fix(validation): enforce strict CDR types and storage ranges

app/common/validation.py

@@ -0,0 +1,49 @@
+"""Shared wire schema. Semantic storage validation belongs to the worker."""
+import json
+import math
+
+STRING_FIELDS = ("raw_id", "device_id", "caller_num", "svc", "lac", "ci")
+INTEGER_FIELDS = ("rat", "epoch_start", "epoch_end", "up_bytes", "down_bytes")
+REQUIRED_FIELDS = STRING_FIELDS + INTEGER_FIELDS
+
+
+class PayloadError(ValueError):
+    def __init__(self, reason, detail, field=None):
+        super().__init__(detail)
+        self.reason = reason
+        self.field = field
+
+
+def parse_payload(data):
+    def reject_constant(value):
+        raise ValueError(f"Non-finite number: {value}")
+    def finite_float(value):
+        number = float(value)
+        if not math.isfinite(number):
+            raise ValueError(f"Non-finite number: {value}")
+        return number
+    try:
+        payload = json.loads(data.decode("utf-8"), parse_constant=reject_constant, parse_float=finite_float)
+    except (UnicodeError, ValueError, RecursionError) as exc:
+        raise PayloadError("invalid_json", str(exc)) from exc
+    if not isinstance(payload, dict):
+        raise PayloadError("invalid_payload", "Expected a JSON object")
+    return payload
+
+
+def validate_structure(payload):
+    for field in REQUIRED_FIELDS:
+        if payload.get(field) is None:
+            raise PayloadError("missing_required_field", f"Field {field} is required", field)
+    for field in STRING_FIELDS:
+        if not isinstance(payload[field], str) or not payload[field].strip():
+            raise PayloadError("invalid_payload", f"Field {field} must be a nonempty string", field)
+        try:
+            payload[field].encode("utf-8")
+        except UnicodeError as exc:
+            raise PayloadError("invalid_payload", f"Field {field} contains invalid Unicode", field) from exc
+    for field in INTEGER_FIELDS:
+        if type(payload[field]) is not int:
+            raise PayloadError("invalid_payload", f"Field {field} must be an integer", field)
+    if payload.get("src_ip") is not None and not isinstance(payload["src_ip"], str):
+        raise PayloadError("invalid_payload", "Field src_ip must be a string or null", "src_ip")