ExcelDna F#和可选参数 [英] ExcelDna F# and optional arguments

问题描述

对于标量(即非数组式)可选参数，我将使用以下模式:

For scalar (i.e. non array-like) optional arguments, I would use this pattern :

[<ExcelFunction(Category= "Test", Description= "Test optional arguments.")>]
let test_test1 ([<ExcelArgument(Description= "Optional. This is a double. Default is 42.0.")>] arg1 : obj) : double = 
    match arg1 with
    | :? ExcelMissing -> 42.0  // the argument was missing
    | :? double as d  -> d     // the argument was a double
    | _               -> -1.0  // otherwise

我不确定这段代码在Excel-Dna/F#中是否是惯用的"，但似乎是可行的".

I am not sure if this code is "idiomatic" within Excel-Dna / F# but it seems to "work".

但是，我不确定如何继续执行类似数组的可选参数.例如:

However I am not sure how to proceed for optional array-like arguments. Eg :

[<ExcelFunction(Category= "Test", Description= "Test optional arguments.")>]
let test_test2 ([<ExcelArgument(Description= "Optional. This is a double. Default is [42, 42].")>] arg1 : obj[]) : double[] = 
    match arg1.[0] with
    | :? ExcelMissing -> [42.0; 42.0] |> List.toArray // the argument was missing OR it was an empty array
    | :? double as d  -> arg1 |> castToDouble         // the argument was an array and its first argument was a double
    | _               -> Array.empty                  // otherwise

以上似乎适用于大多数情况，但不允许正确处理边缘情况:例如，如果arg1是一个空数组. (castToDouble是自定义的obj[] -> double[]转换函数)

The above seems to work for most cases but does not allow to handle the edge-cases properly : eg if arg1 is an empty array. (castToDouble being a custom obj[] -> double[] conversion function)

在F#/Excel-Dna中处理可选双精度数组的正确/惯用方式是什么，然后如何重写test_test2?

What would be the right / idiomatic way to handle optional double arrays in F# / Excel-Dna and how could I then rewrite test_test2?

===========编辑===

=========== EDIT ===

按照Govert的建议，我尝试了以下方法:

Following Govert's advice, I tried the following :

[<ExcelFunction(Category= "Test", Description= "Test optional arguments.")>]
let test_test3 ([<ExcelArgument(Description= "Optional. This is a double. Default is [42, 42].")>] arg1 : obj) : double[] = 
    match arg1 with
    | :? (double[]) as ds  -> [1.0; 2.0] |> List.toArray   // the argument was a array of double elements
    | :? ExcelMissing      -> [42.0; 42.0] |> List.toArray // the argument was missing OR it was an empty array
    | _                    -> Array.empty                  // otherwise        

...但是不幸的是，当我传递一个双精度数组(或其他形式)时，得到了#NUM!输出.只有当我什么都不做时，我才能正确获得[42.0，42.0]数组.

... but unfortunately I get a #NUM! output when I pass an array of doubles (or of anything else). It's only when I pass nothing that I correctly get the [42.0, 42.0] array.

推荐答案

这涵盖了所有可能性:

[<ExcelFunction("describes the input argument")>]
let describe(arg1 : obj) : string =
    match arg1 with
        | :? double as d        -> sprintf "Double: %f" d
        | :? string as s        -> "String: " + s
        | :? bool as b          -> sprintf "Boolean: %b" b
        | :? ExcelError as err  -> sprintf "ExcelError: %A" err
        | :? (obj[,]) as arr    -> sprintf "Array[%i, %i]" (Array2D.length1 arr) (Array2D.length2 arr)
        | :? ExcelEmpty         -> "<<Empty>>"
        | :? ExcelMissing       -> "<<Missing>>"
        | _                     -> "!? Unheard of ?!"

这篇关于ExcelDna F#和可选参数的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！