Reputation: 12859
I generated a (UTF-8) file by an external program for importing into PostgreSQL 9.6.1. Problem is the bytea
field (PWHASH).
Snippet from this file (using TAB as delimiter)
COPY USERS (ID,CODE,PWHASH,EMAIL) FROM stdin;
7 test1 E'\\\\x657B954D27B4AC56FA997D24A5FF2563' [email protected]
\.
When importing with
psql mydb myrole -f test.sql
Everything goes well.
However, if i query the result, the byte array is not 16 bytes, but 37 bytes:
select passwordhash,length(passwordhash) from users;
passwordhash | length
------------------------------------------------------------------------------+--------
\x45275c78363537423935344432374234414335364641393937443234413546463235363327 | 37
What is the correct syntax for this?
Upvotes: 2
Views: 7140
Reputation: 247495
The format of the input file is wrong. It should be like this:
7 test1 \\x657B954D27B4AC56FA997D24A5FF2563 [email protected]
Upvotes: 9
Reputation: 51609
I will have to "prepare" data I believe. Smth like here:
t=# insert into u select 'x657B954D27B4AC56FA997D24A5FF2563';
INSERT 0 1
Time: 5990.809 ms
t=# select b from u;
b
----------------------------------------------------------------------
\x783635374239353444323742344143353646413939374432344135464632353633
(1 row)
Time: 0.234 ms
t=# insert into u select decode('657B954D27B4AC56FA997D24A5FF2563','hex');
INSERT 0 1
Time: 62.767 ms
t=# select b from u;
b
----------------------------------------------------------------------
\x783635374239353444323742344143353646413939374432344135464632353633
\x657b954d27b4ac56fa997d24a5ff2563
(2 rows)
Time: 0.208 ms
So in your case you can:
create table t as select ID,CODE,PWHASH::text,EMAIL from users where false;
COPY t (ID,CODE,PWHASH,EMAIL) FROM stdin;
insert into users select ID,CODE,decode(substr(PWHASH,4),'hex'),EMAIL from t;
Upvotes: 0