With Thanks To:

Mary Woods Scott, author of the first two bibliographies, who gave her blessings for the project.

The North Dakota Geological Survey, for their encouragement.

Allan Ashworth, Chair of the NDSU Geosciences Department, who suggested this project.

Database Construction:

After scanning in numerous printed pages, the resulting text was edited into a form easily handled by other processors: a plain-text file, one record per page, delimited by two asterisks. The asterisks were used because there were many occurances of commas, semicolons, and sometimes unbalanced quotation marks that would have made it difficult to parse any other way.

Once the database was delimited, it was clear that there were going to be some problems handling the data. For example, many of the fields were well over 256 characters - the maximum limit for many conventional database fields. We finally decided to import the data into a relational database called MySQL, which allows large data fields called "blobs". Running under Linux, MySQL is very flexible and powerful.

There is a Perl interface, DBI.pm, which makes accessing the database simple from a Perl script. Essentially, an SQL statement is generated, based on data from an HTML form, and then it is run against the MySQL database, returning matching rows that the same script can handle and display as needed. All of the dynamically generated pages in this database are created this way. If you would like to see the scripts I used, please e-mail me at jury@plains.nodak.edu.