From 72fe6ab3f412fbfd802c6b8694c880db592e0cea Mon Sep 17 00:00:00 2001
From: Bee Webb <bee.webb@opendataservices.coop>
Date: Sun, 30 Jun 2024 15:27:31 +0000
Subject: [PATCH] json_input: If `json_dict` is not a dict, return a useful
 error/warning

https://github.com/OpenDataServices/flatten-tool/issues/442
---
 CHANGELOG.md                         |  4 ++++
 flattentool/json_input.py            |  9 +++++++++
 flattentool/tests/test_json_input.py | 21 +++++++++++++++++++++
 3 files changed, 34 insertions(+)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 5cc79d8..cca8036 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -13,6 +13,10 @@ and this project adheres to [Semantic Versioning](http://semver.org/spec/v2.0.0.
 - Ignore null characters in the input CSV file when reading configuration from the header rows
   https://github.com/OpenDataServices/flatten-tool/pull/446
 
+### Changed
+
+- If `json_dict` is not a dict, return a useful error/warning https://github.com/OpenDataServices/flatten-tool/issues/442
+
 ## [0.24.2] - 2024-06-12
 
 ### Fixed
diff --git a/flattentool/json_input.py b/flattentool/json_input.py
index ab28ec7..8fd8ac9 100644
--- a/flattentool/json_input.py
+++ b/flattentool/json_input.py
@@ -28,6 +28,7 @@
 import zc.zlibstorage
 import ZODB.FileStorage
 
+from flattentool.exceptions import DataErrorWarning
 from flattentool.i18n import _
 from flattentool.input import path_search
 from flattentool.schema import make_sub_sheet_name
@@ -310,6 +311,14 @@ def parse(self):
                 # This is particularly useful for IATI XML, in order to not
                 # fall over on empty activity, e.g. <iati-activity/>
                 continue
+
+            if not isinstance(json_dict, dict):
+                warn(
+                    _(f"The value at index {num} is not a JSON object"),
+                    DataErrorWarning,
+                )
+                continue
+
             self.parse_json_dict(json_dict, sheet=self.main_sheet)
             # only persist every 2000 objects. peristing more often slows down storing.
             # 2000 top level objects normally not too much to store in memory.
diff --git a/flattentool/tests/test_json_input.py b/flattentool/tests/test_json_input.py
index a98186e..b15b3da 100644
--- a/flattentool/tests/test_json_input.py
+++ b/flattentool/tests/test_json_input.py
@@ -112,6 +112,27 @@ def test_parse_basic_json_dict():
     assert parser.sub_sheets == {}
 
 
+def test_parse_not_json_dict(recwarn):
+    parser = JSONParser(root_json_dict=[["test"], {"a": "b"}, "test"])
+    assert list(parser.main_sheet) == ["a"]
+    assert list(parser.main_sheet.lines) == [{"a": "b"}]
+    assert parser.sub_sheets == {}
+
+    assert len(recwarn) == 2
+
+    w = recwarn.pop(UserWarning)
+    assert (
+        repr(w.message)
+        == "DataErrorWarning('The value at index 0 is not a JSON object')"
+    )
+
+    w = recwarn.pop(UserWarning)
+    assert (
+        repr(w.message)
+        == "DataErrorWarning('The value at index 2 is not a JSON object')"
+    )
+
+
 def test_parse_nested_dict_json_dict():
     parser = JSONParser(
         root_json_dict=[