With the format for the CaveXML files, it is possible to link previously separated caves into a pseudo-system. While the
draft DTD I'm working with does not support this, it would be handy to have this feature available.
<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE CaveXML SYSTEM "ver_01b.DTD" >
<CaveXML Version="0.01" xmlns="http://www.CaveXML.org">
<System>
<Name value="Test combination of caves"/>
<Cave>
<Name value="Carlsbad Caverns"/>
<URI value="http:\\www.nps.gov\caca\caca.xml"/>
</Cave>
<Cave>
<Name value="Spider Cave"/>
<URI value="file:\\\e|cave\data\spider.xml"/>
</Cave>
<Cave>
<Name value="Lechuguilla"/>
<URI value="http:\\www.nps.gov\caca\lech.xml"/>
</Cave>
</System>
</CaveXML>
This would allow local and remote files (or either) to be combined.
However, this might cause a problem with collisions of Stations IDs. If all the files were created separately, there is no
reason to assume there isn't a "00001" in caca.xml and spider.xml and lech.xml. If Cave is the lowest unique ID required
(as opposed to System -- higher -- or Survey -- lower), the it seems requiring a Cave ID that is unique would be simpliest.
>From then, other IDs _could_ be assumed as local IDs. This would be somewhat Survex like since dot notation could be
used to dig into the survey tree.
<Survey GUID="0004">
<Name Value="Entrance area"/>
<Data>
<Station GUID="0005" Name="A1"/>
<Shot>
</Shot>
<Station GUID="0006" Name="A2"/>
</Data>
</Survey>
<Survey GUID="0007>
<Data>
<Station GUID="0004.0006" Name="A3"/>
<Shot>
</Shot>
<Station GUID="0008" Name="A4"/>
</Data>
</Survey>
Note that the names assigned during the survey are still separated from the system generated IDs. Instead of having to
export or equate station names, the GUID also can be used to tie surveys together. In this case, it refers to the survey ID
of "0004" with station ID of "0006" (there might have to be a bit more to show the survey is a level higher or maybe it could
be more like directory structure -- "../0004/0006" where ".." means one level higher).
Kind of a big change from using XML IDs but the collision potential seems to be a problem unless there is a repository all
CaveXML files would use to get their IDs (and log used ones).
Comments?
- Richard Knapp
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
This archive was generated by hypermail 2b30 : Mon Sep 03 2001 - 06:00:00 CEST