|
Do you have a data table with Data Or Want to create and fill data table on fly ?
Sreejith Nair
[ My Articles ]
|
|
|
|
|
Hi,
I have a typed dataset with multiple tables, from this I obtain a reference to a table using a DataTable type. I use this reference to add an aditional column to the table/schema/dataset on the fly. I then merge this dataset with another of the same schema (without the aditional column) that is populated with data. I assume that the additional column should provide a value as it is an expression of other columns specified in the typed schema.
Cheers.
...
|
|
|
|
|
I figured out the solution... After I merged the populated dataset with my local dataset that has the extra column I had to set that columns expression again. I find this odd that my function added the new column and set the expression, but when the data is updated I have to respecify the expression for that column, but not re-add the column  .
Could anyone tell me why things work this way?
Cheers.
|
|
|
|
|
is there a limit on the size of the "in memory" database it can hold?
"there is no spoon"
biz stuff about me
|
|
|
|
|
It is limited by the amount of memory it can allocate through .net. I've seen datasets with 300,000 rows lockup processes on servers with 4Gigs of memory. The asp.net process only hit 1.5GB before the admins shut it down, but it wasn't doing anything other then paging at that point.
DataSets are not the most efficient structure on the planet. They have lots and lots of features and it shows. It appears they use roughly an object per cell and sometimes more if you keep track of changes to the data. They do serialize to xml well. Just don't expect them to load schemas fast. I just did a profile of a simple database application using typed datasets and 25% of the time was spent in the DataSet.LoadXml method, which was longer then the time spent actually getting data from the database. 
It should be easy to code, just don't expect it to be blazing fast.
I can imagine the sinking feeling one would have after ordering my book,
only to find a laughably ridiculous theory with demented logic once the book arrives - Mark McCutcheon
|
|
|
|
|
not sure if this the right forum but here goes...
im trying to convert a source database into an xml file ... simple enough u say ... the problems are:
1. the target schema is totally different from the source database schema so some remapping of fields to different tables has to happen ... not sure if a built in solution exists for this or if i have to define my own remapping file and processing code?
2. the output has been given as an xsd file that contains about 7 or more tables ... when i try to read that into a dataset (using readxmlschema()) i get errors ... can a dataset schema only contain one table? if so how do i break the xsd file into single table xsd files?
3. im doing this in vb.net cos thats what the client wants ... are there any technologies in vb.net that can help me with this?
thnx in advance and i hope i didnt completely obfuscate my questions
"there is no spoon"
biz stuff about me
|
|
|
|
|
1. If you can write a sql statement per output table you can put them in a data adapter and you should be able to just fill the dataset.
2. As long as it is a valid XSD schema, readXmlSchema should work fine. DataSets support multiple tables and relationships between them.
3. Vb.net, I'm sorry.  You should be able to load the .xsd file into a vs.net project and generate a typed dataset. This should validate the xsd schema as well.
I can imagine the sinking feeling one would have after ordering my book,
only to find a laughably ridiculous theory with demented logic once the book arrives - Mark McCutcheon
|
|
|
|
|
I am looking for suggestions on maybe the best practices of allowing a user to add their own data fields to the database.
|
|
|
|
|
The safest solution (without the need of SQL script to create new tables and columns) is to create a table for each datatype you need.
When a client needs additional data fields, store this information in a table such as below:
idKey (autonumber) | Field Name | idDataType
1 Pager 1 (i.e. string)
From this table you can then link to a string table:
idKey (FK) | Value | idKey2 (for your own use)
1 9931212
We did this in our intranet. It's a bit more complex than just creating new tables and columns but with this the database structure can be kept intact.
Ed
|
|
|
|
|
I'd say the best practice is don't. If they are manipulating data, let alone the schema in the database outside of your application, they are taking the stability of the application into their hands.
Typically a request like this occurs because something isn't being addressed quickly enough through the current development process. It's usually best to fix the process rather then allow users to grab hold of the heart of your application and fiddle with it.
I can imagine the sinking feeling one would have after ordering my book,
only to find a laughably ridiculous theory with demented logic once the book arrives - Mark McCutcheon
|
|
|
|
|
I am displaying a ComboBox that hopefully contains the contents of a database table (a lookup of sorts). When the form loads, the ComboBox contains the correct values, all sorted.
However, when I try to select a value programitically, I can't (well, I can't select the correct one). 
I create a dataset:
DataSet dsTitles = new DataSet("Titles");<br />
SqlCommand tableCommand = new SqlCommand();<br />
tableCommand.CommandText = "select TitleID, TitleDisplayText FROM tblTitles WHERE TitleHidden <> 1";<br />
<br />
SqlDataAdapter adapter = new SqlDataAdapter();<br />
adapter.SelectCommand = tableCommand;<br />
adapter.Fill(dsTitles, "Titles");
I then bind the dataset to the ComboBox:
comboBoxTitle.BeginUpdate();<br />
comboBoxTitle.DataSource = db.DataSetTitles.Tables["Titles"];<br />
comboBoxTitle.DisplayMember = "TitleDisplayText";<br />
comboBoxTitle.ValueMember = "TitleID";<br />
comboBoxTitle.EndUpdate();
So far, all is well. However, I can't seem to select the correct item. I've tried:
comboBoxTitle.SelectedValue = 2;<br />
Console.WriteLine("Current Title Index: {0} Text: {1} Combo Index: {2} Value: {3} Text: {4}", 2, db.LookupValueTitle(2), comboBoxTitle.SelectedIndex, comboBoxTitle.SelectedValue);<br />
TitleID of 2 matches TitleDisplayText of Mr, but nothing is visibily selected.
Current Title Index: 2 Text: Mr Combo Index: 0 Value: 2 Text: is what gets logged.
Any help would be appreciated.
--G 
|
|
|
|
|
Try to use comboBoxTitle.Text insted of comboBoxTitle.SelectedValue
Sreejith Nair
[ My Articles ]
|
|
|
|
|
Thanks. Tried that -- didn't work. I fixed it by removing the comboBox from the form, then adding it back. Don't understand, but problem solved.
|
|
|
|
|
|
Hello everybody.
Here's something is driving me crazy because I have the feeling it's a simple solution and I'm not getting there yet.
Can some one help me ?
Hee's the problem.
I have a table on SQL Server 2000 with the following records:
<br />
ID Date Flag Description<br />
B1E139CF-C5B8-4C6A-8951-197C0F2BB2AC 2005-09-20 18:00:00 0 Grupo 3<br />
C472A346-A855-4721-8364-71AAA2B0D4A2 2005-09-20 18:00:00 0 Grupo 3<br />
1ADAA758-C975-4C4D-AFFA-069CECDF7BF7 2005-09-21 00:00:00 0 Grupo 1<br />
I need to write a SELECT statement that return me one record for each description for any ID. Here's an example of the final output I need:
<br />
ID Date Flag Description<br />
B1E139CF-C5B8-4C6A-8951-197C0F2BB2AC 2005-09-20 18:00:00 0 Grupo 3<br />
1ADAA758-C975-4C4D-AFFA-069CECDF7BF7 2005-09-21 00:00:00 0 Grupo 1<br />
Is this simple ? Is here anyone that can help me ?
Thanks in advance for all replies
-- modified at 18:29 Tuesday 20th September, 2005
|
|
|
|
|
hi there,
try:
select id, date, flag, distinct description from "ur table"
I'm not sure it's result is correct or not
<< >>
|
|
|
|
|
Thank you for your reply but it doesn't work.
I got an error message saying that I'm using the distinct keywork in an invalid way.
I think that DISTINCT can only be used just after the SELECT keyword.
|
|
|
|
|
yep DISTINCT must be after SELECT keyword as the select query in the previous post
<< >>
|
|
|
|
|
Sure you are correct and I already tried it without sucess. Since distinct applies to distinct records, eliminating all records with ALL fields with the same value. Since the field ID is always diferent I get all records from the table.
I want to make a distinct select by the description field and get one record for each description ignoring what value comes on the ID field.
|
|
|
|
|
Filipe Peixinho wrote:
I want to make a distinct select by the description field and get one record for each description ignoring what value comes on the ID field.
=> play with GROUP BY clause
<< >>
|
|
|
|
|
Thanks you but I probably I don't express myself in a good way.
When I say I want to ingnore the values of the ID I don't mean that I don't want the ID field on the query result. I mean that I want all fields (including the ID field) but I want one record for each description with any ID SQL chooses to return to me.
Is this possible ?
you mention to use the GROUP BY clause. I tired several ways without sucess and also read all documentation on SQL Books online about the group by clause and can't figure out how can group by help me.
Will you explain me better please what are you trying to say on your last post ?
|
|
|
|
|
try this again
select distinct Description, ID, Date, Flag
From "table"
<< >>
|
|
|
|
|
Already tried that and it doesn't work.
It seems sql server is applying the distinct clause to all record and not the description field.
|
|
|
|
|
Do something similar to this
CREATE TABLE #TEMP (
[role] [varbinary] (2048),
[name] [varchar] (51)
)
declare @currentrole varbinary(2048)
declare @lastrole varbinary(2048)
declare @name varchar(20)
declare usr_cusror CURSOR FOR
select roles, [name] from sysusers order by roles
OPEN usr_cusror
FETCH usr_cusror INTO @currentrole, @name
WHILE (@@FETCH_STATUS=0)
BEGIN
IF ( @lastrole <> @currentrole )
BEGIN
INSERT INTO #TEMP VALUES(@currentrole,@name)
END
SET @lastrole = @currentrole
FETCH usr_cusror INTO @currentrole, @name
END
CLOSE usr_cusror
DEALLOCATE usr_cusror
SELECT * from #TEMP
DROP TABLE #TEMP
|
|
|
|
|
In SQL Server (maybe in SQL language in general), if a field of a table has an index on it, and this field is searched with LIKE, its index won't be used if the argument to LIKE starts with the jolly char "%".
I'm trying to speed up the search and i'd like to know if someone here has some ideas.
The customer would accept a search for whole word, so currently the idea would be to create a separate search table, whose record would contain the fields of the primary key of the main table plus another field which would contain a single word from the main table; this field would have an index.
The search table would be populated by a trigger which would parse the content of the field of the main table and would insert one record into the search table for every parsed word from the main table.
Do anyone has other possible solutions to try, or even just some considerations ?
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
