Add test coverage for documented lenient BSON input handling

Issue #5333 documented three intentionally-lenient behaviors of the BSON
reader (any non-zero byte accepted as a boolean `true`, BSON array element
keys not validated against the required decimal sequence, and the payload
of binary subtype 0x02 "old binary" returned as-is including its inner
length prefix), but none of them was pinned by a test, so a future change
could silently regress the documented behavior.

Also add coverage for the out_of_range.412 length-overflow check
(shared by binary, string, and (sub-)document BSON length fields) for
the string and document cases; only the binary case was previously
tested.

Signed-off-by: Niels Lohmann <mail@nlohmann.me>

tests/CMakeLists.txt

@@ -177,24 +177,6 @@ json_test_add_test_for(src/unit-comparison.cpp
     MAIN test_main CXX_STANDARDS ${test_cxx_standards} ${test_force}
 )
 
-# test the parser again with JSON_DIAGNOSTIC_POSITIONS enabled
-json_test_set_test_options(test-class_parser_diagnostic_positions
-    COMPILE_DEFINITIONS JSON_DIAGNOSTIC_POSITIONS=1
-)
-json_test_add_test_for(src/unit-class_parser.cpp
-    NAME test-class_parser_diagnostic_positions
-    MAIN test_main CXX_STANDARDS ${test_cxx_standards} ${test_force}
-)
-
-# test diagnostic positions again without regular diagnostics (JSON pointer paths)
-json_test_set_test_options(test-diagnostic-positions_only
-    COMPILE_DEFINITIONS JSON_DIAGNOSTICS=0
-)
-json_test_add_test_for(src/unit-diagnostic-positions.cpp
-    NAME test-diagnostic-positions_only
-    MAIN test_main CXX_STANDARDS ${test_cxx_standards} ${test_force}
-)
-
 # *DO NOT* use json_test_set_test_options() below this line
 
 #############################################################################

tests/src/unit-bson.cpp

@@ -38,6 +38,26 @@ class huge_binary_t : public std::vector<std::uint8_t>
 using huge_binary_json = nlohmann::basic_json <
                          std::map, std::vector, std::string, bool, std::int64_t, std::uint64_t,
                          double, std::allocator, nlohmann::adl_serializer, huge_binary_t, void >;
+
+// a string type that reports a size beyond INT32_MAX without allocating that
+// much memory, so BSON length overflow can be tested for strings and
+// (embedded) documents as well, following the same idea as huge_binary_t
+class huge_string_t : public std::string
+{
+  public:
+    using std::string::string;
+    huge_string_t(const std::string& s) : std::string(s) {} // NOLINT(google-explicit-constructor,hicpp-explicit-conversions)
+
+    size_type size() const noexcept // NOLINT(readability-convert-member-functions-to-static)
+    {
+        // one byte more than the BSON length field can represent
+        return static_cast<size_type>((std::numeric_limits<std::int32_t>::max)()) + 1;
+    }
+};
+
+using huge_string_json = nlohmann::basic_json <
+                         std::map, std::vector, huge_string_t, bool, std::int64_t, std::uint64_t,
+                         double, std::allocator, nlohmann::adl_serializer, std::vector<std::uint8_t>, void >;
 } // namespace
 
 TEST_CASE("BSON")
@@ -105,10 +125,36 @@ TEST_CASE("BSON")
 
     SECTION("lengths exceeding INT32_MAX cannot be serialized to BSON")
     {
-        huge_binary_json j;
-        j["b"] = huge_binary_json::binary(huge_binary_t{});
+        // out_of_range.412 is thrown from a single shared helper
+        // (to_bson_length) that guards the BSON length fields of binary
+        // values, strings, and (embedded) documents alike
+        SECTION("binary")
+        {
+            huge_binary_json j;
+            j["b"] = huge_binary_json::binary(huge_binary_t{});
 
-        CHECK_THROWS_WITH_AS(huge_binary_json::to_bson(j), "[json.exception.out_of_range.412] BSON length 2147483661 exceeds maximum of 2147483647", huge_binary_json::out_of_range&);
+            CHECK_THROWS_WITH_AS(huge_binary_json::to_bson(j), "[json.exception.out_of_range.412] BSON length 2147483661 exceeds maximum of 2147483647", huge_binary_json::out_of_range&);
+        }
+
+        SECTION("string")
+        {
+            huge_string_json j;
+            j["s"] = huge_string_json::string_t("value");
+
+            CHECK_THROWS_WITH_AS(huge_string_json::to_bson(j), "[json.exception.out_of_range.412] BSON length 4294967308 exceeds maximum of 2147483647", huge_string_json::out_of_range&);
+        }
+
+        SECTION("document")
+        {
+            // an oversized string nested one level deep makes the
+            // *embedded* document's own length exceed INT32_MAX as well
+            huge_string_json nested;
+            nested["s"] = huge_string_json::string_t("value");
+            huge_string_json j;
+            j["nested"] = nested;
+
+            CHECK_THROWS_WITH_AS(huge_string_json::to_bson(j), "[json.exception.out_of_range.412] BSON length 6442450963 exceeds maximum of 2147483647", huge_string_json::out_of_range&);
+        }
     }
 
     SECTION("string length must be at least 1")
@@ -193,6 +239,23 @@ TEST_CASE("BSON")
             CHECK(json::from_bson(result, true, false) == j);
         }
 
+        SECTION("non-empty object with bool from a non-0/1 byte (lenient parsing)")
+        {
+            // documented lenient behavior (see gh-5333): any non-zero byte
+            // is accepted as `true`, not just 0x01
+            std::vector<std::uint8_t> const input =
+            {
+                0x0D, 0x00, 0x00, 0x00, // size (little endian)
+                0x08,               // entry: boolean
+                'e', 'n', 't', 'r', 'y', '\x00',
+                0x02,           // value = 0x02 (neither 0x00 nor 0x01)
+                0x00                    // end marker
+            };
+
+            const json expected = { { "entry", true } };
+            CHECK(json::from_bson(input) == expected);
+        }
+
         SECTION("non-empty object with double")
         {
             json const j =
@@ -499,6 +562,29 @@ TEST_CASE("BSON")
             CHECK(json::from_bson(result, true, false) == j);
         }
 
+        SECTION("array elements with non-conforming keys (lenient parsing)")
+        {
+            // documented lenient behavior (see gh-5333): BSON array element
+            // keys are not checked against the required decimal sequence
+            // "0", "1", "2", ... - elements are taken in encoded order
+            std::vector<std::uint8_t> const input =
+            {
+                0x26, 0x00, 0x00, 0x00, // size (little endian)
+                0x04, 'e', 'n', 't', 'r', 'y', '\x00', // entry: embedded array
+
+                0x1A, 0x00, 0x00, 0x00, // size (little endian)
+                0x10, '5', 0x00, 0x0A, 0x00, 0x00, 0x00, // key "5" (bogus)      -> 10
+                0x10, 'x', 0x00, 0x14, 0x00, 0x00, 0x00, // key "x" (non-numeric) -> 20
+                0x10, '1', 0x00, 0x1E, 0x00, 0x00, 0x00, // key "1" (out of order) -> 30
+                0x00, // end marker (embedded array)
+
+                0x00 // end marker
+            };
+
+            const json expected = { { "entry", json::array({10, 20, 30}) } };
+            CHECK(json::from_bson(input) == expected);
+        }
+
         SECTION("non-empty object with binary member")
         {
             const size_t N = 10;
@@ -594,6 +680,31 @@ TEST_CASE("BSON")
             CHECK(json::from_bson(result, true, false) == j);
         }
 
+        SECTION("binary member with subtype 0x02 (old binary) keeps its inner length prefix (lenient parsing)")
+        {
+            // documented lenient behavior (see gh-5333): the payload for
+            // binary subtype 0x02 ("old binary") is returned as-is,
+            // including its own inner 4-byte length prefix; it is not
+            // stripped or reinterpreted
+            std::vector<std::uint8_t> const input =
+            {
+                0x17, 0x00, 0x00, 0x00, // size (little endian)
+                0x05, 'e', 'n', 't', 'r', 'y', '\x00', // entry: binary
+
+                0x06, 0x00, 0x00, 0x00, // size of binary (little endian)
+                0x02, // "old binary" subtype
+                0x02, 0x00, 0x00, 0x00, // inner length prefix (part of the old-binary payload)
+                0x68, 0x69, // payload ('h', 'i')
+
+                0x00 // end marker
+            };
+
+            // the inner length prefix is part of the (unmodified) payload
+            const std::vector<std::uint8_t> expected_payload = {0x02, 0x00, 0x00, 0x00, 0x68, 0x69};
+            const json expected = { { "entry", json::binary(expected_payload, 0x02) } };
+            CHECK(json::from_bson(input) == expected);
+        }
+
         SECTION("Some more complex document")
         {
             json const j =

tests/src/unit-class_parser.cpp

@@ -344,50 +344,6 @@ void trailing_comma_helper(const std::string& s)
     }
 }
 
-#if JSON_DIAGNOSTIC_POSITIONS
-/**
- * Validates that the generated JSON object is the same as expected
- * Validates that the start position and end position match the start and end of the string
- *
- * This check assumes that there is no whitespace around the json object in the original string.
- */
-void validate_generated_json_and_start_end_pos_helper(const std::string& original_string, const json& j, const json& check)
-{
-    CHECK(j == check);
-    CHECK(j.start_pos() == 0);
-    CHECK(j.end_pos() == original_string.size());
-}
-
-/**
- * Parses the root object from the given root string and validates that the start and end positions for the nested object are correct.
- *
- * This checks that whitespace around the nested object is included in the start and end positions of the root object.
- */
-void validate_start_end_pos_for_nested_obj_helper(const std::string& nested_type_json_str, const std::string& root_type_json_str, const json& expected_json, const json::parser_callback_t& cb = nullptr)
-{
-    json j;
-
-    // 1. If callback is provided, use callback version of parse()
-    if (cb)
-    {
-        j = json::parse(root_type_json_str, cb);
-    }
-    else
-    {
-        j = json::parse(root_type_json_str);
-    }
-
-    // 2. Check if the generated JSON is as expected
-    // Assumptions: The root_type_json_str does not have any whitespace around the json object
-    validate_generated_json_and_start_end_pos_helper(root_type_json_str, j, expected_json);
-
-    // 3. Get the nested object
-    const auto& nested = j["nested"];
-    // 4. Check if the start and end positions are generated correctly for nested objects and arrays
-    CHECK(nested_type_json_str == root_type_json_str.substr(nested.start_pos(), nested.end_pos() - nested.start_pos()));
-}
-#endif
-
 } // namespace
 
 TEST_CASE("parser class")
@@ -1823,228 +1779,6 @@ TEST_CASE("parser class")
         CHECK_THROWS_WITH_AS(_ = json::parse("/a", nullptr, true, true), "[json.exception.parse_error.101] parse error at line 1, column 2: syntax error while parsing value - invalid comment; expecting '/' or '*' after '/'; last read: '/a'", json::parse_error);
         CHECK_THROWS_WITH_AS(_ = json::parse("/*", nullptr, true, true), "[json.exception.parse_error.101] parse error at line 1, column 3: syntax error while parsing value - invalid comment; missing closing '*/'; last read: '/*<U+0000>'", json::parse_error);
     }
-
-#if JSON_DIAGNOSTIC_POSITIONS
-    // Macro for all test cases for start_pos and end_pos
-#define SETUP_TESTCASES() \
-    SECTION("with callback") \
-    { \
-        SECTION("filter nothing") \
-        { \
-            json::parser_callback_t const cb = [](int /*unused*/, json::parse_event_t /*unused*/, json& /*unused*/) noexcept \
-            { \
-                return true; \
-            }; \
-            validate_start_end_pos_for_nested_obj_helper(nested_type_json_str, root_type_json_str, expected, cb); \
-        } \
-        SECTION("filter element") \
-        { \
-            json::parser_callback_t const cb = [](int /*unused*/, json::parse_event_t event, json& j) noexcept \
-            { \
-                return (event != json::parse_event_t::key && event != json::parse_event_t::value) || j != json("a"); \
-            }; \
-            validate_start_end_pos_for_nested_obj_helper(nested_type_json_str, root_type_json_str, filteredExpected, cb); \
-        } \
-    } \
-    SECTION("without callback") \
-    { \
-        validate_start_end_pos_for_nested_obj_helper(nested_type_json_str, root_type_json_str, expected); \
-    }
-
-    SECTION("retrieve start position and end position")
-    {
-        SECTION("for object")
-        {
-            // Create an object with spaces to test the start and end positions. Spaces will not be included in the
-            // JSON object, however, the start and end positions should include the spaces from the input JSON string.
-            const std::string nested_type_json_str =  R"({    "a":       1,"b"      : "test1"})";
-            const std::string root_type_json_str =  R"({    "nested": )" + nested_type_json_str + R"(, "anotherValue": "test2"})";
-            auto expected = json({{"nested", {{"a", 1}, {"b", "test1"}}}, {"anotherValue", "test2"}});
-            auto filteredExpected = expected;
-            filteredExpected["nested"].erase("a");
-
-            SETUP_TESTCASES()
-        }
-
-        SECTION("for array")
-        {
-            const std::string nested_type_json_str =  R"(["a", "test", 45])";
-            const std::string root_type_json_str =  R"({   "nested": )" + nested_type_json_str + R"(, "anotherValue": "test" })";
-            auto expected = json({{"nested", {"a", "test", 45}}, {"anotherValue", "test"}});
-            auto filteredExpected = expected;
-            filteredExpected["nested"] = json({"test", 45});
-            SETUP_TESTCASES()
-        }
-
-        SECTION("for array with objects")
-        {
-            const std::string nested_type_json_str =  R"([{"a": 1, "b": "test"}, {"c": 2, "d": "test2"}])";
-            const std::string root_type_json_str =  R"({   "nested": )" + nested_type_json_str + R"(, "anotherValue": "test" })";
-            auto expected = json({{"nested", {{{"a", 1}, {"b", "test"}}, {{"c", 2}, {"d", "test2"}}}}, {"anotherValue", "test"}});
-            auto filteredExpected = expected;
-            filteredExpected["nested"][0].erase("a");
-            SETUP_TESTCASES()
-
-            auto j = json::parse(root_type_json_str);
-            auto nested_array = j["nested"];
-            const auto& nested_obj = nested_array[0];
-            CHECK(nested_type_json_str.substr(1, 21) == root_type_json_str.substr(nested_obj.start_pos(), nested_obj.end_pos() - nested_obj.start_pos()));
-            CHECK(nested_type_json_str.substr(24, 22) == root_type_json_str.substr(nested_array[1].start_pos(), nested_array[1].end_pos() - nested_array[1].start_pos()));
-        }
-
-        SECTION("for two levels of nesting objects")
-        {
-            const std::string nested_type_json_str =  R"({"nested2": {"b": "test"}})";
-            const std::string root_type_json_str =  R"({   "a": 2, "nested": )" + nested_type_json_str + R"(, "anotherValue": "test" })";
-            auto expected = json({{"a", 2}, {"nested", {{"nested2", {{"b", "test"}}}}}, {"anotherValue", "test"}});
-            auto filteredExpected = expected;
-            filteredExpected.erase("a");
-            SETUP_TESTCASES()
-
-            auto j = json::parse(root_type_json_str);
-            auto nested_obj = j["nested"]["nested2"];
-            CHECK(nested_type_json_str.substr(12, 13) == root_type_json_str.substr(nested_obj.start_pos(), nested_obj.end_pos() - nested_obj.start_pos()));
-        }
-
-        SECTION("for simple types")
-        {
-            SECTION("no nested")
-            {
-                SECTION("with callback")
-                {
-                    json::parser_callback_t const cb = [](int /*unused*/, json::parse_event_t /*unused*/, json& /*unused*/) noexcept
-                    {
-                        return true;
-                    };
-
-                    // 1. string type
-                    std::string json_str =  R"("test")";
-                    auto j = json::parse(json_str, cb);
-                    validate_generated_json_and_start_end_pos_helper(json_str, j, "test");
-
-                    // 2. number type
-                    json_str =  R"(1)";
-                    j = json::parse(json_str, cb);
-                    validate_generated_json_and_start_end_pos_helper(json_str, j, 1);
-
-                    // 3. boolean type
-                    json_str =  R"(true)";
-                    j = json::parse(json_str, cb);
-                    validate_generated_json_and_start_end_pos_helper(json_str, j, true);
-
-                    // 4. null type
-                    json_str =  R"(null)";
-                    j = json::parse(json_str, cb);
-                    validate_generated_json_and_start_end_pos_helper(json_str, j, nullptr);
-                }
-
-                SECTION("without callback")
-                {
-                    // 1. string type
-                    std::string json_str =  R"("test")";
-                    auto j = json::parse(json_str);
-                    validate_generated_json_and_start_end_pos_helper(json_str, j, "test");
-
-                    // 2. number type
-                    json_str =  R"(1)";
-                    j = json::parse(json_str);
-                    validate_generated_json_and_start_end_pos_helper(json_str, j, 1);
-
-                    json_str = R"(1.001239923)";
-                    j = json::parse(json_str);
-                    validate_generated_json_and_start_end_pos_helper(json_str, j, 1.001239923);
-
-                    json_str = R"(1.123812389000000)";
-                    j = json::parse(json_str);
-                    validate_generated_json_and_start_end_pos_helper(json_str, j, 1.123812389);
-
-                    // 3. boolean type
-                    json_str =  R"(true)";
-                    j = json::parse(json_str);
-                    validate_generated_json_and_start_end_pos_helper(json_str, j, true);
-
-                    json_str =  R"(false)";
-                    j = json::parse(json_str);
-                    validate_generated_json_and_start_end_pos_helper(json_str, j, false);
-
-                    // 4. null type
-                    json_str =  R"(null)";
-                    j = json::parse(json_str);
-                    validate_generated_json_and_start_end_pos_helper(json_str, j, nullptr);
-                }
-            }
-
-            SECTION("string type")
-            {
-                const std::string nested_type_json_str =  R"("test")";
-                const std::string root_type_json_str =  R"({ "a": 1,   "nested": )" + nested_type_json_str + R"(, "anotherValue": "test" })";
-                auto expected = json({{"nested", "test"}, {"anotherValue", "test"}, {"a", 1}});
-                auto filteredExpected = expected;
-                filteredExpected.erase("a");
-                SETUP_TESTCASES()
-            }
-
-            SECTION("number type")
-            {
-                const std::string nested_type_json_str =  R"(2)";
-                const std::string root_type_json_str =  R"({ "a": 1,   "nested": )" + nested_type_json_str + R"(, "anotherValue": "test" })";
-                auto expected = json({{"nested", 2}, {"anotherValue", "test"}, {"a", 1}});
-                auto filteredExpected = expected;
-                filteredExpected.erase("a");
-                SETUP_TESTCASES()
-            }
-
-            SECTION("boolean type")
-            {
-                const std::string nested_type_json_str =  R"(true)";
-                const std::string root_type_json_str =  R"({ "a": 1,   "nested": )" + nested_type_json_str + R"(, "anotherValue": "test" })";
-                auto expected = json({{"nested", true}, {"anotherValue", "test"}, {"a", 1}});
-                auto filteredExpected = expected;
-                filteredExpected.erase("a");
-                SETUP_TESTCASES()
-            }
-
-            SECTION("null type")
-            {
-                const std::string nested_type_json_str =  R"(null)";
-                const std::string root_type_json_str =  R"({ "a": 1,   "nested": )" + nested_type_json_str + R"(, "anotherValue": "test" })";
-                auto expected = json({{"nested", nullptr}, {"anotherValue", "test"}, {"a", 1}});
-                auto filteredExpected = expected;
-                filteredExpected.erase("a");
-                SETUP_TESTCASES()
-            }
-        }
-        SECTION("with leading whitespace and newlines around root JSON")
-        {
-            const std::string initial_whitespace = R"(
-                
-            )";
-            const std::string nested_type_json_str = R"({
-                "a": 1,
-                "nested": {
-                    "b": "test"
-                },
-                "anotherValue": "test"
-            })";
-            const std::string end_whitespace = R"(
-                
-            )";
-            const std::string root_type_json_str = initial_whitespace + nested_type_json_str + end_whitespace;
-
-            auto expected = json({{"a", 1}, {"nested", {{"b", "test"}}}, {"anotherValue", "test"}});
-
-            auto j = json::parse(root_type_json_str);
-
-            // 2. Check if the generated JSON is as expected
-            CHECK(j == expected);
-
-            // 3. Check if the start and end positions do not include the surrounding whitespace
-            CHECK(j.start_pos() == initial_whitespace.size());
-            CHECK(j.end_pos() == root_type_json_str.size() - end_whitespace.size());
-        }
-    }
-#undef SETUP_TESTCASES
-#endif
 }
 
 // this test relies on parse errors being thrown, so it is skipped when

tests/src/unit-diagnostic-positions-only.cpp

@@ -0,0 +1,44 @@
+//     __ _____ _____ _____
+//  __|  |   __|     |   | |  JSON for Modern C++ (supporting code)
+// |  |  |__   |  |  | | | |  version 3.12.0
+// |_____|_____|_____|_|___|  https://github.com/nlohmann/json
+//
+// SPDX-FileCopyrightText: 2013-2026 Niels Lohmann <https://nlohmann.me>
+// SPDX-License-Identifier: MIT
+
+#include "doctest_compatibility.h"
+
+#ifdef JSON_DIAGNOSTICS
+    #undef JSON_DIAGNOSTICS
+#endif
+
+#define JSON_DIAGNOSTICS 0
+#define JSON_DIAGNOSTIC_POSITIONS 1
+#include <nlohmann/json.hpp>
+
+using json = nlohmann::json;
+
+TEST_CASE("Better diagnostics with positions only")
+{
+    SECTION("invalid type")
+    {
+        const std::string json_invalid_string = R"(
+        {
+            "address": {
+                "street": "Fake Street",
+                "housenumber": "1"
+            }
+        }
+        )";
+        json j = json::parse(json_invalid_string);
+        CHECK_THROWS_WITH_AS(j.at("address").at("housenumber").get<int>(),
+                             "[json.exception.type_error.302] (bytes 108-111) type must be number, but is string", json::type_error);
+    }
+
+    SECTION("invalid type without positions")
+    {
+        const json j = "foo";
+        CHECK_THROWS_WITH_AS(j.get<int>(),
+                             "[json.exception.type_error.302] type must be number, but is string", json::type_error);
+    }
+}

tests/src/unit-diagnostic-positions.cpp

@@ -8,9 +8,7 @@
 
 #include "doctest_compatibility.h"
 
-#ifndef JSON_DIAGNOSTICS
-    #define JSON_DIAGNOSTICS 1
-#endif
+#define JSON_DIAGNOSTICS 1
 #define JSON_DIAGNOSTIC_POSITIONS 1
 #include <nlohmann/json.hpp>
 
@@ -29,13 +27,8 @@ TEST_CASE("Better diagnostics with positions")
         }
         )";
         json j = json::parse(json_invalid_string);
-#if JSON_DIAGNOSTICS
         CHECK_THROWS_WITH_AS(j.at("address").at("housenumber").get<int>(),
                              "[json.exception.type_error.302] (/address/housenumber) (bytes 108-111) type must be number, but is string", json::type_error);
-#else
-        CHECK_THROWS_WITH_AS(j.at("address").at("housenumber").get<int>(),
-                             "[json.exception.type_error.302] (bytes 108-111) type must be number, but is string", json::type_error);
-#endif
     }
 
     SECTION("invalid type without positions")
@@ -81,12 +74,7 @@ TEST_CASE("Better diagnostics with positions")
         // (/foo/bar); the position of that parent is reported in the message
         const json doc = json::parse(R"({"foo":{"bar":"a string"}})");
         const json patch = json::parse(R"([{"op":"add","path":"/foo/bar/baz","value":1}])");
-#if JSON_DIAGNOSTICS
         CHECK_THROWS_WITH_AS(doc.patch(patch),
                              "[json.exception.out_of_range.411] (/foo/bar) (bytes 14-24) cannot add value: the JSON Patch 'add' target's parent is of type string, but must be an object or array", json::out_of_range);
-#else
-        CHECK_THROWS_WITH_AS(doc.patch(patch),
-                             "[json.exception.out_of_range.411] (bytes 14-24) cannot add value: the JSON Patch 'add' target's parent is of type string, but must be an object or array", json::out_of_range);
-#endif
     }
 }