Re: Embedded Ruby and Tag Libs

I agree with DHH for the most part (I have to read it in more detail  
but for now, I'll grant the argument). However, he's talking about  
web applications that follow a MVC pattern. As it turns out, that's  
not the ONLY type of problem to solve with ruby.

My problem is primarily one of content management. I have a client  
who sells widgets. They're not selling widgets online they're just  
doing promotional marketing pieces. Normally you'd just write the  
HTML directly and be done with it. However, sometimes, as a web  
developer, you have to write fairly complex markup in order to make  
things work out the way you want to. Unless you want to maintain the  
site yourself, a general pain in the neck, you need some mechanism to  
hide complexity from your clients yet allow them to edit the content  
themselves. In some ways this is what textile does (hiding the  
complexity of HTML from authors) but I need to be able to create more  
complex behaviors than what textile does.

Actually DHH says something that I think supports my point:

> Extracting complex view logic is painful: With tag libraries, it's  
> possible to be a good programmer and rid the views of complex logic  
> by extracting code and replace it by tags. That is a noble and  
> right path to follow. The problem is when the effort required to do  
> the right thing is so intense that its basically a project in  
> itself. Then its not something you're gently invited to do. It's a  
> huge barrier and easy source of procrastination ("I'll extract  
> later...") and guilt ("If only I had extracted sooner...").

I simply have a different view of complexity. I want to extract those  
things that are simple from my point of view but complex from the  
point of view of content authors.

Adam

On 26-Sep-05, at 2:53 PM, Devin Mullins wrote:

> Hi, Adam,
>
> First, DHH's opinion on this matter might be of interest to you:
> http://www.loudthinking.com/arc/000408.html
> (DHH = The Rails Guy, btw)
>
> Second, the best way I can think of to achieve something like what  
> you're talking about in ERB would look like this:
>
> class AvTable
>  def initialize(out)
>    @out = out; @side = :odd
>  end
>  def trow
>    @out << %Q{<tr class="#@side">}
>    yield
>    @out << "</tr>\n"
>    @side = (@side == :odd) ? :even : :odd
>  end
> end
>
> def av_table(out)
>  out << "<table>\n"
>  yield AvTable.new(out)
>  out << "</table>\n"
> end
>
> a = ERB.new <<RHTML
> <% av_table(_erbout) do |av| %>
>  <% av.trow do %><td>...</td><% end %>
>  <% av.trow do %><td>...</td><% end %>
>  <% av.trow do %><td>...</td><% end %>
>  <% av.trow do %><td>...</td><% end %>
>  <% av.trow do %><td>...</td><% end %>
> <% end %>
> RHTML
>
> a.result.display
>
> Not sure how you would take care of "last", though (without writing  
> an internal iterator for your data), but I'm no expert, and this is  
> probably a crappy way of achieving what you're after. (I, for  
> instance, am not a fan of passing around _erbout.)
>
> Devin
>
> Adam Van Den Hoven wrote:
>
>
>> Hey guys,
>>
>> I've been dabbling with ruby on and off for a year or so and  
>> recently  more often. At work we do Java development and I keep  
>> thinking that  it would be nice to be able to bring some of the  
>> stuff I designed  there into my own work, which I want to do in ruby.
>>
>> I was wondering if anyone had thought about adding something like  
>> the  Java TagLibs to ruby (I guess it would embedded ruby). In my  
>> case, it  would make long term maintenance of files MUCH easier as  
>> I can give  it to a non-technical person without much  
>> documentation. In most  cases people who don't understand  
>> programming do understand parts of  HTML markup. If I could hide  
>> what is currently inline code behind  things that look like tags,  
>> its easier to give maintenance over to  non-technical folks.
>>
>> Here's a simple example. Often you want to do tiger striping on   
>> tables (alternate background color on each row). Because there  
>> isn't  sufficiently wide spread support for CSS2 and CSS3  
>> selectors, you  tend to do the following:
>>
>> <table>
>>         <tr class="odd"><td>...</td></tr>
>>         <tr class="even"><td>...</td></tr>
>>         <tr class="odd"><td>...</td></tr>
>>         <tr class="even"><td>...</td></tr>
>>         <tr class="odd"><td>...</td></tr>
>>         <tr class="even last"><td>...</td></tr>
>> </table>
>>
>> Unfortunately, if you want to add something between rows 4 and 5  
>> you  need to reset all your classes. This can get complicated.  
>> However, if  I had a taglib I could ask my content authors to  
>> write the following:
>>
>> <av:table>
>>     <av:trow><td>...</td></av:trow >
>>     <av:trow><td>...</td></av:trow >
>>     <av:trow><td>...</td></av:trow >
>>     <av:trow><td>...</td></av:trow >
>>     <av:trow><td>...</td></av:trow >
>>     <av:trow><td>...</td></av:trow >
>> </av:table>
>>
>> And leave the generation up to the code.
>>
>> I can think of hundreds of specific tags I might create, depending  
>> on  the situation, which would make authoring content easier and  
>> would  give me better control over the rendering of a page. It  
>> could also  make writing code easier in general. It would also be  
>> useful to have  many of the tags defined in the JSTL.
>>
>> Just a thought I had.
>> Adam
>>
>>
>>
>
>
>