That was it! Thanks! On Mon, Apr 27, 2009 at 2:44 PM, Artem Pavlenko <[email protected]> wrote:
> ---------- Forwarded message ---------- > From: Artem Pavlenko <[email protected]> > Date: 2009/4/27 > Subject: Re: [Mapnik-users] bad lexical cast: > To: James McManus <[email protected]> > > > Hi James, > > I fixed extent parsing in postgis.input in svn trunk. The original > parser was expecting extent values with no spaces e.g : x0,y0,x1,y1 . > Latest version should handle spaces correctly , please, let me know if > it works for you. > > Regards, > Artem > > > 2009/4/27 James McManus <[email protected]>: > > Dane, > > > > +init=epsg:2264 works. I had not been putting it in Layer. However, I > am > > still getting the lexical problem. I'm using estimate_extent">false and > > then defining extent. My xml files is at: > > > > http://www.piedmontgeographic.com/mapdata/wake.xml > > > > Could it be a problem with my installation of mapnik? When I configure > > mapnik I get the following message: > > > > However, these optional dependencies were not found: > > - cairo > > - cairomm > > - boost system > > - pycairo > > > > Further up I have: > > > > Checking for Boost version >= 1.33... yes > > Found boost lib version... 1_34_1 > > Checking for C++ library boost_system-mt... no > > Could not find optional header or shared library for boost system > > Checking for C++ library boost_filesystem-mt... yes > > Checking for C++ library boost_regex-mt... yes > > Checking for C++ library boost_iostreams-mt... yes > > Checking for C++ library boost_program_options-mt... yes > > Checking for C++ library boost_thread-mt... yes > > > > I do not have the boost_system-mt library. Could this be the problem? > I'm > > using Fedora Core 9, and according to YUM I have all the boost libraries > > installed, but this one does not appear to be there. If this is possibly > > the cause of the problem, I could install new boost libraries in > /usr/local > > > > Thanks > > Jim > > On Fri, Apr 24, 2009 at 8:02 PM, Dane Springmeyer <[email protected]> > > wrote: > >> > >> Jim, > >> On Apr 24, 2009, at 11:53 AM, James McManus wrote: > >> > >> Dane, > >> > >> My tables do have SRID set in the geometry columns, but I have been > having > >> problems using EPSG codes with mapnik. When I use src=+init=epsg:2264 > in my > >> xml file I get a blank map, so I have been using src=+proj=lcc > +datum=NAD83 > >> instead. > >> > >> you mean 'srs' not 'src' right? > >> Also as long as you can do this in an interpreter without error you > should > >> be fine using the Proj4 integer codes: > >> >>> from mapnik import * > >> >>> Projection('+init=epsg:2264') > >> Projection('+init=epsg:2264') > >> >>> > >> otherwise use the proj literal from: > >> http://spatialreference.org/ref/epsg/2264/mapnik/ > >> If you are getting a blank map with the right projection then I'd > confirm > >> that the 'srs' is properly set for you layers as well. > >> > >> This works, but it looks like it may be the cause of the bad lexical > cast > >> problem. > >> > >> > >> Okay, hard to say. > >> > >> Looking at other emails on this list, I see that mapnik has a place to > set > >> epsg codes ( allowedepsgcodes) in ogcserver.conf. Do I have to use this > >> method, when using epsg codes? > >> > >> > >> Yes, otherwise the OGCServer will throw an exception. > >> > >> In my current configureation I'm using mapnik with openlayers > (Layer.WMS) > >> and tilecache (type=MapnikLayer), > >> > >> Okay, although with type=Mapnik TileCache is using Mapnik's python > >> bindings and I find it is easier to make TMS requests via OpenLayers. > >> > >> with eather mod_python (must restart server, after each edit) or just > cgi > >> when editing maps. > >> > >> Okay. > >> > >> Is there currently a conventional way to setup mapniks ogcserver to work > >> with tilecache? > >> > >> Just set up the ogcserver as normal (I prefer WSGI) and use type=wms in > >> TileCache. > >> > >> I currently I have epsg:2264 defined in both tilecache.cfg and in > >> Openlayers. > >> > >> TileCache ignores the 'srs' parameter when using the type=Mapnik. (i've > >> been forgetting to submit a patch for this) > >> Therefore you need to supply: > >> projection=+proj=lcc +lat_1=36.16666666666666 +lat_2=34.33333333333334 > >> +lat_0=33.75 +lon_0=-79 +x_0=609601.2192024384 +y_0=0 +ellps=GRS80 > >> +datum=NAD83 +to_meter=0.3048006096012192 +no_defs > >> > >> Possibly I could just incorporate ogcserver's wms.py into my > tilecache.py > >> or tilecache.cgi scripts? > >> > >> > >> Why? Sounds tricky :) > >> Dane > >> > >> > >> Thanks > >> Jim > >> > >> On Thu, Apr 23, 2009 at 6:59 PM, Dane Springmeyer <[email protected]> > >> wrote: > >>> > >>> Hi James, > >>> > >>> The lexical cast error is coming from a boost function that is used to > >>> convert data types being pulled from postgis. Essentially what is > happening > >>> is likely due to problems involved in calculating extents or the table > srid, > >>> where data is not able to be correctly cast to a new type required by > the > >>> PosGIS plugin. > >>> > >>> The I've seen this occur when querying an empty table and having > >>> estimate_extent=True, because its impossible to convert null extents to > the > >>> required types. We should likely insert more friendly error checking, > but > >>> until then I would assume that lexical cast errors indicate a problem > in the > >>> data or parameters used to query postgis. > >>> > >>> So, in your example below it looks like you are providing a manual > >>> extent, so the above does not directly apply. But does your table have > an > >>> SRID set in the geometry columns? Is that extent value correct/valid > for > >>> your data projection? Does your user have privileges to access all the > >>> tables including the geometry columns table? > >>> > >>> Dane > >>> > >>> On Apr 23, 2009, at 11:25 AM, James McManus wrote: > >>> > >>>> I'm using mapnik with postgis and tilecache. Everything appears to be > >>>> working, except I am getting the error message "bad lexical cast: > source > >>>> type value could not be interpreted as target" in my httpd/error_log, > even > >>>> when I have debug set to off. I would like to resolve this, so my > error_log > >>>> does not fill up. Below is part of the error_output from the > error_log: > >>>> > >>>> [Thu Apr 23 13:57:06 2009] [error] [client 71.120.222.68] > >>>> datasource=0xbf8420 type=1 > >>>> [Thu Apr 23 13:57:06 2009] [error] [client 71.120.222.68] size = 7 > >>>> [Thu Apr 23 13:57:06 2009] [error] [client 71.120.222.68] dbname=wake > >>>> [Thu Apr 23 13:57:06 2009] [error] [client 71.120.222.68] > >>>> estimate_extent=false > >>>> [Thu Apr 23 13:57:06 2009] [error] [client 71.120.222.68] > >>>> extent=1947126.12, 596061.63, 2258533.48, 891583.72 > >>>> [Thu Apr 23 13:57:06 2009] [error] [client 71.120.222.68] > host=localhost > >>>> [Thu Apr 23 13:57:06 2009] [error] [client 71.120.222.68] > >>>> table=wakepublicopenspace0902 > >>>> [Thu Apr 23 13:57:06 2009] [error] [client 71.120.222.68] type=postgis > >>>> [Thu Apr 23 13:57:06 2009] [error] [client 71.120.222.68] user=apache > >>>> [Thu Apr 23 13:57:06 2009] [error] [client 71.120.222.68] bad lexical > >>>> cast: source type value could not be interpreted as target > >>>> [Thu Apr 23 13:57:06 2009] [error] [client 71.120.222.68] borrow > >>>> 0xbbb8f0 > >>>> [Thu Apr 23 13:57:06 2009] [error] [client 71.120.222.68] unknown > >>>> type_oid=408585 > >>>> [Thu Apr 23 13:57:06 2009] [error] [client 71.120.222.68] return > >>>> 0xbbb8f0 > >>>> > >>>> It appears to not like how I'm referenceing user? > >>>> > >>>> Jim > >>>> > >>>> _______________________________________________ > >>>> Mapnik-users mailing list > >>>> [email protected] > >>>> https://lists.berlios.de/mailman/listinfo/mapnik-users > >>> > >> > >> > > > > > > _______________________________________________ > > Mapnik-users mailing list > > [email protected] > > https://lists.berlios.de/mailman/listinfo/mapnik-users > > > > > _______________________________________________ > Mapnik-users mailing list > [email protected] > https://lists.berlios.de/mailman/listinfo/mapnik-users >
_______________________________________________ Mapnik-users mailing list [email protected] https://lists.berlios.de/mailman/listinfo/mapnik-users
