[Gdal-dev] census 2000 tiger 108th org2ogr errors

[Heitzso]

Because the large majority of the data records
were read in without an error message being thrown
I assumed the code was picking up the format okay.

I haven't actually started to work with the postgis
tables generated so cannot say for certain that
the data was loaded into them correctly.

Let me know if there's anything I can do to help.


Heitzso


Frank Warmerdam wrote:
> Heitzso wrote:
> 
>> The files came from census 2000 108th congressional district tiger files
>> for Georgia and are found at
>>     ftp://ftp2.census.gov/geo/tiger/tgrcd108/GA
>> as tgr13___.zip files.  So, from my short table above, the
>> county '-001' files came from tgr13001.zip.
> 
> 
> Heitzso,
> 
> The files in question have a version code of 1002 (October 2002) on the
> basis of which I classify them as being in Tiger 2002 format.  
> Unfortunately,
> as you suggest, they are in fact in Tiger 2000 format and so things fail
> quite dramatically for some kinds of features, such as the FeatureIds,
> where the format differs in substantial ways.
> 
> I am not exactly sure what to do about this.  There are definately 2002 
> files
> produced in this version range so I am wondering if it is practical to
> determine the version only on the basis of the version code.  It may end up
> that I need to inspect the RT5 record length or other similar "hacks" to
> determine the correct version code.  Yuck.
> 
> I have created a bug report, please feel free to add yourself to the cc:
> list.
> 
>   http://bugzilla.remotesensing.org/show_bug.cgi?id=400
> 
> I will discuss this with Mark Phillips and see if he has any advice.  He 
> did
> much of the recent work on OGR's TIGER/Line translator.
> 
> Best regards,
>