I think one of the most useful things that would deserve the name "design pattern" is the "phantom types" technique. It is a more-or-less-systematic way to encode static information in a type parameter. See some examples :
You might look at the design advice for Haskell, and specifically, some of the more recent research papers on embedding strong properties into the type system:
Putting Curry-Howard to Work Tim Sheard, Proceedings of the 2005 ACM SIGPLAN workshop on Haskell. Tallinn, Estonia, 74 - 85, 2005
Language-Based Program Verification via Expressive Types Martin Sulzmann and Razvan Voicu. In Programming Languages meet Program Verification (PLPV'06)
Typed Contracts for Functional Programming Ralf Hinze, Johan Jeuring and Andres Loh.
A Library for Light-weight Information-Flow Security in Haskell A Russo, K Claessen, J Hughes .
There's no standard methodologies here yet, though there are many many nice examples of individual problems.
发布评论
评论（2）
我认为最有用的、值得称为“设计模式”的东西之一是“幻像类型”技术。这是一种在类型参数中编码静态信息的或多或少系统的方法。请参阅一些示例：
您可能还对相关的 Oleg Wondertrick 感兴趣：轻量级静态功能（与单忠杰）。
I think one of the most useful things that would deserve the name "design pattern" is the "phantom types" technique. It is a more-or-less-systematic way to encode static information in a type parameter. See some examples :
You may also be interested in a related Oleg wondertrick : Lightweight static capabilities (with Chung-chieh Shan).
您可以查看 Haskell 的设计建议，特别是一些关于将强属性嵌入到类型系统中的最新研究论文：
Tim Sheard，2005 年 ACM SIGPLAN Haskell 研讨会论文集。爱沙尼亚塔林，74 - 85，2005
马丁·苏尔兹曼和拉兹万·沃伊库。在编程语言中满足程序验证（PLPV'06）
拉尔夫·辛兹 (Ralf Hinze)、约翰·杰林 (Johan Jeuring) 和安德烈斯·洛 (Andres Loh)。
A·鲁索，K·克莱森，J·休斯。
尽管有很多关于个别问题的很好的例子，但这里还没有标准的方法。
You might look at the design advice for Haskell, and specifically, some of the more recent research papers on embedding strong properties into the type system:
Tim Sheard, Proceedings of the 2005 ACM SIGPLAN workshop on Haskell. Tallinn, Estonia, 74 - 85, 2005
Martin Sulzmann and Razvan Voicu. In Programming Languages meet Program Verification (PLPV'06)
Ralf Hinze, Johan Jeuring and Andres Loh.
A Russo, K Claessen, J Hughes .
There's no standard methodologies here yet, though there are many many nice examples of individual problems.