我正在尝试将文件从S3加载到Redshift。该文件由Pipe分隔,但是有些值包含Pipe和其他特殊字符,但是如果值包含Pipe,则将其用双引号引起来。
示例:
Field1|Field2
"abc|dh"|123
efh@ih|233
我尝试使用以下命令,但出现错误消息,指出无效数字。这是由于复制命令正在将值中的管道视为定界符。
copy table
from 's3'
iam_role 'arn'
region 'us-east-1'
MAXERROR AS 10 NULL AS '(null)'
'ESCAPE "'
IGNOREHEADER AS 1
DELIMITER '|' timeformat 'auto' GZIP;
答案 0 :(得分:1)
您正在寻找REMOVEQUOTES
参数。 https://docs.aws.amazon.com/redshift/latest/dg/copy-parameters-data-conversion.html#copy-removequotes
ESCAPE
要求您的数据准备有嵌入的转义符。例如,如果您的转义字符为\
,则需要准备数据,以使内容为"abc\|dh"
。
示例:
DROP TABLE IF EXISTS public.quote_test;
CREATE TABLE IF NOT EXISTS public.quote_test (col_a VARCHAR(10), col_b VARCHAR(10));
SELECT * FROM quote_test;
echo '"a|b"|"c|d"' > ~/simple_quotes.txt
aws s3 cp ~/simple_quotes.txt s3://my-bucket/simple_quotes.txt
--Will fail
COPY quote_test FROM 's3://my-bucket/simple_quotes.txt'
CREDENTIALS 'aws_iam_role=arn:aws:iam::012345678901:role/redshift-cluster'
DELIMITER '|' REGION 'us-west-2';
--Succeeds with REMOVEQUOTES
COPY quote_test FROM 's3://benchmark-files/simple_quotes.txt'
CREDENTIALS 'aws_iam_role=arn:aws:iam::012345678901:role/redshift-cluster'
REMOVEQUOTES DELIMITER '|' REGION 'us-west-2';
SELECT * FROM quote_test;
-- col_a | col_b
-- -------+-------
-- a|b | c|d