From: Phlip <phlip_cpp@...>
Date: 2005-04-17T23:44:35+09:00
Subject: Re: Html Tag Generation

Karsten Meier wrote:

> I'm working on a project which translates programming examples from the
> perl cookbook to several other languages
> (http://pleac.sourceforge.net/)

Then why you are not using Perl?

<a beat>

;-)

> To create the following html:
> # <OL><LI>red</LI> <LI>blue</LI> <LI>green</LI></OL>
>
> in perl you can write:
> use CGI qw(:standard :html3)
> print ol( li([ qw(red blue green)]) );

Ruby/CGI might have a bigger problem. I could not find a way to make it
produce XHTML. A project should test its complex features, and XHTML tests
super-easy when XPath queries out target details. But Ruby/CGI writes tags
like <input>, not <input/>.

For MiniRubyWiki, I just rolled my own:

class XhtmlGenerator

 def initialize(stream)
  @stream = stream
 end

 def balance_(tag, attributes, pretty = "\n")
  write(pretty)
  write("<")
  write(tag)
  write(" " + attributes)  if ! attributes.nil? and attributes != ''
  write(">")
  write(pretty)

        begin
      yield()
        ensure
      write(pretty)
      write("</")
      write(tag)
      write(">")
      write(pretty)
        end
 end

 def small(attributes = '')
  balance_('small', attributes, '') { yield() }
 end

 def em(attributes = '')
  balance_('em', attributes, '') { yield() }
 end

 def table(attributes = '')
  balance_('table', ' summary="cosmetic table"' + attributes) { yield() }
 end

 def form(attributes = '')
  balance_('form', attributes) { yield() }
 end

 def th(attributes = nil)
  balance_('th', attributes) { yield() }
 end

 def td(attributes = nil)
  balance_('td', attributes, '') { yield() }
 end

 def tr(attributes = nil)
  balance_('tr', attributes) { yield() }
 end

 def h1(attributes = nil)
  balance_('h1', attributes, '') { yield() }
 end

 def body(attributes = nil)
  balance_('body', attributes) { yield() }
 end

 def head(attributes = nil)
  balance_('head', attributes) { yield() }
 end

 def title(attributes = nil)
  balance_('title', attributes, '') { yield() }
 end

 def xhtml()
  write('<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
"http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
')
  attributes = 'xmlns="http://www.w3.org/1999/xhtml"'
  balance_('html', attributes) { yield() }
 end

 def a_href(path, stuff = nil)
  attributes = "href=\"#{path}\""
  attributes += " " + stuff  if stuff
  balance_('a', attributes, '') { yield() }
 end

 def textarea(attributes)
  balance_('textarea', attributes, '') { yield() }
 end

 def nbsp()
  write('&nbsp;')
 end

 def space()
  write(' ')
 end

 def endl()
  write("\n")
 end

    def write(q)
  @stream.write(q)
    end

end

That provides this:

    x = XhtmlGenerator.new()
....
  x.xhtml() {
   titleStuff()

            style = "style=\"margin-left: 0.5cm; margin-right: 0.5cm;
margin-top: 0.2cm\""

   x.body("bgcolor=\"#eeeeee\" text=\"black\" " + style) {
                formatFileBody(x, searchTags) {
                    wikiFrame(x, contents)
                    }
    }
   }

> There ruby solution has two problems:
> 1) because the methods like CGI::li() take a block as parameter, we can
not
> use a list as content parameter:
> cgi.li{%w(red blue green)} returns   "<LI>redbluegreen</LI>" instead of
> "<LI>red</LI> <LI>blue</LI> <LI>green</LI>"
>
> 2) we need to write cgi.li{..} instead of just li(..)

myList.each{|n| x.li(n)}

> The result is that the scripts looks more confusing than the solution
> with perls cgi module.

Perl uses the design technique Yet Another Parser Hack to distinguish arrays
and scalars as parameters. You could enhance my def li() to detect array
arguments (using clean elegant reflection, not YAPH), and then iterate thru
the array.

-- 
  Phlip
  http://industrialxp.org/community/bin/view/Main/TestFirstUserInterfaces