refactor(clp-s): Use simdjson string_builder for JSON string escaping in marshalling.

components/core/CMakeLists.txt

@@ -211,7 +211,7 @@ if(CLP_NEED_NLOHMANN_JSON)
 endif()
 
 if(CLP_NEED_SIMDJSON)
-    find_package(simdjson REQUIRED)
+    find_package(simdjson 4.6.2 REQUIRED)
     if(simdjson_FOUND)
         message(STATUS "Found simdjson ${simdjson_VERSION}")
     endif()

components/core/src/clp_s/CMakeLists.txt

@@ -387,6 +387,7 @@ if(CLP_BUILD_CLP_S_ARCHIVEREADER)
                 clp_s::timestamp_pattern
                 msgpack-cxx
                 nlohmann_json::nlohmann_json
+                simdjson::simdjson
                 ystdlib::error_handling
                 PRIVATE
                 Boost::url

components/core/src/clp_s/Utils.cpp

@@ -8,6 +8,7 @@
 
 #include <boost/url.hpp>
 #include <fmt/format.h>
+#include <simdjson.h>
 #include <spdlog/spdlog.h>
 #include <string_utils/string_utils.hpp>
 
@@ -197,6 +198,17 @@ bool UriUtils::get_last_uri_component(std::string_view const uri, std::string& n
 }
 
 void StringUtils::escape_json_string(std::string& destination, std::string_view const source) {
+    simdjson::builtin::builder::string_builder json_string_builder{};
+    json_string_builder.clear();
+    json_string_builder.escape_and_append(source);
+    std::string_view escaped_source;
+    auto const status = json_string_builder.view().get(escaped_source);
+    if (simdjson::SUCCESS == status) {
+        destination.append(escaped_source);
+        return;
+    }
+
+    // Preserve the original implementation in case simdjson fails.
     // Escaping is implemented using this `append_unescaped_slice` approach to offer a fast path
     // when strings are mostly or entirely valid escaped JSON. Benchmarking shows that this offers
     // a net decompression speedup of ~30% compared to adding every character to the destination one

docs/src/dev-docs/components-core/index.md

@@ -50,7 +50,7 @@ The task will download, build, and install (within the build directory) the foll
 | [mongo-cxx-driver](https://github.com/mongodb/mongo-cxx-driver)       | r4.1.1         |
 | [msgpack-cxx](https://github.com/msgpack/msgpack-c/tree/cpp_master)   | v7.0.0         |
 | [nlohmann_json](https://github.com/nlohmann/json)                     | v3.11.3        |
-| [simdjson](https://github.com/simdjson/simdjson)                      | v3.13.0        |
+| [simdjson](https://github.com/simdjson/simdjson)                      | v4.6.2         |
 | [spdlog](https://github.com/gabime/spdlog)                            | v1.15.3        |
 | [SQLite3](https://www.sqlite.org/download.html)                       | v3.36.0        |
 | [utfcpp](https://github.com/nemtrif/utfcpp)                           | v4.0.6         |

taskfiles/deps/main.yaml

@@ -483,8 +483,8 @@ tasks:
             - "-DCMAKE_BUILD_TYPE=Release"
             - "-DCMAKE_INSTALL_MESSAGE=LAZY"
           LIB_NAME: "simdjson"
-          TARBALL_SHA256: "07a1bb3587aac18fd6a10a83fe4ab09f1100ab39f0cb73baea1317826b9f9e0d"
-          TARBALL_URL: "https://github.com/simdjson/simdjson/archive/refs/tags/v3.13.0.tar.gz"
+          TARBALL_SHA256: "c240d4bffcccda4fe3a2bba2872718d96fc92e56d2615bfac4f9b2bad89a6386"
+          TARBALL_URL: "https://github.com/simdjson/simdjson/archive/refs/tags/v4.6.2.tar.gz"
 
   spdlog:
     internal: true