Post Release TODOS

[gerrymanoim]

• [x] count for array
• [x] index for array
• [x] benchmarks (look at overhead of access vs python directly)
• [ ] Other compiler flags?
• [ ] string padding?
• [ ] extra functions
  • [ ] validate_utf8
  • [ ] load_many/parse_many
  • [ ] load_many/parse_many with threads?
  • [ ] case insensitive at?
• [ ] Add reprs to Object and Array
• [x] Expose versions of libpy_simdjson and simdjson

Note: Right now when iterate we convert everything automatically to python objects. Should we be smarter? Or is that not expected?

[llllllllll]

Note: Right now when iterate we convert everything automatically to python objects.

I think that now that we are using ranges, it would be possible to write a transform that has a shared pointer to the parser so it can convert the Objects and Arrays into our rich objects.

[gerrymanoim]

Could we use disambiguate_result plus the shared pointer inside the transform?