Add RapidJSON and nlohmann_json SAX to distinct_user_id benchmark (#1593)

* Add rapidjson_sax for distinct_user_id

* Add nlohmann_json_sax.h for distinct_user_id

* Add flags for RapidJSON.

* Fix revisions.

* Fix revisions again.

* Replace strcpy with memcpy. Increase performance fix.

benchmark/bench_ondemand.cpp

@@ -55,7 +55,9 @@ SIMDJSON_POP_DISABLE_WARNINGS
 #include "distinct_user_id/yyjson.h"
 #include "distinct_user_id/sajson.h"
 #include "distinct_user_id/rapidjson.h"
+#include "distinct_user_id/rapidjson_sax.h"
 #include "distinct_user_id/nlohmann_json.h"
+#include "distinct_user_id/nlohmann_json_sax.h"
 
 #include "find_tweet/simdjson_dom.h"
 #include "find_tweet/simdjson_ondemand.h"

benchmark/distinct_user_id/nlohmann_json_sax.h

@@ -0,0 +1,59 @@
+#pragma once
+
+#ifdef SIMDJSON_COMPETITION_NLOHMANN_JSON
+
+#include "distinct_user_id.h"
+#include <string.h>
+namespace distinct_user_id {
+
+using json = nlohmann::json;
+
+struct nlohmann_json_sax {
+    struct Handler : json::json_sax_t
+    {
+        std::vector<uint64_t>& result;
+        bool user = false;
+        bool user_id = false;
+        Handler(std::vector<uint64_t> &r) : result(r) { }
+
+        bool key(string_t& val) override {
+            // Assume that valid user/id pairs appear only once in main array of user objects
+            if (user) { // If already found user object, find id key
+                if (val.compare("id") == 0) { user_id = true; }
+            }
+            else if (val.compare("user") == 0) { user = true; } // Otherwise, find user object
+            return true;
+        }
+        bool number_unsigned(number_unsigned_t val) override {
+            if (user_id) {
+                result.emplace_back(val);
+                user = false;
+                user_id = false;
+            }
+            return true;
+        }
+        // Irrelevant events
+        bool null() override { return true; }
+        bool boolean(bool val) override { return true; }
+        bool number_float(number_float_t val, const string_t& s) override { return true; }
+        bool number_integer(number_integer_t val) override { return true; }
+        bool string(string_t& val) override { return true; }
+        bool start_object(std::size_t elements) override { return true; }
+        bool end_object() override { return true; }
+        bool start_array(std::size_t elements) override { return true; }
+        bool end_array() override { return true; }
+        bool binary(json::binary_t& val) override { return true; }
+        bool parse_error(std::size_t position, const std::string& last_token, const json::exception& ex) override { return false; }
+    }; // Handler
+
+    bool run(simdjson::padded_string &json, std::vector<uint64_t> &result) {
+        Handler handler(result);
+        json::sax_parse(json.data(), &handler);
+
+        return true;
+    }
+}; // nlohmann_json_sax
+BENCHMARK_TEMPLATE(distinct_user_id, nlohmann_json_sax)->UseManualTime();
+} // namespace distinct_user_id
+
+#endif // SIMDJSON_COMPETITION_NLOHMANN_JSON

benchmark/distinct_user_id/rapidjson_sax.h

@@ -0,0 +1,61 @@
+#pragma once
+
+#ifdef SIMDJSON_COMPETITION_RAPIDJSON
+
+#include "distinct_user_id.h"
+#include <string.h>
+namespace distinct_user_id {
+
+using namespace rapidjson;
+
+struct rapidjson_sax {
+    struct Handler {
+        std::vector<uint64_t>& result;
+        bool user = false;
+        bool user_id = false;
+        Handler(std::vector<uint64_t> &r) : result(r) { }
+
+        bool Key(const char* key, SizeType length, bool copy) {
+            // Assume that valid user/id pairs appear only once in main array of user objects
+            if (user) { // If already found user object, find id key
+                if ((length == 2) && memcmp(key,"id",2) == 0) { user_id = true; }
+            }
+            else if ((length == 4) && memcmp(key,"user",4) == 0) { user = true; } // Otherwise, find user object
+            return true;
+        }
+        bool Uint(unsigned i) {     // id values are treated as Uint (not Uint64) by the reader
+            if (user_id) {  // Getting id if previous key was "id" for a user
+                result.emplace_back(i);
+                user_id = false;
+                user = false;
+            }
+            return true;
+        }
+        // Irrelevant events
+        bool Null() { return true; }
+        bool Bool(bool b) { return true; }
+        bool Double(double d) { return true; }
+        bool Int(int i) { return true; }
+        bool Int64(int64_t i) { return true; }
+        bool Uint64(uint64_t i) { return true; }
+        bool RawNumber(const char* str, SizeType length, bool copy) { return true; }
+        bool String(const char* str, SizeType length, bool copy) { return true; }
+        bool StartObject() { return true; }
+        bool EndObject(SizeType memberCount) { return true; }
+        bool StartArray() { return true; }
+        bool EndArray(SizeType elementCount) { return true; }
+    }; // handler
+
+    bool run(simdjson::padded_string &json, std::vector<uint64_t> &result) {
+        Reader reader;
+        Handler handler(result);
+        InsituStringStream ss(json.data());
+        reader.Parse<kParseInsituFlag | kParseValidateEncodingFlag | kParseFullPrecisionFlag>(ss,handler);
+        return true;
+    }
+
+}; // rapid_jason_sax
+BENCHMARK_TEMPLATE(distinct_user_id, rapidjson_sax)->UseManualTime();
+} // namespace distinct_user_id
+
+#endif // SIMDJSON_COMPETITION_RAPIDJSON