1

我有一个读取 CSV 文件并构建数据表的函数。我不会声称这是我自己的，因为它不是。我不记得我从哪里得到它，但它是来自几个不同来源的组合。

我的函数工作正常，直到我需要处理一些值中有“，”逗号的 CSV。谁能帮我解决这个问题来处理这个问题？

非常感谢 ..

示例 CSV 文件

    FirstName, LastName, Comment, "address, just city",
    John, Smith, "you are very good, but ugly", London,
    Britney, Spear, "I am a singer, and beautiful", New York,

我的功能

    Public Function BuildDataTable() As DataTable

    Dim myTable As DataTable = New DataTable("MyTable")
    Dim i As Integer
    Dim myRow As DataRow
    Dim fieldValues As String()
    Dim myReader As IO.StreamReader
    Dim csv2xml As New csv2xml

    Try
        'Open file and read first line to determine how many fields there are.
        myReader = File.OpenText(_fileFullPath)
        fieldValues = myReader.ReadLine().Split(_seperator)
        'Create data columns accordingly
        If _hasheader = False Then
            For i = 0 To fieldValues.Length() - 1
                myTable.Columns.Add(New DataColumn("Column(" & i & ")"))
            Next
        Else
            'if the file has header, take the first row as header for datatable
            For i = 0 To fieldValues.Length() - 1
                myTable.Columns.Add(New DataColumn(fieldValues(i).Replace(" ", "")))
            Next
        End If

        'Adding the first line of data to data table
        myRow = myTable.NewRow

        'if the csv file has not got a column header. defined by radio button list on first page by user
        'if csv file has header, then not need to read the first line
        If _hasheader = False Then
            For i = 0 To fieldValues.Length() - 1
                myRow.Item(i) = fieldValues(i).ToString
            Next
            myTable.Rows.Add(myRow)
        End If

        'Now reading the rest of the data to data table
        While myReader.Peek() <> -1
            fieldValues = myReader.ReadLine().Split(_seperator)
            myRow = myTable.NewRow
            For i = 0 To fieldValues.Length() - 1
                myRow.Item(i) = fieldValues(i).Trim.ToString
            Next
            'check if there are empty rows in csv, ignore empty rows

            If Not csv2xml.AreAllColumnsEmpty(myRow) = True Then
                myTable.Rows.Add(myRow)
            End If

        End While
    Catch ex As Exception
        'MsgBox("Error building datatable: " & ex.Message)
        Dim oError As ErrorLog = New ErrorLog
        oError.LogError(_strWebsiteName, _
            loginID, _
            ex.Source.ToString, _
            ex.Message.ToString, _
            , _
            ex.StackTrace.ToString)
        oError = Nothing
        Return New DataTable("Empty")
        'Server.Transfer(CustomErrorPage)
    Finally
        csv2xml = Nothing
        myRow = Nothing
    End Try
    myReader.Close()
    Return myTable

    End Function
4

2 回答 2

6

只需使用 VB TextFieldParser类，并将 HasFieldsEnclosedInQuotes 属性设置为 True：

于 2012-07-05T15:37:00.413 回答
2

您的问题是您似乎通过简单地用逗号分隔每个行字符串作为分隔符来解析列。该方法不适用于所有 CSV 文件，因为如果它是包含逗号的字符串，如您所示，整个字段值将被引号括起来。因此，您需要先用引号分隔，然后在剩下的部分用逗号分隔，然后从字段的值中删除引号，这是一个更复杂的算法。您可以自己编写类似的内容，或使用正则表达式来执行此操作，但那时，我认为您最好使用 ADO 加载 CSV 文件：

Public Function LoadCsvFile(filePath As String, hasHeader As Boolean) As DataTable
    Dim folderPath As String = Path.GetDirectoryName(filePath)
    Dim fileName As String = Path.GetFileName(filePath)
    Dim hdr As String = Nothing
    If hasHeader Then
        hdr = "Yes"
    Else
        hdr = "No"
    End If
    Dim connectionString As String = String.Format("Driver={Microsoft Text Driver (*.txt; *.csv)};Dbq={0};Extended Properties=""Text;HDR={1};FMT=Delimited""", folderPath, hdr)
    Dim connection As New OdbcConnection(connectionString)
    Dim adapter As New OdbcDataAdapter("SELECT * FROM [" + fileName + "]", connection)
    Dim table As New DataTable()
    adapter.Fill(table)
    Return table
End Function
于 2012-07-05T13:05:06.140 回答