[GRASSLIST:4606] Re: Use of -i option in m.in.e00

[Michel Wurtz]

Bruce Saunders wrote:
> Hello,
> 
> I have imported a number of Arc/Info vector E00 coverages into GRASS 5.0, 
> mostly without problems.  However, I am not certain if I should or should 
> not use the -i option to change the indexing term assigned to polygon 
> and line objects. 
> 
> The E00 format documentation (v7) suggests some feature attributes are 
> linked by "feature id" and others by "feature #" in the E00 file.  I 
> could not unravel the ball of string which links various EOO data 
> together, and I could not find much detail in the m.in.e00 manual page
> or previous posts to this list.

In fact each object has a feature-ID and a feature-# ident. Both are intergers.
The feature-ID is choosed when digitizing (two objects may have the same
feature-ID). The feature-# is an internal creation of Arc/Info and is an
unique identifier for each object.  It may change after some operations, so
it is unsafe to use it for linkink *external* object to an Arc/Info coverage.
Internally (i.e. in .e00 files), everything is coherent, so use of feature-#
is safe (and recommended).

On most e00 files, each object has a different ID number, so you can use
ID or # with quite the same result. The only diff will be the contents of
dig_att and dig_cat : the grass attribute number is normaly the feature-#,
but will be the feature-ID when the -i flag is given.  Some peoples prefers
that.

I personnaly think that the -i flag may be suppressed (it is mostly here
for historical reasons when my knowledge of e00 format was pretty poor)

> Can anyone briefly explain how GRASS object indexing works, how m.in.e00 
> converts Arc/Info indexing to GRASS indexing, and when the -i option 
> matters?
> 
> Thanks for any pointers,
> 
>  - Bruce Saunders <b.saunders at mindspring.com>
> 
>