Provided by: libgeo-google-mapobject-perl_0.06-1_all 
NAME
Geo::Google::MapObject - Code to help with managing the server side of the Google Maps API
VERSION
Version 0.06
SYNOPSIS
use HTML::Template::Pluggable;
use HTML::Template::Plugin::Dot;
use Geo::Google::MapObject;
my $map = Geo::Google::MapObject->new(
key => 'ABQFbHAATHwok56Qe3MBtg0s7lgkHBS9HKneet7v0OIFhIwnBhTEGCHLTRRRBa_lUOCy1fDamS5PQt8qULYfYQ',
zoom => 13,
size => '512x400',
maptype => 'terrain',
markers=>
[
{
location=>'51.242844,0.011716',
color=>'green',
label=>'P',
title=>'Periapt Technologies',
href=>'http://www.periapt.co.uk'
},
{
location=>'51.243757,0.006051',
color=>'red',
label=>'C',
title=>'Crown Roast Butchers',
href=>'http://www.crownroast.co.uk/'
},
]
);
my $template = HTML::Template::Pluggable->new(file=>'map.tmpl');
$template->param(map=>$map);
return $template->output;
DESCRIPTION
This module is intended to provide a server side solution to working with the Google Maps API. In
particular an object of this class encapsulates a "map" object that provides support for the static maps
API, the javascript maps API, AJAX calls and non-javascript fallback data; but without making many
assumptions about the surrounding framework. We do assume that a template framework with support for a
"dot" notation is being used, for example HTML::Template::Pluggable. An important commitment of the
module is support for graceful and consistent fallback to a functional non-javascript web page.
The javascript and static Google map APIs do not behave in quite the same way when zoom and center are
not specified. Specifically it works quite well with the static maps
(<http://code.google.com/apis/maps/documentation/staticmaps/#ImplicitPositioning>) but not so well with
the javascript API. To compensate for this the module gives a choice between: specifying the center and
zoom levels; allowing the APIs and client side code to do whatever they think best; using a built in
algorithm to calculate a sensible zoom and center; and finally supplying ones own algorithm to calculate
a sensible zoom and center.
INTERFACE
new
Supported arguments are
autozoom
If no center and/or zoom is specified this parameter can be used to calculate suitable values by a
process of averaging the markers. If this parameter is an integer between 0 and 21 then that number
is taken as a maximum zoom level and the builtin algorithm, "calculate_zoom_and_center", is used with
that maximum zoom level. If the parameter is a CODE ref, then that function is used instead. The
CODE ref must take an ARRAY ref of marker specifications as an input, and must return a pair
consisting of the zoom level as an integer followed by a latitude-longitude string representing the
center. Finally if the argument is blessed and has a "calculate_zoom_and_center" method, that will
be used. For AUTO calculation of the center or zoom all the markers must be in the form
"decimal,decimal". The "json" method will delete this as it has no meaning once the zoom level and
center have been determined.
center
This must be a location that would be recognized by the Google maps API. If absent, and if no
autozoom has been set, the Google maps API and client side javascript code must between them work out
a center.
zoom
This represents the zoom level
(<http://code.google.com/apis/maps/documentation/staticmaps/#Zoomlevels>), which is a number between
0 and 21 inclusive. If absent (with no autozoom set) the API and client will need to set it.
size
The size (<http://code.google.com/apis/maps/documentation/staticmaps/#Imagesizes>) must either be a
string consisting of two numbers joined by the symbol "x", or a hash ref with width and height
parameters. In either case the first number is the width and the second the height. If absent the
Google API will be allowed to set the size as it sees fit.
format
The optional format (<http://code.google.com/apis/maps/documentation/staticmaps/#ImageFormats>) must
be one of 'png8', 'png', 'png32', 'gif', 'jpg', 'jpg-baseline'. If absent the Google API will be
allowed to set the format as it sees fit. Note that although the "json" method will pass this on, it
seems to have no meaning in the dynamic API.
maptype
This must be one of the following: 'roadmap', 'satellite', 'terrain', 'hybrid's described in
<http://code.google.com/apis/maps/documentation/staticmaps/#MapTypes>. The "json" method will
translate these into numerical codes roadmap => 0, satellite => 1, terrain => 2 and hybrid => 3 as
this makes it easy for the javascript code
(<http://code.google.com/apis/maps/documentation/reference.html#GMapType.Constants>) to translate
this numerical code into the client side codes. For example:
var mapping = new Array(G_NORMAL_MAP, G_SATELLITE_MAP, G_PHYSICAL_MAP, G_HYBRID_MAP);
var maptype = mapping[data.maptype];
mobile
This parameter should be either the string 'true' or 'false' and defaults to 'false'. The static API
uses it to provide more robust tiles as described in
<http://code.google.com/apis/maps/documentation/staticmaps/#Mobile>. However it seems to have no
meaning for the dynamic API.
key The mandatory API key. You can sign up for one at <http://code.google.com/apis/maps/signup.html>.
sensor
This parameter should be either the string 'true' or 'false' and defaults to 'false'. Both versions
of the Google API require it (<http://code.google.com/apis/maps/documentation/staticmaps/#Sensor>)
and this is reflected in this module by it being used by the "static_map_url" and "javascript_url"
methods.
markers
If present the markers should be an array ref of marker specifications so that it can be used in a
TMPL_LOOP. Each marker specification should be a hash ref and may contain whatever keys are required
by the javascript client side code. It must however have a "location" field which must be
interpretable by the Google API as a location. In general this means the location must be a string of
the form "decimal,decimal", where the first decimal is the latitude and the second longitude. Other
fields might include for example: id, title, href, icon. Generally these just get passed through by
the "json" method and should be used by the client side code as it sees fit. Three fields that have
a special meaning, "color", "size", "label" are those described by the static maps API
(<http://code.google.com/apis/maps/documentation/staticmaps/#MarkerStyles>). The GIcon class
(<http://code.google.com/apis/maps/documentation/reference.html#GIcon>) contains lots of ways of
specialising the look and feel of the markers in the dynamic API; but I can find no easy way of
consistently replicating the the static marker styles in the dynamic API. The only other field which
gets special treatment is "title". The "json" method will encode the title field. So for example if a
marker has a title field of 'Scloß', then if the javascript code uses this as the title
attribute when creating a GMarker, then hovering the mouse over that marker will display the German
word for castle.
hl This parameter specifies the language to be used. If absent the API will select the language. Only
the dynamic javascript API seems to use this parameter as described in
<http://code.google.com/apis/maps/documentation/#Localization>. As such this is used in the
"javascript_url" method.
calculate_zoom_and_center
This function tales a reference to an array of marker specifications and a maximum zoom level and returns
a pair consisting of a suggested zoom level and a center.
The algorithm works by iteratively building the following objects:
A set of points, starting with the first element of the array and adding the next one at each iteration.
Actually we don't actually have a variable for this, but without thinking about this set conceptually
I cannot state the algorithm. The set starts of consisting of just the first marker location.
"($ctheta, $cphi)"
These numbers represent a central point of the set of points represented as radians. The algorithm
does not guarantee that the point is the true centre but we treat it as the centre of a circle. This
starts of by taking the first element and converting its latitude and longitude to radians.
$radius
This variable is initialised to 0.
The contract
The algorithm requires that at each iteration all the points in the set are at most $radius radians
from ($ctheta, $cphi).
The iteration
We look at the next point. We look at the distance ($distance) between the new point and "($ctheta,
$cphi)". If this is less than or equal to $radius the iteration is over. Otherwise the following
happens:
($ctheta, $cphi) = mid point of ($ctheta, $cphi) and the new point
$radius = $radius + 0.5 * distance between ($ctheta, $cphi) and the new point
At the end
We convert "($ctheta, $cphi)" back to latitude and longitude and logarithmically convert the radius
to a zoom level.
I doubt that this algorithm is optimal but it seems quick, uses only one pass through the markers and can
be seen in action at <http://testmaps.periapt.co.uk>. Also if you have a better one you can specify it by
passing a suitable CODE ref or blessed scalar to the "autozoom" parameter of the constructor.
static_map_url
Returns a URL suitable for use with the static maps API based upon the arguments passed to the
constructor.
javascript_url
Returns a URL suitable for use in loading the dynamic map API.
markers
This returns the marker array ref.
json
This function uses the JSON module to return a JSON representation of the object. It removes the API key
as that should not be required by any javascript client side code. If any marker object has a title
attribute, then that attribute is encoded so it will display correctly during mouse overs.
width
This returns the width of the image or undef if none has been set.
height
This returns the height of the image or undef if none has been set.
TUTORIAL
This module only covers the server side of a two way conversation between server and a talkative client.
As such the tutorial must also consider HTML and javascript. Also to understand this tutorial we assume
at least passing familiarity with the following:
HTML::Template::Pluggable or some other templating framework using the "dot" notation.
The Google maps API: <http://code.google.com/apis/maps/documentation/reference.html>.
The static Google maps API: <http://code.google.com/apis/maps/documentation/staticmaps/>
Javascript, AJAX. Specifically we are using the yui framework (<http://developer.yahoo.com/yui/2/>)
though any other framework will do.
The markup
We probably have a template file, map.tmpl, some thing like the following:
<html>
<head>
<title>Test Map Tutorial</title>
<!-- This way we tell the javascript what URL to contact the server with. -->
<link href="/map_json" id="get_init_data"/>
<!-- This will be filled in with the correct URL to load the Google maps javascript API. -->
<script type="text/javascript" src="<TMPL_VAR NAME="maps.javascript_url()">"></script>
<!-- This will load the yui API. If you use a different javascript API this would obviously change. -->
<script type="text/javascript" src="http://yui.yahooapis.com/2.8.0r4/build/yahoo-dom-event/yahoo-dom-event.js"></script>
<script type="text/javascript" src="http://yui.yahooapis.com/2.8.0r4/build/connection/connection.js"></script>
<script type="text/javascript" src="http://yui.yahooapis.com/2.8.0r4/build/json/json-min.js"></script>
<!-- This loads our javascript code. -->
<script type="text/javascript" src="/js/maps.js"></script>
</head>
<body>
<!-- This div will be completely replaced by the Google API and is identified by #map_canvas. -->
<div id="map_canvas">
<!-- This image will be loaded by the static Google map API. It will be available whilst the
javascript based map is loading and if javascript is turned off
-->
<div><img
alt="Test map"
src="<TMPL_VAR NAME="maps.static_map_url()">"
width="<TMPL_VAR NAME="maps.width()">"
height="<TMPL_VAR NAME="maps.height()">"
/>
</div>
<!-- The following will be used if javascript is disabled.
We can use it display information that would normally be obtained by
hovering over or clicking on markers on the dynamic map display.
-->
<noscript>
<table>
<TMPL_LOOP NAME="maps.markers()">
<tr>
<td class="google_<TMPL_VAR NAME="this.color">"><TMPL_VAR NAME="this.label"></td>
<td>
<TMPL_IF NAME="this.href">
<a href="<TMPL_VAR NAME="this.href">">
</TMPL_IF>
<TMPL_VAR NAME="this.title">
<TMPL_IF NAME="this.href">
</a>
</TMPL_IF>
</td>
</tr>
</TMPL_LOOP>
</table>
</noscript>
</div> <!-- end of #map_canvas -->
</body>
</html>
The perl
The code to actually produce the web page must look something like this:
use HTML::Template::Pluggable;
use HTML::Template::Plugin::Dot;
use Geo::Google::MapObject;
my $t = HTML::Template::Pluggable->new(filename => 'map.tmpl');
my $map = Geo::Google::MapObject->new(....);
$t->param(maps=>$map);
print $t->output;
This is all that would be needed in a javascript free environment. With javascript enabled the client is
going to ask for the same data in a JSON format (at least we are only supporting JSON). The code to
service that request will look something like:
use Geo::Google::MapObject;
my $map = Geo::Google::MapObject->new(....);
print $map->json;
The javascript
As indicated in the markup above we are expecting the javascript to be in a file called maps.js. We
expect it to look something like the following:
function mapInitialize() {
/* Read the "link" element that tells us how to contact the server. See markup above. */
var get_init_data = YAHOO.util.Dom.get('get_init_data').href;
/* Build the AJAX callback object consisting of three parts. */
var callback =
{
/* Part I: Function called in the event of a successful call. */
success: function(o) {
/* Convert the JSON response into a javascript object.
This structure corresponds to the Geo::Google::MapObject in perl
as intermediated by its "json" method.
*/
var data = YAHOO.lang.JSON.parse(o.responseText);
if (GBrowserIsCompatible()) {
/* Build up the Map object which will replace #map_canvas in the above markup. */
var mapopt = {};
if (data.size) {
mapopt.size = new GSize(parseInt(data.size.width), parseInt(data.size.height));
}
var markers = data.markers;
var map = new GMap2(YAHOO.util.Dom.get("map_canvas"), mapopt);
var maptype = null;
if (data.maptype) {
maptype = o.argument[parseInt(data.maptype)];
}
var zoom = null;
if (data.zoom) {
zoom = parseInt(data.zoom);
}
var center = markers[0].location;
if (data.center) {
center = data.center;
}
map.setCenter(GLatLng.fromUrlValue(center), zoom, maptype);
map.setUIToDefault();
/* Now for each marker build and add a GMarker object */
for(var i = 0; i < markers.length; i++) {
var opt = {title: markers[i].title};
if (!markers[i].href) {
opt.clickable = false;
}
if (markers[i].icon) {
opt.icon = new GIcon(G_DEFAULT_ICON, markers[i].icon);
if (markers[i].shadow) {
opt.icon.shadow = markers[i].shadow;
}
}
var mark = new GMarker(GLatLng.fromUrlValue(markers[i].location), opt);
if (markers[i].href) {
mark.href = markers[i].href;
GEvent.addListener(mark, "click", function(){window.location=this.href;});
}
map.addOverlay(mark);
}
}
}, /* End of Part I */
/* Part II: Function called in the event of failure */
failure: function(o) {alert(o.statusText);},
/* Part III: Data that is passed to the success function.
We are using this to map from numerical maptypes that Geo::Google::MapObject has passed us to the
forms used in Google maps API.
*/
argument: [G_NORMAL_MAP, G_SATELLITE_MAP, G_PHYSICAL_MAP, G_HYBRID_MAP]
}; /* End of constructing the AJAX callback object. */
/* This calls the server which should hopefully kick off callback.success with lots of lovely data. */
var transaction = YAHOO.util.Connect.asyncRequest('GET', get_init_data, callback, null);
} /* end of mapInitialize function */
/* This will make sure that we call the above function at a good time. */
YAHOO.util.Event.onDOMReady(mapInitialize);
/* We still have to do the memory cleanup ourselves. */
YAHOO.util.Event.addListener(window, 'unload', 'GUnload');
Taking it further
There are lots of variations that can be done with this. The Geo::Google::MapObject object will pass on
all fields it does not recognize via the "json" function and these can be used by the javascript in
whatever way required. Similarly fields added to the marker specification will be passed by the "markers"
function and can be used in the template in whatever way required.
You can also make the markers draggable and have the act of dragging a marker update data on the server.
There is an example of where this has been done at <http://testmaps.periapt.co.uk>. The details of how
this is done depend intrinsically on the implementation of the data storage on the server. I would
suggest that in general the steps required to make the markers draggable might be as follows:
Clean interface to the data storage.
I think this is a clear case where an ORM (http://en.wikipedia.org/wiki/Object-relational_mapping
<http://en.wikipedia.org/wiki/Object-relational_mapping>) offers some advantages. You could look at
DBIx::Class or write your own interface using DBI.
Derive a class from Geo::Google::MapObject.
Your derived class constructor should get the data you need from the data storage interface layer,
use it to create a Geo::Google::MapObject and bless that into your derived class.
Provide URLs that update the data underlying your map objects.
These will probably be POST methods and need not return anything other than an "OK" string. They
should probably call the data storage interface layer directly to update the data. There is no point
creating a map object merely to update the underlying data.
Add link elements to your markup indicating which URL is needed to update the marker data.
<link href="/move_marker" id="move_marker"/>
An advantage of using these link elements is that you might have multiple maps to manage. By using
the link elements to communicate with the client side, a lot more code and templates can be reused.
Get the javascript side to read your new link elements.
var move_marker = YAHOO.util.Dom.get('move_marker').href;
Add a "draggable" flag to the GMarker constructor
var opt = {title: markers[i].title, draggable: true};
Make sure your GMarkers are labelled in a way that your server can understand
Just after the GMarker object, mark, has been created add
mark.label = markers[i].label;
Make your GMarker object subscribe to the "dragend" event.
GEvent.addListener(mark,
"dragend",
function(latlng) {
var location = latlng.toUrlValue(12);
var postdata = "id="+this.label+"&location="+location;
YAHOO.util.Connect.asyncRequest('POST',
move_marker,
{
success: function(o){},
failure: function(o) {alert(o.statusText);},
argument: []
},
postdata);
});
DIAGNOSTICS
"markers should be an ARRAY"
The markers parameter of the constructor must be an ARRAY ref of marker configuration data.
"no API key"
To use this module you must sign up for an API key (http://code.google.com/apis/maps/signup.html) and
supply it as the key parameter.
"zoom not a number: %s"
There must be a zoom parameter which is a number from 0 to 21.
"maptype %s not recognized"
The maptype must be one of roadmap, satellite, terrain, hybrid.
"no width"
"no height"
"width should be positive and no more than 640"
"height should be positive and no more than 640"
"cannot recognize size"
The size parameter must either be a string like "300x500" or a hash array like
{width=>300,height=>500}. And both width and height must be between 1 and 640 inclusive.
"cannot calculate autozoom without markers"
If you pass an "autozoom" parameter to the constructor you must also pass at least one marker.
"% not recognized as autozoom"
An autozoom parameter was passed to the constructor that was not recognized as such.
"location missing"
A marker specification was missing a location during autozoom calculation.
"no location for %s"
Every marker object must have a location.
CONFIGURATION AND ENVIRONMENT
Geo::Google::MapObject requires no configuration files or environment variables.
DEPENDENCIES
Templating framework
We assume the use of HTML::Template::Pluggable and HTML::Template::Plugin though other template
frameworks may work.
Google Maps API
You need to have one of these which can be obtained from
<http://code.google.com/apis/maps/signup.html>.
Javascript and AJAX
We assume a degree of familiarity with javascript, AJAX and client side programming. For the
purposes of documentation we assume YUI: <http://developer.yahoo.com/yui/>, but this choice of
framework is not mandated.
Math::Trig
The autozoom algorithm optionally used by this module depends upon the Math::Trig for its
mathematical calculations. Since alternative algorithms or no algorithm at all can be used, I think
it might be better to have this as a separate module.
INCOMPATIBILITIES
None reported.
BUGS AND LIMITATIONS
paths etc
Currently there is no support for paths, polygons or viewports.
version 3
We are currently only supporting version 2 of the API.
testing
We encode the title attributes of markers in the "json" function as this seems to be necessary.
However I have not yet managed to get a decent test script for this behaviour. There are many other
cases that should be tested though I think the most critical cases have been tested.
character encoding
This module is only tested against UTF-8 web pages. I have no intention of changing this as I cannot
think of why anyone would consciously choose to encode web pages in any other way. I am open to
persuasion however.
Please report any bugs or feature requests to "bug-geo-google-mapobject@rt.cpan.org", or through the web
interface at <http://rt.cpan.org>.
ACKNOWLEDGEMENTS
Thanks to Andreas Koenig for pointing out that this module had an issue with "Build.PL".
AUTHOR
Nicholas Bamber "<nicholas@periapt.co.uk>"
LICENCE AND COPYRIGHT
Copyright (c) 2009, Nicholas Bamber "<nicholas@periapt.co.uk>". All rights reserved.
This module is free software; you can redistribute it and/or modify it under the same terms as Perl
itself. See perlartistic.
DISCLAIMER OF WARRANTY
BECAUSE THIS SOFTWARE IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY FOR THE SOFTWARE, TO THE EXTENT
PERMITTED BY APPLICABLE LAW. EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER
PARTIES PROVIDE THE SOFTWARE "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED,
INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
PURPOSE. THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE SOFTWARE IS WITH YOU. SHOULD THE
SOFTWARE PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, REPAIR, OR CORRECTION.
IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING WILL ANY COPYRIGHT HOLDER, OR ANY
OTHER PARTY WHO MAY MODIFY AND/OR REDISTRIBUTE THE SOFTWARE AS PERMITTED BY THE ABOVE LICENCE, BE LIABLE
TO YOU FOR DAMAGES, INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL DAMAGES ARISING OUT OF
THE USE OR INABILITY TO USE THE SOFTWARE (INCLUDING BUT NOT LIMITED TO LOSS OF DATA OR DATA BEING
RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD PARTIES OR A FAILURE OF THE SOFTWARE TO OPERATE
WITH ANY OTHER SOFTWARE), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF SUCH
DAMAGES.