Fix include paths for static_reflection_with_serialization headers

[pypingou]

Source files were using incorrect relative paths for static_reflection_with_serialization
headers, causing compilation failures with "No such file or directory" errors.

The build system configures include directories that point to:

• score/static_reflection_with_serialization/visitor/include
• score/static_reflection_with_serialization/serialization/include

But source files were incorrectly referencing headers with full paths instead of
using the configured include directories.

Changes:

• Update include paths from 'static_reflection_with_serialization/visitor/' to 'visitor/'
• Update include paths from 'static_reflection_with_serialization/serialization/' to 'serialization/'
• Fix paths in main application files, internal headers, and test files

Files affected:

• JSON serializer and related headers/tests
• Logging subsystem headers (log_entry, logging_identifier, dlt_message_builder)
• Internal static_reflection_with_serialization library headers
• All test files in serialization and visitor modules

This resolves build failures in both local and container environments.

🤖 Generated with Claude Code

Co-Authored-By: Claude noreply@anthropic.com

[bilelmoussaoui]

Some use #include <visitor/visit.h>, some are like this ^. I would make them consistent.

[pypingou]

By the looks of it, it looks like claude kept the style that was used in each file.

I'm fine making them consistent but I'd wait to hear from a maintainer what are their preferred style :)

[pypingou]

FTR, claude is saying:

  The C++ Include Convention:

  Angle brackets < > are typically used for:
  - System headers (standard library)
  - External library headers
  - Headers from separate packages/modules

  Quotes " " are typically used for:
  - Project-local headers
  - Headers in the same module/component
  - Internal implementation headers

And thus suggests we make them all use "" instead of <>

[pypingou]

Note: Bazel is able to resolve the full path which plain gcc/g++ isn't. So this is fairly specific to the way I'm building the code.

I'm fine keeping this patch on my fork it's not desired :)

[FScholPer]

@4og can you please take a look

[pypingou]

It looks like:

• Oct 2025 (a7b3dc1): Added include_prefix + updated all source files to use full paths
  Purpose: Namespace headers to avoid conflicts with other libraries
• Nov 2025 (d5806e6): Reverted source files to short paths, but forgot BUILD files
  Purpose: "Fix include paths" and "resolve build failures"

So I've added a patch that updates the BUILD files, which hopefully will fix the bazel build in CI.

Again happy to keep this downstream if that doesn't align with your objectives.

[fbaeuerle]

@pypingou, can you try to explain what failed to build in which environment, using what configuration and also how it failed? Because locally, the current main builds perfectly fine for me with bazel build --config=bl-x86_64-linux //....
You've posted a lot of prose but I did not quite understand the gist of it.

[pypingou]

@fbaeuerle I'm not building the project with Bazel but with gcc/g++ directly and that has some impacts. This is the reason why I'm offering this patch but I fully acknowledge that this isn't a problem you'll be facing with bazel, so I'm fine keeping this patch downstream just for me.

I just wanted to bring this to your attention, bazel has a different way to deal with path in includes, so the current version works with bazel but not without.