legend scale dependency issues

[percy]

never mind! I have determined that I am doing something wrong on my end :-)

now to figure out what...

thanks again!
percy

percy wrote:
> I've stripped the code down to the bare minimum. The errant behavior 
> persists!
> 
> Geologic faults turn on in the map (and show up in the DIV resulting 
> from a map->draw()) at a scale of 500000. If I then zoom in sort of 
> slowly, drawing a zoom rectangle just slightly smaller than the screen 
> each time, they'll show up the legend at around 250000. The legend gets 
> redrawn on every zoom.
> 
> If I use our pre-defined zoom levels, the faults don't show up at 
> 250000, but they do show up at 100000. It sometimes seems like the 
> legend is a step behind.
> 
> Same behavior if I use a normal legend instead of legend template.
> 
> Just for fun (!) I tried turning on legend items (naming the classes) 
> for all of the other layers to see how they behaved. I had legend items 
> coming and going at just weird times. This was what led me to think that 
> it might be known bug in MapServer version 4.8.0-rc2
> 
> Below is the stripped down code, it get's called by a little AJAX-y 
> javascript function that stuffs the results into the innerhtml on a 
> popup tab. Hmmmmmmm, I wonder if the AJAX-y stuff is interfering 
> somehow... :-)
> 
> <?php // legend.php
> // set error reporting
> ini_set("display_errors", 1);
> ini_set("display_startup_errors", 1);
> error_reporting (E_ALL);
> 
> header("Content-type: text/html charset=utf-8");
> 
> if (isSet ($_REQUEST['extent']) &&
>     isSet ($_REQUEST['mapfile']) &&
>     isset($_REQUEST['layers'])
>   ) {
>     $extent = $_REQUEST['extent'];
>     $mymap = './map/' . $_REQUEST['mapfile'];
>     $layers=$_REQUEST['layers'];
> } else {
>     print "<p>Invalid query string: ".$_SERVER['QUERY_STRING']."</p>";
>     exit;
> }
> //first set the exyent, this should take care of the scale dependencies
> $map = ms_newMapObj($mymap);
>     if ($extent['x1'] && $extent['y1'] &&
>         $extent['x2'] && $extent['y2']) {
>         // set the extent
>         $map->setExtent($extent['x1'], $extent['y1'], $extent['x2'], 
> $extent['y2']);
>     }
> //now handle turning the layers on or off
> if (count($layers)>0) {
>   for ($i=0; $i<$map->numlayers; ++$i) {
>     $oLayer=$map->getLayer($i);
>     if (in_array($oLayer->name,$layers)) {
>       $oLayer->set("status",MS_ON);
>     } else {
>       $oLayer->set("status",MS_OFF);
>     }
>   }
> }
> $my_html=$map->processLegendTemplate(null);
> echo $my_html;
> ?>
> 
> Daniel Morissette wrote:
>> The code that you included in your last email seems to be missing the 
>> stuff that handles the zoom. i.e. a comment suggests that you create a 
>> $extent_obj for this, which is either initialized from the values in 
>> the $extent[] array, or from the $map->extents which have just been 
>> set from the contents of the $extent[] array, so in both case you get 
>> more or less the same result minus the width/height ratio adjustment 
>> that MapServer does automatically on $map->setExtent()... and there is 
>> nothing to adjust the extents to reflect the last zoom operation and 
>> pass that to $map->setExtents().
>>
>> What does the $extent[] array contain when this code starts being 
>> executed? Is it the extents before or after applying the zoom?
>>
>> Daniel
>>
>> percy wrote:
>>> Thanks Daniel, but I've been testing this pretty hard as you can see 
>>> from the code snippet below (testing parts are commented out). The 
>>> layers are getting turned on and off correctly in the section right 
>>> before the call to processlegendtemplate...
>>>
>>> $map = ms_newMapObj($mymap);
>>>     if ($extent['x1'] && $extent['y1'] &&
>>>         $extent['x2'] && $extent['y2']) {
>>>         // set the extent
>>>         $map->setExtent($extent['x1'], $extent['y1'], $extent['x2'], 
>>> $extent['y2']);
>>>
>>>         // set up extent object for use in zoom (!?!?)
>>>         $extent_obj = ms_newRectObj();
>>>         $extent_obj->setExtent($extent['x1'], $extent['y1'], 
>>> $extent['x2'], $extent['y2']);
>>>     } else {
>>>         $extent_obj = ms_newRectObj();
>>>
>>> $extent_obj->setExtent ($map->extent->minx, $map->extent->miny, 
>>> $map->extent->maxx, $map->extent->maxy);
>>>     }
>>> //now handle turning the layers on or off
>>> if (count($layers)>0) {
>>> //echo "I see more than 0 layers\n";
>>>   for ($i=0; $i<$map->numlayers; ++$i) {
>>>     $oLayer=$map->getLayer($i);
>>> //    echo "got layer number".$i;
>>>     if (in_array($oLayer->name,$layers)) {
>>>       $oLayer->set("status",MS_ON);
>>> //      echo "turning on layer  ".$oLayer->name;
>>> //      echo $oLayer->status;
>>>     } else {
>>>       $oLayer->set("status",MS_OFF);
>>> //      echo "turning OFF layer  ".$oLayer->name;
>>>     }
>>>   }
>>> }
>>> $my_html=$map->processLegendTemplate(null);
>>> //echo $_SERVER['QUERY_STRING'];
>>> echo $my_html;
>>>
>>>
>>>
>>> Daniel Morissette wrote:
>>>> percy wrote:
>>>>>
>>>>> When you zoom in on the map, the scale dependencies all work great 
>>>>> for the map, but the legend is always out of whack. I am using 
>>>>> php/mapscript to generate a processlegendtemplate() request. For 
>>>>> debugging this issue, I have also tried just a straight 
>>>>> drawlegend() request and I get similarly odd behavior. The geologic 
>>>>> faults, which are set to show up at a scale of 1:500,000 don't show 
>>>>> up in the legend until zoomed in to around 1:100,000. (though 
>>>>> sometimes if I zoom in small increments, they'll show up sooner...)
>>>>>
>>>>
>>>> The problem could be related to the order in which you make calls in 
>>>> your PHP script. The legend object uses the last calculated scale in 
>>>> its tests to decide which layer/classes to include in the legend, so 
>>>> if you call processlegendtemplate() before the extents and scale 
>>>> have been updated to reflect the last zoom operation then you could 
>>>> get this kind of behavior.
>>>>
>>>> I just did a quick check in the code and $map->setExtent() does 
>>>> update the scale value, so if you call that with the right extents 
>>>> before processing the legend then you should be fine.
>>>>
>>>> Daniel
>>>
>>
>>
> 

-- 
David Percy
Geospatial Data Manager
Geology Department
Portland State University
http://gisgeek.pdx.edu
503-725-3373