parse( ptr, len ) ?!?

[boris-999]

i have a uint8_t* (pointing at UTF-8) and a size_t len. how do i pass that to parse() ? i DONT have a null terminated string. to save you some time, im going to preempt some reply about iterators; if so, please explain the syntax enough to actually use them, the docs have been overly complicated on the matter, and the examples seem to be lacking.

[gregmarr]

Have you tried auto j = json::parse(pointer, pointer + len);?

[boris-999]

thanks. that did it. :) feel free to ignore the following, as problem solved... bbut.. if you can direct me to why this is in the docs that would help me in future. why did i fail to find this in docs?
also, somewhere i read about (ptr,ptr+len) being deprecated, not sure where.. was in one of the 27 web pages i had open trying to find this answer. (vaguely recall it might have been in a recent release notes on github)

[nlohmann]

The parse function is documented here: https://json.nlohmann.me/api/basic_json/parse/. The deprecation is for from_cbor etc. where a pair of ptr and len is passed as {ptr, len}. That is:

• new: basic_json::from_cbor(ptr, ptr+len)
• deprecated: basic_json::from_cbor({ptr, len})

I am currently streamlining the documentation. Who would have known too much documentation can be an issue ;)

[boris-999]

i shall try to explain my confusion: the examples on that page don't show an example for using 2 iterators. template parameters paragraph say IteratorType: a compatible iterator type, well, i wasn't expecting an invalid one! the comment is kinda useless. maybe the idea that a ptr to the data is magically one of these valid IteratorTypes might be obvious to you, it's far from obvious to me.

I'm not sure why there are many examples of null-terminted inputs and so few of non-null terminated inputs; in my experience the acquisition of JSON text is rarely null terminated.. either by file read or network packet.

[nlohmann]

The page states

The value_type of the iterator must be a integral type with size of 1, 2 or 4 bytes, which will be interpreted respectively as UTF-8, UTF-16 and UTF-32.

which is true in your case. But you're right. I shall add some examples for iterator ranges.

[nlohmann]

I added a PR: #3100 - feedback welcome!