[Kst] Easy bug with ASCII field names

[Peter Kümmel]

Brisset, Nicolas wrote:
> Hi,
> 
>  
> 
> I know it's unfair as there are people who are just waking up and others
> have already warmed up quite a bit, but let's see who can fix this bug
> the fastest J I think it would even be within my reach, but I really
> can't afford to spend time on it right now…
> 
> So, the issue is with ASCII files and the construction of the field
> list. If you use a file that has comma-separated field names (for
> example, exported from Excel) or field names with a pre-defined length,
> the data wizard (i.e. the ASCII data source) does not get it right. It
> seems that although the code tries to use the right approach, something
> is misconfigured somewhere and field names are always separated with
> whitespace.
> 
> Attached are two extremely simple files to test the problem. The first
> has comma-separated values and the other one has 14-character wide columns.
> 

This is fixed.

>  
> 
> Then the next, maybe not so easy problem with the datawizard is that on
> the first page it restores the last filename in the lineedit (useful, we
> have to keep it), and instantiates the corresponding datasource
> immediately (painful, if you wanted to look at another file and that one
> happens to be big). But one thing at a time…

Could you test if attached patch works?

Peter
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: update.patch
Url: http://mail.kde.org/pipermail/kst/attachments/20091116/b09bcff1/attachment.bat