Combining multiple csv files from folder, each with different number and/or order of columns

an attempt at a quick and very rough solution

The scenario:

3 file csv in the directory: csv1,csv2,csv3

function load in 4 file: the fourth raises error since doesn't exist.

then the query concaTable put the tables together as required(?):

in this form some parameters are hard coded, but it is not difficult to make them dynamic.
I just wanted to propose an idea of how to set up a solution .. if interested, one can work to introduce the management of other aspects.

the code:

function readCsv:

let
    csvNum=(m)=>
    let
    n=Text.From(m), 
    Source = Csv.Document(File.Contents("C:\Users\abcdefg\OneDrive - TIM\MyD2020\BI\loc0" & n & ".csv"),[Delimiter=";",  Encoding=65001, QuoteStyle=QuoteStyle.None]),
    csv = Table.PromoteHeaders(Source, [PromoteAllScalars=true])
    in Table.AddColumn(csv, "idx", each "csv"& n)
  
in
   csvNum

query concaTable:

let
    Source = List.Transform({1..4}, each readCsvs(_)),
    cols=List.Union(List.Transform({1..4},each try Table.ColumnNames(Source{_-1}) otherwise {})),
    t= Table.FromRecords(List.Combine(List.Transform({0..2},each Table.ToRecords(Source{_}))),cols,MissingField.UseNull)
    
in
    t

#######edited########

I have removed this line of code which was a step of a previous idea, but is no longer needed here

head=Record.FromList(List.Repeat({""}, List.Count(cols)),cols ),