Just for fun: Get your daily Dilbert comic and display it. If you provide a directory name as argument, it'll be saved to ''dirname/dilbert'''yyyymmdd'''.gif'' where yyyymmdd is todays date.

----

 package require Tk
 package require http
 package require base64

 set um [http::geturl http://www.dilbert.com/]
 set b [split [http::data $um] "\n"]
 #set ln [lsearch -inline -glob $b "*ALT=?Today's Dilbert Comic*"]
 set ln [lsearch -inline -glob $b "*ALT=?Today's Comic*"]
 
 regexp {^.*IMG SRC=\"([^"]+)\".*$} $ln --> picurl
 http::cleanup $um
 set um [http::geturl http://www.unitedmedia.com$picurl]
 if {$argc && [file isdirectory [lindex $argv 0]]} {
   set today [clock format [clock seconds] -format "dilbert%Y%m%d.gif"]
   set fd [open [file join [lindex $argv 0] $today] w]
   fconfigure $fd -translation binary -encoding binary
   puts $fd [http::data $um]
   close $fd
 }
 set pic [base64::encode [http::data $um]]
 http::cleanup $um
 image create photo dilbert -data $pic
 label .l -image dilbert
 pack .l
----
[Michael Jacobson]: Also see [Comic Server] for some code to get a bunch of comic strips from uComics.  Or you can check out (on windows only ~ sorry) in a [TclKit] application that servers the comics and display them in a embededd MSIE frame using [optcl].  Get the application here ... http://mywebpages.comcast.net/jakeforce/COMon!ics.kit (96kb).  (I use this code in [NewzPoint])
----
02. Aug. 2005: The Text for the Daily Dilbert comic changed.
Replace the lsearch line with the following ''(This has already been done; the [if 0 {]
wrapper was added to make the version grabbed by [wish-reaper] operate correctly)'':
 if 0 {
   set ln [lsearch -inline -glob $b "*ALT=?Today's Comic*"]
 }
[RLH] 20050802: I get the error:
 if 0 {
   "Can't read "picurl": no such variable while executing "http::geturl http://www.unitedmedia.com$picurl"
 }
[RS] Have you applied the fix above? It worked for me after I did it.

[RLH] I had read the post wrong and thought it was fixed above. Works now and I have commented out the old line above and added the one that works.

''[escargo] 3 Aug 2005'' - Touched up the previous included text to make compatible with [wish-reaper].

[RLH] Why would you put "if 0" around the error message I was getting?

''[escargo] 3 Aug 2005'' - The reaping tools ([wiki-reaper], [wish-reaper], and [XotclIDE]) can extract
what they think is '''code''' from the wiki pages.  The definition of '''code''' is simple; too simple
in several cases. One such case is the error message you copied into the page.  When reaped, that error
message becomes a line of code, the last one in the file.  When wish tries to run the code, the error message
causes wish to throw an error.  By using the [if 0 {] wrapper, wish will ignore it.  The text can stay on
the page, somewhat distorted, but still there; the reaped page will still execute, though it has some dead
code.  It's a compromise.

[RLH] Cool. I wasn't complaining just wondering and now the cat is satisfied.  :)

CJL 11 July 2006 - The presence (and validity) of the optional directory name argument determined whether a file was written, but didn't affect where it was written. Fixed.
----
''See also:''
   * [Web Scraping]
----
[Category Community] | [Category Example] | [Category Internet]