I have a database where is supposed to have only one register with 2 value columns that can be filled by my web application. The values to one of the columns where given to me in an excel and we had to put it into the database. The person who did that, should had used an "update if exists else insert" but he didn't. Now, we have for some data, duplicate lines, one having just the column "valor_realizado_oficial" filled (with the column adt_login filled with 'talend' and with the key columns filled too), and another with the other column filled by the application.
So:
If exists two lines, I would like to copy the value of the column "valor_realizado_oficial" from the line with adt_login like 'talend' to the other line and delete this line.
If exists just one line, do nothing.
I tried to perform the copy part with:
update indicador_val iv
set valor_realizado_oficial=carga.valor_realizado_oficial
from (
select valor_realizado_oficial, ano, municipio_fk, indicador_fk, und_federativa_fk
from indicador_val
where adt_login like 'talend' and ano=2013 ) carga
where iv.ano=2013 and iv.municipio_fk=carga.municipio_fk
and iv.indicador_fk=carga.indicador_fk and iv.ano=carga.ano
and iv.und_federativa_fk = carga.und_federativa_fk;
But 0 rows where affected. Here's an example of a pair of lines:
id; adt_login; ano; valor_estimado, valor_realizado_oficial, indicador_fk, municipio_fk, und_federativa_fk
313885; "talend";2013;;888;2;2202;
291998;"suagenda";2013;900;;2;2202;
And I would like to have just the second, with values:
291998;"suagenda";2013;900;888;2;2202;
What I did wrong? Thanks.
In the sample data, there are empty columns. For example, the last one: und_federativa_fk.
Some of these columns are used in your joining conditions in the correlated UPDATE. If these empty values translate to the SQL NULL value, you should realize that NULL=NULL as a condition is going to be NULL, which means false in this context.
That alone might explain why no row get updated.
The solution is to use IS NOT DISTINCT FROM as the comparison operator for values that may be null.
Example:
where iv.ano=2013 and iv.municipio_fk IS NOT DISTINCT FROM carga.municipio_fk
and iv.indicador_fk IS NOT DISTINCT FROM carga.indicador_fk
etc...
Related
I have a table I wish to query. It has a string variable called comment which contains an ID along with other things. (i.e. "123456;varA;varB")
rowNo
comment
1
"123456;varA;varB"
2
"987654;varA;varB"
I want to filter based on the first substring in the comment variable.
That is, I want to filter the table on rows where the first substring of comment is "123456" (which in the example would return the first row)
How do I do this?
I was thinking something along the lines of the code below, using the "string_split" function, but it doesn't work.
SELECT *,
FROM table
WHERE (SELECT value FROM STRING_SPLIT(comment,';',1)="123456")
Does anyone have any ideas?
Note, I am querying in SQL in SAS, and this is on a large dataset, so I don't want to create a new table with a new column to then query on instead. Ideally I'd want to query on the existing table directly.
You can use the SCAN() function to parse a string.
WHERE '123456'=scan(comment,1,';')
When I run the query:
select * from table
I get 300 results. 2 of which have col_name with the same value: 10348-0. It's of type text.
When I run
select * from table where col_name = '10348-0'
I expect to be returned two rows return because two rows meet that criteria.
Instead I only get one row who's ID is one less than the other row. I.e. returned row id: 4556 unreturned row id: 4557. The ID column is of type serial integer.
Why am I only getting one row return when I add a "Where" clause to the query?
Presumably there are hidden characters or look-alike characters in the name -- so what looks the same is not necessarily the same.
The most common issue are leading or trailing spaces. I would suggest that you try finding these using:
col_name like '%10348-0%'
Another common problem are em-dashes versus en-dashes. You might also try:
col_name like '10348_0'
The _ is the LIKE wildcard to match exactly one character.
Assume I have a query that returns a result set of columns A and B from table First_Table. I want to limit the result set to those columns if the value of column X in table Second_Table is 0, and I want to add column C from table First_Table if the value of column X is 1.
The problem is easily resolved using a Python for example whereby I just have a variable as an empty string if value in column X is 0 or it would be equal to the string 'First_Table.ColumnC AS [Dynamic Value],', and I just format the sql in the script accordingly.
If Else solution is not an elegant way because I have multiple columns to add dynamically depending on multiple values...
I am just looking for some ideas on directions.. I have been looking at this for a while, might be bogged up
Dynamic sql is the best way to resolve this as suggested in the comments.
So I am comparing two Oracle databases by grabbing random rows in database A, and searching for these rows in database B based off their key columns. Then I compare the rows which are returned in java.
I am using the following query to find rows in database B using the key columns from database A:
select * from mytable
Where (Key_Column_A,Key_Column_B,Key_Column_C)
in (('1','A', 'cat'),('2','B', 'dog'),('3','C', ''));
This works just fine for the first two sets of keys, but the third key('3','C', '') does not work because there is a null value in the third column. Changing the statement to ('3','C', NULL) or changing the SQL to
select * from mytable
Where (Key_Column_A,Key_Column_B,Key_Column_C)
in ((('1','A', 'cat'),('2','B', 'dog'),('3','C', ''))
OR (Key_Column_A,Key_Column_B,Key_Column_C) IS NULL);
will not work either.
Is there a way to include a null column in an IN clause? And if not, is there a way to efficiently do the same thing? (My only solution currently is to create a check to make sure there are no nullable columns in my keys which would make this process rather unefficient and somewhat messy).
You can use it this way. I think it would work.
select * from mytable
Where (NVL(Key_Column_A,''),NVL(Key_Column_B,''),NVL(Key_Column_C,''))
in (('1','A', 'cat'),('2','B', 'dog'),('3','C', ''));
I am not sure about this (Key_Column_A,Key_Column_B,Key_Column_C) IS NULL. Wouldn't this imply that all of the columns (A,B,C) are NULL ?
I have 2 sqlite databases, and I'm trying to insert data from one database to another. For example, "db-1.sqlite" has a table '1table' with 2 columns ('name', 'state'). Also, "db-2.sqlite" has a table '2table' with 2 columns ('name', 'url'). Both tables contain a list of 'name' values that are mostly common with each other but randomized, so the id of each row does not match.
I want to insert the values for the 'url' column into the db-1's table, but I want to make sure each url value goes to its corresponding 'name' value.
So far, I have done this:
> sqlite3 db-1.sqlite
sqlite> alter table 1table add column url;
sqlite> attach database 'db-2.sqlite' as db2;
Now, the part I'm not sure about:
sqlite> insert into 1table(url) select db2.2table.url from db2.2table where 1table.name==db2.2table.name
If you look at what I wrote above, you can tell what I'm trying to accomplish, but it is incorrect. If I can get any help on the matter, I'd be very grateful!!
The equality comparison operator in SQL is =, not ==.
Also, I suspect that you should be updating 1table, rather then inserting in it.
Finally, your table names start with digits, so you need to escape them.
This SQL should work better:
update `1table`
set url = (select db2.`2table`.url
from db2.`2table`
where `1table`.name = db2.`2table`.name);