Use std::error_code instead of exceptions

[ruipacheco]

Is your feature request related to a problem? Please describe.
I'm currently trying to use this library in a codebase that should be compiled without exceptions.

Describe the solution you'd like
Replace throwing exceptions with std::error_code. Since exceptions seem to be only used in the detail:: namespace this should not break compatibility with the std:: implementation.

Describe alternatives you've considered
Write my own?

[gulrak]

I'm sorry, but it would break the standard. As an example see 'Exceptions' for std::filesystem::absolute at cppreference.com, and sure enough the implementation of ghc::filesystem::absolute without error_code throws an exception in case of an error and an implementation is required to have that version of absolute too. The idea of giving the user the choice to select a call with or without exception error handling is everywhere in std::filesystem and I would not take that away.

Error handling ist described in [fs.err.report] (30.10.7 in C++17):

Functions not having an argument of type error_code& handle errors as follows, unless otherwise specified:
— When a call by the implementation to an operating system or other underlying API results in an error that prevents the function from meeting its specifications, an exception of type filesystem_- error shall be thrown. For functions with a single path argument, that argument shall be passed to the filesystem_error constructor with a single path argument. For functions with two path arguments, the first of these arguments shall be passed to the filesystem_error constructor as the path1 argument, and the second shall be passed as the path2 argument. The filesystem_error constructor’s error_code argument is set as appropriate for the specific operating system dependent error.

So a std::filesystem implementation without exceptions would break the specification. And as stated in the readme, I am open for changes but none breaking compatibility with code expecting a filesystem implementation to throw.

So while the standard says nothing about disabling exceptions and handling disabled exceptions inside their standard library is a compiler implementation specific issue, I could go down the route to try to test if exceptions are enabled or introduce a new option to disable exceptions and and simply call std::abort() in case of an error that would lead to a throw if exceptions where enabled (what GCC does when disabling them). For the large number of compilers currently supported, implementing automatic detection would be more work than I currently could put into it for the rare use-case of having a filesystem but no exception support (e.g. __cpp_exceptions is not supported by some compilers).

If I get a clean PR that uses std::abort instead of throwing only when a new option macro is enabled (e.g. GHC_NO_EXCEPITONS) and works on the systems tested by the CI scripts, I might integrate it, but there is no chance of generally never-throwing-by-default change.

[ruipacheco]

My bad!

[gulrak]

As #61 is a PR that implemented a variant of this, current WIP version 1.3.3 (6b30995) on master and upcoming release v1.3.4 are supporting environments without exceptions.