-
…by update. The behavior of update is to create a row if not exist or update it if exists. But this behavior introduce a bug I didn't notice until now. By calling only update_row function the trigger belonging to create function are never applied. So the core was updated to distinguish the creation or update and then apply the correct trigger.
-
Add a comment to replace it, because the trigger are not managed here.
-
[~] in function insert_or_update() , add print debug info on failure
-
[+] db_get_Int(db, clause, binds). [+] db_get_ids_by_full_clause(db, table_name, id_col, full_clause, binds).
-
…t the DB_id other than the first "id"
-
…xt with reference on table_name, column to edit and id of the row
-
…e these types are to big. The good solution is to have a diff function on these type to show the difference
-
[+] add get_db_ids_by_full_clause which allow to make query on any table, set the column id, give the full clause. Return a list of DB_id. [+] add get_all_by_full_clause which allow to make query on any table, set the column id, give the full clause and provide the extractor for this column data type.
-
…rent type anymore because it was useless.
-
…example this help to get list of foreign keys
-
… the database. This is useful because the name is predictable and the CKEDITOR can be destroyed easily
-
move css file in the new path (hayamiki/css/ or hayamiki/js/)