Historically, backups of an activity using questions would include all
questions for the entire context, where any question had been used. This
could lead to very large backup files, and exacerbated problems with
question duplication.
This changes the process for annotating categories to include in the
backup to find specific categories, rather than whole contexts. It
includes 3 groups of categories:
1. Those that belong to the backed up activity's context.
2. Those that contain a question used directly in a question_reference
belonging to the activity.
3. Those that contain a question used via a question_set_reference
belonging to the activity.
It also adds in any parent categories that are not included by the
above, so we have a complete hierarchy.
For group 1, the backup will include all questions belonging to these
categories, as before. For groups 2 and 3, it will only include
questions for question bank entries that have been annotated by the
activity.
While MDL-43938 introduced badge importing, two tasks were left incomplete.
1) Creating the backup_import_badges setting for default import behaviour.
2) Preventing the inclusion of badges during activity duplication.
This commit implements them.
This commit adds two new levels to the restore course structure form:
subsection (a delegated section that belongs to a course module) and
subactivity (a course module in a subsection).
Restore form can only use information from the backup file. To allow
activities to know if they are inside a subsection, the backup now
incorporates an "insubsection" attribute. This attribute is used only
for the form display but not for the restore logic.
The patch adds two new levels to the backup form structure: subsection
(a delegated section which parent is a course module) and subactivity
(an activity inside a subsection). Those new elements are displayed as a
inner section below the parent activitiy.
For now the UI is the same as a regular section but future issues will
define how the UX should be. The new two levels have their own classes
to represents settings and have extra dependencies.
This commit adds all the logic to allow delegate sections backup and
restore. The backup and restore process is quite complex and it is not
designed to have task hierarchy. To solve the subsection problem, the
backup/restore planners do not include delegated sections at a course
level, but they are included when the activity with delegated section is
processed.
To allow restoing, the activity is responsible to store the
component/itemid mapping in the backup_structure_dbops. This way, when
the delegated section is restored (delegated sections are always
processed right after the parent activity) it can use the itemid
mapping.
This patch adds asynchronous backup and restore functionality
to Moodle. This is an optional feature and is not enabled by
default. It can be enabled by site administrators.
Asynchronous backup and restores are actioned by the Moodle
adhoc task API. The progress of backup and restores is
displayedin the Moodle UI. Users can also be sent a message
when a backup or restore operation completes via the
Moodle messaging API.
The unserialize() function will not always return an object
especially when as can happen the controller field of a
record in the backup_controllers table is empty.
This change ensures that in this case the script will
not cause a fatal error, but instead throw an error.
Adds 'Preparing page display' progress bars for user interface pages
if they take a long time to display.
Also adds changes where other parts of the backup progress timed out on
long backups. After this change, and MDL-41838, it is finally possible on
my dev server to successfully back up the 'XL' test course.
encode_backup_temp_info() and decode_backup_temp_info() have been
introduced to keep the info field encoding in one place.
Many locations used get_backup_ids_record() to obtain info, that
makes lots of calls to get_backup_ids_record() which can be slow.
We now complete all those inline by adding the info field to the query.
To reduce memory usage, all queries of that nature have been changed
to use get_recordset_*. gzcompress was introduced if available to minimize
traffic to/from the database and to decrease the memory required for caching.
The compression time is saved by the benefits in other places of having smaller data.
There were still some differences between the php code
used in the issue and the php code generated by the
xmldb editor for the template tables. This commit put
both 100% in sync.
When we import courses or duplicate activities, the deduplicating nature of
the file_storage system means that we don't actually have to include the
real files in the backup - just their metadata from the file table.
The restoration process needs to know not to expect files from the backup
phase so a flag is set in moodle_backup metadata.
Some modes of backup cannot be preloaded with general backup
settings because they may conflict with the definitions required/set
by the mode (for example, import backups are, by definition, without
users always, so we cannot apply one default in the opposite direction).
So, after the patch, current general_backup setting will be only applied
to general backups, import/hub modes won't be using any default and
automated backup defaults should be loaded here (TODO)
Mark Johnson