Type Providers: Making Data First Class
summary
F# type providers connect to a live CSV/XML/DB, run IDE-time schema generation to produce typed APIs and Intellisense, and surface errors as squiggles before compilation; most providers emit erased types => no runtime cost so performance stays native but runtime reflection is limited. Examples include SQLProvider with LINQ-style queries, converting SQL nulls to option types to catch missing data, handling multiple samples for XML to avoid surprises, and known limits like schema drift, missing geo types, and debugger opacity.
AI-written summary · about
chapters (11)
- 0:00Start
- 0:30Definition of What a Type Provider
- 6:07Downcast Operator
- 7:30Type Safety at the Ide
- 10:15Idea behind Type Providers
- 12:05Intellisense
- 22:39Type Inference
- 22:53Units of Measure
- 24:58Logic Errors
- 26:31Explain What a Type Provider Is
- 38:56Recap
related talks
56mHow To "Dot" Into Anything, Anywhere, Anytime (Facilitating Data-First Programming)
38mThe Gamma: Programming Tools for Data Journalism
30mThe Fun of Runtime Type Systems (or "Funtime Type Systems")
25mLeveraging (algebraic data) types to make your UI rock solid
24mLet the Scala compiler work for you
31mWrite your Own Domain Specific Language with F#
15mLightning Talk: My Journey Building a Dynamic Provider
48mTDD, Like in Type Driven Development