+        },
+        "port": {
+            "type": "integer"
+        },
+        "whois_description": {
+            "type": "string"
+        },
+        "asn": {
+            "type": "string"
+        },
+        "asn_country_code": {
+            "type": "string"
+        },
+        "ptr": {
+            "type": "string"
+        },
+        "abuse_mail": {
+            "type": "string"
+        },
+        "domain": {
+            "type": "string"
+        },
+        "timestamp_in_utc": {
+            "type": "string"
+        },
+        "user_presentation": {
+            "type": "object",
+            "properties": {
+                "description": {
+                    "type": "string"
+                },
+                "data": {
+                    "type": "object"
+                }
+            }
+        }
+    },
+    "required": [
+        "document_version",
+        "ip",
+        "port",
+        "timestamp_in_utc",
+        "user_presentation"
+    ]
+}
+
+
+def validate_collector_data(json_blob):
+    try:
+        jsonschema.validate(json_blob, schema)
+    except jsonschema.exceptions.ValidationError as e:
+        print(f'Validation failed with error: {e}')
+        return False
+
+    return True
+
+
+if __name__ == '__main__':
+    with open('example_data.json') as fd:
+        json_data = json.loads(fd.read())
+
+    validate_collector_data(json_data)