Version 23 of Simple Tkhtml web page displayer

Updated 2003-03-05 16:50:56

Purpose: to provide Neil Madden's five line Tkhtml based web page displayer.

This app fetches a web page, formats it, and displays it in a tk scrollable widget. It currently does not handle redirecting URLs, making the web pages active, or ease of changing fonts, etc.

Perhaps some variant of this would be a useful proc for tklib.

Also see the One-line web browser in Tcl


 package require Tkhtml;package require http;pack [scrollbar .vsb  -orient vertical -command {.html yview}] -side right -fill y;pack  [html .html -bg white -yscrollcommand {.vsb set}] -fill both -expand 1;set t [http::geturl http://mini.net/tcl/976.html];.html parse [http::data $t];http::cleanup $t

A much more readable version for those just interested in a quick look:

 package require Tkhtml
 package require http
 pack [scrollbar .vsb  -orient vertical -command {.html yview}] -side right -fill y
 pack  [html .html -bg white -yscrollcommand {.vsb set}] -fill both -expand 1
 set t [http::geturl http://mini.net/tcl/976.html]
 .html parse  [http::data $t]
 http::cleanup $t

Latest and greatest (thanks to DKF and others for comments). No -bg white, but fits in 4 lines/80 chars and adds hyperlinks and error handling (ignoring)!:

 package r Tkhtml;package r http;pack [scrollbar .v -o v -co {.h yv}] -s right -f y;pack [html .h -ys {.v set}] -f both -e 1;bind .h.x <1> {eval g [.h href %x %y]};proc g u {set t [http::geturl $u];.h cl;.h p [http::data $t];http::cleanup $t;.h co -base $u};g http://mini.net/tcl/976.html;proc bgerror args {};# NEM :-)

NEM - Redirecting URLs aren't too hard. Perhaps there should be an option in the http package though? http::config -followredirects 1. The HTTP codes that indicate a redirect are 301 and 302, so the following should do:

 proc geturl_followredirects {url} {
     set t [http::geturl $url]
     while {([http::ncode $t] == 301) || ([http::ncode $t] == 302)} {
         upvar #0 $t state
         array set meta $state(meta)
         http::cleanup $t
         set t [http::geturl $meta(Location)]
     }
     return $t
 }

Next, we probably want images too, so the following can be useful:

 image create photo default -data {
    R0lGODdhJAAkAPEAAACQkADQ0PgAAAAAACwAAAAAJAAkAAACmISPqcsQD6OcdJqKM71PeK15
    AsSJH0iZY1CqqKSurfsGsex08XuTuU7L9HywHWZILAaVJssvgoREk5PolFo1XrHZ29IZ8oo0
    HKEYVDYbyc/jFhz2otvdcyZdF68qeKh2DZd3AtS0QWcDSDgWKJXY+MXS9qY4+JA2+Vho+YPp
    FzSjiTIEWslDQ1rDhPOY2sXVOgeb2kBbu1AAADv/
 }

This creates a nice default image for us (a red cross if I remember correctly). Then we need a proc for fetching the images:

 proc FetchImage {src w h args} {
     # Fetch the image
     if {[catch {
         http::geturl $src -timeout 10000
     } token]} {
         return default
     }

     set data [http::data $token]
     http::cleanup $token
     # Hack needed to make sure the data is binary:
     binary scan $data {}
     set name [image create photo]
     if {[catch {$name put $data} ret]} {
         return default
     }
     return $name
 }

You could add caching to this, but I leave that as an exercise. So, we can put this all together:

 .html configure -imagecommand FetchImage -hyperlinkcommand  geturl

Although, the -hyperlink command doesn't seem to work for me, so try:

 proc HrefBinding {x y} {
     set new [.html href $x $y]
     set new [string trim $new {{}}]
     if {[string length $new]} {
         geturl $new
     }
 }
 bind .html <Button-1> [list HrefBinding %x %y]

Add a wrapper for the hyperlinks:

 proc geturl {url} {
     .html configure -cursor watch
     set t [geturl_followredirects $url]
     .html clear
     set data [http::data $t]
     # Get the url, incase we followed redirects
     upvar #0 $t state
     set url $state(url)
     http::cleanup $t
     # Get the bgcolor and stuff. 
     if {[regexp -nocase {<title>(.*)</title>} $data -> title]} {
        wm title .   $title
     }
     if {[regexp -nocase {<body[^>]+bgcolor=([^>]+)>} $data -> bgcolor]} {
         set bgcolor [string trim [lindex $bgcolor 0] \"]
         catch {.html configure -bg $bgcolor}
     }
     .html configure -base $url
     .html parse $data
     .html configure -cursor xterm
 }

Missing items -

 package require Img

goes a long way in helping images work.

  • You need some sort of resolution for URLs as a lot of them are not properly formed when passed from Tkhtml, so http packages croaks on them. Usually just a case of adding a http:// to the front.
  • Fonts can be done with the -fontcommand configuration option.
  • configuring -scriptcommand script, and then setting 'script' as a proc that returns nothing, can solve some problems with not updating properly.
  • This does very little error checking.
  • Applets using TclBlend would be cool.
  • URL resolution using the uri package would work for images.
  • A binding to change the cursor over links:
 bind .html.x <Motion> {
    set url [string trim [.html href %x %y] {{}}]
    if {[string length $url]} {
        .html configure -cursor hand2
    } else {
        .html configure -cursor xterm
    }
 }

So, the final stage:

 proc script {args} {}
 .html configure -scriptcommand script
 geturl http://www.slashdot.org

OK - I've created a package that wraps up this and some more functionality. The package now has a sourceforge page: http://sourceforge.net/projects/tkbrowser

The basic usage of the package is:

 package require browser
 pack [scrollbar .vsb -orient vertical -command {.html yview}] -side right -fill y
 pack [browser::browser .html -bg white -yscrollcommand {.vsb set}]  -side top -fill both -expand 1
 pack [label .l -bg gray -textvariable ::browser::browser(.html,status)] -side bottom -fill x
 proc updateTitle {var1 var2 op} {
     wm title .   $::browser::browser($var2)
 }
 trace variable ::browser::browser(.html,title) w updateTitle
 browser::geturl .html http://slashdot.org

This deals with getting the page and all images (including a rudimentary cache), and handling links etc. It even will handle forms, but this doesn't seem to be working correctly in my Tkhtml build. License is normal BSD/Tcl license. I'll contribute it to tklib once I've polished it up a bit. Requires Tkhtml and http, with the optional addition of Img (recommended). I need to sort out the URL handling as well - may change to the uri package from tcllib, which handles various URLs better. You can add https support if you have TLS, too (see the http package docs).

For details, or if the above URLs still aren't working, send me an email: mailto:[email protected]


escargo 5 Mar 2003 - I think the choice of the package name browser is a bad one because it's too generic. There are web browser and file browsers. There are also likely to be multiple implementations of each kind, and just grabbing the generic browser package name seems a bit hasty.


Category Application, Tkhtml, http