Hot to sort datatable by length of a column ?

[Ontrack16]

In a datatable I have 2 columns with text.

The data looks like this
"Onderkast", "Base unit"
"Onderkast met 1 deur", "Base unit with 1 door"
"Onderkast met 1 lade", "Base unit with 1 drawer"
...

Now I want to translated a sentence eg
"Onderkast met 1 deur en 2 glazen pottenwagens"

To do that, I would start by sorting the datatable by length of the first column. So it would be :
"Onderkast met 1 deur", "Base unit with 1 door"
"Onderkast met 1 lade", "Base unit with 1 drawer"
"Onderkast", "Base unit"
...

I tried with the command
tblDataTable.Select("Nederlands like 'onderkast%', "Nederlands DESC")

That however doesn't give the right result.
When I try the expression "Length(Nederlands) DESC" then there is -of course- a message that column "Length(Nederlands)" cannot be found.

 

[cjard]

http://msdn.microsoft.com/en-us/library/system.data.datacolumn.expression(VS.71).aspx says the function is called LEN(), not LENGTH()

You can make an additional column that is LEN() and then sort by that column in the bindingsource

I wouldnt do this in the client side. Add it to the SQL command.

 

[Ontrack16]

With the sql statement I was able to get the datatable sorted the right way

Unfortunately when I run a select on the datatable, it gets out of the right order again

Even when I don't add the sortorder parameter.

I guess a workaround -like you propose- is needed then.
Too bad because I need to lookup lots of strings and every extra step means extra time.

I guess I will have to tell the users : patience is a virtue

 

[cjard]

With the sql statement I was able to get the datatable sorted the right way

Unfortunately when I run a select on the datatable, it gets out of the right order again

Of course.. it is nOT the data models job to return or store rows in any order at all. Pray tell why are you using Select() at all?

Too bad because I need to lookup lots of strings and every extra step means extra time.

If you plan on using this as a lookup, please do not use Select. It works by actually creating and compiling a comparator for every time it is run and this step is very, very slow. Please use Find() and set a sensible primary key

 

[Ontrack16]

Solved it as proposed :
- I read in the datatable (dt1, 2 columns : "From" and "To")
- I created a new datatable (dt2, 3 columns : "length(From)", "From", "To")
- I looked up stuff in dt2 by using select("From Like ...", "Length DESC")

I couldn't use find because I don't have unique values.

A word can be replaced by several others in another language.
So I have a datatable with :
WordA -> XWordA
WordA -> XWordB
WordA -> XWordC
...

And it even doesn't go that slow