I have some vendor data that has the SOH (ASCII character 1) as the field delimiter and STX (ASCII character 2) as the record delimiter. Is it possible to load this data with LOAD DATA INFILE without pre-processing the file and replacing those characters with something more common?
I got it.
LOAD DATA LOCAL INFILE 'myfile.txt' INTO TABLE my_table
CHARACTER SET UTF8
FIELDS TERMINATED BY X'01'
LINES TERMINATED BY X'02'
(col1, col2, col3);
You might try FIELDS TERMINATED BY _ascii 0x02
. I don't know if it will work for LOAD DATA INFILE
, but it works in SELECT
(i.e., SELECT _ascii 0x61
yields 'a').
If you're using mysqlimport the format for hex values in fields-terminated-by and lines-terminated-by etc is:
mysqlimport --local --user=username --password=secret --ignore-lines=4 --default-character-set=UTF8 --fields-terminated-by=0x01 --lines-terminated-by=0x02 --verbose databasename thefiletoimport
You can try just sending the ascii char directly inside the string literal.. if your connection doesn't have a charset or encoding assigned, then mysql may simply accept it as a valid string. You'd have to do it through a network connection, or piping data to the mysql client. I don't think you're going to be able to type that in at a console.
FIELDS TERMINATED BY X'01'
works for me
来源:https://stackoverflow.com/questions/1493650/using-mysql-load-data-infile-with-nonprintable-character-delimiters