simdjson

simdjson copied to clipboard

Swap key_value_pair for std::pair.

11

The work in #1050 is great, and enables array and object iterators to be used with many algorithms expecting `std::iterator`. However, the return value of `object` is a `key_value_pair` instead...

TkTech

Build a multithreaded version (for very large documents)

2

@Somefive and others have begun work on such a project... https://github.com/Somefive/MercuryJson We may parallelize stage 1 and stage 2 as well as do backward/forward parsing.

lemire

parser.on_invalid_number

2

Create a parser.on_invalid_number hook. cc @jkeiser

lemire

Detect if any keys have non-ASCII characters

18

Non-ascii keys seem pretty weird, but I guess they must be out there. Normally, there are no keys like this and so you can deal with strings in a fast...

michaeleisel

Scan input (stage 1) in fixed-size batches

16

simdjson's stage 2 (DOM creation) should call stage 1 in *batches*, capping the buffer size passed to stage 1 to some constant. and process the structural indexes in that batch...

jkeiser

Add logging capability

1

If we want to be able to help downstream users, we need a way for them to send us logs. * Compiled in by default, configurable by environment variable *...

jkeiser

Discussion regarding a new 16-byte tape (comments invited)

11

Here is the first sketch of a proposal to have a (wider and flatter) 16-byte tape up from our current hybrid 8-byte tape. (This is issue https://github.com/lemire/simdjson/issues/361) In this new...

lemire

Support conversion of simdjson_result lvalue to dom::element

12

`simdjson_result` is a very transparent structure. It can be implicitly treated as an array with `at()` and `begin()/end()` members, or it can be treated as an object with `operator[]` member....

mmatrosov

Make it convenient to parse JSON from compressed inputs

10

Currently, to parse a compressed JSON document, we expect folks to fully uncompressed the document first. For obvious reasons, it might be more profitable to allow the user to work...

lemire

Faster bulk loading and total streaming for large machine generated JSON

47

I'm building a totally demand streaming parser - everything but field names can be streamed, and it can all run comfortably in 4kB of ram if you want it to....

codewitch-honey-crisis