⚡ Bolt: Optimize file reading performance

.jules/bolt.md

@@ -2,12 +2,6 @@
 
 ## YYYY-MM-DD - Optimized logging message formatting
 **Learning:** std::format is significantly faster than std::ostringstream for string formatting, especially with chrono types.
-<<<<<<< Updated upstream
-=======
-<<<<<<< Updated upstream
-**Action:** Use std::format instead of std::ostringstream for log message formatting in future optimizations.
-=======
->>>>>>> Stashed changes
 **Action:** Use std::format instead of std::ostringstream for log message formatting in future optimizations.
 
 ## 2026-02-02 - IPv4 parser pointer optimization
@@ -16,13 +10,12 @@
 
 ## 2026-02-02 - Log level check lock-free early exit
 **Learning:** Replacing mutex-protected level checks with a lock-free atomic read removes frequent mutex acquisitions on filtered log paths. Microbenchmark (`tests::LogFilterPerformance`, 100k iterations) measured ~10ms for filtered debug calls on Linux x86_64.
-<<<<<<< Updated upstream
-**Action:** Prefer atomic operations on small shared flags that are read frequently and updated rarely (e.g., log level). Add similar microbenchmarks when optimizing hot logging paths.
-=======
 **Action:** Prefer atomic operations on small shared flags that are read frequently and updated rarely (e.g., log level). Add similar microbenchmarks when optimizing hot logging paths.
 
 ## 2026-02-03 - Avoid temporary std::string allocations on filtered logs
 **Learning:** Adding `const char*` overloads for log methods and checking the level before constructing a `std::string` avoids temporary allocations when logging string literals that are filtered by log level. This reduces allocation overhead on hot, filtered paths without changing the API behavior for callers using `std::string`.
-**Action:** Add `const char*` overloads for common logging entry-points in other modules if needed and add microbenchmarks to `tests/` to track regression."
->>>>>>> Stashed changes
->>>>>>> Stashed changes
+**Action:** Add `const char*` overloads for common logging entry-points in other modules if needed and add microbenchmarks to `tests/` to track regression.
+
+## 2026-02-03 - Optimize filesystem read_file with direct string construction
+**Learning:** Replacing stringstream-based file reading with direct string construction from istreambuf_iterators eliminates intermediate buffering overhead, providing measurable performance improvement for file I/O operations.
+**Action:** Use direct iterator-based string construction for other stream-to-string conversions in the codebase to avoid unnecessary intermediate objects.

include/filesystem.hpp

@@ -105,9 +105,7 @@ class FileSystem
   static std::string read_file(const std::string &path)
   {
     std::ifstream file(path, std::ios::binary);
-    std::stringstream buffer;
-    buffer << file.rdbuf();
-    return buffer.str();
+    return std::string(std::istreambuf_iterator<char>(file), std::istreambuf_iterator<char>());
   }
 
   static bool write_file(const std::string &path, const std::string &content)