if 0 {phk 2003-08-18 Let's assume your application is generating html pages.
tdom can help in a nice way to test the output.
Let's get all options from a html select tag:}
package require tdom
package require http
# get the html page
set token [http::geturl http://aspn.activestate.com/ASPN/Cookbook/Tcl/]
set data [http::data $token]
# parse the html
set doc [dom parse -html $data]
set root [$doc documentElement]
# get all option nodes
set optionList [$root selectNodes {//select/option}]
set result {}
# loop through all the options
foreach option $optionList {
set text [[$option nextSibling] nodeValue]
set value [$option getAttribute value]
lappend result [list $text $value]
}
puts $resultif 0 {which shows all the options
{{this section} Subsection}
{{all ASPN} ASPN}
{Products Products}
{Recipes Recipes}
{News NewsFeeds}
{Modules Modules}
{{Mailing Lists} Archive}
{{The Perl Journal} TPJ}
{Reference Reference}
from this html code fragment
...
<select name="type">
<option value="Subsection">this section</option>
<option value="ASPN">all ASPN</option>
<option value="Products">Products</option>
<option value="Recipes">Recipes</option>
<option value="NewsFeeds">News</option>
<option value="Modules">Modules</option>
<option value="Archive">Mailing Lists</option>
<option value="TPJ">The Perl Journal</option>
<option value="Reference">Reference</option>
</select>
...
The result can be used in a tcltest proc or however.
of course can code can be shorter, but I think it explains more this way.
This is my first wiki contribution, any feedback is appreciated
DMG 20-Aug-2003 asks: Offhand (and this is a general tdom/XML query) why use:
set text [[$option nextSibling] nodeValue]
versus
set text [$option text]
??
}