Max Bachmann

I updated my implementation to find the correct results in all cases even for needles with a length over 64. In addition I improved the performance. E.g. the similarity for...

Multiprocessing in the 1:n scenario is not really implemented in `process.cdist`. Currently it uses multiprocessing for the outer loop, so it has basically no effect when used on 1:n. So...

For the `max`/`sorted` functions it would make sense to store both the score and the corresponding index in the results. However I am unsure how this should be stored in...

I am not sure yet. I think I could add a simple implementation in the close future. When matching long sequences you would probably want to use a more optimized...

After looking at the paraseil python bindings I found them way to hard to use. E.g. > Be careful using the attributes of the Result object - especially on Result...

After thinking about this a bit more I think stuffing this into `token_set_ratio` is confusing, since I try to keep the whole `fuzz` module as similar in behavior to fuzzywuzzy...

@DamonBlais it should work when you add it to your /root/.bashrc ``` sudo su echo 'export PATH="$HOME/.cargo/bin:$PATH"' >> /root/.bashrc source /root/.bashrc pip3 install snips-nlu ```

This converts: ``` CYTHON_UNUSED int a = 0; void func(CYTHON_UNUSED int a) {} ``` to ``` int a = 0; CYTHON_UNUSED_VAR(a); void func(int a) {CYTHON_UNUSED_VAR(a);} ``` this should work for...

@scoder this is ready for review

@scoder after thinking about this again, maybe the following implementation is more appropriate: ```c #ifndef CYTHON_UNUSED_VAR_DECL # if defined(__GNUC__) # if !(defined(__cplusplus)) || (__GNUC__ > 3 || (__GNUC__ == 3...