|
1
|
INTRODUCTION:
|
|
2
|
This file describes how to use the sql scripts in the /metacat/test/tools directory
|
|
3
|
to create different versions of the metacat database schema. This is useful for testing
|
|
4
|
the metacat auto-upgrade functionality.
|
|
5
|
|
|
6
|
These scripts were extracted from the ddl for metacat using code tags. The scripts
|
|
7
|
were retrieved back to version 1.4. For earlier scripts, you will need to go to the
|
|
8
|
SVN repository.
|
|
9
|
|
|
10
|
Note that database scripts currently only support postgresql and oracle databases.
|
|
11
|
|
|
12
|
PURPOSE AND NAMING CONVENTIONS:
|
|
13
|
You will find three major categories of sql scripts in this directory. Here we list
|
|
14
|
their usage and naming conventions.
|
|
15
|
|
|
16
|
1) xmltables scripts - these are used to create a database schema for a particluar
|
|
17
|
version of metacat. The naming format is:
|
|
18
|
|
|
19
|
<version_number>_xmltables-<database_type>.sql
|
|
20
|
|
|
21
|
2) loaddtdschema scripts - these are use to populate eml schema information into
|
|
22
|
the database. The nameing format is:
|
|
23
|
|
|
24
|
<version_number>_loaddtdschema.sql
|
|
25
|
|
|
26
|
Note that there is no qualification for database type. Currently, all database types
|
|
27
|
use the same eml schema scripts. If this should change in the future, we would need
|
|
28
|
to add the database type qualification.
|
|
29
|
|
|
30
|
3) cleanall scripts - these clear out all data and metacat schema information. This
|
|
31
|
should remove all tables, sequences and triggers for the given version. The naming
|
|
32
|
format is:
|
|
33
|
|
|
34
|
<version_number>_cleanall-<database_type>.sql
|
|
35
|
|
|
36
|
HOW TO RUN SCRIPTS:
|
|
37
|
|
|
38
|
If you are testing against a postgresql database, you should start with an empty
|
|
39
|
database. You will run the xmltables script first and then the loaddtdschema script.
|
|
40
|
Change to the postgres user and run the script by typing:
|
|
41
|
|
|
42
|
psql -f <file_name> <database_name>
|
|
43
|
|
|
44
|
So for instance, if you have an empty postgres database named "metacat" and you wanted
|
|
45
|
to install the 1.4 metacat schema, you would type:
|
|
46
|
|
|
47
|
sudo su - postgres
|
|
48
|
psql -f 1.4_xmltables-postgres.sql
|
|
49
|
psql -f 1.4_loaddtdschema.sql
|
|
50
|
|
|
51
|
If you are testing against an oracle database, you should log into the database
|
|
52
|
using sqlplus. A typical scenario is that you have downloaded and installed the
|
|
53
|
development XE version of oracle. You may need to 'locate sqlplus' to find the sqlplus
|
|
54
|
program.
|
|
55
|
|
|
56
|
the command to log into sqlplus looks like:
|
|
57
|
|
|
58
|
sqlplus <db_user_name>@<db_sid>
|
|
59
|
|
|
60
|
Once you are at the sqlplus command line, you can run a sript by referencing it with a
|
|
61
|
"@" and the script name. So if you have an empty oracle XE development license database
|
|
62
|
with a "metacat" user, that you want to install with the 1.4 metacat schema, you would
|
|
63
|
run the scripts by typing:
|
|
64
|
|
|
65
|
sqlplus metacat@XE
|
|
66
|
SQL> @1.4_xmltables-oracle.sql
|
|
67
|
SQL> @1.4_loaddtdschema.sql
|
|
68
|
|
|
69
|
|
|
70
|
CLEANING THE DATABASE:
|
|
71
|
Whatever version your database schema is at, you can get it back to an empty state
|
|
72
|
by running the script named:
|
|
73
|
|
|
74
|
<version_number>_cleanall-<database_type>.sql
|
|
75
|
|
|
76
|
Follow the same steps shown above to run the script depending on which database
|
|
77
|
you have installed.
|
|
78
|
|
|
79
|
|
|
80
|
|
|
81
|
|
|
82
|
|
|
83
|
|
|
84
|
|
|
85
|
|
|
86
|
|
|
87
|
|
|
88
|
|
|
89
|
|