For all this examples to work you first have to do:

% package require TclCurl

Basic examples

To download a webpage

        package require TclCurl
        curl::transfer -url http://www.curl.com

You can also get information about the transfer:

	curl::transfer -url http://www.curl.com     \
	        -infohttpcode     httpCode          \
	        -infocontenttype  contentType       \
	puts "Http code: $httpCode"
	puts "Content type: $contentType"

The curl::transfer command is used for simple transfers where you don't want to use persistant connections. For more complex operations:

        set curlHandle [curl::init]
        $curlHandle configure -url http://www.curl.com \
                -file "index.html"
        $curlHandle perform

        $curlHandle configure -header 1                \
                -file "anotherfile.html"
        $curlHandle perform

        $curlHandle cleanup

As you can see in this example, it is easy to reuse connections.

Maybe you want to know which files are in a directory of a ftp server:

        curl::transfer -url "ftp://ftp.kde.org/pub/" \
                -ftplistonly 1

You may also want to to process the data yourself:

        proc writeToFile {readData} {
            puts "writeToFile called [incr ::i]"
            puts -nonewline $::outFile $readData

            return
        }

        set i 0

        set outFile [open "cosa.tar" w+]
        fconfigure $outFile -translation binary

        curl::transfer -url "127.0.0.1/~andres/cosa&co.tar" -writeproc writeToFile

        close $outFile

Or perhaps something more unusual:

        curl::transfer -url "dict://dict.org/m:curl"

For a better idea of what you can do, you can read the cURL manual page and The art of HTTP scripting.