Tobias Reif wrote:

> Michael Schuerig wrote:
> 
> 
>>   Read in both versions of the file
>>   For each "dubious" entry in the current file
>>     Find a corresponding (based on title and year) entry
>>     in the original file
>>     Fix the current entry
>>   Dump a fixed file
> 
> 
> Perhaps you want to post tiny representative fragments.

Sure.

Old:

<biblio>
  <entry type="Book">
    <author>Thomas, David; Hunt, Andrew</author>
    <language>english</language>
    <publisher>Addison-Wesley</publisher>
    <title>Programming Ruby. The Pragmatic Programmer's Guide</title>
    <year>2000</year>
  </entry>
</biblio>

Current (broken, authors missing):

  <entry type="Book">
    <language>english</language>
    <publisher>Addison-Wesley</publisher>
    <title>Programming Ruby. The Pragmatic Programmer's Guide</title>
    <type>Book</type>
    <year>2000</year>
  </entry>


What I want (in the next step):

<biblio>
  <entry type="Book">
    <author>Thomas, David</author>
    <author>Hunt, Andrew</author>
    <language>english</language>
    <publisher>Addison-Wesley</publisher>
    <title>Programming Ruby. The Pragmatic Programmer's Guide</title>
    <type>Book</type>
    <year>2000</year>
  </entry>
</biblio>

I've considered munging the entries to DocBook bibliographic format; 
but so far I'm not convinced it's worth the effort (apart from getting 
more practice with XSLT and possibly Ruby XML tools).

Thanks

Michael

-- 
Michael Schuerig                   GPG Fingerprint
mailto:schuerig / acm.org            DA28 7DEB 5856 3365 BED9
http://www.schuerig.de/michael/    8365 0A30 545A 82D2 05D7