[GRASS5] Re: v.in.shp on solaris2.7,5.0.0pre3

[David D Gray]

John Gillette wrote:

> David,
> 
> Here is a listing of what happens.
> 
> One other question.  I am trying to compile your arctest1
> program to be able to extract the shp files. Possibly
> an interim solution until v.in..shape works.
> But I get an undefined symbol G_fatal_error and G_warning.
> I am using Sun's "make" and I assume that the problem is I am
> not including something. Do you know where G_fatal_error and
> G_warning reside, I mean as in what library or object file
> these are in? (I know that v.in.shape compiles...)
> 


They are declared in src/include/gisdefs.h. Defined in 
src/libes/gis/error.c and (I assume) included in libgis.a. But a 
standard setup should pick up these    without any problems. Probably, 
the module should be built in the GRASS environment, as this will be 
needed to configure the various include and lib directories needed for 
any GRASS specific code.


> gcc -g -L/usr/lib -lm -o arctest1 shpopen.o dbfopen.o shp2dig.o 
> dbutils.o writelin.o gbtree.o arctest1.o
> 
> Thanks,
> 
> John
> ---------------------------------------------------------------------------
>  v.in.shape input=/scratch/grassdata/gishome/n784903L.shp verbose=9
> logfile=ZZZlogfile label=TAG
>  Extracting all records.
>  Creating vector network...
> 
>     Processing shape       1 of     328.
>     Processing shape       2 of     328.
>     [...  more of the same]
>     Processing shape     328 of     328.
>  Number of shapes registered is 0
>  Number of parts registered is 328
>  Number of valid records is 0 of 0
>  Extracting and storing area edges...
> 
>     Tracking line       1.
>     Tracking line       2.
>     [...  more of the same]
>     Tracking line       7.
>  Segmentation fault
> 

So, it's a good old segfault. Is it possible for you to send me this 
file if it's not confidential?

David