Issues with autoincrement

[timh]

An element in a dataset I am working with is an integer datatype, with autoincrement set to "True". The idea is that this field is a unique ID for each entry in the database, which is saved as an XML file. This is not an SQL database.

If I open my application with the relevant XML loaded into the dataset, the autonumber field picks up as expected, however if I switch XML sources (load a different set of data based on the same schema) whilst the application is running, the ID field for the new entry starts from the last number of the previous dataset. I am unable to reset the autoincrement seed value manually (by resetting the autoincrement seed value etc), though I am clearing the dataset ( ds.clear() ) prior to loading in the new XML data.

Short of ditching the autoincrement feature altogether and manually allocating a unique ID to each record, is there any other way of achieving the desired result, i.e an autoincrement value that picks up from the last record? All the advice I have found seems to relate to using autoincrement with an SQL based system, which I can't seem to relate to my application.

A workaround would be to automatically restart the application when I load a new XML database, but that seems like the equivalent of using a hammer to do precision work...

Thanks,

Tim

 

[JohnH]

DataColumn.AutoIncrementSeed Property (System.Data)

Generally without setting a seed, the column will automatically continue counting from highest existing value in table - if source is not ordered that means it will count from highest value plus number of subsequent rows. You can't set any lower seed than that.

 

[timh]

DataColumn.AutoIncrementSeed Property (System.Data)

Generally without setting a seed, the column will automatically continue counting from highest existing value in table - if source is not ordered that means it will count from highest value plus number of subsequent rows. You can't set any lower seed than that.

Thanks. I understand that, but the issue is this...

Say I have a dataset loaded with 19 records. When I add a new record, the new record is given the ID number 20, which is fine.

If however I then load a new dataset, without first closing the application, with say 3 records, when I add a new record to this dataset, the new record is given the ID number 24.

When I load the new dataset, I first empty the old data by

 ds.clear()

What clearing the dataset doesn't seem to do is reset the autoincrement value. I mistakenly thought that if I redefined the seed value, it would work out.

 ds.Tables("entrant").Columns("ID").AutoIncrementSeed=1

I set a break point at these two lines, and found that there is a property "AutoIncrementCurrent", that holds the next number in sequence. However, this doesn't appear to be accessible to me? I don't really understand why clearing the dataset doesn't reset this value to zero, though I'm sure there's a very good reason for it.

 

[JohnH]

Clear method only removes all data rows from table, nothing else. If you can clear all rows in all tables in the whole dataset, why can't you create a new DataSet object? DataSet.Reset method and reloading schema will also do.

 

[timh]

Got there in the end thanks, although having created a new object, I also needed to recreate all the bindingsources associated with it...

Is there are a quicker way than going through each bindingsource and setting the datasource and datamember again? These were initially set in the design window.

 

[JohnH]

Changing the datasource object will have that side effect, although DataMember is just a string in this case and don't have to be set again.

If you don't create a new object and instead do DataSet.Reset, I found the opposite to be happening, you don't set BindingSource.DataSource now since it's the same object, but table objects has changed and resetting+setting DataMember was necessary for the grid to rebind to this.

 

[timh]

That saves a few lines of code! Thanks very much for your help.