Sample data

[William Grant]

We should have a set of sample data that can be loaded into the IVLE DB for
testing or development.

[Matt Giuca (mgiuca)]

Just going to dump the DB and create a hand-modified SQL script to load the sample data. Also write a script bin/ivle-load-sample-data, which automatically populates the DB.

[William Grant (wgrant)]

'pg_dump --schema=public --disable-triggers -a -D -O -d ivle' will do a dump of the data.

[Matt Giuca (mgiuca)]

Expanding all the options to pg_dump, since my version doesn't support some of the short forms:
pg_dump --schema=public --disable-triggers --data-only --column-inserts --inserts --no-owner ivle

[Matt Giuca (mgiuca)]

- Unmount the jails
- Test if there is any data, if not fail unless --delete-all is given
- Drop and recreate the database
- Populate with the scheme
- Populate with the sample data
- Run the filesystem refresh script
- Check in sample files into the Subversion repository for some users

We won't write to the sample user's workspaces, we'll just leave them unchecked out.

[Matt Giuca (mgiuca)]

Almost done. Just need to create some sample user files and have them automatically checked into the repository by ivle-loadsampledata. This is difficult because you need to get the Subversion path, make a checkout, copy the files, then commit. One tricky thing is the date stamps on the commits. How to get submissions done before the due dates if the due dates are fixed, but the commits happen at the current time?

Also need to populate with more sample data, such as submissions and worksheets, but this can be done incrementally (it isn't blocking this bug from being resolved).

[Matt Giuca (mgiuca)]

Almost almost done. Can now check files into a student repository (but we still need the capability of copying files into the repo, rather than having them hard-coded in ivle-loadsampledata).

Also need at least one worksheet, to encompass the exercise we put in today.

[Matt Giuca (mgiuca)]

There is now what I would deem to be a sufficient amount of sample data, as well as sample data framework and documentation.