3.1. MapSever interaction using CGI

To map through the Mapfile configuration file, you have understood the basic ideas of maps, layers and classes, and know how the various parts of a Mapfile are combined. Now move on to an app that will offer full pan and zoom functionality, as well as the ability to toggle layers on and off.

So far, the map can only be viewed when the map is created. Creating a web mapping application usually means creating maps that can be changed by the user (of the application). That is, a user can change the content (or information) of the map. To do this interactively, we use the MapServer HTML template.

MapServer CGI is template based. When first executed in response to a web request, it reads a configuration file (Mapfile) that describes the map’s layers and other components. Then it draws and saves the map. Next, it reads one or more HTML template files identified in the Mapfile. Each template consists of traditional HTML markup tags and special MapServer replacement strings. For example, these strings are used to specify paths to map images created by MapServer, identify layers to render, and specify zoom levels and orientations. MapServer replaces the current values of these strings and sends the data stream to the web server, which then forwards it to the browser. When the requester changes any form element on the page (for example, by changing the zoom direction or zoom value) and clicks the submit button, MapServer will receive requests with these new values from the web server. Then the cycle starts again.

3.1.1. Technical note on network access of MapServer

Your original purpose is to generate maps in a CGI environment where users connect to the Apache web server through a web browser. In this environment, Apache calls MapServer, which can pass various forms of variables from the browser.

Using this information, MapServer generates maps and web pages, which Apache returns to the browser. Of course, it is not enough for MapServer to generate a map only from the browser’s information.

In fact, CGI MapServer network applications have four components: Map files, HTML initialization forms, one or more template files, and spatial data (libraries).

Note: Web applications can provide fixed content for static pages, as well as dynamic content through scripts in response to web forms, as well as database queries and other functions.

The standard that determines the interaction between these scripts and the Web server is called the universal gateway interface, which is usually abbreviated to CGI.

These four components are discussed in this chapter. The premise is that the user’s web browser has already displayed the initialized file.

MapServer, like all web applications, is based on a stateless protocol - that is, on each call, it only knows what information the browser communicates to it. Stateless protocols preclude using programs to answer anything other than the last question. However, some clever code can provide a stateful environment that enables web applications to do more complex work. For example, state can be maintained between calls by storing state information in a variable in a hidden form, such as in a URL or a cookie. But some methods need to bootstrap the program, so it contains the information needed for the first call.

In MapServer, this is done by initializing the file. In CGI MapServer applications, initialization files are traditional HTML forms that have initialization information that converts hard-coding into form variables. Almost all values used by MapServer can be set in the initialization file.

When Apache first calls MapServer using HTML’s initialization form, the form variable is used to specify the name of the map file (the extension is usually .map ).

It then finds the fonts, symbols, templates, and spatial data of the file. The map file also specifies the size and geographical range of the final map, and whether the map is in GIF,JPEG or PNG format. After reading the map file, MapServer renders one or more images: the map itself, legends and scale bars, and possibly a reference map. It stores these images in a specified location in the map file.

In order to show the results, MapServer needs to design the map format and organize the relevant elements to make it a web page. The program itself does not generate HTML; instead, it scans the HTML template to replace the string. String substitution can refer to files, details of map geometry, layer specifications, or scaling factors. They can also be constants of CGI variables, such as image size, name of the map file, map range, and so on.

MapServer replaces the string with the appropriate value and returns the modified HTML to the browser that requested the information.

In this chapter, you will build a map application to illustrate the simplicity of using MpaServer and how the components work together.

The following chapters will introduce the features of MapServer (projection, tile retrieval, notation, query, etc.), which will make the capabilities of MapServer more obvious.

3.1.2. MapServer application development through CGI

MapServer provides two development interfaces: CGI and MapScript. CGI is the most basic method for MapServer application development. In CGI mode, MapServer uses the GD library to convert spatial data into geographic graphics and transmit them back to the user’s browser. Users do not need to install any specific browsing program to operate on it through the browser.

The user can send a request directly to the MapServer server through a URL similar to the following, and MapServer creates the appropriate map image according to the requested parameters, and returns the result to the user’s browser:

http://hostname/cgi-bin/mapserv?map=/ms4w/apps/txitorial/htdocs/example1-1.map&layer=states&mode=map

where /cgi-bin/mapserv is the MapServer CGI program. ? is followed by a query string consisting of MapServer CGI parameters such as map , layer , and mode . The map parameter specifies the map file to be processed by MapServer, the layer parameter tells mapserv to display the layer, and the mode parameter tells mapserv to generate a static map image.

Alternatively, dynamic, interactive maps can be generated by specifying the value of the mode parameter to the browser. At this time, when the user clicks on the map, MapServer will generate a zoom-in, zoom-out or repositioning map centered on the user’s clicked location.

Generating a dynamic map requires an HTML template with a form to interact with the user.

A simple form is as follows:

<!-START OF MAPSERVER FORM->
<form name=”mapserv” method=”GET" action="/cgi-bin/mapserv.exe">
<!-HIDDENMAPSERVERCGIVARIABLES->
<input type=Mhidden"name="map"value=n[map]">
<input type="hidden"name="imgext"value="[mapext]n>
<input type="hidden"name="imgxy"value="199.5149.5">
<input type="hidden"name="zoom"value=,T'>
<input type="hidden"name="mode"value="browse">
<div align="center">
<table border='T'cellpadding="0"cellspacing=”0">
<tr>
<td>
<!-THE INTERACTIVE,DYNAMICALLY CREATEDMAP->
<input type="image"name="img"src="[img]"
width=”400”height=”300”>
</td>
</tr>
</table>
</div>
</form>

A MapServer HTML template is an HTML-formatted file with MapServer tags. MapServer tags are MapServer CGI variables enclosed in square brackets ([]), such as [map], [img], and [mapext], etc., which represent Map file, map image URL, and extent of the map. Through HTML templates, MapServer CGI programming can be greatly simplified.

When the MapServer CGI program receives the request from the MapServer HTML template, The parameters are first parsed, the request is processed and the necessary output is generated, then some of these outputs are replaced with the corresponding MapServer tags (variables) in the HTML template, and finally the HTML (template) file with the output results is returned to the client browser. Since the map in the template is returned as a form element, the user can interact with it.

MapServer provides many CGI variables for Web mapping. Almost all keywords in MapServer’s map files can be defined as a variable, but there are not many CGI variables in the core of mapping. In addition, users can also customize variables, such as users can define a variable named root, which is used to represent the root directory of the application.

MMap browsing can only rely on MapServer to provide the basic functions of Web mapping services. In addition, MapServer also provides powerful spatial and attribute query functions. In CGI mode, HTML templates are used extensively to construct queries and organize the information returned by queries, which often involves complex interactions with map files, templates, and MapServer CGI programs.

MapServer has many modes ( mode, and browse ( browse ) mode is the default mode of MapServer. There are 18 modes related to queries. By using these query modes, MapServer can implement points such as: querying points based on mouse position, querying points based on input coordinate values, querying attributes by inputting an expression, and querying based on the serial number of elements.

MapServer creates legends and scales (configurable in mapfile) and generates reference maps. The reference map displays the context of the currently displayed map. Zooming and panning are controlled by the user.

3.1.3. MapServer HTML template

With the MapServer HTML template, you need to add the MapServers template declaration to each HTML template file. Otherwise, when using GetFeatureInfo (with an info_format collection) as the text/html request, you will get the following error message:

Content-type: text/xml isValidTemplate(): Web application error.
Missing magic string, template-file doesn't look like a MapServer template.

You need to add at the beginning of each template:

<!-- MapServer Template -->

Example the exemplar template query footer.html is:

<!-- MapServer Template -->

A MapServer HTML template is actually a MapServer-specific HTML file – these table tags are MapServer CGI variables enclosed in []. When a MapServer CGI program processes an application, it first parses the query_string and map file, and produces the necessary output. These outputs need to be written in an HTML template file, which must be defined in the Mapfile by using the web template keyword (or in a separate HTML initialization file). The CGI program will replace all variables in the HTML template with appropriate values before sending it to the web browser. If you view the contents of the HTML template directly on a web browser, there will be no maps available, but empty images and other characters of unknown meaning.

MapServer’s network mapping provides some variables – img variables, which have been examples before, where several commonly used CGI variables are designed as part of the initialization of the map interface, but actually all MapFile parameters can be defined as variables.

An explicit reference to CGI variables is at: http://www.mapserver.org/cgi/index.html

The CGI variables used by MapServer can also be defined by the developer and MapServer will pass it to the application. For example, we could create a variable called root to represent our root directory for this tutorial – the value of root will be read by /tutorial . When the MapServer CGI value processes our HTML template, it will replace every instance of the [root] tag and /tutorial. You’ll see this in action in each of the following examples.