W3C home > Mailing lists > Public > public-poiwg@w3.org > January 2012

Re: POI XML validator for testing

From: Raj Singh <rsingh@opengeospatial.org>
Date: Sun, 8 Jan 2012 19:02:33 -0500
Cc: "public-poiwg@w3.org W3C" <public-poiwg@w3.org>
Message-Id: <6840B8D7-E90C-4569-9D07-1ADBDF483377@opengeospatial.org>
To: Alex Hill <ahill@gatech.edu>
1) Yes POITermTypes must have a term property according to the data model. if we make term an optional property, we weaken the likelihood that it will be used, but I see your point.
2) Yes the validator should do this. My laziness (or eagerness to get a version out) is at fault here. Should be fixed now.
3) We said we require URIs for IDs. I still think this is good policy. I do however think #StataCenter should be legal if a base URI can be identified (but the validator doesn't make any attempt to do this right now).
4) If above #3 is true, then target POI can refer to a fragment. Good treatment of the topic in the draft spec will help.

On Jan 7, at 2:09 PM, Alex Hill wrote:

> I attempted to convert the existing XML example in the core draft to the spec.
> Doing so brought up some interesting questions from the validator results:
> <point> must have a term property.
> Points must have exactly 2 coordinates. This posList has 3.
> <point> must have a term property.
> Invalid URI of 'mainpoint' in <point>'s id property.
> <polygon> must have a term property.
> Invalid URI of 'StataCenter' in <poi>'s id property.
> <label> must have a term property.
> <label> must have a term property.
> <label> must have a term property.
> <label> must have a term property.
> <category> must have a term property.
> Invalid URI of '#StataCenter' in <relationship>'s targetPOI attribute.
> <relationship> must have a term property.
> Invalid URI of 'w3c' in <poi>'s id property.
> <label> must have a term property.
> <label> must have a term property.
> 
> 1) must label, point and other POITermTypes have a term property?
> The data model PNG suggests this but I don't think it makes sense.
> 2) should the validator recognize srsName 4979 and allow coordinates with 3 points?
> 3) can we not use a simple character string as the id property?
> What exactly does "using fragments" mean here?
> Is an XML base be a requirement before a fragment can be used?
> 4) can targetPOI refer to a fragment (i.e. #StataCenter)?
> What is the relationship to XML base for this reference?
> 
> On Jan 6, 2012, at 12:20 PM, Raj Singh wrote:
> 
>> Hi group. I wrote a little XML validation tool for POIs. 
>> http://www.rajsingh.org/poiwg/poixmlvalidator.html
>> 
>> Hope this can encourage more people to experiment with creating POI files!
>> 
>> -----
>> Raj Singh
>> rsingh@opengeospatial.org
>> +1 (617) 642-9372
>> The OGC: Making location count...
>> http://www.opengeospatial.org/ogc/organization/staff/rsingh
>> 
>> 
>> 
>> 
>> 
> 
> Alex Hill Ph.D.
> www.alexshill.com
> 
Received on Monday, 9 January 2012 00:03:03 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 9 January 2012 00:03:03 GMT