is this a bug in db.py?

[BjornT]

Look at the different results when I switch types of quotes in
SQLLiteral:

>>> web.insert('foo', wtf=web.SQLLiteral("SHA1('yomomma')"),joe='bob',a=2,_test=True)
<sql: "INSERT INTO foo (a, joe, wtf) VALUES (2, 'bob',
SHA1('yomomma'))">

>>> web.insert('foo', wtf=web.SQLLiteral('SHA1("yomomma")'),joe='bob',a=2,_test=True)
<sql: 'INSERT INTO foo (a, joe, wtf) VALUES (2, \'bob\',
SHA1("yomomma"))'>

Is this a bug or am I doing something dumb?

I think if there is a problem it's because I have DBUtils or this:

sql_query = SQLQuery("INSERT INTO %s (%s) VALUES (%s)" % (
tablename,
", ".join(values.keys()),
', '.join([aparam() for x in values])
), values.values())

Might have something to do with two string formats being created here?
The first explicitly and the next passed to SQLQuery or maybe the
issue is in SQLQuery, perhaps here:

return self.s % tuple([sqlify(x) for x in self.v])

The use of sqlify?

I'm not sure. :o

[Aaron Swartz]

> Look at the different results when I switch types of quotes in
> SQLLiteral:

These aren't different results; the backslashes are just put there by
Python's repr.

[BjornT]

Oh right, I was just going to say that. I noticed this:

>>> n
' bob SHA1("yomomma") '
>>> web.db.sqlify(n)
'\' bob SHA1("yomomma") \''
>>> n2 = " bob SHA1('yomomma') "
>>> web.db.sqlify(n2)
'" bob SHA1(\'yomomma\') "'
>>> print web.db.sqlify(n)
' bob SHA1("yomomma") '
>>> print web.db.sqlify(n2)
" bob SHA1('yomomma') "

Thanks.

[BjornT]

Seems like something is not ok though, check it out:

>> d = web.insert('users', email='bjorn....@foo.com',password=web.SQLLiteral('SHA1("foo")'),level=1,joindate=web.SQLLiteral('NOW()'),laston=web.SQLLiteral('NOW()'), _test=True)

>>> print d
INSERT INTO users (password, joindate, email, laston, level) VALUES
(SHA1("foo"), NOW(), 'bjorn....@foo.com', NOW(), 1)

now I copy and paste that string in MySQL:

mysql> INSERT INTO users (password, joindate, email, laston, level)
VALUES (SHA1("foo"), NOW(), 'bjorn....@foo.com', NOW(), 1)
Query OK, 1 row affected (0.06 sec)

mysql> select * from users;

+----+-------------------------
+------------------------------------------+---------------------
+---------------------+-------+
| id | email |
password | joindate |
laston | level |
+----+-------------------------
+------------------------------------------+---------------------
+---------------------+-------+
| 5 | bjorn....@ask.com |
b3d27fe823942946dbbbc16e62537743a458131a | 2008-04-07 23:45:41 |
2008-04-07 23:45:41 | 1 |
+----+-------------------------
+------------------------------------------+---------------------
+---------------------+-------+

But if I use web.insert check it out:

d = web.insert('users',
email='bjorn.t...@foo2.com',password=web.SQLLiteral('SHA1("foo")'),level=1,joindate=web.SQLLiteral('NOW()'),laston=web.SQLLiteral('NOW()'))

mysql> select * from users;
+----+-------------------------
+------------------------------------------+---------------------
+---------------------+-------+
| id | email |
password | joindate |
laston | level |
+----+-------------------------
+------------------------------------------+---------------------
+---------------------+-------+
| 5 | bjorn....@foo.com |
b3d27fe823942946dbbbc16e62537743a458131a | 2008-04-07 23:45:41 |
2008-04-07 23:45:41 | 1 |
| 6 | bjorn.t...@foo2.com | SHA1("foo")
| 0000-00-00 00:00:00 | 0000-00-00 00:00:00 | 1 |
+----+-------------------------
+------------------------------------------+---------------------
+---------------------+-------+

It didn't create an SHA1 hash.

On Apr 7, 11:07 pm, "Aaron Swartz" <m...@aaronsw.com> wrote:

[Anand Chitipothu]

On Tue, Apr 8, 2008 at 12:22 PM, BjornT <bjorn....@gmail.com> wrote:
>
> Seems like something is not ok though, check it out:
>
> >> d = web.insert('users', email='bjorn....@foo.com',password=web.SQLLiteral('SHA1("foo")'),level=1,joindate=web.SQLLiteral('NOW()'),laston=web.SQLLiteral('NOW()'), _test=True)
>

Yes, that is a bug. It has been fixed in 0.3.