On Sat, 2006-10-28 at 08:59 -0600, Jack Daniels (Butch) wrote: > I work with library cataloguing records and am trying to accomplish this > amazing feat. > > Trying is no problem, amazing feat is eluding me > > I need to find all records with the same ISBN number (field 020)save all > matching records which will result in more than 1 record with the same ISBN, > then somehow magically end up with only 1 record containing all 856 fields.
If there are not a huge number of isbn's then load a hash with an array for each ISBN. Then access each element of the hash loop through the array through the different input streams and use business logic to merge the two inputs. Simplest approach is if field is empty use the content provided. More system oriented approach might be to write a quick parser of the input files which adds the ISBN as the first key and a second key that is the source. sort all files together by ISBN then input files. This would allow you to prioritise where data comes. This whole process can get very complex. My work has megalithic programs at work that merges direct mail address lists together. -- Ken Foskey FOSS developer -- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED] <http://learn.perl.org/> <http://learn.perl.org/first-response>
