SQL 批量插入

Question

我有一个发送给我的 csv 文件。我已将该文件放入 Notepad++，它将 ROWTERMINATOR 显示为 CRLF。这意味着我在 SQL 中使用 \r\n 作为我的 ROWTERMINATOR 对吗？

这是文件格式（每行末尾有 CRLF）：

2020/02/10 03:00:00,2020/02/11 02:59:59,2
101,1,5,1,7,10950,0,10950,0,1429
101,1,5,2,7,28878,5500,28878,0,0
101,1,5,3,4,9525,1200,9525,0,0
101,1,5,4,1,7686,0,7686,0,0
101,1,5,7,3,9094,1300,9094,0,0

我基本上需要从第一个单元格中提取日期，然后忽略该行的其余部分，我已经这样做了：

SET @sql =
'BULK INSERT #date  
FROM ' + '''' + @fileName + '''' + '
WITH
(
    FIRSTROW = 1,
    LASTROW = 1,
    ROWTERMINATOR = ' + '''' + ',' + '''' + ',
    FIELDTERMINATOR = ' + '''' + ',' + '''' + ',
    CODEPAGE=' + '''' + '65001' + '''' + ',
    KEEPNULLS
)'

然后拉取剩余的 5 行数据，这不起作用，因为它将所有数据包装到一行中：

SET @sql =
'BULK INSERT #fileColumns  
FROM ' + '''' + @fileName + '''' + '
WITH
(
    FIRSTROW = 2,
    FIELDTERMINATOR = ' + '''' + ',' + '''' + ',
    ROWTERMINATOR = ' + '''' + '\r\n' + '''' + ',
    CODEPAGE=' + '''' + '65001' + '''' + ',
    KEEPNULLS
)'

我也试过 \n 作为 ROWTERMINATOR，它给我最后 4 行，因为第一行数据被包装到第一行中。

Answer 1

我不太喜欢我必须这样做的方式，但它确实有效：

SET @sql =
'BULK INSERT #fileColumns  
FROM ' + '''' + @fileName + '''' + '
WITH
(
    FIRSTROW = 2,
    ROWTERMINATOR = ' + '''' + '\n' + '''' + ',
    CODEPAGE=' + '''' + '65001' + '''' + ',
    KEEPNULLS
)'

EXEC (@sql)

WITH Split AS (
SELECT  Column1,
        Split.value,
        ROW_NUMBER() OVER(PARTITION BY Column1 ORDER BY (SELECT NULL)) AS RowNumber
    FROM #fileColumns
    CROSS APPLY STRING_SPLIT(Column1, ',') AS Split
)

SELECT  [1] AS ProfitCenterId,
        [2] AS CheckTypeId,
        [3] AS MealPeriodId,
        [4] AS TenderId,
        [5] AS NumberOfTenders,
        [6] AS TenderAmount,
        [7] AS TipAmount,
        [8] AS ReceivedAmount,
        [9] AS BreakageAmount,
        [10] AS ChangeGiven
        FROM Split
        PIVOT
        (
            MAX(value)
            FOR RowNumber IN ([1], [2], [3], [4], [5], [6], [7], [8], [9], [10])
        ) AS piv

绝对不漂亮，但它会做。

SQL 批量插入

SQL Bulk Insert

sql

sql-server

bulkinsert