I'm using SQL Server 2005 and I want to synchronize two tables which have the same definition but exist in different databases. MERGE INTO only exists in 2008 and I'd prefer a syntax where I don't have to specify columns in the UPDATE. So I stumbled upon various posts using the following syntax:
UPDATE Destination FROM (Source INTERSECT Destination)
INSERT INTO Destination FROM (Source EXCEPT Destination)
But when I try to execute it I get:
Incorrect syntax near the keyword 'FROM'.
How can I get this working? I have multiple tables which I need to synchronize and I don't want to specify all the columns in every statement.
Thanks for any hint!
According to Books Online the update command requires the set keyword, and it must come before the optional from keyword. The insert command doesn't have a stand alone from keyword, the from only exists as part of a select statement either as a derived table source or within a common table expression.
The link you reference is not showing valid SQL Server 2005 syntax.
"How can I get this working? I have multiple tables which I need to synchronize and I don't want to specify all the columns in every statement."
For update, you must specify all the columns. For insert if the source and destination have the same struture then you can use insert into TARGTET_TABLE_NAME select * from SOURCE_TABLE_NAME BUT that is not recommended for production code, if the source or destination change, the statement would break. If source and destination differ, then you must specify columns on at least one side of the insert.
I'm sorry if this doesn't answer your question, but assuming the whole reason for this is in the interest of saving time, can't you just right-click the source table and generate the INSERT script, then right-click the destination table and generate a blank SELECT script, then combine the two? This will only work if a kill-and-fill is acceptable in your environment.
Related
I have a table with a 'query' field containing a select sql and another 'parameters' field containing the sql parameters. I have merged these two fields into a new field containing a correct select sql statement. Now I need to execute this new field containing select sql, get the return from select (the output fields) and generate an excel file.
Use Table-Input if you are interested in a query result set. Table-Input supports SQL parameters, so no need to build the statement yourself using e.g. Replace-In-String, and tripping over escapes on your way. Also, there's variable substitution, just in case you can't live with a single template.
Update 21:14 GMT
I'm not very fond of the way you try to prepare the SELECT statement, but here we go, assuming it's a single statement we have:
Create a job with a Start entry and 2 Transformation entries (T1, T2). Let T1 produce the field containing your SELECT statement and use a Set-Variables step to make the statement available to T2 as variable SELECT. In T2 use a Table-Input step referencing ${SELECT} in the SQL statement text area. Don't forget to enable option "Replace variables in script".
From now on it's a matter of taste. I would prefer to create a CSV file using Text-File-Output. Using the right field separator Excel will open the file after double-clicking it. The advantage of Text-File-Output is that you don't have to specify the fields you don't know at design-time anyway. An empty field list will just handle all fields coming in. Comparable to the total projection in a Table-Input which will create the necessary fields from the retrieved columns downstream.
If you must produce an Excel workbook, you'll have to learn about metadata injection. That would be a separate project for a beginner, though. There are samples in your Kettle installation folder. And there is a very active community if you find yourself in trouble.
I am writing a database abstraction layer that also abstracts some of the different query types. One of them is called "field_exists" - its purpose should be pretty self-explanatory.
And I want to implement that for SQLite.
The problem I am having is that I need to use one query that either returns a row confirming that the field exists or none if it doesn't. Thus, I cannot use the PRAGMA approach.
So, what query can I use to check whether a field exists in SQLite, that fulfills the above criteria?
EDIT: I should add that the query needs to be able to run in PHP code (using PDO).
Also, the query should look something like this (which only works with MySQL):
SHOW COLUMNS FROM table LIKE 'field'
Trying to select a field that doesn't exist will return an exception, then you can catch it and return nothing.
Use the .schema TABLENAME command. It will tell you the command that was issued to create the table. For more info chekcout the SQLite command shell documentation.
If you don't have access to the sqlite command line, you can always query the sqlite_master table. Let's say you want to know the command used to create the table MyTable. You'd issue this:
select sql from sqlite_master where name='MyTable';
This then gives you the sql command that was used to create the table. Then just grep through that output and see if the column you're looking for is in the command used to create the table.
UPDATE 2:
Actually better than the sql I posted above, you can use this:
PRAGMA table_info(*table_name*)
This will show you all the columns in a given table along with their types and other info.
I was wondering if there was a way that I can do the folloing in one sql statement.
I am parsing an csv file for product insertion into the database. is there a way that I can check to see if an entry in a table with X equaling to N, if so update the rest f the columns else insert it?
regards
Phil
This type of operation is sometimes called an "UPSERT". The SQL standard way to do it is to use the MERGE statement, but unfortunately it is not widely implemented yet.
Some databases have added their own ways to do it, such as the non-standard MySQL extensions REPLACE and INSERT ... ON DUPLICATE KEY UPDATE.
The MERGE statement will get you there, but it's only available on SQL Server 2008 R2. You would otherwise need to treat this as a Slowly Changing Dimension and either bring the data into a temp table and compare or use an integration services package to do the work.
Would something like this work for you?
if(update myTable where x='n')
else{
insert into myTable(x,y,z) values(1,2,3)
}
The update query will run regardless, and will return false if it couldn't update that record, causing it to do the insert.
I'm having to support an Access .mdb file that someone else has written. One of the button functions in this .mdb calls out to delete some data in an external MSSQL database. All very straightforward, but this syntax isn't something I've seen before:
DELETE
tblEquipmentConnections.SourceEquip,
tblEquipmentConnections.EquipmentConnectionID
FROM tblEquipmentConnections
WHERE
tblEquipmentConnections.SourceEquip = [Forms]![frmEquipment]![EquipmentID];
Is that any different than this?
DELETE
FROM tblEquipmentConnections
WHERE
tblEquipmentConnections.SourceEquip = [Forms]![frmEquipment]![EquipmentID];
I can't find a case where specifying specific columns does anything - but I don't spend much time in Access, so I'm not sure how different the SQL syntax is...
Thanks!
Specifying the column names makes no difference. It's just an Access thing.
The reason they might be there is because Access used to generate DELETE statements that way (not sure if it still does).
The second form without columns names is obviously preferable.
I think the query has been built directly into Access query editor.
And generally we begin by building a select query. Then we change the query type from "Select query" to "Delete query". Then we display the query source by selecting "SQL Mode" where we copy / paste a sql statement like this one :
DELETE qc_Boxes.idBox, qc_Boxes.idScreen, qc_Boxes.title
FROM qc_Boxes;
This is absolutely redundant. The place between DELETE and FROM is used only when the deletion is performed based on a multi-table condition, but even in this case it contains table names and not field names. Also it can contain * which is also redundant. In MySQL, for example it's an incorrect syntax.
I've been attempting to write embedded SQL statements for DB2 that ultimately gets compiled in C.
I couldn't find a tutorial or manual on the embedded SQL syntax for C for reference. One case I would like to do is to insert data into a table. I know most embedded sql statements need the initalizer EXEC SQL, but that's the extent of my knowledge generally. I'm doing this for an assignment and would appreciate if there are more information regarding this or solution.
Example of a statement to query the database:
EXEC SQL SELECT SNAME, AGE into :sname, :sage
FROM ONE.SAILOR
WHERE sid = :sid;
I like to see what statement allows me to INSERT into the database. I've tried something like the following, but it doesn't work.
EXEC SQL INSERT ....
See IBM's Embedded SQL manual.
Embedded SQL is largely the same no matter what the host language is.
The four dots aren't syntactically valid :-D
The reliable way is the same as with any other INSERT statement: list the columns and the values.
EXEC SQL INSERT INTO SomeTable(Col1, Col2, Col3) VALUES(:hv1, :hv2, :hv3);
Here, the :hv1, :hv2 and :hv3 represent three host variables of types appropriate to the columns in the table. Note that the table could contain other columns than these three as long as those columns have a default specified or accept NULL (which is really just a default default in this case). The unreliable way does not list the columns:
EXEC SQL INSERT INTO SomeTable VALUES(:hv1, :hv2, :hv3);
Now you are dependent on getting the sequence right, and you must provide a value for each column -- there cannot be extra columns in SomeTable.
I just started using sqllite. Besides the good documentation for C++, SQLlist might be a nice thing to have because you can unit-test your code without being dependent on DB2 and it's really easy add with your code.