Optimize `btreemap::range(...).count()`

[peterpeterparker]

Features

It would be nice to expose a more performant way to get to know the number of entries in the stable structure btreemap where keys belong to the specified range than my_btree_map.range().count().

Context

In Juno, while I still need to perform quite some work to improve my code's usage of stable structures, I have exposed a few "count documents and assets" functions. These functions inform the developer about the number of elements in a collection. Therefore, it would be handy for my use case if such information could be found in a performant way.

Notes

• Follow-up of forum thread.

• This suggestion is not a blocker on my side.

[ielashi]

Thanks @peterpeterparker. @dragoljub-duric will be taking a look at this.

[ielashi]

@peterpeterparker with #190, range(...)count() should be much faster. Here are our benchmarks:

---------------------------------------------------

Benchmark: btreemap_iter_count_small_values
  total:
    instructions: 10.12 M (improved by 60.45%)
    heap_increase: 0 pages (no change)
    stable_memory_increase: 0 pages (no change)

---------------------------------------------------

Benchmark: btreemap_iter_count_10mib_values
  total:
    instructions: 497.38 K (improved by 97.89%)
    heap_increase: 0 pages (no change)
    stable_memory_increase: 0 pages (no change)

---------------------------------------------------

Once we release a version containing #190, we can close this issue. Thanks @dragoljub-duric for taking care of this!

[peterpeterparker]

Aweome 🔥

[dragoljub-djuric]

This can be found in the new 0.6.3 release of Stable Structures.