Add "incomplete database" concept

[aik099]

As Code-Inside code is updated we might end up in a situation (e.g. introduced by a #37 or #38), where the sync command will discover more information about the code, than before. Since the actual scanned code base is the same, then:

• the sync command won't rescan the code to pull missing data;
• the bc command won't notice new/changed information in the database.

TODO:

1. in a PHP/SQL migration script:
   1. figure out in which table the data is now incomplete (e.g. ClassMethods)
   2. if a FileId column is found in that table, then use its value
   3. otherwise trace back that table to the table with a FileId column (e.g. ClassMethods > Classes) and get the FileId value
   4. set 0 or -1 to the Files.Size column for files with the above-found IDs
2. next time the sync runs it will reparse these files, because of a filesize difference and add all missing info to the database
3. add the public KnowledgeBase::isIncomplete method, which would count records in the Files table that has Size=-1 or 0 and return true, when such records were found or false otherwise
4. the bc command would call isIncomplete method on both databases and if throw an exception asking ti resync