[GRASS5] GRASS 5.3 real problem with s.in.ascii

[Hamish]

> I tried s.in.ascii with binaries from 27 March 2004.

Ok, that should have the new version (23 Mar) in it.

 
> I had the same result as before. Attached is the file I tested as a 
> *.tar.gz

Short version: your file is corrupt.


upon un-tar.gz'ing it I get the following error message:

$ tar xzf sites_test.txt.tar.gz --to-stdout
tar: Read 1536 bytes from sites_test.txt.tar.gz
501175 3836254 1 "siteA"
503264 3839817 2 "siteB"
502621 3842789 3 "siteC"
503184 3844798 4 "siteD"
501175 3842602 5 "siteE"
505943 3838799 6 "siteF"
tar: A lone zero block at 3


and the resulting "sites_test.txt" file is only 150 bytes long (the
.tgz is 245 bytes). Running hexdump on the resulting file doesn't show
anything odd, but I think tar already took care of getting rid of it.
The remaining 6 lines import fine..

[cc grass5 if anyone knows what this means]

Perhaps a null value (0x00) snuck into the file somehow? Could you have
a look with 'hexdump -C' or fix with a 'hexedit' program?



FWIW, this is the resulting site_lists/ file for the data that did make it:

name|barton
desc|s.in.ascii sites=barton input=sites_test.txt fs=space
501175|3836254|#1 %1 @siteA
503264|3839817|#2 %2 @siteB
502621|3842789|#3 %3 @siteC
503184|3844798|#4 %4 @siteD
501175|3842602|#5 %5 @siteE
505943|3838799|#6 %6 @siteF



?
Hamish