[mapserver-dev] RFC-93: UTFGrid output

[Even Rouault]

Le lundi 26 août 2013 13:06:18, Daniel Morissette a écrit :
> Hi Devs,
> 
> RFC-93 has been updated by François to reflect his implementation of
> UTGGrid output in MapServer over the summer. Please have a look and
> share your comments so that he has time to make any required changes
> before the end of his project:
> 
>   http://mapserver.org/development/rfc/ms-rfc-93.html
> 
> A sample app and docs are available at:
> 
>   http://msgsoc.mapgears.com/
> 
> And his code is in branch utfgridgsoc of his github fork:
> 
>   https://github.com/fdesj/mapserver/tree/utfgridgsoc

Nice work.

A few points and ideas :

* http://msgsoc.mapgears.com/html/en/output/utfgrid.html mentions "UTFGrids 
can be set on multiple layers but only one UTFGrid can be rendered at a time" 
but the RFC not. Perhaps this could be added in it too.

* I'm wondering if the JSON escaping of values expanded in the "keys" (from 
UTFITEM)  and "data" (from UTFDATA) has been taken into account (in particular 
escaping of " and \ ). My quick review of the code seems to indicate no, but I 
might be wrong. 

* perhaps there could be a mode (for the lazy folks ;-) ) in which UTFDATA 
could be generated automatically from the "gml_include_items" and 
"gml_exclude_items" layer metadata ?

* No validation of UTFRESOLUTION is done (line 310). If set to values < 1, 
this could lead to divizon by zero or failed mallocs due to negative values. 
Such values shoud likely be refused.

* r->duplicates = strcmp("false", msGetOutputFormatOption(format, 
"DUPLICATES", "true")); at line 320 is a convoluted way of writing :
  
 r->duplicates = EQUAL("true",  msGetOutputFormatOption(format, "DUPLICATES", 
"true"));

or !EQUAL("false", ... )

* MSVC will not be happy with wchar_t string[img->width/renderer-
>utfresolution + 1]; at line 378


Even

-- 
Geospatial professional services
http://even.rouault.free.fr/services.html