I've wrapped the Google Maps API some time ago.
It's a component that contains a TWebBrowser that loads a local html file that contains some javascript helper functions and sets up a basic Google maps page. The HTML file is embedded into the exe as a resource for easy deployment.
It can use some work, but it has done some nice jobs for me already. It expands on the idea that was posed here:
http://www.stevetrefethen.com/blog/UsingGoogleMapsFromVCLSampleApplication.aspx
I've put a demo executable and sourcecode online here:
http://www.xs4all.nl/~niff/GoogleMaps.zip
On this screenshot I've switched to the Google Earth view:
Here are a few examples of how you can use it:
Center the map to some shithole in the Netherlands, with nicely rounded coordinates:
GoogleMaps1.SetCenter(52,5,True);
Load a KML or KMZ file:
GeoXML := TGGeoXML.Create('http://mywebsite.com/mykml.kmz');
GeoXML.GoogleMaps := GoogleMaps1;
GoogleMaps1.AddOverlay( GeoXML );
Switch to the embedded version of Google Earth, for fast and smooth 3d action:
GoogleMaps1.MapType := MT_SATELLITE_3D;
Keep a list of overlays, and manage it via Delphi structures:
GoogleMaps1.Overlays[2].hide; // hide overlay 2
GoogleMaps1.RemoveOverlayByIndex(3); // delete overlay 3
It lets you create polygons using fast Delphi arrays and then plot that in GMaps;
MyPolygon := TGPolygon.Create(MyPointArray);
MyPolygon.Color := clBlue;
MyPolygon.Name := 'Awesome Polygon 1';
GoogleMaps1.AddPolygon( MyPolygon );
It doesn't intend to wrap the full API; it's just to make life easier when dealing with Google maps.
The unit DouglasPeuckers is not really needed. It's used to simplify a polygon when you run out of resources.
Good luck, and let me know if you've created something useful with it.
{—————————————————————————————————————————————————————————————————————————}
{ Project : uGoogleMaps.pas }
{ Comment : Google Maps API wrapper }
{ }
{ Date : 14 mrt 2008 }
{ Author : Wouter van Nifterick }
{—————————————————————————————————————————————————————————————————————————}
{ The aim of this unit is to wrap the Google Maps API in such a way that }
{ users don't need to know that map operations end up being rendered by }
{ a browser component and JavaScript. }
{ }
{ JavaScript classes have a Delphi counterpart, and each Delphi class }
{ takes care of proper JavaScript rendering and execution. }
{ }
{ For many things, like constructing detailed polygons, this provides a }
{ major performance boost and adds compile-time type checking. }
{ }
{ A big limitation so far is that I didn't find a way to directly pass }
{ complex types from and to the JavaScript engine in IE via COM, so for }
{ now, everything needs to be (de)serialized to and from strings. :( }
{—————————————————————————————————————————————————————————————————————————}
{ Last modified }
{ Date : }
{ Author : }
{—————————————————————————————————————————————————————————————————————————}
{$M+}
unit uGoogleMaps;
interface
uses
Controls,
Dialogs,
ActiveX,
StdCtrls,
ExtCtrls,
SysUtils,
Classes,
Contnrs,
Forms,
SHDocVw,
MSHTML,
StrUtils,
DouglasPeuckers
// , uGoogleEarth_intf
;
const
GoogleMapsFileName = 'GoogleMaps.html';
WGS84_MULT_FACT = 100000; // multiply lat/lon values by this value in order to fit them into integers
DEFAULT_SIMPLIFY_TOLERANCE = 0.5;
{$R GoogleMaps_html.res}
type
TGoogleMapControl = (MC_NONE=1,MC_SMALL,MC_LARGE);
TGoogleMapType = (MT_NORMAL=1,MT_SATELLITE,MT_HYBRID,MT_PHYSICAL,MT_SATELLITE_3D);
TGoogleMaps = class; // forward declaration
GIcon = class end; // to be implemented
IJsClassWrapper=interface(IInterface)
function JsClassName:String;
function GetJsVarName:String;
procedure SetJsVarName(const aVarName:String);
property JsVarName:String read GetJsVarName write SetJsVarName;
function ToJavaScript:String;
end;
IHidable=interface(IInterface)
procedure hide; // Hides the object if the overlay is both currently visible and the overlay's supportsHide() method returns true. Note that this method will trigger the respective visibilitychanged event for each child overlay that fires that event (e.g. GMarker.visibilitychanged, GGroundOverlay.visibilitychanged, etc.). If no overlays are currently visible that return supportsHide() as true, this method has no effect. (Since 2.87)
function isHidden : Boolean; // Returns true if the GGeoXml object is currently hidden, as changed by the GGeoXml.hide() method. Otherwise returns false. (Since 2.87)
procedure show; // Shows the child overlays created by the GGeoXml object, if they are currently hidden. Note that this method will trigger the respective visibilitychanged event for each child overlay that fires that event (e.g. GMarker.visibilitychanged, GGroundOverlay.visibilitychanged). (Since 2.87)
function supportsHide : Boolean; //
end;
// marker class
GMarkerOptions=record
icon : GIcon; // Chooses the Icon for this class. If not specified, G_DEFAULT_ICON is used. (Since 2.50)
dragCrossMove : Boolean; // When dragging markers normally, the marker floats up and away from the cursor. Setting this value to true keeps the marker underneath the cursor, and moves the cross downwards instead. The default value for this option is false. (Since 2.63)
title : String; // This string will appear as tooltip on the marker, i.e. it will work just as the title attribute on HTML elements. (Since 2.50)
clickable : Boolean; // Toggles whether or not the marker is clickable. Markers that are not clickable or draggable are inert, consume less resources and do not respond to any events. The default value for this option is true, i.e. if the option is not specified, the marker will be clickable. (Since 2.50)
draggable : Boolean; // Toggles whether or not the marker will be draggable by users. Markers set up to be dragged require more resources to set up than markers that are clickable. Any marker that is draggable is also clickable, bouncy and auto-pan enabled by default. The default value for this option is false. (Since 2.61)
bouncy : Boolean; // Toggles whether or not the marker should bounce up and down after it finishes dragging. The default value for this option is false. (Since 2.61)
bounceGravity : Integer; // When finishing dragging, this number is used to define the acceleration rate of the marker during the bounce down to earth. The default value for this option is 1. (Since 2.61)
autoPan : Boolean; // Auto-pan the map as you drag the marker near the edge. If the marker is draggable the default value for this option is true. (Since 2.87)
// to implement:
// zIndexProcess : Function; // This function is used for changing the z-Index order of the markers when they are overlaid on the map and is also called when their infowindow is opened. The default order is that the more southerly markers are placed higher than more northerly markers. This function is passed in the GMarker object and returns a number indicating the new z-index. (Since 2.98)
end;
TGPoint=class
end;
TGLatLng=class(TInterfacedObject,IJsClassWrapper)
private
FLat,
FLng:Double;
FJsVarName: String;
function GetJsVarName: String;
procedure SetJsVarName(const Value: String);
published
constructor Create(aLat,aLng:Double);
property Lat:Double read FLat write FLat;
property Lng:Double read FLng write FLng;
function ToJavaScript:String;
function Equals(const AGLatLng:TGLatLng):Boolean;
function ToString:String;
function JsClassName:String;virtual;
property JsVarName:String read GetJsVarName write SetJsVarName;
end;
TGBounds=class(TInterfacedObject,IJsClassWrapper)
private
FJsVarName: String;
FMinX, FMinY, FMaxX, FMaxY:Double;
FMin,FMax,FMid:TGLatLng;
function GetMax: TGLatLng;
function GetMid: TGLatLng;
function GetMin: TGLatLng;
procedure SetJsVarName(const Value: String);
function GetJsVarName: String;
published
destructor Destroy;override;
property minX : Double read FMinX write FMinX;
property minY : Double read FMinY write FMinY;
property maxX : Double read FMaxX write FMaxX;
property maxY : Double read FMaxY write FMaxY;
function ToString:String;
function Equals(aGBounds:TGBounds):Boolean;
property Min:TGLatLng read GetMin;
property Mid:TGLatLng read GetMid;
property Max:TGLatLng read GetMax;
function JsClassName:String;virtual;
property JsVarName:String read GetJsVarName write SetJsVarName;
function ToJavaScript:String;
end;
TGLatLngBounds=class
private
procedure setNorthEast(const Value: TGLatLng);
procedure setSouthWest(const Value: TGLatLng);
published
constructor Create(sw,ne:TGLatLng);
destructor Destroy;override;
function contains(aLatLng:TGLatLng):Boolean; deprecated; // Returns true iff the geographical coordinates of the point lie within this rectangle. (Deprecated since 2.88)
function containsLatLng(aLatLng:TGLatLng):Boolean; // Returns true iff the geographical coordinates of the point lie within this rectangle. (Since 2.88)
function intersects(aGLatLngBounds:TGLatLngBounds):Boolean;
function containsBounds(aGLatLngBounds:TGLatLngBounds):Boolean;
procedure extend(aLatLng:TGLatLng); // Enlarges this rectangle such that it contains the given point. In longitude direction, it is enlarged in the smaller of the two possible ways. If both are equal, it is enlarged at the eastern boundary.
function toSpan() : TGLatLng; // Returns a GLatLng whose coordinates represent the size of this rectangle.
function isFullLat() : Boolean ; // Returns true if this rectangle extends from the south pole to the north pole.
function isFullLng() : Boolean ; // Returns true if this rectangle extends fully around the earth in the longitude direction.
function isEmpty() : Boolean ; // Returns true if this rectangle is empty.
function getCenter() : TGLatLng; // Returns the point at the center of the rectangle. (Since 2.52)
function getSouthWest() : TGLatLng; // Returns the point at the south-west corner of the rectangle.
function getN
