webgen users - Oct 2010 - regenerating a page when another file changes

Hello to everyone,
I''m using webgen to build a website where some information (in
particular the
news) are read from a YAML file I have in the src directory. For example

mysite/src/news.YAML:

2010-10-27:
 title: a news
 content: description of the news
2010-10-24:
 title: an older news
 description: description of the older news

In the news.page file I have written some ERB code which reads the data 
contained in news.yaml and generates the appropriate HTML for it. The only 
problem is that I don''t know how to tell webgen to rebuild the
news.html page
from news.page whenever news.yaml changes. I looked at the webgen 
documentation and tried to search google but found nothing. Currently,
I''m
deleting webgen.cache everytime that news.yaml changes, but I don''t
like this
solution very much. Is there another way to do what I want?

Thanks in advance

Stefano

On 2010-10-27 12:23 +0200 Stefano Crocco wrote:
> Hello to everyone,
> I''m using webgen to build a website where some information (in
> particular the news) are read from a YAML file I have in the src
> directory. For example
> 
> mysite/src/news.YAML:
> 
> 2010-10-27:
>  title: a news
>  content: description of the news
> 2010-10-24:
>  title: an older news
>  description: description of the older news
> 
> In the news.page file I have written some ERB code which reads the
> data contained in news.yaml and generates the appropriate HTML for
> it. The only problem is that I don''t know how to tell webgen to
> rebuild the news.html page from news.page whenever news.yaml changes.
> I looked at the webgen documentation and tried to search google but
> found nothing. Currently, I''m deleting webgen.cache everytime that
> news.yaml changes, but I don''t like this solution very much. Is
there
> another way to do what I want?
You can use the meta information `used_nodes` (see
http://webgen.rubyforge.org/documentation/reference_metainfo.html#usednodes).

However, for this to work the file src/news.YAML would need to be
processed by a source handler. This can be done by making the copy
handler aware of the file by adding the following to config.yaml:

    patterns:
      Copy:
        add: [**/news.YAML]

Then you just need two more things:

1. The news.YAML should not actually be copied to the output directory.
   Create a file src/metainfo with the following contents:

       news.YAML:
         no_output: true

2. The page file src/news.page should depend on changes in news.YAML:

       ---
       title: News
       used_nodes: [news.YAML]
       ---
       content here

   I.e. you need to insert the `used_nodes` meta information key.

After that every time you change the news.YAML file, the news.page file
gets automatically regenerated.

-- Thomas

On Thursday 28 October 2010, Thomas Leitner wrote:
> |On 2010-10-27 12:23 +0200 Stefano Crocco wrote:
> |> Hello to everyone,
> |> I''m using webgen to build a website where some information
(in
> |> particular the news) are read from a YAML file I have in the src
> |> directory. [...]
> |> The only problem is that I don''t know how to tell webgen to
> |> rebuild the news.html page from news.page whenever news.yaml changes.
> |You can use the meta information `used_nodes` (see
> |http://webgen.rubyforge.org/documentation/reference_metainfo.html#usednode
> |s).
> |
> |However, for this to work the file src/news.YAML would need to be
> |processed by a source handler. This can be done by making the copy
> |handler aware of the file by adding the following to config.yaml:
> |
> |    patterns:
> |      Copy:
> |        add: [**/news.YAML]
> |
> |Then you just need two more things:
> |
> |1. The news.YAML should not actually be copied to the output directory.
> |   Create a file src/metainfo with the following contents:
> |
> |       news.YAML:
> |         no_output: true
> |
> |2. The page file src/news.page should depend on changes in news.YAML:
> |
> |       ---
> |       title: News
> |       used_nodes: [news.YAML]
> |       ---
> |       content here
> |
> |   I.e. you need to insert the `used_nodes` meta information key.
> |
> |After that every time you change the news.YAML file, the news.page file
> |gets automatically regenerated.
> |
> |-- Thomas
Thank you very much for your answer. Your suggestion worked perfectly.

Stefano

Hi Stefano, is your code visible somewhere?

I have a [blogging extension for webgen][1] that has a similar problem
and I''d like to compare :)
I think I''m setting the used_nodes metadata in the code, but probably
incorrectly?

[1]: http://github.com/cdlm/website/blob/master/ext/blog.rb


On Wed, Oct 27, 2010 at 12:23, Stefano Crocco <stefano.crocco at alice.it>
wrote:
> Hello to everyone,
> I''m using webgen to build a website where some information (in
particular the
> news) are read from a YAML file I have in the src directory. For example
>
> mysite/src/news.YAML:
>
> 2010-10-27:
> ?title: a news
> ?content: description of the news
> 2010-10-24:
> ?title: an older news
> ?description: description of the older news
>
> In the news.page file I have written some ERB code which reads the data
> contained in news.yaml and generates the appropriate HTML for it. The only
> problem is that I don''t know how to tell webgen to rebuild the
news.html page
> from news.page whenever news.yaml changes. I looked at the webgen
> documentation and tried to search google but found nothing. Currently,
I''m
> deleting webgen.cache everytime that news.yaml changes, but I
don''t like this
> solution very much. Is there another way to do what I want?
>
> Thanks in advance
>
> Stefano
> _______________________________________________
> webgen-users mailing list
> webgen-users at rubyforge.org
> http://rubyforge.org/mailman/listinfo/webgen-users
>


-- 
Damien Pollet
type less, do more [ | ] http://people.untyped.org/damien.pollet

On Thursday 28 October 2010, Damien Pollet wrote:
> |Hi Stefano, is your code visible somewhere?
> |
> |I have a [blogging extension for webgen][1] that has a similar problem
> |and I''d like to compare :)
> |I think I''m setting the used_nodes metadata in the code, but
probably
> |incorrectly?
> |
> |[1]: http://github.com/cdlm/website/blob/master/ext/blog.rb
My code is much more basic than yours, so I don''t think you''d
find anything
useful in it. However, here''s what it looks like after following
Thomas''s
instructions:

# config.yaml

default_processing_pipeline:
  Page: erb,tags,textile,blocks,fragments

patterns:
 Copy:
  add: [**/news.yaml]


# src/news.page

---
title: News
in_menu: true
sort_info: 25
used_nodes: [news.yaml]
--- name:content pipeline:erb,tags,redcloth

h2. News

<% RuberNews.news_data.each_with_index do |news, idx| %>

<div <%= idx > 0 ? ''class="rule"
style="padding-top:10px"'' : ''''%> >
h3(#<%=news.id%>). <%=news.title%>

p{margin-bottom:10px;font-size:1.2em;color:#FF7E80x}. *<%= news.date %>*

<%= news.contents %>
</div>
<%end %>


#src/data.metainfo

news.yaml:
 no_output: true


#src/news.yaml

- title: first news
  date: 2010-10-28
  contents: some text

Stefano

On Thu, Oct 28, 2010 at 17:02, Stefano Crocco <stefano.crocco at alice.it>
wrote:
> My code is much more basic than yours, so I don''t think
you''d find anything
> useful in it. However, here''s what it looks like after following
Thomas''s
> instructions:
Simplicity is interesting as well, thanks :)
So I suppose your RuberNews class is just loading the news.yaml file
into some Structs ?



-- 
Damien Pollet
type less, do more [ | ] http://people.untyped.org/damien.pollet

On Thursday 28 October 2010, Damien Pollet wrote:
> |On Thu, Oct 28, 2010 at 17:02, Stefano Crocco <stefano.crocco at
alice.it>
wrote:
> |> My code is much more basic than yours, so I don''t think
you''d find
> |> anything useful in it. However, here''s what it looks like
after
> |> following Thomas''s
> |
> |> instructions:
> |Simplicity is interesting as well, thanks :)
> |So I suppose your RuberNews class is just loading the news.yaml file
> |into some Structs ?
That''s about it. It seems I have forgot to add the the contents of
ext/init.rb
to the last message, so here it is:

#ext/init.rb

module RuberNews
  
  News = Struct.new :title, :date, :contents
  
  class News
    
    def id
      "#{title.gsub(/[^\w_-]/, ''-'')}-#{date}"
    end
    
    def format_date
      date.strftime ''%d %B, %G''
    end
    
  end
  
  def self.news_data
    data = begin YAML.load(File.read(''src/news.yaml'')) || []
    rescue ArgumentError, SystemCallError
      []
    end
    data.map! do |d|
      News.new d[''title''], d[''date''],
d[''contents'']
    end
    data.sort do |x, y|
      y.date <=> x.date
    end
  end
  
end

As you can see, RuberNews is a module which define a News class (a Struct with 
a couple of convenience methods) and a class method to read the YAML file, 
create News objects from the entries it contains and sort them from the most 
recent to the oldest. In case the news file doesn''t exist, is empty or
isn''t a
valid YAML file, an empty array is returned.

Stefano

Hi again!

So, inspired by this, I hacked together my version :)

The idea is that news items go in the metainfo of some page, then
there''s a tag to generate a small list of the most recent news. My
use-case is for keeping a record of conferences I attended, talks I
did, etc, so items will never be longer than a few words or a
sentence.

Anyway, code is here, comments appreciated:
http://github.com/cdlm/website/blob/master/ext/ticker.rb

-- 
Damien Pollet
type less, do more [ | ] http://people.untyped.org/damien.pollet