I have been importing data from a CSV file into elog which has worked.

However I'd like to import data from the CSV file, directly to the 'text' 
field, rather than just to one an attribute fields. Is it possible to do 
this?

Paul

> However I'd like to import data from the CSV file, directly to the 'text' 
> field, rather than just to one an attribute fields. Is it possible to do 
> this?

Not yet, but I can implement it. Do you always have a column "text" in your
CSV data or would you like to be able to select which column goes into the
"text" field? Can you apend some sample CSV data so that I can have a look at it?

> > However I'd like to import data from the CSV file, directly to the 'text' 
> > field, rather than just to one an attribute fields. Is it possible to do 
> > this?
> 
> Not yet, but I can implement it. Do you always have a column "text" in your
> CSV data or would you like to be able to select which column goes into the
> "text" field? Can you apend some sample CSV data so that I can have a look at it?

I've been trying an alternative approach which is basically to :

1. Import the data (not including the text field) into elog from a CSV file
2. Stop the elog server
3. Use a perl script to add each text entry to the appropriate record within the 
logbook file. 
(I.e. between the =============================== of the appropriate entry
and the next  '

I implemented the feature to import the main entry text body from a CSV column labeled
"text". Please see the newest release
(http://midas.psi.ch/cgi-bin/cvsweb/elog/src/elogd.c?rev=1.319) and let me know if this
works for you.

I had to add code to break apart long text bodies after 78 characters in order not to
get too long lines. Please note that the maximum body size in the CSV import is limited
to 500 characters.

I'm not sure if this is a configuration problem or a bug, but running
v2.5.4subver1.413, elogd runs as user ROOT (UID0) even though the following
is in the elogd.cfd file:

Usr = nobody
Grp = essadm

All other directives added to the cfg file work, so I know eLog is reading
the cfg file.  

The elogd binary *is not* setuid 0.

Thanks

> I'm not sure if this is a configuration problem or a bug, but running
> v2.5.4subver1.413, elogd runs as user ROOT (UID0) even though the following
> is in the elogd.cfd file:
> 
> Usr = nobody
> Grp = essadm
> 
> All other directives added to the cfg file work, so I know eLog is reading
> the cfg file.  
> 
> The elogd binary *is not* setuid 0.

I could not reproduce your problem with the current version (Revision 1.460)
under Linux. I guess you made sure that user "nobody" and group "essadm" exist.
Try to run elogd interactively, if you see any error message (without "-D"
flag). In the most recent version (1.460), I added some more debugging code
which tells you if elogd successfully fell back to another user, if you use the
"-v" (verbose) flag.

If all that does not help, I guess it's some peculiarity of Solaris. Maybe
someone else using Solaris has some idea. All elogd does is a call to 

setuser("<user>");

I see no reason why this should not work on Solaris.

> > I'm not sure if this is a configuration problem or a bug, but running
> > v2.5.4subver1.413, elogd runs as user ROOT (UID0) even though the following
> > is in the elogd.cfd file:
> > 
> > Usr = nobody
> > Grp = essadm
> > 
> > All other directives added to the cfg file work, so I know eLog is reading
> > the cfg file.  
> > 
> > The elogd binary *is not* setuid 0.
> 
> I could not reproduce your problem with the current version (Revision 1.460)
> under Linux. I guess you made sure that user "nobody" and group "essadm" exist.
> Try to run elogd interactively, if you see any error message (without "-D"
> flag). In the most recent version (1.460), I added some more debugging code
> which tells you if elogd successfully fell back to another user, if you use the
> "-v" (verbose) flag.
> 
> If all that does not help, I guess it's some peculiarity of Solaris. Maybe
> someone else using Solaris has some idea. All elogd does is a call to 
> 
> setuser("<user>");
> 
> I see no reason why this should not work on Solaris.

Ok, just checking.  I will fiddle around with running it interactively and see what
I get, plus I'll have a look at the setuser function under Solaris.

Just for grins, what version of compiler are you using under Linux?

Thanks again!

> Just for grins, what version of compiler are you using under Linux?

gcc 3.2.2 (from RedHat Linux 9.0)