From: Alex Gutteridge <alexg@...>
Date: 2007-08-06T11:08:14+09:00
Subject: Re: sgub stretching over several lines

On 6 Aug 2007, at 10:12, Jan Ask wrote:

> Alex & Sebastian,
>
> Thanks for taking the time to reply. The string.gsub(/start.*end/m,
> 'some_value') did indeed help, but I am afraid my problem is a bit  
> more
> complicated.
>
> I am basically trying to cleanup a long xml file. A typical part of  
> the
> string looks like this:
>
> &lt;div class=&quot;field field-type-text field-field- 
> audience&quot;&gt;
>
>   &lt;h3 class=&quot;field-label&quot;&gt;audience&lt;/h3&gt;
>
>   &lt;div class=&quot;field-items&quot;&gt;
>
>           &lt;div class=&quot;field-item&quot;&gt;Public&lt;/div&gt;
>
>       &lt;/div&gt;
>
> &lt;/div&gt;
>
> &lt;div class=&quot;field field-type-text field-field- 
> creator&quot;&gt;
>   &lt;h3 class=&quot;field-label&quot;&gt;creator&lt;/h3&gt;
>   &lt;div class=&quot;field-items&quot;&gt;
>           &lt;div class=&quot;field-item&quot;&gt;Tom Jones&lt;/ 
> div&gt;
>       &lt;/div&gt;
> &lt;/div&gt;
>
> I am trying to format it like this:
> <audience>Public</audience>
> <creator>Tom Jones</creator>
>
> So the problem is that the values in the xml change throughout the
> string, so I cannot do a pattern match for them directly. Any ideas
> would be hugely appreciated!
>
> Jan
> -- 
> Posted via http://www.ruby-forum.com/.

Without knowing the whole problem it is difficult to say what the  
best solution is, but for the string you post above, I would clean it  
up and parse with something like Hpricot:

require 'rubygems'
require 'hpricot'

string = DATA.read #read in string

string.gsub!(/&lt;/,'<') #Convert lt and gt symbols to real <>
string.gsub!(/&gt;/,'>')
string.gsub!(/&quot;/,'"') #Put in quotes

doc = Hpricot(string) #Parse with Hpricot

fields = ['audience','creator'] #Create array of 'fields' to extract

fields.each do |f| #For each field...
   el = doc.search("//div[@class='field field-type-text field-field-# 
{f}']") #...find appropriate divs
   el.each do |e| # for each field div...
     puts "<#{f}>" + e.at("//div[@class='field-item']").inner_html +  
"</#{f}>" #print data
   end
end

__END__
&lt;div class=&quot;field field-type-text field-field-audience&quot;&gt;

   &lt;h3 class=&quot;field-label&quot;&gt;audience&lt;/h3&gt;

   &lt;div class=&quot;field-items&quot;&gt;

           &lt;div class=&quot;field-item&quot;&gt;Public&lt;/div&gt;

       &lt;/div&gt;

&lt;/div&gt;

&lt;div class=&quot;field field-type-text field-field-creator&quot;&gt;
   &lt;h3 class=&quot;field-label&quot;&gt;creator&lt;/h3&gt;
   &lt;div class=&quot;field-items&quot;&gt;
           &lt;div class=&quot;field-item&quot;&gt;Tom Jones&lt;/div&gt;
       &lt;/div&gt;
&lt;/div&gt;

Alex Gutteridge

Bioinformatics Center
Kyoto University