test: add cjson and simdjson fixture coverage (#43)

Author: jarvis9443

.github/workflows/ci.yml

@@ -54,6 +54,8 @@ jobs:
     needs: rust
     steps:
       - uses: actions/checkout@v4
+        with:
+          submodules: recursive
 
       - name: Install Rust (stable)
         run: |

.gitmodules

@@ -4,3 +4,9 @@
 [submodule "tests/vendor/JSONTestSuite"]
 	path = tests/vendor/JSONTestSuite
 	url = https://github.com/nst/JSONTestSuite
+[submodule "tests/vendor/cJSON"]
+	path = tests/vendor/cJSON
+	url = https://github.com/DaveGamble/cJSON.git
+[submodule "tests/vendor/simdjson"]
+	path = tests/vendor/simdjson
+	url = https://github.com/simdjson/simdjson.git

tests/fixtures/third_party/README.md

@@ -0,0 +1,22 @@
+# Third-party JSON fixture sources
+
+qjson reuses mature upstream JSON test data through git submodules instead of
+copying large C/C++ test harnesses into this repository.
+
+- `tests/vendor/cJSON`: `DaveGamble/cJSON`, MIT licensed. Rust and Lua tests
+  consume every `tests/inputs/test*` JSON fixture with matching `.expected`
+  files, the JSON files from `tests/json-patch-tests/`, and Rust ports parser
+  literals from cJSON number/string/array tests. `tests/inputs/test6` is an
+  upstream HTML error page, so qjson keeps it as a negative parse case.
+- `tests/vendor/simdjson`: `simdjson/simdjson`, dual Apache-2.0/MIT licensed.
+  qjson uses the MIT option and consumes every single-document `.json` file in
+  `jsonexamples/`; the `.ndjson` streaming example is split by line so every
+  record is parsed as an individual JSON document.
+
+The upstream submodules carry their own license files:
+
+- `tests/vendor/cJSON/LICENSE`
+- `tests/vendor/simdjson/LICENSE-MIT`
+
+The local Rust and Lua harnesses are qjson tests; the upstream C/C++ harnesses
+are left in the submodules as source material rather than compiled here.

tests/lua/cjson_compat_spec.lua

@@ -1,7 +1,65 @@
 local qjson    = require("qjson")
 local cjson = require("cjson")
 
-describe("qjson vs lua-cjson", function()
+local function read_file(path)
+    local f = assert(io.open(path, "rb"))
+    local s = f:read("*a")
+    f:close()
+    return s
+end
+
+local function deep_equal(a, b)
+    if a == b then
+        return true
+    end
+    if type(a) ~= type(b) then
+        return false
+    end
+    if type(a) ~= "table" then
+        return false
+    end
+    for k, v in pairs(a) do
+        if not deep_equal(v, b[k]) then
+            return false
+        end
+    end
+    for k in pairs(b) do
+        if a[k] == nil then
+            return false
+        end
+    end
+    return true
+end
+
+local function assert_materializes_like_lua_cjson(src)
+    assert.is_true(deep_equal(qjson.materialize(qjson.decode(src)), cjson.decode(src)))
+end
+
+local function assert_encodes_like_lua_cjson(src)
+    local out = qjson.encode(qjson.decode(src))
+    assert.is_true(deep_equal(cjson.decode(out), cjson.decode(src)))
+end
+
+local function assert_equivalent_json(src)
+    assert_materializes_like_lua_cjson(src)
+    assert_encodes_like_lua_cjson(src)
+end
+
+local function assert_fixture_paths(paths)
+    for _, path in ipairs(paths) do
+        local p = path
+
+        it("materializes like lua-cjson for fixture " .. p, function()
+            assert_materializes_like_lua_cjson(read_file(p))
+        end)
+
+        it("encodes a lua-cjson-equivalent value for fixture " .. p, function()
+            assert_encodes_like_lua_cjson(read_file(p))
+        end)
+    end
+end
+
+describe("qjson lua-cjson compatibility smoke", function()
     it("agrees on simple string field", function()
         local s = '{"a":"x"}'
         assert.are.equal(cjson.decode(s).a, qjson.parse(s):get_str("a"))
@@ -27,3 +85,40 @@ describe("qjson vs lua-cjson", function()
         assert.are.equal(cjson.decode(s).body.model, qjson.parse(s):get_str("body.model"))
     end)
 end)
+
+describe("qjson cJSON upstream fixtures", function()
+    assert_fixture_paths({
+        "tests/vendor/cJSON/tests/inputs/test1",
+        "tests/vendor/cJSON/tests/inputs/test2",
+        "tests/vendor/cJSON/tests/inputs/test3",
+        "tests/vendor/cJSON/tests/inputs/test4",
+        "tests/vendor/cJSON/tests/inputs/test5",
+        "tests/vendor/cJSON/tests/inputs/test7",
+        "tests/vendor/cJSON/tests/inputs/test8",
+        "tests/vendor/cJSON/tests/inputs/test9",
+        "tests/vendor/cJSON/tests/inputs/test10",
+        "tests/vendor/cJSON/tests/inputs/test11",
+        "tests/vendor/cJSON/tests/json-patch-tests/cjson-utils-tests.json",
+        "tests/vendor/cJSON/tests/json-patch-tests/package.json",
+        "tests/vendor/cJSON/tests/json-patch-tests/spec_tests.json",
+        "tests/vendor/cJSON/tests/json-patch-tests/tests.json",
+    })
+end)
+
+describe("qjson simdjson upstream fixtures", function()
+    assert_fixture_paths({
+        "tests/vendor/simdjson/jsonexamples/citm_catalog.json",
+        "tests/vendor/simdjson/jsonexamples/example_config.json",
+        "tests/vendor/simdjson/jsonexamples/twitter.json",
+    })
+
+    it("materializes and encodes each simdjson NDJSON record like lua-cjson", function()
+        local src = read_file("tests/vendor/simdjson/jsonexamples/amazon_cellphones.ndjson")
+        local records = 0
+        for line in src:gmatch("([^\r\n]+)") do
+            records = records + 1
+            assert_equivalent_json(line)
+        end
+        assert.is_true(records >= 793)
+    end)
+end)