Daniel Lemire
Daniel Lemire
> I'm not entirely sure I understand what the purpose of a BSI is. I cannot be certain, but it seems that what you have implemented is the same data...
Let us run tests!
@gabrieldemarmiesse This looks like great work. Have you tried benchmarking it, if only privately (for yourself)? A file that I like to use for benchmarking is canada.txt: https://github.com/lemire/simple_fastfloat_benchmark/blob/master/data/canada.txt I can...
@gabrieldemarmiesse We can do comparative benchmarking with other systems. That's why I was offering to help. We can know how good it can be, because we have other implementations. So...
> too slow (5 min+) :-/
@soraros I suspect that @gabrieldemarmiesse refers to `_get_w_and_q_from_float_string`. You need to profile this code, but it is possible that it could become a bottleneck. I have not studied it carefully,...
@gabrieldemarmiesse My comments are not an objection to this PR, and I do recommend it gets merged... I was only elaborating on the point you raised in the PR comment,...
Currently, `b64decode` does not appear to handle white-space characters. I would have expected the following to print 'Bonjour', it does not: ``` from base64 import b64decode def main(): var data...
> We have such algorithms in simdutf, and I am working on porting them to C#. Juste an update. So we have SIMD base64 decoding with spaces (skipping spaces) and...
We already support comma-separated documents. So all that is needed is to strip the starting `[` and the ending `]` which is nearly trivial. Hmmm. So yeah, this could be...