SQL Server 2005 - Bulk Insert failing

Question

I have a txt file that contains 1600 rows and 82 columns of comma delineated data that I am trying to import into a table. I get the following error on every row on the very last field:

Msg 4864, Level 16, State 1, Line 1 Bulk load data conversion error (type mismatch or invalid character for the specified codepage) for row 1, column 81 (DB252D20C8).

The import statement is

BULK
INSERT  [ENERGY].[dbo].[READINGS1]
       from 'c:\readings2.txt'
with
(
DATAFILETYPE='widechar',
FIELDTERMINATOR = ',',
ROWTERMINATOR = '\n'
)
GO

The table structure is as follows, the top and bottom of the script:

USE [ENERGY]
GO
/****** Object:  Table [dbo].[READINGS1]    Script Date: 05/13/2013 20:00:30 ******/
SET ANSI_NULLS ON
GO
SET QUOTED_IDENTIFIER ON
GO
CREATE TABLE [dbo].[READINGS1](
[DateAndTime] [datetime] NOT NULL,
[DB240D4C7] [float] NULL,
[DB240D8C7] [float] NULL,
[DB240D12C7] [float] NULL,
[DB240D16C7] [float] NULL, 



[DB252D12C8] [float] NULL,
[DB252D16C8] [float] NULL,
[DB252D20C8] [float] NULL,
 CONSTRAINT [READINGS1DataTimeStamp] PRIMARY KEY CLUSTERED 
(
[DateAndTime] ASC
)WITH (PAD_INDEX  = OFF, IGNORE_DUP_KEY = OFF) ON [PRIMARY]
) ON [PRIMARY]

The text file is as follows:

2013-02-19 00:00:00.000,6,945,1886,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,22,2040,6299,0,0,6,567,1248,0,0,251,8859,8655,0,0,10,316,1786,0,0,7,180,1206,0,0,1,16,56,0,0,368,18953,36949,0,0,NULL,NULL
2013-02-19 01:00:00.000,6,147,1886,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,22,1516,6299,0,0,3,115,1248,0,0,250,5077,8655,0,0,9,219,1786,0,0,5,147,1206,0,0,1,15,56,0,0,362,8907,36949,0,0,NULL,NULL

Answer 1

Alright so what you need to do is alter your statement so that after the end of the file you use KEEPNULLS . This informs SQL server that you wish to keep your null values. Currently it's trying to convert NULL as a string into your FLOAT COLUMN . Alter your statment to look like this.

BULK
INSERT  [ENERGY].[dbo].[READINGS1]
       from 'c:\readings2.txt'
with
(
DATAFILETYPE='widechar',
FIELDTERMINATOR = ',',
ROWTERMINATOR = '\n',
KEEPNULLS
)
GO

There is an article on BOL about this. .

Otherwise you can always build a Integration Services package to handle this. That is an easy fast way to import information from flat file sources.

Answer 2

事实证明，表的输入文本文件中有太多字段。