CREATING A VERSION OF THE APPLICATION FOR A NEW DATABASE

The app assumes that the database will be in Prolog format. Name entries in the Regulus grammar are automatically extracted from the database by a script. There is an example of a correctly formatted database in the file Prolog/database_im2.pl.

This document should contain all the information you need if you want to create a version of the system using a new database. Specifically, it describes 1) the format of the database, 2) which other files need to be edited or created.

FORMAT OF DATABASE

The database needs to define a module called 'database' exporting the following predicates:

meeting/7 person/6 attends/2 location/5

NOTE: THE EXACT FORM OF THE FIELDS CAN BE IMPORTANT, SINCE SOME OF THEM ARE USED TO EXTRACT ENTRIES FOR THE REGULUS GRAMMAR.

FIELDS CONTAINING INFORMATION THAT WILL APPEAR IN THE LEXICON MUST BE PROLOG ATOMS WHOSE PRINT NAMES CONTAIN ONLY LOWER-CASE LETTERS AND UNDERSCORES.

In detail, the format of the database relations is as follows:

meeting/7 Pattern: meeting(ID, Day, Month, Year, StartTime, EndTime, LocID). Example: meeting(meeting_1, 29, 5, 2007, 9:30, 12:0, nikos_room_1).

• ID is an arbitrary Prolog atom
• Day is a number (1-31) representing day of month
• Month is a number (1-12) representing a month
• Year is a number (2000-2100) representing a year
• StartTime is two numbers (0-23, 0-59) separated by a colon, representing a time
• EndTime is two numbers (0-23, 0-59) separated by a colon, representing a time
• LocID is an arbitrary Prolog atom.

person/6 Pattern: person(ID, FirstName, LastName, Affiliation, Phone, Email). Example: person(pierrette_bouillon, pierrette, bouillon, geneva, "+41 22 379 8679", "Pierrette.Bouillon@issco.unige.ch").

• ID is an arbitrary Prolog atom
• FirstName is a Prolog atom whose print name may only contain letters and underscores.
• LastName is a Prolog atom whose print name may only contain letters and underscores.
• Affiliation is a Prolog atom whose print name may only contain letters and underscores.
• Phone is an arbitrary Prolog string
• Email is an arbitrary Prolog string

attends/2 Pattern: attends(PersonID, MeetingID). Example: attends(pierrette_bouillon, meeting_1).

• PersonID is a Prolog atom that appears in the first field of some person record
• MeetingID is a Prolog atom that appears in the first field of some meeting record

location/5 Pattern: location(ID, Name, Country, City, Organisation). Example: location(nikos_room_1, nikos_s_room, switzerland, geneva, geneva_university).

• ID is an arbitrary Prolog atom
• Name is a Prolog atom whose print name may only contain letters and underscores.
• Country is a Prolog atom whose print name may only contain letters and underscores.
• City is a Prolog atom whose print name may only contain letters and underscores.
• Organisation is a Prolog atom whose print name may only contain letters and underscores.

DECLARING THE NEW DATABASE

You need to create or edit the following additional files:

1. A top-level config file. Use the one in $REGULUS/Examples/Calendar/scripts/calendar_im2.cfg as a model. Edit all the fields that include the string 'im2' and replace them with corresponding values.
2. scripts/Makefile: add make targets corresponding to the ones for im2. This will include adding a file corresponding to scripts/build_name_lexicon_im2, to create the new name lexicon.
3. scripts directory: create files corresponding to all those containing the string 'im2'. Modify the content accordingly.