Bug 115662

Created attachment 13271 [details]
Small test case

Don't forget to configure the ASCII datasource in kst settings to read field
names from line 0 and data in free format from line 1 to try the suggested
example.

On Friday 04 November 2005 05:23, Nicolas Brisset wrote:
> It seems the -F commandline switch does not work as "advertised" in "kst
> --help": "kst -x INDEX -y VEC1 data.txt" works
> while
> "kst -x INDEX -y VEC1 -F data.txt" does not load anything
>
> Either it is a bug in the -F commandline switch handling, or misleading
> documentation in "kst --help". In any case, it would be nice to fix it.


   The documentation in --help says:
  {"F <dataFile>", I18N_NOOP("Specify data file: used to override a kst file 
default"), "|"},

   It's used to override the data file in a kst file.  How should we change it 
to make it more clear?

OK, so this is a documentation issue... Well, I think your explanation is clear enough, so we should use it ("Specify data file: used to override the data file in a kst file"). Of course, when more than one datasource (datafile) is used in the .kst, this messes things up quite a bit: should we warn against that ? (If it is complicated to do, I'd suggest not doing anything as a user using this commandline switch probably knows what he's doing)

On Monday 07 November 2005 04:10, Nicolas Brisset wrote:
> 10:10 ------- OK, so this is a documentation issue... Well, I think your
> explanation is clear enough, so we should use it ("Specify data file: used
> to override the data file in a kst file"). Of course, when more than one
> datasource (datafile) is used in the .kst, this messes things up quite a
> bit: should we warn against that ? (If it is complicated to do, I'd suggest
> not doing anything as a user using this commandline switch probably knows
> what he's doing)


  I thought about this as well.  We don't have any consideration for files 
with > 1 data source.  Barth, any comments?  Should we just leave it as-is?

I can modify the behavior so that it behaves more forgivingly....

So that we're clear. :-)

SVN commit 487589 by netterfield:

BUG: 115662

Improve documentation for -F option.



 M  +1 -1      main.cpp  


--- trunk/extragear/graphics/kst/kst/main.cpp #487588:487589
@@ -48,7 +48,6 @@
 static QStringList startupErrors;
 
 static KCmdLineOptions options[] = {
-  {"F <dataFile>", I18N_NOOP("Specify data file: used to override a kst file default"), "|"},
   { "y <Y>",  I18N_NOOP("Field for Y axis (multiple allowed)"), 0 },
   { "z <Z>",  I18N_NOOP("Field for a Z image (multiple allowed)"), 0 },
   { "ye <equation>",  I18N_NOOP("Equation for Y axis (multiple allowed)"), 0 },
@@ -72,6 +71,7 @@
   { "print <file>",  I18N_NOOP("Print to file and exit"),"<none>"},
   { "png <file>",  I18N_NOOP("Save as a png file and exit"),"<none>"},
   { "nq",     I18N_NOOP("Bypass the quickstart dialog"), 0},
+  {"F <dataFile>", I18N_NOOP("Override a *.kst file's data file with <datafile>"), "|"},
   { "+[Files]", I18N_NOOP("Data files (if -y given) or *.kst file"), 0},
   KCmdLineLastOption
 };

I am tempted to reopen that one, as there are cases where we'd like to override more than one datafile. Something like -F1 file1.dat -F2 file2.dat would be very cool and sounds pretty much user-friendly to me. Is that OK ?
(A new option that would just parse the .kst file to list datasources in the right order would be a nice complement, but maybe that's a separate issue ?)

Can I reopen this one ? I have had many users ask for this again, and I think it would be nice to do what I suggest in the first part of comment #8. This is not critical (they currently use sed on the .kst !!) but if there's a chance...