Aug 18, 2014
The SPARK platform requires a few conventions that the data model should adhere to.
The main requirement is that the primary key is an int or long identity field with a name 'Id'.
If the existing database is in compliance you can immediately start the solution. In other words, there is no difference whether this is a new database or an existing one (as long as the conventions are followed).
Currently there is no separate document explaining how to setup a solution with an existing database. We are planning to make setting up the solution somewhat easier in the next release, possibly with a separate guide.
Hope this helps.
Mar 25, 2015
Isn't there also a requirement that the existing table structures DO NOT end in the letter "s"?. It seems that when I ran Spark against my existing database I got naming collisions between the Domain and the Repository because of this situation. Can you please confirm this? I'm interested in starting a new database on my project but want to conform to Spark's expected database design patterns.
Mar 26, 2015
It is not so much that names cannot end with the letter 's' but the PRO code generator assumes that tables names are singular. It then attempts to pluralize these names and uses these are part of the Repository. If a name is the same whether it is singular or plural (like Deer) then that may cause a conflict.
If there is a conflict the best approach (right now) is to rename the table and this should solve it. The next release of the PRO code generator will have the option to specify how individual names are pluralized.
BTW: internally the PRO code generator uses the same Pluralize method that is used by the Entity Framework. It is very good, but not perfect.
Hope this helps.