Opened 14 years ago
Closed 14 years ago
#804 closed defect (invalid)
ST_AsGML: srsName not compliant ?
Reported by: | strk | Owned by: | coliver |
---|---|---|---|
Priority: | medium | Milestone: | PostGIS 2.0.0 |
Component: | postgis | Version: | master |
Keywords: | Cc: |
Description
select st_asgml(3, 'SRID=3819;LINESTRING(0 0, 1 0)'::geometry, 0, 1) <gml:Curve srsName="urn:ogc:def:crs:EPSG::3819"> <gml:segments> <gml:LineStringSegment> <gml:posList srsDimension="2">0 0 1 0</gml:posList> </gml:LineStringSegment> </gml:segments> </gml:Curve>
Note the srsName: "urn:ogc:def:crs:EPSG::3819"
Isn't there a column too much ?
Change History (5)
comment:1 by , 14 years ago
comment:2 by , 14 years ago
You are right.
in srsName="urn:ogc:def:crs:EPSG::3819" there is a colon too much.
comment:3 by , 14 years ago
Funny enough the C code contains comment that also get it wrong:
/* * Retrieve an SRS from a given SRID * Require valid spatial_ref_sys table entry * * Could return SRS as short one (i.e EPSG:4326) * or as long one: (i.e urn:ogc:def:crs:EPSG::4326) */
Now I wonder how's right and who's wrong...
comment:4 by , 14 years ago
The double colon seems to be correct according to this page:
http://www.oostethys.org/best-practices/best-practices-ogc-urns
It would mean a blank (empty) version part
comment:5 by , 14 years ago
Resolution: | → invalid |
---|---|
Status: | new → closed |
Alright, I take it as not-a-bug. Added a regression test to avoid breaking it in r6697
Note:
See TracTickets
for help on using tickets.
I meant "colon", not "column", of course.