In SQL Server, I can use IF conditional structure to execute some statements if a condition is true. According to this and this, there seem to be no such structure in SQLite.
I want to check if a table exist, if it does, do nothing, if not, do a lot of things including creating tables, inserting and deleting data from other tables and updating as well:
CASE WHEN ((SELECT COUNT(*) FROM sqlite_master WHERE type = 'table' AND name = 'TraitsSwap') = 1) THEN
-- 50 lines of code, including CREATE, DROP, INSERT, DELETE and UPDATE statements, with random() in used
ELSE
-- Do nothing
END
Is there anyway I can achieve this? The code includes usage of random() and it requires consistent result (i.e, only random in the first time). I am sorry if this sounds unreasonable, but this is in context of game modding, so I cannot really change the backend code to run separated transaction code.
I think there may be an alternative if there is a function in SQLite that can execute a string/statement block and return a result. For that, I can transform the query into
SELECT CASE WHEN ((SELECT COUNT(*) FROM sqlite_master WHERE type = 'table' AND name = 'TraitsSwap') = 1) THEN
ExecuteCode("Code; RETURN 1;")
ELSE
0
END
I tried
SELECT CASE WHEN ((SELECT COUNT(*) FROM sqlite_master WHERE type = 'table' AND name = 'TraitsSwap') = 1) THEN
SELECT 1;
INSERT INTO Foo(Test) VALUES("");
SELECT "A";
ELSE
SELECT 1;
SELECT 2;
SELECT "A";
END
but it's unsuccessful, the error is
near "SELECT": syntax error: SELECT CASE WHEN ((SELECT COUNT(*) FROM
sqlite_master WHERE type = 'table' AND name = 'TraitsSwap') = 1) THEN
SELECT
Related
I want to execute DELETE instruction before the end of my procedure. First I get some data in my CURSOR. On the loop of this CURSOR, for each record of my CURSOR I concatenate id values in variable named "final_list". At the end of the loop I would like to execute DELETE instruction like this : DELETE FROM my_table where my_field IN final_list. But not working.
create or replace PROCEDURE TEST_PURGE is
CURSOR clients IS SELECT DISTINCT client_id
FROM client
WHERE client_description LIKE 'Test%';
client clients%ROWTYPE;
id_log client.client_id%type;
final_list VARCHAR(100);
BEGIN
final_list:='(';
OPEN clients;
LOOP
FETCH clients INTO client;
EXIT WHEN clients%notfound;
SELECT log_id INTO id_log
FROM (SELECT log_id
FROM log
WHERE log_client_id = client.client_id
AND client_description LIKE 'Test%'
ORDER BY log_date DESC)
WHERE ROWNUM < 2;
final_list:=concat(final_list, id_log || ',');
END LOOP;
CLOSE clients;
final_list:=SUBSTR(final_list, 0, LENGTH(final_list) - 1);
final_list:=concat(final_list, ')');
DBMS_OUTPUT.PUT_LINE('Id list: ' || final_list);
DELETE FROM contrh_client_log WHERE contrh_client_log_id IN final_list;
COMMIT;
END TEST_PURGE;
DELETE instruction not working but no error message. When I execute the same DELETE instruction in classic SQL sheet with value of "final_list" variable, it's work.
Does anyone have an idea ?
Yes: the best way of doing this is not by looping through a cursor, running a select statement and then a delete statement; instead, you can do it all in a single delete statement, such as:
DELETE FROM contrh_client_log
WHERE contrh_client_log_id IN (SELECT log_id
FROM (SELECT l.log_client_id,
l.log_id,
row_number() OVER (PARTITION BY l.log_client_id ORDER BY log_date DESC) rn
FROM LOG l
INNER JOIN client c ON l.log_client_id = c.client_id
WHERE l.client_description LIKE 'Test%'
AND c.client_description LIKE 'Test%')
WHERE rn = 1);
(N.B. untested.)
You can then put this in a procedure:
CREATE OR REPLACE PROCEDURE test_purge AS
BEGIN
DELETE FROM contrh_client_log
WHERE contrh_client_log_id IN (SELECT log_id
FROM (SELECT l.log_client_id,
l.log_id,
row_number() over(PARTITION BY l.log_client_id ORDER BY log_date DESC) rn
FROM log l
INNER JOIN client c
ON l.log_client_id = c.client_id
WHERE l.client_description LIKE 'Test%'
AND c.client_description LIKE 'Test%')
WHERE rn = 1);
COMMIT;
END;
/
Doing it like this will be
faster,
easier to understand,
easier to debug (you can just run the delete statement on its own outside of the procedure with little or no changes to it)
easier to maintain in the future.
My current query:
select timestamp from messagesTable
where partner_jid='" + lastUserJid + "' AND msg='.roll'
order by timestamp DESC LIMIT 1 OFFSET 1;
This works fine... unless the values don't exist in the database.
If values do not exist in database, then it should Select * messagesTable; or do nothing if possible.
Is there a way to add a check for that within the same query? It has to be the same query unfortunately because I need to execute things through adb shell. I've been trying things out with CASE but I do not really understand much about SQL.
You can just append a second query, with a WHERE filter that checks whether the first query did not return anything:
SELECT *
FROM (SELECT timestamp
FROM messagesTable
WHERE partner_jid = ?
AND msg = '.roll'
ORDER BY timestamp DESC
LIMIT 1 OFFSET 1)
UNION ALL
SELECT -1 -- or "timestamp FROM msgTab", or whatever
WHERE NOT EXISTS (SELECT timestamp
FROM messagesTable
WHERE partner_jid = ?
AND msg = '.roll');
I have the following code which is not executing correctly. I have data stored in date_tmp (varchar) that includes dates and nondates. I want to move the dates in that column to date_run (date) and data that is not a date, will be moved to a comments (varchar) column. When I run the following code, the entire set of data gets moved to comments. It runs fine when I edit out the insert statement and just run the dbms_outputline line. What might I be doing incorrectly?
DECLARE
CURSOR getrow IS
SELECT a.id, a.date_tmp
FROM mycolumn a
WHERE a.id < 1300;
v_date date;
BEGIN
FOR i in getrow LOOP
BEGIN
v_date := to_date(i.date_tmp, 'mm/dd/yy');
INSERT INTO mycolumn a(a.date_run)
VALUES(i.date_tmp);
EXCEPTION
WHEN OTHERS THEN
--dbms_output.put_line(i.date_tmp);
update mycolumn a
SET a.comments = i.date_tmp
where a.id = i.id;
END;
END LOOP;
END;
You try to insert varchar i.date_tmp into a date field. Instead insert v_date.
...
INSERT INTO mycolumn a (a.date_run)
VALUES(v_date);
...
But actually your requirement is a move. That calls for an update actually. So I think what you really want to do is:
...
update mycolumn a
SET a.date_run = v_date
where a.id = i.id
...
And actually you could have a function that checks if you have a valid date or not and then you might be able to handle the whole task using a simple update statement.
create or replace function is_a_date(i_date varchar2, i_pattern varchar2)
return date
is
begin
return to_date(i_date, i_pattern);
exception
when others return null;
end is_a_date;
With that function you could write two update statements
update mycolumn
set date_run = to_date(date_tmp,'dd/mm/yy')
where is_a_date(date_tmp, 'dd/mm/yy') is not null;
update mycolumn
set comment = date_tmp
where is_a_date(date_tmp, 'dd/mm/yy') is null;
I designed the function in a way that you could use it in various ways as it returns you a date or null but no exception if the varchar does not conform to the date pattern.
You have an insert where it looks like you need an update, like you have in the exception handler. So just change it to:
v_date := to_date(i.date_tmp, 'mm/dd/yy');
update mycolumn
set date_run = v_date
where id = i.id;
or you could shorten it to:
update mycolumn
set date_run = to_date(i.date_tmp, 'mm/dd/yy')
where id = i.id;
#hol solution is the best approach for me.
Avoid always you can loops and procedures if you can do it with simple SQL statments, your code will be more faster.
Also, if you have always have a data fixed format , you can ride of the PL/SQL function is_a_date function and do it everything with SQL... but the code gets a little uglier with something like this:
update mycolumn
set date_run = to_date(date_tmp,'dd/mm/yy')
where substr(date_tmp,1,2) between '1' and '31'
and substr(date_tmp,4,2) between '1' and '12'
and substr(date_tmp,7,2) between '00' and '99';
If you need more speed in your query or you have a huge amount of data in date_tmp, as function is_a_date is deterministic (always returns the same value given the same values for X, Y,), you can create an index for it:
create index mycol_idx on mycolumn(is_a_date(date_tmp));
And when you use the function, Oracle will use your index, like in those selects:
SELECT a.id, a.date_tmp
FROM mycolumn a
WHERE a.id < 1300
and is_a_date(a.date_tmp) is not null;
SELECT a.id, a.date_tmp
FROM mycolumn a
WHERE a.id < 1300
and (is_a_date(a.date_tmp) is not null and is_a_date(a.date_tmp)>sysdate-5);
I am executing the following query
DELETE FROM List,Tree WHERE List.CatID = Tree.CatID AND List.ID = '1' AND Tree.Cat = '332'
but I run into following error
near ",": syntax error
Correct syntax for delete statement is this
DELETE FROM table_name WHERE somecolumn=somevalue
So you cannot use 2 tables in a single delete query by separating them with a comma..
You need to do something like
DELETE something FROM table_name INNER JOIN...
You can delete only from one table at a time, and you have to rewrite the join as a subquery:
DELETE FROM List
WHERE ID = '1'
AND CatID IN (SELECT CatID
FROM Tree
WHERE Cat = '332')
I am having trouble getting a block of pl/sql code to work. In the top of my procedure I get some data from my oracle apex application on what checkboxes are checked. Because the report that contains the checkboxes is generated dynamically I have to loop through the
APEX_APPLICATION.G_F01
list and generate a comma separated string which looks like this
v_list VARCHAR2(255) := (1,3,5,9,10);
I want to then query on that list later and place the v_list on an IN clause like so
SELECT * FROM users
WHERE user_id IN (v_list);
This of course throws an error. My question is what can I convert the v_list to in order to be able to insert it into a IN clause in a query within a pl/sql procedure?
If users is small and user_id doesn't contain commas, you could use:
SELECT * FROM users WHERE ',' || v_list || ',' LIKE '%,'||user_id||',%'
This query is not optimal though because it can't use indexes on user_id.
I advise you to use a pipelined function that returns a table of NUMBER that you can query directly. For example:
CREATE TYPE tab_number IS TABLE OF NUMBER;
/
CREATE OR REPLACE FUNCTION string_to_table_num(p VARCHAR2)
RETURN tab_number
PIPELINED IS
BEGIN
FOR cc IN (SELECT rtrim(regexp_substr(str, '[^,]*,', 1, level), ',') res
FROM (SELECT p || ',' str FROM dual)
CONNECT BY level <= length(str)
- length(replace(str, ',', ''))) LOOP
PIPE ROW(cc.res);
END LOOP;
END;
/
You would then be able to build queries such as:
SELECT *
FROM users
WHERE user_id IN (SELECT *
FROM TABLE(string_to_table_num('1,2,3,4,5'));
You can use XMLTABLE as follows
SELECT * FROM users
WHERE user_id IN (SELECT to_number(column_value) FROM XMLTABLE(v_list));
I have tried to find a solution for that too but never succeeded. You can build the query as a string and then run EXECUTE IMMEDIATE, see http://docs.oracle.com/cd/B19306_01/appdev.102/b14261/dynamic.htm#i14500.
That said, it just occurred to me that the argument of an IN clause can be a sub-select:
SELECT * FROM users
WHERE user_id IN (SELECT something FROM somewhere)
so, is it possible to expose the checkbox values as a stored function? Then you might be able to do something like
SELECT * FROM users
WHERE user_id IN (SELECT my_package.checkbox_func FROM dual)
Personally, i like this approach:
with t as (select 'a,b,c,d,e' str from dual)
--
select val
from t, xmltable('/root/e/text()'
passing xmltype('<root><e>' || replace(t.str,',','</e><e>')|| '</e></root>')
columns val varchar2(10) path '/'
)
Which can be found among other examples in Thread: Split Comma Delimited String Oracle
If you feel like swamping in even more options, visit the OTN plsql forums.