Skip to content

Latest commit

 

History

History
executable file
·
41 lines (26 loc) · 1.97 KB

TODO.md

File metadata and controls

executable file
·
41 lines (26 loc) · 1.97 KB

#add JSON to db code.

#abstract some fixed constants (shapefile_tiles, .tif)

#question of whether source rasters can also be added to the db. (probably should be able to) #n.b. polys are not an issue here. source is already poly. #have worked around this for now by providing calc_passthrough.bf

possibly write a script to do the transform/index on source geometry automatically.

use comment on and see notes on extracting descriptions.

possibly have 'pre-rbuild' and 'post-rbuild' scripts optionally - e.g. to allow rule values to be transformed.

rem to drop table in qgis etc before adding with shp2pgsql. add error output?

check dependencies more carefully. at each stage, check the stage has what it needs? deps = other progs. and versions. check all vars needed are present.

if ever using this via the web - beware of sql injection attacks on the rbuild stages that extract data via sql (rasterize, import to postgis)

Should really do something tidier with program output in non-verbose mode. It's hard because set -e will abort as soon as there is a failure.

So how do you print out the last piece of output in that situation?

add demos of 'passthrough' and other program aspects.

possibly:

write a wrapper (rbuild, rbuild_main) to catch failure of rbuild and report on it (e.g. cat $WORK_DIR/output?)

call build_functions as independent programs?

add tiling code to the polygonize function for sources? (really, should this even be a feature?)

gdal_trace_outline is an extra dependency ; maybe use gdal_polygonize or provide it as an option

also; I think trace_outline produces the column 'value' whereas polygonize produces 'dn'

in the polygonize function, require that all-map raster has been built if

working without tiles.

OR

write code to always work with tiles, then merge tiled shps optionally.

check settings in the poly building action files

possibly offer a gdal_polygonize alternative to dan's scripts, even though gdal_polygonize is slower.