Chia-Network / clvm_rs / 14007036343

push

github

[CHIA-2052] optimize compressed CLVM serialization (#562)

* TreeCache to speed up serialization with backrefs

* use SHA-1 to deduplicate sub trees

* use bump allocator (bumpalo) to speed up allocations for partial paths, in find_path()

* extend comments

* address review feedback

* PathBuilder::better_than() is not used

* Update src/serde/tree_cache.rs

Co-authored-by: Richard Kiss <him@richardkiss.com>

* rename VisitedNodes -> BitSet

* rename node_entry -> node_entries

* rename cache -> node_map

* update comments

* rename TreeUndoState -> TreeCacheCheckpoint

* Update src/serde/tree_cache.rs

Co-authored-by: Richard Kiss <him@richardkiss.com>

---------

Co-authored-by: Richard Kiss <him@richardkiss.com>

580 of 594 new or added lines in 5 files covered. (97.64%)

6851 of 7254 relevant lines covered (94.44%)

28492641.04 hits per line