algorithm - 将 Excel 列号转换为 Matlab 中的列名

Question

我正在使用 Excel 2007，它支持最多 16,384 列的列。我想获取列名对应的列号。

目前，我正在使用以下代码。但是此代码最多支持 256 列。如果列号大于 256，知道如何获取列名。

function loc = xlcolumn(column)

    if isnumeric(column)
        if column>256
            error('Excel is limited to 256 columns! Enter an integer number <256');
        end
        letters = {'A','B','C','D','E','F','G','H','I','J','K','L','M','N','O','P','Q','R','S','T','U','V','W','X','Y','Z'};
        count = 0;
        if column-26<=0
            loc = char(letters(column));
        else
            while column-26>0
                count = count + 1;
                column = column - 26;
            end
            loc = [char(letters(count)) char(letters(column))];
        end

    else
        letters = ['A','B','C','D','E','F','G','H','I','J','K','L','M','N','O','P','Q','R','S','T','U','V','W','X','Y','Z'];
        if size(column,2)==1
            loc =findstr(column,letters);
        elseif size(column,2)==2
            loc1 =findstr(column(1),letters);
            loc2 =findstr(column(2),letters);
            loc = (26 + 26*loc1)-(26-loc2);
        end
    end

谢谢

Answer 1

作为一种消遣，这里有一个全函数句柄示例，(几乎)不需要基于文件的函数。这是基于 dec2base 函数，因为 Excel 列名称(几乎)是 base 26 数字，令人沮丧的区别是没有“0”字符。

注意:总的来说，这可能是一个糟糕的想法，但它确实有效。在文件交换的其他地方可能会找到更好的解决方案。

首先，我无法绕过的一个基于文件的函数，用于执行任意深度函数组合。

function result = compose( fnHandles )
%COMPOSE Compose a set of functions
%    COMPOSE({fnHandles}) returns a function handle consisting of the
%    composition of the cell array of input function handles.  
%
%    For example, if F, G, and H are function handles with one input and
%    one output, then: 
%        FNCOMPOSED = COMPOSE({F,G,H});
%        y = FNCOMPOSED(x);
%    is equivalent to 
%        y = F(G(H(x)));
if isempty(fnHandles)
    result = @(x)x;
elseif length(fnHandles)==1
    result = fnHandles{1};
else
    fnOuter     = fnHandles{1};
    fnRemainder = compose(fnHandles(2:end));
    result = @(x)fnOuter(fnRemainder(x));
end

然后，将 base26 值转换为正确字符串的奇怪的、人为的路径

%Functions leading to "getNumeric", which creates a numeric, base26 array
remapUpper = @(rawBase)(rawBase + (rawBase>='A')*(-55));          %Map the letters 'A-P' to  [10:26]
reMapLower = @(rawBase)(rawBase + (rawBase<'A')*(-48));          %Map characters  '0123456789' to [0:9]
getRawBase = @(x)dec2base(x, 26);

getNumeric = @(x)remapUpper(reMapLower(getRawBase(x)));

%Functions leading to "correctNumeric"
%    This replaces zeros with 26, and reduces the high values entry by 1.
%    Similar to "borrowing" as we learned in longhand subtraction
borrowDownFrom   = @(x, fromIndex) [x(1:(fromIndex-1))  (x(fromIndex)-1) (x(fromIndex+1)+26)  (x((fromIndex+2):end))];
borrowToIfNeeded = @(x, toIndex)   (x(toIndex)<=0)*borrowDownFrom(x,toIndex-1) + (x(toIndex)>0)*(x);  %Ugly numeric switch

getAllConditionalBorrowFunctions = @(numeric)arrayfun(@(index)@(numeric)borrowToIfNeeded(numeric, index),(2:length(numeric)),'uniformoutput',false);
getComposedBorrowFunction = @(x)compose(getAllConditionalBorrowFunctions(x));

correctNumeric = @(x)feval(getComposedBorrowFunction(x),x);

%Function to replace numerics with letters, and remove leading '@' (leading
%zeros)
numeric2alpha = @(x)regexprep(char(x+'A'-1),'^@','');

%Compose complete function
num2ExcelName = @(x)arrayfun(@(x)numeric2alpha(correctNumeric(getNumeric(x))), x, 'uniformoutput',false)';

现在使用一些压力转换进行测试:

>> num2ExcelName([1:5 23:28 700:704 727:729 1024:1026 1351:1355 16382:16384])
ans = 
'A'
'B'
'C'
'D'
'E'
'W'
'X'
'Y'
'Z'
'AA'
'AB'
'ZX'
'ZY'
'ZZ'
'AAA'
'AAB'
'AAY'
'AAZ'
'ABA'
'AMJ'
'AMK'
'AML'
'AYY'
'AYZ'
'AZA'
'AZB'
'AZC'
'XFB'
'XFC'
'XFD'