From 73489502c59053a0073f72d709b102eaf7e3661f Mon Sep 17 00:00:00 2001
From: hofmockel <82425084+hofmockel@users.noreply.github.com>
Date: Sun, 12 Jul 2026 11:05:12 -0700
Subject: [PATCH] fix: to_graphml crashes on dict/list-valued graph, node, and
 edge attributes

nx.write_graphml only accepts scalar attribute values. to_graphml already
coerces None -> "" (fixing a prior ValueError, #1502), but a dict or list
value still raises TypeError: GraphML does not support type <class 'dict'>
(or 'list') as data values. This hits real graphs in practice: a per-node
"metadata" dict, or attach_hyperedges()'s graph-level "hyperedges" list,
both crash `graphify export graphml` outright.

Fixed by extending the existing scrub step to also JSON-serialize any
dict/list-valued attribute (graph-level, node, and edge) before calling
nx.write_graphml, alongside the existing None handling.

Regression test: test_to_graphml_tolerates_dict_and_list_attribute_values
(tests/test_export.py), mirroring the existing None-value test. Verified
against a synthetic graph with a node metadata dict, an edge list attribute,
and a graph-level hyperedges list, and against a real 2,300+ node project
graph that previously crashed on all three.
---
 graphify/export.py   | 18 ++++++++++++++----
 tests/test_export.py | 19 +++++++++++++++++++
 2 files changed, 33 insertions(+), 4 deletions(-)

diff --git a/graphify/export.py b/graphify/export.py
index be44a9d03..217c1013e 100644
--- a/graphify/export.py
+++ b/graphify/export.py
@@ -912,14 +912,24 @@ def to_graphml(
         for k in [k for k in attrs if k.startswith("_")]:
             del attrs[k]
     # nx.write_graphml raises ValueError on None attribute values; replace with "".
+    # It also only accepts scalar attribute values, so dict/list-valued
+    # attributes (e.g. per-node "metadata", graph-level "hyperedges") must be
+    # serialized to a JSON string first, or the writer raises TypeError.
+    def _scrub(val):
+        if val is None:
+            return ""
+        if isinstance(val, (dict, list)):
+            return json.dumps(val)
+        return val
+
+    for key, val in list(H.graph.items()):
+        H.graph[key] = _scrub(val)
     for node_id in H.nodes():
         for key, val in list(H.nodes[node_id].items()):
-            if val is None:
-                H.nodes[node_id][key] = ""
+            H.nodes[node_id][key] = _scrub(val)
     for u, v in H.edges():
         for key, val in list(H.edges[u, v].items()):
-            if val is None:
-                H.edges[u, v][key] = ""
+            H.edges[u, v][key] = _scrub(val)
     nx.write_graphml(H, output_path)
 
 
diff --git a/tests/test_export.py b/tests/test_export.py
index 7194d8b78..b6ab84860 100644
--- a/tests/test_export.py
+++ b/tests/test_export.py
@@ -99,6 +99,25 @@ def test_to_graphml_tolerates_none_attribute_values():
         content = out.read_text()
         assert "<graphml" in content
 
+def test_to_graphml_tolerates_dict_and_list_attribute_values():
+    """nx.write_graphml only accepts scalar attribute values and raises
+    TypeError on a dict or list — e.g. a node's "metadata" dict, or a
+    graph-level "hyperedges" list (attach_hyperedges()). to_graphml must
+    JSON-serialize these instead of passing them through unmodified."""
+    G = make_graph()
+    communities = cluster(G)
+    a_node = next(iter(G.nodes()))
+    G.nodes[a_node]["metadata"] = {"kind": "file"}
+    if G.number_of_edges():
+        u, v = next(iter(G.edges()))
+        G.edges[u, v]["tags"] = ["a", "b"]
+    G.graph["hyperedges"] = [{"nodes": [a_node], "label": "x"}]
+    with tempfile.TemporaryDirectory() as tmp:
+        out = Path(tmp) / "graph.graphml"
+        to_graphml(G, communities, str(out))  # must not raise
+        content = out.read_text()
+        assert "<graphml" in content
+
 def test_to_html_creates_file():
     G = make_graph()
     communities = cluster(G)