Internet Programming

Internet - File
Transfers

In this tutorial I will
cover how to perform file transfers between your PC and a web server. The topics
of web site management, dynamic generation of web pages, and control/script
inclusion in web pages is covered in a different tutorial.

In the Professional and Enterprise Editions of VB
are the two controls which provide Internet features - the Internet Transfer
Control and the Web Browser Control. The Internet Transfer Control (ITC)
provides both FTP and HTTP file transfer controls and is the subject of this
tutorial - as is the use of the wininet.dll, which also provides file transfer
capabilities that you can access from within your VB applications.

The wininet.dll file is simply a library of
functions that you can use to do file transfers. It is not a part of VB6,
but is installed along with the Microsoft Internet Explorer. This means
you can safely assume that it is available on most PCs.

The reason for using the ITC is simplicity. The
control provides a very straightforward, fairly simple interface to use in your
VB programs. The down side is that the control is about 126K in size, increasing
the size of the installation files your application users will have to download.

By using the existing wininet.dll file,
you eliminate the increased application distribution file size, plus the
wininet.dll offers greater control over the file transfer process. As you would
expect, the wininet.dll is harder to learn and is not documented as part of the
VB documentation package.

So, I'll focus
first on the ITC, then end with enough detail on wininet.dll to show you how to
use it as an alternative to the ITC.

Protocols

For our purposes, only
two of the many protocols matter - HTTP and FTP. Protocols are simply rules
which programmers have agreed upon, and documented so that anyone using the
protocol can be assured that their software can communicate with other programs
which also use that protocol.

For the
purposes of file transfer as discussed in this tutorial, It's not really that
important to understand the details of either protocol, but there are a few
facts which are important to know. The key point to remember right now is that
the code for using the ITC for an FTP or for an HTTP file transfer are slightly
different.

Most web sites are accessible
by both FTP and HTTP (i.e., servers typically run both FTP and HTTP server
software for accessing content), so you can usually chose which approach to
take.

In general, it really makes little
difference which protocol you use. Many of the books I've read recommend the FTP
protocol because it is more flexible than HTTP (i.e., read that as FTP has more
features which can be controlled by code than does HTTP). I generally agree with
that recommendation, but will confess that my own freeware applications now have
an online update feature which is based on the HTTP protocol.

Here's the tradeoff that drove me to that
decision:

      
• FTP - Many ISPs do not allow
         anonymous (i.e., no username and no password) FTP connections to a
         website. But, I do not want to put my username/password into a
         distributed application for fear of compromising security on my
         web site. I could password protect and expose just a particular
         directory on the web site, but I've chosen to take no risks in
         that area.

      
• HTTP - The ITC code for an HTTP
         file transfer is extremly simple (the FTP code is not that
         complicated, it's just that the HTTP code is simpler).
        

One of the key drawbacks of using the ITC for
file transfer (regardless of the protocol that is used) is that it does not
provide any built-in capability to identify how many bytes of the transfer are
complete at any point in time. All you can tell from within your VB program is
whether the file transfer is in progress, or that it has stopped (because of a
successful transfer or some error that stopped the file transfer process). This
is one of the key reasons you might be interested in looking at one of the 3rd
party file transfer OCXs.

OpenURL Method of
File Transfer

Now that I've gotten the introductory comments out of the way, let's talk about
the details of the ITC. The two most important things to know about the ITC is
that there are two methods of downloading files from a web site - the OpenURL
method and the Execute method. Both support the FTP and HTTP file transfer
protocols.

The OpenURL method is
very simple. You put in a file name to download and tell the program whether the
file is all text or binary. The code looks for an HTTP transfer of a text file
looks like this:

text1.text = inet1.OpenURL ("http://www.vbinformation.com/badclick.htm", icString)

The code for an HTTP transfer of a binary file
looks like this:

Dim bData() as Byte
bData() = inet1.OpenURL ("http://www.vbinformation.com/badclick.htm", icByteArray)

Since all files (text or binary) can be
transferred as a binary file, I used the same file name in both examples. Note
that in the first case, the downloaded file content is placed in a textbox named
'text1'. In the second case, the downloaded file content is saved in a Byte
array whose upper bound is set by the number of bytes downloaded by the OpenURL
method. Also, note that both examples use HTTP URLs, but FTP URLs could have
been used just as readily.

In case you
don't remember, an easy way to save the bData byte array is:

Open "filename" for Binary as #1
Put #1, , bData()
Close #1

This is really all there is to successfully
downloading a file by using the OpenURL method. I'll cover the question of
errors (such as when the server is down, or the file is not there) later in this
tutorial.

You should note that the
OpenURL method is synchronous - which simply means that any code that follows
the OpenURL statement will not be executed until the file transfer is completed,
or until the file transfer is stopped by the occurence of an error or by a user
command (I'll show how to do this later).

Execute Method of
File Transfer

The second method for downloading a file is the Execute method. As you'll see it
provides more features, but is definitely more complicated to code. The one key
difference that you'll want to be aware of is that with the Execute method the
bytes of data are sometimes, but not always, kept within the ITC itself (in a
memory buffer). When the ITC does keep the downloaded bytes in its buffer, you
must use another method called GetChunk to extract the dowloaded bytes. Whether
the memory buffer is used varies with the arguments used in calling the Execute
method. I'll give more detail on that later.

Another key difference that you should know about
is the Execute method is asynchronous - meaning that it will download the file
in the background and that any code following the Execute statement will be
executed immediately.

Finally, to
complicate the discussion a bit more, the arguments you use in the Execute
method differ depending on whether you want to use FTP or HTTP for the file
transfer.

Here's the general syntax for
the Execute method:

inet1.Execute (url, operation, data, requestheaders)

For an FTP file transfer, only the first two
arguments are used. For an HTTP file transfer, all four arguments may be used.

Here's an example of the code you would
use to start a transfer using the Execute method and the FTP protocol:

inet1.Execute ("ftp://www.microsoft.com", "DIR")

This command transfers the directory listing of
the Microsoft ftp site. Note than while the OpenURL method returns data to a
variable or an array, the Execute method does not! The data returned by the
Execute method will either be kept within the ITC's buffer, or be directed to a
file according to the specifics of the command it is given.

The Execute method actually supports 14 FTP
commands (which are placed in the 'operation' argument), but there are primarily
three (CD, GET, and PUT) which you will use most often:

      
• inet1.Execute
         ("ftp://www.microsoft.com", "CD newdirectory"

      
• inet1.Execute
         ("ftp://www.microsoft.com", "GET remotefile localfile"
        

      
• inet1.Execute
         ("ftp://www.microsoft.com", "PUT localfile remotefile"
        

The first of these three allow you to make the
connection to the FTP server and to navigate to the directory where the files
are located. The second shows how to GET a file from the server and put it on
your PC, while the third shows how to PUT a local file from your PC onto the FTP
server (in the directory to which you navigated to using the CD command).

Also, you will note that the GET and PUT
commands create a file on either the local or remote computers. In these cases,
the ITC memory buffer is not used. However, the ITC memory buffer must be
accessed in order to get the output of the 'DIR' command.

In order to discuss how the ITC memory buffer is
accessed, we have to talk first about the StateChanged Event. Statechanged is
the only event the ITC control has, and it provides a variable called 'State'
which must be read to determine the status of a pending Execute method.

The State values are:

      
• icNone (0)

      
• icHostResolvingHost (1)
        

      
• icHostResolved (2)

      
• icConnecting (3)

      
• icConnected (4)

      
• icRequesting (5)

      
• icRequestSent (6)

      
• icReceivingResponse (7)
        

      
• icResponseReceived (8)
        

      
• icDisconnecting (9)

      
• icDisconnected (10)

      
• icError (11)

      
• icResponseCompleted (12)
        

Typically, Select Case code is used within the
StateChanged event to determine what action to take. In general, only actions 8,
11, and 12 are used to generate a code response. The others are used mostly to
decide what message to display in a status label/toolbar.

For State=12, where the file transfer is complete,
the action you take is entirely up to you. This would usually be a simple popup
message telling the user that the file transfer is complete.

For State=11, which indicates that an error has
occurred, you would have to generate code necessary to correct or ignore the
error condition.

Generally, you simply
wait for State=12 to indicate that the transfer is complete. But, in some cases
you may want to begin extracting data before the transfer is complete. For
example, the HTTP header information is received first, but is not included in
the ITC download buffer. To get that information you use the .GetHeader method.
You can use the State=8 to determine when the header information is available.

In those cases where the ITC buffer is
used to temporarily store the downloaded information, the .GetChunk method is
used. Here's the code for the case where string data is being downloaded and a
State=12 has been received to indicate that the transfer is complete:

Do
 DoEvents
 bData = inet1.GetChunk (1024, icString)
 AllData = AllData & bData
Loop Until bData = ""

In the case where a State=8 has been received,
it is possible that no actual data is in the ITC buffer (such as when only
header information has been received). So, if the above code is used following a
State=8 event, the condition of bData="" may not indicate completion of the data
transfer.

Finally, remember that you may
have to set the .UserName and .Password properties if you are using the FTP
commands within the Execute method. If the FTP site you are accessing allows
'anonymous' logon's, then you will not have to set these properties.

Properties/Methods/Events

As you saw in the sample
code, the basics of file transfer don't require knowledge of all 19 properties,
5 methods, and one event exposed by the ITC. The following table list all of the
ITC interface elements, but as you will see in the discussion that follows, you
will not use but a few of these in most applications:

      
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
        
      

Properties	Methods	Events
AccessType	Cancel	StateChanged
Document	Execute
hInternet	GetChunk
Password	GetHeader
Protocol	OpenURL
Proxy
RemoteHost
RequestTimeout
ResponseCode
ResponseInfo
StillExcuting
URL
UserName

In addition, the normal control properties of
.Index, .Left, .Top, .Tag, .Parent and .hInternet are available for the ITC.

This table can be digested more easily
if you think in terms of how the properties/methods/events are used. Here's the
way I've grouped them in my notes:

      
• Basic
  
         The .OpenURL and .Execute methods are the heart of using the ITC.
         Everything you do requires the use of one of these two methods.
         The .GetChunk method is used to capture data downloaded by the
         .Execute method.

      
• Working
  
         The .URL, .Cancel, .StillExecuting, .ResponseCode, .ResponseInfo,
         .Cancel, .GetChunk, and .GetHeader interface elements are used
         extensively during program execution.

      
• Startup
  
         The .AccessType, .Proxy, .Protocol, .RequestTimeout, .RemoteHost,
         .UserName, .Password, and .RemotePort are very basic properties
         which you set once or use the default - then you're done with
         them.

The bottom line of this tutorial section is
that file transfers can be made very easily using the ITC and just a handful of
the properties, methods and events supported by the control.

As a final note, in case you've been watching
carefully you will notice that I left out any discussion on the use of the
.Execute method with HTTP commands. This was strictly for lack of time/space in
this tutorial. The .Execute method can be used equally with either FTP or HTTP
commands, but the FTP options are generally more extensive so FTP is the normal
choice for programmers.

WinInet.dll - File Transfer Alternative