您可以使用 ADO（ODBC/JET/OLEDB 文本驱动程序）将“体面”的 .txt/.csv/.tab/.flr 文件视为 SQL 数据库中每种支持 COM 语言的表。然后可以使用 SQL 的强大功能进行比较（DISTINCT、GROUP、(LEFT) JOINS、...）。
添加了关于您的评论：
这是你的问题，我不想把你推到你不想去的地方。但是如果您需要比较表格数据，SQL 是一个很好的（最好的？）工具。作为发现两个 .txt 文件中差异的脚本输出的证据：
======= The .txt files to play with
------- file1.txt
"AC";"AM"
40000;-19083,00
40100;20000,00
40200;350004,00
40300;3498,99
------- file2.txt
"AC";"AM"
40000;-19083,00
40300;3498,99
40105;-234567,00
40200;350,00
======= Some diagnostic SQL
------- <NULL> indicates: In F1 but not in F2 (LEFT JOIN)
SELECT T1.AC, T1.AM, T2.AM FROM [file1.txt] AS T1 LEFT JOIN [file2.txt] AS T2 ON (T1.AC =
T2.AC)
------- Result
AC File1 File2
40000 -19083 -19083
40100 20000 <NULL>
40200 350004 350
40300 3498,99 3498,99
------- <NULL> indicates: Not in the other file (LEFT JOIN, UNION)
SELECT T1.AC, T1.AM, T2.AM FROM [file1.txt] AS T1 LEFT JOIN [file2.txt] AS T2 ON (T1.AC =
T2.AC) UNION SELECT T2.AC, T1.AM, T2.AM FROM [file2.txt] AS T2 LEFT JOIN [file1.txt] AS T1
ON (T1.AC = T2.AC)
------- Result
AC File1 File2
40000 -19083 -19083
40100 20000 <NULL>
40105 <NULL> -234567
40200 350004 350
40300 3498,99 3498,99
------- the problems: missing, different values
SELECT T1.AC, T1.AM, T2.AM FROM [file1.txt] AS T1 LEFT JOIN [file2.txt] AS T2 ON (T1.AC =
T2.AC) WHERE T2.AM IS NULL OR T1.AM <> T2.AM UNION SELECT T2.AC, T1.AM, T2.AM FROM [file2.
txt] AS T2 LEFT JOIN [file1.txt] AS T1 ON (T1.AC = T2.AC) WHERE T1.AM IS NULL OR T1.AM <>
T2.AM
------- Result
AC File1 File2
40100 20000 <NULL>
40105 <NULL> -234567
40200 350004 350
进一步补充：
这个article 处理ADO 和文本文件；查找文件 adoNNN.chm
（NNN=版本号，例如 210）在您的计算机上；这是good book 关于
ADO。
您可以使用 Access 或 OpenOffice Base 来试验 SQL 语句
应用于链接/引用（未导入！）文本数据库。
在您掌握了最初的障碍后，脚本/程序将变得很容易：连接
到数据库，即到包含文件和 schema.ini 的文件夹
文件来定义 files=tables 的结构。
上面的输出是由以下生成的：
Const adClipString = 2
Dim oFS : Set oFS = CreateObject( "Scripting.FileSystemObject" )
Dim sDir : sDir = oFS.GetAbsolutePathName( ".\txt" )
Dim oDB : Set oDb = CreateObject( "ADODB.Connection" )
oDB.Open "Provider=Microsoft.Jet.OLEDB.4.0;Data Source=" & sDir & ";Extended Properties=""text"""
Dim sSQL
Dim sFiNa
WScript.Echo "=======", "The .txt files to play with"
For Each sFiNa In Array( "file1.txt", "file2.txt" )
WScript.Echo "-------", sFiNa
WScript.Echo oFS.OpenTextFile( "txt\" & sFiNa ).ReadAll()
Next
WScript.Echo "=======", "Some diagnostic SQL"
Dim aSQL
For Each aSQL In Array( _
Array( "<NULL> indicates: In F1 but not in F2 (LEFT JOIN)" _
, Join( Array( _
"SELECT T1.AC, T1.AM, T2.AM FROM" _
, "[file1.txt] AS T1" _
, "LEFT JOIN [file2.txt] AS T2 ON (T1.AC = T2.AC)" _
), " " ) ) _
, Array( "<NULL> indicates: Not in the other file (LEFT JOIN, UNION)" _
, Join( Array( _
"SELECT T1.AC, T1.AM, T2.AM FROM" _
, "[file1.txt] AS T1" _
, "LEFT JOIN [file2.txt] AS T2 ON (T1.AC = T2.AC)" _
, "UNION" _
, "SELECT T2.AC, T1.AM, T2.AM FROM" _
, "[file2.txt] AS T2" _
, "LEFT JOIN [file1.txt] AS T1 ON (T1.AC = T2.AC)" _
), " " ) ) _
, Array( "the problems: missing, different value" _
, Join( Array( _
"SELECT T1.AC, T1.AM, T2.AM FROM" _
, "[file1.txt] AS T1" _
, "LEFT JOIN [file2.txt] AS T2 ON (T1.AC = T2.AC)" _
, "WHERE T2.AM IS NULL OR T1.AM <> T2.AM" _
, "UNION" _
, "SELECT T2.AC, T1.AM, T2.AM FROM" _
, "[file2.txt] AS T2" _
, "LEFT JOIN [file1.txt] AS T1 ON (T1.AC = T2.AC)" _
, "WHERE T1.AM IS NULL OR T1.AM <> T2.AM" _
), " " ) ) _
)
sSQL = aSQL( 1 )
WScript.Echo "-------", aSQL( 0 )
WScript.Echo sSQL
Dim oRS : Set oRS = oDB.Execute( sSQL )
WScript.Echo "------- Result"
WScript.Echo Join( Array( "AC", "File1", "File2" ), vbTab )
WScript.Echo oRS.GetString( adClipString, , vbTab, vbCrLf, "<NULL>" )
Next
oDB.Close
如果你删除/忽略胖子（创建 SQL 语句、诊断输出），它就会沸腾
最多 6 行
Dim oDB : Set oDb = CreateObject( "ADODB.Connection" )
oDB.Open "Provider=Microsoft.Jet.OLEDB.4.0;Data Source=" & sDir & ";Extended Properties=""text"""
sSQL = "..."
Dim oRS : Set oRS = oDB.Execute( sSQL )
WScript.Echo oRS.GetString( adClipString, , vbTab, vbCrLf, "<NULL>" )
oDB.Close
它可以很容易地“移植”到每一种支持 COM 的语言，因为 ADO
物体完成所有繁重的工作。 .GetString 方法很方便，当你
想要保存结果集：只需旋转分隔符/分隔符/Null 参数
并将其转储到文件中
oFS.CreateTextFile( ... ).WriteLine oRS.GetString( _
adClipString, , ",", vbCrLf, ""
)
（不要忘记将该表的定义添加到您的 schema.ini）。的
当然你也可以使用“SELECT/INSERT INTO”，但这样的语句可能不会
易于正确/通过 ADO 文本驱动程序的解析器。
加法 wrt 计算：
从一个 5 x 2 主/批准文件开始，其中包含：
Num0 Num1 Num2 Num3 Num4
7,6 6,1 3,8 0,9 8,9
0,9 9,4 4,7 8,8 9,9
将其转换为expected.txt
Num0 Num1 Num2 Num3 Num4 Spot
7,6 6,1 3,8 0,9 8,9 1
0,9 9,4 4,7 8,8 9,9 2
通过附加 Spot 列使其符合
[expected.txt]
ColNameHeader=True
CharacterSet=1252
Format=Delimited(;)
Col1=Num0 Float
Col2=Num1 Float
Col3=Num2 Float
Col4=Num3 Float
Col5=Num4 Float
Col6=Spot Integer
在您的 schema.ini 文件中。同样，转换一个测量文件，如：
Num0 Num1 Num2 Num3 Num4
7,1 1,1 3,8 0,9 8,9
0,9 9,4 4,7 8,8 9,9
到measured.txt
Num0 Num1 Num2 Num3 Num4 Spot
7,1 1,1 3,8 0,9 8,9 1
0,9 9,4 4,7 8,8 9,9 2
申请
sSQL = Join( Array( _
"SELECT E.Num0 - M.Num0 AS Dif0" _
, ", E.Num1 - M.Num1 AS Dif1" _
, ", E.Num2 - M.Num2 AS Dif2" _
, ", E.Num3 - M.Num3 AS Dif3" _
, ", E.Num4 - M.Num4 AS Dif4" _
, ", E.Spot AS Spot" _
, "FROM [expected.txt] AS E" _
, "INNER JOIN [measured.txt] AS M" _
, "ON E.Spot = M.Spot" _
), " " )
将结果集写入differences.txt
aFNames = Array("Num0", ... "Spot" )
oFS.CreateTextFile(sFSpec).Write _
Join( aFNames, sFSep ) & sRSep & oRS.GetString( adClipString, , sFSep, sRSep, "" )
你会得到：
Num0 Num1 Num2 Num3 Num4 Spot
0,5 5 0 0 0 1
0 0 0 0 0 2