Re: file insertion ?

[David Ronis <ronis ronispc chem mcgill ca>]

I played around a bit, and while it's not elegant, I was able to get
what I want by putting in the appropriate =[FILENAME]SHEETNAME!CELL into
each cell of the report spreadsheet (actually, copying one does all of
the cell arithmetic automatically).  

Nonetheless there are problems:

1.  This doesn't work unless the source sheets are opened within the
same instance of gnumeric.  If I simply open the report, I see the
[FILENAME]SHEETNAME!CELL references literally.

2.  Some of the source data is blank, which is converted to 0 when being
transferred to the report.  This messes up some of the statistical
analysis I do later.  I tried the following:

  =if( [FILENAME]SHEETNAME!CELL != 0 , [FILENAME]SHEETNAME!CELL , "" )

but gnumeric kept giving me syntax errors (I tried different variations
of the condition as well).

3.  Here's the worst problem.  Basically, I circulate my report by
e-mail, usually by saving it as an excel file.  This no longer works.
When I open the result in, say openoffice, all the file references are
gone.  Is there a tool to convert all the FILENAME references to local
numbers?

David

P.S., I'm running the CVS HEAD version of gnumeric.



On Wed, 2006-09-20 at 17:41 -0400, David Ronis wrote:
> I have to routinely prepare reports based on several csv data files 
> generated by another program.  The way I currently do this is to open 
> all the csv files and the last gnumeric report and simply cut and past 
> each of the csv files into the appropriate section of my report (I also 
> recenter each of the cell entries).  Basically I want do something like 
> [CSV]'sheet1'!A1:N30->[GNUMERIC]'sheet1'!A10:N41 (unfortunately this 
> doesn't work as an = option, just individual cells seem to).  In 
> addition there are other columns in the report.
>
> Is there some way to automate this?   Alternately, is there a way to 
> define a gnumeric keystroke script (e.g., as you can do in emacs) that 
> can be rerun each time my CSV files change?
>
> Thanks in advance.  I don't subscribe to the list and would appreciate a 
> CC of any reply.
>
> David