>>>>> "Neal" == Neal Holtz <nholtz@stripped> writes:
Neal> MySQL Version 3.22.19b (and 3.22.15-gamma):
Neal> The LOAD DATA INFILE query does not seem to complain about
Neal> duplicate records, as my reading of Section 7.15 says it should
Neal> (there are REPLACE and IGNORE options if you want that behaviour).
Neal> Currently it just silently ignores duplicates. I create and load
Neal> a table with:
Neal> ----------------
Neal> DROP TABLE IF EXISTS dups;
Neal> CREATE TABLE dups (
Neal> uid INT PRIMARY KEY,
Neal> name CHAR(10)
Neal> );
Neal> LOAD DATA LOCAL INFILE 'test-data.txt'
Neal> INTO TABLE dups
Neal> FIELDS TERMINATED BY ','
Neal> OPTIONALLY ENCLOSED BY '"'
Neal> ;
Neal> ----------------
Neal> And the data file 'test-data.txt' is:
Neal> ----------------
Neal> 1,one
Neal> 2,two
Neal> 1,one again
Neal> 3,three
Neal> ----------------
Neal> I think an error message should be produced by the third record,
Neal> as the primary key is a duplicate of that in record 1. Currently,
Neal> it is silently dropped.
Neal> cheers
Neal> neal
Hi!
The problem is that when you are using LOAD DATA LOCAL INFILE, then
^^^^^
the DEFAULT behaveour is to ignore all duplicates. The problem is
that the server/protocol for dumping files from the client to the
server doesn't include any possibility for the server to stop the
transmission in the middle.
Regards,
Monty
