search

chicoreus / cco_poc

Proof of concept files for an implementation of an information model for complex natural science collections objects.
GNU General Public License v2.0
1 stars 3 forks source link

Pauls full #7

Open Inkimar opened 7 years ago

Inkimar commented 7 years ago

Hi Paul,

The following has been done:

  1. Using a Makefile for running maven ('mvn') and additional commands , instead of the up.sh-file 1.1 the standard way to do things 1.2 NB: you can create a MariaDB/MYSQL by sending arguments to a recipe in the makefile (see below) $ make create_db DB_USER=my_user DB_PSW=my_password

  2. tables.sql has been updated. 2.1 divided into additional 'changesets' - making the footprint of each 'changeset' smaller (could even be made even smaller) 2.2 removed the table-attribute 'ENGINE=InnoDB' and 'DEFAULT CHARSET=utf8;' 2.3 NB. ENGINE=myisam is still there

  3. exampledata.sql has been updated. 3.1 was lacking '-- liquibase formatted sql' and did not contain any 'changesets'

  4. additional_alter_statements.sql has been added 4.1 four 'alter table' due to problems with insert from the exampledata.sql - lacking 'default'-value 4.2 NB : Should be moved to tables.sql

  5. db.changelog-master.xml has been updated. 5.1 includes a reference to tables.sql + additional_alter_statements.sql + exampledata.sql

Test:

mysql> select count(*) from DATABASECHANGELOG;
+----------+
| count(*) |
+----------+
|      199 |
+----------+
1 row in set (0.00 sec)

ToDo, when it comes to 'Running with Liquibase'

  1. Add a recipe to the Makefile, for creating and dropping a postgreSQL-database
  2. is there any documentaion on functions.sql, optional.sql and triggers.sql - we are not using those right now
  3. Go through the changesets in tables.sql , making them even more coherent