[XML4Lib] Getting MARCXML into a relational database

[Eric Lease Morgan]

On Jun 11, 2009, at 4:56 PM, Gong, Dao Rong wrote:

> Has anyone had successful experience importing MARCXML file into a  
> relational database?

The short answer to your question is, "Yes, many of us have experience  
doing this sort of work."

The long answer is, "What is the problem you are trying to solve?"  
Putting a single MARCXML file containing a single MARCXML record into  
a single text field of a (relational or flat file) database is  
straight-forward. If the single MARCXML file contains a collection of  
many MARCXML records, then the text field might need to be rather  
large -- megabytes and megabytes in size.

If you want to parse each MARCXML record into distinct fields (title,  
author, subject, etc.), then you will probably want to apply some sort  
of XSL processing against the file. If you wanted a challenge, then  
you could convert the MARCXML into "real" MARC records and parse it  
that way. Pulling the data out in a cursory way is easy. Pulling it  
out in a more finely grained way is more difficult because there are  
literally thousands of options. Then of course, to what degree do you  
want to exploit relational database techniques to your problem? Join  
tables are fun and productive, and the design of databases like these  
are not difficult, but inserting data into them requires the creation  
of keys and the use of SQL to do inserts. More programming.

Again, what is the real problem you are trying to solve?

-- 
Eric Lease Morgan
University of Notre Dame