Few Versatile vs. Many Specialized Collections: How to design a collection library for exploratory programming?

While an integral part of all programming languages, the design of collection libraries is rarely studied. This work briefly reviews the collection libraries of 14 languages to identify possible design dimensions. Some languages have surprisingly few but versatile collections, while others have large libraries with many specialized collections. Based on the identified design dimensions, we argue that a small collection library with only a sequence, a map, and a set type are a suitable choice to facilitate exploratory programming. Such a design minimizes the number of decisions programmers have to make when dealing with collections, and it improves discoverability of collection operations. We further discuss techniques that make their implementation practical from a performance perspective. Based on these arguments, we conclude that languages which aim to support exploratory programming should strive for small and versatile collection libraries.

Transcript

1. Few Versatile vs. Many Specialized Collections How to design a collection library for exploratory programming? Stefan Marr, Benoit Daloze PX/18 10 April 2018 http://stefan-marr.de/papers/px-marr-daloze-few-versatile-vs-many-specialized-collections/
2. 2 Opinion Paper Worst case: you simply agree
3. How to design a collection library? ? 3 ? ?
4. What do existing systems do? 4 C/C++ C#
5. Five* Design Dimensions • Collection types • Language style • Properties • Operation design • Algorithms, data structures, implementation choices 5 * Probably more, but a start
6. Exploratory Programming Programming ✓ Domain ??? 6 Tax Codes Routing Algorithms Machine Learning Protein Folding ? ? ?
7. A Design for Exploratory Programming Collection Types sequences, maps, sets 7 SEQ[1, 2, 3] MAP{a: 1, b: 2} SET{1, 2, 3} array, list, vector multimap stack, queue
8. A Design for Exploratory Programming Language Style OO, FP, …, typed/untyped, templates, … Defined by Language, but: Adaptive optimizations better than explicit choices Types: might require hiding of superfluous details 8
9. A Design for Exploratory Programming Properties Choose Defaults variable sized, ordered, thread safe Offer opt-in/opt-out for special cases! 9 array list SEQ[1, 2, 3]
10. A Design for Exploratory Programming Operation Design All you can need! Optimized for discoverability via completion/suggestions 10
11. A Design for Exploratory Programming Algorithms, data structures, implementation choices Abstraction is bliss! Delegate choices to VM 11 ArrayList LinkedList SEQ[1, 2, 3]
12. For Exploratory Programming: 12 Few but Versatile! SEQ[1, 2, 3] MAP{a: 1, b: 2} SET{1, 2, 3} Defaults + Opt In/Out All operations you can need! Abstraction is bliss!