Regina Obe [Mon, 30 May 2011 01:51:37 +0000 (01:51 +0000)]
Move Bborie's new cool stat functions into their own section. Add documentation for st_valuecount -- more coming for other raster functions. macro replace hasnodata with exclude_nodata_value which hopefully will be clearer to everyone. The raster codebase needs to be changed to agree with new names (mostly for those utilizing named arguments in PostgreSQL 9.0+). In enhnced section --put a note that hasnodata was renamed to exclude_nodata_value.
Regina Obe [Sat, 28 May 2011 15:07:19 +0000 (15:07 +0000)]
fix for #996 also had to revise least and greatest helper functions to deal with to and froms that can't be converted to integers. Also added missing indexes to base tables (helps the planner with inherited tables though it never really uses those indexes)
Regina Obe [Sat, 28 May 2011 11:18:24 +0000 (11:18 +0000)]
fix for #995 handling lower level fips. Also many fixes for #908 sh tiger loader profile. Testing now and will close out if confirmed its fixed. Also put in a note in the .sh and .bat that for PostgreSQL 9.1+ have to install fuzzystrmatch with the CREATE EXTENSION syntax
Bborie Park [Fri, 27 May 2011 23:35:07 +0000 (23:35 +0000)]
- remove OUT parameters on those functions that only return a single value.
- updated regression tests that made use of those functions returning single values with OUT parameters
Sandro Santilli [Fri, 27 May 2011 08:34:23 +0000 (08:34 +0000)]
Add an optional force_new parameter to topology.ST_AddFace to force creation of a new face when one already exists. Regress test and document. [RT-SIGTA]
Bborie Park [Fri, 27 May 2011 00:02:59 +0000 (00:02 +0000)]
- Added default values for function parameters of ST_ValueCount and ST_ValuePercent. Removed duplicative functions of the same.
- Tweaked the regression tests for ST_ValueCount and ST_ValuePercent as one question for each function type was too ambiguous.
Bborie Park [Thu, 26 May 2011 20:04:12 +0000 (20:04 +0000)]
Addition of ST_ValuePercent, sibling of ST_ValueCount. This function provides the ability to determine the percentage of a raster's band that is of a user-specified value.
Assocated ticket is #990
Also specified STRICT for functions with "searchvalue" (the singular) argument.
Bborie Park [Thu, 26 May 2011 15:43:26 +0000 (15:43 +0000)]
Changed the datatype for "count" from integer to bigint. This should allow the "count" return value to handle large numbers coming from the various coverage functions.
Sandro Santilli [Thu, 26 May 2011 07:26:56 +0000 (07:26 +0000)]
Error handling refinements in topology.ST_AddEdgeNewFaces (#988)
Check for given edge having two distinct vertices before getting a new
edge id from sequence, and before testing it for crossing nodes (or it'd
give a confusing error message). Refine error message about no-segment
edges to not talk about nodes. [RT-SIGTA]
Regina Obe [Tue, 24 May 2011 18:50:35 +0000 (18:50 +0000)]
experiment with methodsynopsis evidentally seems to support default parameters and conditional args, we need to move from funcsynopsis (designed for C) to methodsynopsis. This is what PHP uses -- https://doc.php.net/php/dochowto/chapter-skeletons.php
Bborie Park [Tue, 24 May 2011 14:27:08 +0000 (14:27 +0000)]
- Added test testgdalraster.in which properly tests the output of ST_AsGDALRaster. This test makes use of rtgdalraster.py found in raster/scripts/python, which itself requires psycopg2.
- Refactored rt_asgdalraster.sql, rt_astiff.sql, rt_asjpeg.sql and rt_aspng.sql to no longer attempt to validate the output of the respective functions with an MD5 checksum but rather be a test of the functions' behaviors.
Sandro Santilli [Tue, 24 May 2011 10:45:40 +0000 (10:45 +0000)]
Add another (successful) test for ST_AddEdgeNewFaces when creating a new face in the universal face by closing a ring around an existing non-isolated edge [RT-SIGTA].
Sandro Santilli [Tue, 24 May 2011 08:45:55 +0000 (08:45 +0000)]
ST_ChangeEdgeGeom: check existence of given edge, reduce scans used to check start/end point and simplify the code, regress test the fix. Fixes bug #979.
Sandro Santilli [Tue, 24 May 2011 06:55:29 +0000 (06:55 +0000)]
Do not let ST_AddIsoNode add non-isolated nodes. Allow specifying 0 as the isolated node face and check it is correct. If containing_face is given as null then compute it. Patch by Andrea Peri. [RT-SIGTA]
Regina Obe [Tue, 24 May 2011 03:03:41 +0000 (03:03 +0000)]
Document ST_ChangeEdgeGeom, revise function to confirm to new plpgsql coding standard. Get rid of redundant && check since its already encapsulated in ST_Intersects and ST_Within
Sandro Santilli [Mon, 23 May 2011 13:13:21 +0000 (13:13 +0000)]
Fix test for ST_AddIsoEdge verifying "geometry intersects an edge" exception. Fix exception message raised when attempting to insert a closed edge. Update regress test accoringly.
Sandro Santilli [Mon, 23 May 2011 12:49:51 +0000 (12:49 +0000)]
ST_AddIsoEdge: set containing_face of nodes connected by the newly added edge to null as the nodes are not isolated anymore. Regress test it. Closes ticket #976.
Regina Obe [Thu, 19 May 2011 04:53:22 +0000 (04:53 +0000)]
support for 2D points, add multipoint 2d / 3d to cunit. Add a mapping table to documentation to clarify how we map PostGIS geometries to X3D equivalent.
Bborie Park [Wed, 18 May 2011 23:33:05 +0000 (23:33 +0000)]
- taking a suggestion from strk in #958, all regression tests for functions related to ST_SummaryStats has been rounded to 3 decimal places
- added additional argument checks for pgraster to rt_pg.c
- changed floating point comparisons in rt_pg.c and rt_api.c
Bborie Park [Wed, 18 May 2011 14:18:15 +0000 (14:18 +0000)]
ST_SummaryStats returns the sum as part of the summary stats.
_ST_SummaryStats(rastertable, rastercolumn) function changed to make use of the sum and compute a straight mean (sum / count) rather than computing a weighted mean.
Sandro Santilli [Wed, 18 May 2011 13:59:03 +0000 (13:59 +0000)]
topology.AddFace(): properly update left/right face for edges _contained_ in the newly registered face and containing_face for isolated nodes in it, refine documentation about it [RT-SIGTA]
Regina Obe [Wed, 18 May 2011 13:32:19 +0000 (13:32 +0000)]
damn mathml you mock me! anyrate put in defaults as Pierre asked for ST_ASTiff (hope www... fixes their damn mathml). We really got to stop downloading that file.
projects / postgis / log