templatized output adapters

[FrancoisChabot]

This is a draft Fix for #2172.

The output_adapter_protocol is replaced by templatizing the serializer on the output adapter.

The PR also adds the basic_json::dump_to() method, which allows for outputing json to arbitrary user-defined sinks.

Valid sinks are:

• std::basic_string<>, This will append to the string very efficiently
• any type that passes std::iterator_traits<T>::iterator_tag == std::output_iterator_tag
• Any type for which std::back_inserter(T) works. In practice, that means having a push_back(char) function.

If all else fails, the output will be assumed to be:

• a type that has a ostream-like interface (put(c) and write(ptr, len))

TODO:

• minor cleanup
• test cases
• documentation
• before and after benchmarks. Done: Benchmarks haven't moved.

---

Pull request checklist

Read the Contribution Guidelines for detailed information.

• Changes are described in the pull request, or an existing issue is referenced.
• The test suite compiles and runs without error.
• Code coverage is 100%. Test cases can be added by editing the test suite.
• The source code is amalgamated; that is, after making changes to the sources in the include/nlohmann directory, run make amalgamate to create the single-header file single_include/nlohmann/json.hpp. The whole process is described here.

[coveralls]

Coverage remained the same at 100.0% when pulling a3fe169 on FrancoisChabot:2172-templated-output-adapter into 7444c7f on nlohmann:develop.

[FrancoisChabot]

A comment in the reddit post about the 3.8.0 release caught my attention:

https://www.reddit.com/r/cpp/comments/h8w7e0/json_for_modern_c_380/fuuozqb?utm_source=share&utm_medium=web2x

That and the dummy hash implementation that allocates (!) probably does more harm than good.

This would actually be easy to fix by having the hash<>::operator() do a dump_to() to an interface that does the hashing as passthrough.

[nlohmann]

A comment in the reddit post about the 3.8.0 release caught my attention:

https://www.reddit.com/r/cpp/comments/h8w7e0/json_for_modern_c_380/fuuozqb?utm_source=share&utm_medium=web2x

That and the dummy hash implementation that allocates (!) probably does more harm than good.

This would actually be easy to fix by having the hash<>::operator() do a dump_to() to an interface that does the hashing as passthrough.

I also saw that comment. I think we can also implement a small recursive function that collects and combines the individual hashes. Using the output of dump and hashing the result string was lazy and naive...

[nlohmann]

A comment in the reddit post about the 3.8.0 release caught my attention:

https://www.reddit.com/r/cpp/comments/h8w7e0/json_for_modern_c_380/fuuozqb?utm_source=share&utm_medium=web2x

That and the dummy hash implementation that allocates (!) probably does more harm than good.

This would actually be easy to fix by having the hash<>::operator() do a dump_to() to an interface that does the hashing as passthrough.

FYI: I tried to fix this in #2269 - comments welcome!

[nlohmann]

@FrancoisChabot Any idea how to proceed here?

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.