[GRASS-dev] [GRASS GIS] #2986: r.mapcalc with same variable on LHS and RHS

[GRASS GIS]

#2986: r.mapcalc with same variable on LHS and RHS
--------------------------+-------------------------
  Reporter:  mankoff      |      Owner:  grass-dev@…
      Type:  defect       |     Status:  closed
  Priority:  major        |  Milestone:  7.0.4
 Component:  Docs         |    Version:  7.0.3
Resolution:  fixed        |   Keywords:  r.mapcalc
       CPU:  Unspecified  |   Platform:  Unspecified
--------------------------+-------------------------

Comment (by glynn):

 Replying to [comment:5 mankoff]:
 > Just to be clear, {{{LHS = RHS}}} **does** work. I've never actually had
 it crash or return NULL or what appears to be invalid data when I've done
 this. But perhaps it isn't guaranteed to work...
 It works by coincidence rather than by design.

 The data (cell/fcell) file and null bitmap are written to temporary files,
 which are renamed into place when the map is closed. The metadata files
 (cellhd, quant, histogram, categoriess, colour table, history, etc), are
 stored in memory and written when the map is closed. So the output map
 shouldn't change while it's being read.

 A specific case which won't quite work is "r.mapcalc 'map = map'", because
 that copies the categories, colour table and history from the input to the
 output, but closing the output map will delete or replace those before the
 copy occurs. If the logic is ever extended to cover more cases, those will
 similarly fail.

 Actually, I'm not sure that it will work on Windows, because at the point
 that the data and null files are renamed, the original files are still
 open for read (r.mapcalc never actually closes the input maps). That
 should be easy to fix, though (call close_maps() from within execute()
 before the loop which closes the output maps).

--
Ticket URL: <https://trac.osgeo.org/grass/ticket/2986#comment:6>
GRASS GIS <https://grass.osgeo.org>