From: Paul Lutus <nospam@...>
Date: 2006-12-10T09:40:04+09:00
Subject: Re: text wrap

ishamid wrote:

/ ...

> Perhaps my case is different. I am converting OOo xml to TeX for
> further processing.

Do you mean an OpenOffice content.xml file, unpacked from a .odt document?
In that case, why not just insert some conveniently placed linefeeds to put
the tags on separate lines?

It happens I do this regularly, because I am constantly playing with
OpenOffice XML content. Here is my script to beautify content.xml files:

-------------------------------------------

#!/usr/bin/ruby -w

# beuatifies an XML file, usually only for analysis,
# because splitting one up and indenting its lines may
# make it unusable to the originating program.

def beautifyXML(data)
   tab = 0
   xml = ""
   data.gsub!(%r{<},"\n<")
   data.gsub!(%r{>},">\n")
   data.gsub!(%r{\n+},"\n")
   data.split("\n").each { |record|
      record.strip!
      outc = record.scan(%r{(</|/>)}).length
      inc = record.scan(%r{<\w}).length
      net = inc - outc
      tab += (net < 0)?net:0
      xml += ("   " * tab) + record + "\n"
      tab += (net > 0)?net:0
   }
   if(tab != 0) 
      $stderr.puts "Error: tag mismatch: #{tab}"
   end
   xml
end

# stream in/stream out

print beautifyXML(readlines.join("\n"))

-------------------------------------------

I want to emphasize this is a quickie script, not a thoroughly tested
application, and its sole purpose is to aid in understanding the syntax of
an XML file. I should also say that it may not work if the output is put
back into an OpenOffice document after beautification.

To use the script:

$ (script name) < content.xml > output.xml

This idea of beautifying an XML file so it is easier to understand and
process is quite different than your original inquiry. In many cases, XML
files are perfectly fine after they have been indented and made readable,
and nothing is lost.

But ... between some of the XML tag pairs emitted by this script may be
long, long lines of text. Those lines should not be broken up without first
considering the implications (the essence of my first post).

Also, one defect in the above script is that the first line of the script,
the XML identifying header, is not on the first line. This is something I
don't care about, because I don't try to use the output of this script for
anything but analysis. This error should easy to fix if you plan to use the
output as legitimate XML.

-- 
Paul Lutus
http://www.arachnoid.com