https://github.com/tkvogt/streamly-judy

Stream the key/value-pairs of a judy array into an arbitrary data structure
https://github.com/tkvogt/streamly-judy

Last synced: 7 days ago
JSON representation

Stream the key/value-pairs of a judy array into an arbitrary data structure

• Host: GitHub
• URL: https://github.com/tkvogt/streamly-judy
• Owner: tkvogt
• License: bsd-3-clause
• Created: 2022-01-20T12:41:36.000Z (about 3 years ago)
• Default Branch: master
• Last Pushed: 2023-11-12T16:45:55.000Z (about 1 year ago)
• Last Synced: 2023-11-12T17:57:12.966Z (about 1 year ago)
• Language: Haskell
• Size: 8.79 KB
• Stars: 0
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • Changelog: ChangeLog.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# streamly-judy

Stream the key/value-pairs of a judy array, dumping it to a file.

In the [judy](https://hackage.haskell.org/package/judy-0.4.1) library the functions

```bash

keys :: JudyImmutable a -> IO [Key]

elems :: JE a => JudyImmutable a -> IO [a]

toList :: JE a => JudyImmutable a -> IO [(Key, a)] 

```

return lists, which can use quite a lot of memory. This is problematic when the array is big.

It makes much more sense to stream a judy array into an arbitrary data structure.

This way much bigger judy arrays can be loaded from/stored into a file, which is also in my use case faster than the construction of the judy array and reduces startup time.

Instead of adding these functions to [judy](https://hackage.haskell.org/package/judy-0.4.1/docs/Data-Judy.html) or [streamly](https://hackage.haskell.org/package/streamly-0.8.1),

we make a new library to not force new dependencies on either of the two.