gdalindex question

[Frank Warmerdam]

Ed McNierney wrote:
> Frank -
>
> It's entirely possible that I haven't been looking, but I'm not aware of a clear specification on how relative paths are SUPPOSED to work.  I am under the impression that relative paths are somewhat vulerable to how the Web server is configured and how the MapServer CGI is referenced relative to that Web server.
>
> In the absence of a SHAPEPATH and/or IMAGEPATH statement, it's been my impression that the behavior of relative paths is not specified.  There have been recent changes to MapServer that have changed this behavior (at least in some cases I've observed) but I can't say that I thought either the earlier or the newer behavior was "wrong".
>
> If we intend to have relative paths behave in some consistent, predictable way (e.g. always relative to the directory containing the mapfile) we should be sure that's well-specified before trying to fix the code.  I apologize if that's already been done and I've overlooked it, but I'm not aware of it.

Ed,

In the code substantial change were made (for 4.0 I think, or possibly
4.2) to use msBuildPath() to build paths to files.  It basically "tries"
to find the file relative to the shapepath, or failing that relative to the
map or failing that taking it as an absolute path.  And there is an additional
version for taking names in tile indexes relative to the tile index file
I think.

So, I think, at the least real files should work if accessed relative to the
SHAPEPATH location.  That should include tileindexes.  If at least that much
isn't working then I would like to know.

Best regards,

--
---------------------------------------+--------------------------------------
I set the clouds in motion - turn up   | Frank Warmerdam, warmerdam at pobox.com
light and sound - activate the windows | http://pobox.com/~warmerdam
and watch the world go round - Rush    | Geospatial Programmer for Rent