From a4ed5a9171ca2ca4491f2a5844ce00b91b8fd4e1 Mon Sep 17 00:00:00 2001
From: Weaselbot <weaselbot@weaselab.dev>
Date: Tue, 23 Jun 2026 12:27:05 -0400
Subject: [PATCH] schemagen: reject duplicate unknown keys in non-strict
 objects

Track unknown keys in a per-object unordered_set so that permissive
objects (additionalProperties absent/true) still reject duplicate keys,
matching the README guarantee.

- Add std::unordered_set<std::string> to Frame.
- Insert unknown keys in cbKeyData and reject duplicates before skipping.
- Add a permissive "loose" subobject to example.schema.json.
- Test single unknown key accepted and duplicate unknown/known keys rejected.
---
 contrib/schemagen/example.schema.json     |  8 ++++++++
 contrib/schemagen/test_gen.cpp            | 14 ++++++++++++++
 contrib/schemagen/weaseljson_schemagen.py |  8 +++++++-
 3 files changed, 29 insertions(+), 1 deletion(-)

diff --git a/contrib/schemagen/example.schema.json b/contrib/schemagen/example.schema.json
index e1c0f6b..a0ff867 100644
--- a/contrib/schemagen/example.schema.json
+++ b/contrib/schemagen/example.schema.json
@@ -52,6 +52,14 @@
         }
       }
     },
+    "loose": {
+      "type": "object",
+      "properties": {
+        "known": {
+          "type": "string"
+        }
+      }
+    },
     "friends": {
       "type": "array",
       "items": {
diff --git a/contrib/schemagen/test_gen.cpp b/contrib/schemagen/test_gen.cpp
index 577e983..c52d9bc 100644
--- a/contrib/schemagen/test_gen.cpp
+++ b/contrib/schemagen/test_gen.cpp
@@ -186,6 +186,20 @@ int main() {
   expectReject(R"([1,2,3])", "array where object expected (root)");
   expectReject(R"({"name":"x","age":1,)", "truncated / invalid json");
 
+  // ---- duplicate keys in permissive (additionalProperties allowed) object
+  // ----
+  {
+    RootBuilder b;
+    WeaselJsonStatus s =
+        parseStrided(b, R"({"name":"x","age":1,"loose":{"unknown":1}})");
+    CHECK(s == WeaselJson_OK);
+    printf("ok   single unknown key in non-strict object\n");
+  }
+  expectReject(R"({"name":"x","age":1,"loose":{"unknown":1,"unknown":2}})",
+               "duplicate unknown key in non-strict object");
+  expectReject(R"({"name":"x","age":1,"loose":{"known":"a","known":"b"}})",
+               "duplicate known key in non-strict object");
+
   if (failures == 0) {
     printf("\nALL TESTS PASSED\n");
     return 0;
diff --git a/contrib/schemagen/weaseljson_schemagen.py b/contrib/schemagen/weaseljson_schemagen.py
index 11b268b..3143cc3 100644
--- a/contrib/schemagen/weaseljson_schemagen.py
+++ b/contrib/schemagen/weaseljson_schemagen.py
@@ -531,6 +531,7 @@ class Emitter:
 #include <optional>
 #include <string>
 #include <string_view>
+#include <unordered_set>
 #include <vector>
 
 #include "weaseljson.h"
@@ -885,6 +886,7 @@ private:
     void *dest;
     int field = -1;              // object: selected field (-1 want key, -2 skip)
     std::vector<uint64_t> seen;  // populated field bitset (object frames)
+    std::unordered_set<std::string> unknown;  // unknown keys seen in non-strict objects
   }};
   static constexpr int kWantKey = -1;
   static constexpr int kSkip = -2;
@@ -968,12 +970,16 @@ private:
     scratch_.append(buf, len);
     if (!done) return;
     int idx = matchKey(f.kind, scratch_);
-    scratch_.clear();
     if (idx < 0) {{
       if (isStrict(f.kind)) {{ reject(); return; }}
+      if (!f.unknown.insert(scratch_).second) {{
+        reject(); return;  // duplicate unknown key
+      }}
+      scratch_.clear();
       f.field = kSkip;
       return;
     }}
+    scratch_.clear();
     if (f.seen[idx >> 6] & (1ull << (idx & 63))) {{
       reject(); return;  // duplicate key
     }}