The Randomized Content Block XBlock only randomizes the selection of the children blocks and has unpredictable randomization of the order of the selected child blocks due to the usage of sets, which are unordered, for storing the selected blocks. This becomes apparent when all the available child blocks in a library are chosen for a Randomized Content Block, to randomize just the order of the child blocks and not just the selection of the blocks. The order of the selected blocks ends up being similar for multiple learners. This change modifies the XBlock to store the selected child blocks in a list, instead of a set, after randomly shuffling them. A new block structure transformer, ContentLibraryOrderTransformer has been added to transform the order of the selected children XBlocks to match the order of the selections made in the ContentLibraryTransformer. This ensures that same correct randomized order of blocks is also returned by the course blocks API.
246 lines
9.9 KiB
Python
246 lines
9.9 KiB
Python
"""
|
|
Content Library Transformer.
|
|
"""
|
|
|
|
|
|
import json
|
|
import logging
|
|
|
|
import six
|
|
from eventtracking import tracker
|
|
|
|
from lms.djangoapps.courseware.models import StudentModule
|
|
from openedx.core.djangoapps.content.block_structure.transformer import (
|
|
BlockStructureTransformer,
|
|
FilteringTransformerMixin
|
|
)
|
|
from track import contexts
|
|
from xmodule.library_content_module import LibraryContentModule
|
|
from xmodule.modulestore.django import modulestore
|
|
|
|
from ..utils import get_student_module_as_dict
|
|
|
|
logger = logging.getLogger(__name__)
|
|
|
|
|
|
class ContentLibraryTransformer(FilteringTransformerMixin, BlockStructureTransformer):
|
|
"""
|
|
A transformer that manipulates the block structure by removing all
|
|
blocks within a library_content module to which a user should not
|
|
have access.
|
|
|
|
Staff users are not to be exempted from library content pathways.
|
|
"""
|
|
WRITE_VERSION = 1
|
|
READ_VERSION = 1
|
|
|
|
@classmethod
|
|
def name(cls):
|
|
"""
|
|
Unique identifier for the transformer's class;
|
|
same identifier used in setup.py.
|
|
"""
|
|
return "library_content"
|
|
|
|
@classmethod
|
|
def collect(cls, block_structure):
|
|
"""
|
|
Collects any information that's necessary to execute this
|
|
transformer's transform method.
|
|
"""
|
|
block_structure.request_xblock_fields('mode')
|
|
block_structure.request_xblock_fields('max_count')
|
|
block_structure.request_xblock_fields('category')
|
|
store = modulestore()
|
|
|
|
# needed for analytics purposes
|
|
def summarize_block(usage_key):
|
|
""" Basic information about the given block """
|
|
orig_key, orig_version = store.get_block_original_usage(usage_key)
|
|
return {
|
|
"usage_key": six.text_type(usage_key),
|
|
"original_usage_key": six.text_type(orig_key) if orig_key else None,
|
|
"original_usage_version": six.text_type(orig_version) if orig_version else None,
|
|
}
|
|
|
|
# For each block check if block is library_content.
|
|
# If library_content add children array to content_library_children field
|
|
for block_key in block_structure.topological_traversal(
|
|
filter_func=lambda block_key: block_key.block_type == 'library_content',
|
|
yield_descendants_of_unyielded=True,
|
|
):
|
|
xblock = block_structure.get_xblock(block_key)
|
|
for child_key in xblock.children:
|
|
summary = summarize_block(child_key)
|
|
block_structure.set_transformer_block_field(child_key, cls, 'block_analytics_summary', summary)
|
|
|
|
def transform_block_filters(self, usage_info, block_structure):
|
|
all_library_children = set()
|
|
all_selected_children = set()
|
|
for block_key in block_structure:
|
|
if block_key.block_type != 'library_content':
|
|
continue
|
|
library_children = block_structure.get_children(block_key)
|
|
if library_children:
|
|
all_library_children.update(library_children)
|
|
selected = []
|
|
mode = block_structure.get_xblock_field(block_key, 'mode')
|
|
max_count = block_structure.get_xblock_field(block_key, 'max_count')
|
|
|
|
# Retrieve "selected" json from LMS MySQL database.
|
|
state_dict = get_student_module_as_dict(usage_info.user, usage_info.course_key, block_key)
|
|
for selected_block in state_dict.get('selected', []):
|
|
# Add all selected entries for this user for this
|
|
# library module to the selected list.
|
|
block_type, block_id = selected_block
|
|
usage_key = usage_info.course_key.make_usage_key(block_type, block_id)
|
|
if usage_key in library_children:
|
|
selected.append(selected_block)
|
|
|
|
# Update selected
|
|
previous_count = len(selected)
|
|
block_keys = LibraryContentModule.make_selection(selected, library_children, max_count, mode)
|
|
selected = block_keys['selected']
|
|
|
|
# Save back any changes
|
|
if any(block_keys[changed] for changed in ('invalid', 'overlimit', 'added')):
|
|
state_dict['selected'] = selected
|
|
StudentModule.save_state(
|
|
student=usage_info.user,
|
|
course_id=usage_info.course_key,
|
|
module_state_key=block_key,
|
|
defaults={
|
|
'state': json.dumps(state_dict),
|
|
},
|
|
)
|
|
|
|
# publish events for analytics
|
|
self._publish_events(
|
|
block_structure,
|
|
block_key,
|
|
previous_count,
|
|
max_count,
|
|
block_keys,
|
|
usage_info.user.id,
|
|
)
|
|
all_selected_children.update(usage_info.course_key.make_usage_key(s[0], s[1]) for s in selected)
|
|
|
|
def check_child_removal(block_key):
|
|
"""
|
|
Return True if selected block should be removed.
|
|
|
|
Block is removed if it is part of library_content, but has
|
|
not been selected for current user.
|
|
"""
|
|
if block_key not in all_library_children:
|
|
return False
|
|
if block_key in all_selected_children:
|
|
return False
|
|
return True
|
|
|
|
return [block_structure.create_removal_filter(check_child_removal)]
|
|
|
|
def _publish_events(self, block_structure, location, previous_count, max_count, block_keys, user_id):
|
|
"""
|
|
Helper method to publish events for analytics purposes
|
|
"""
|
|
|
|
def format_block_keys(keys):
|
|
"""
|
|
Helper function to format block keys
|
|
"""
|
|
json_result = []
|
|
for key in keys:
|
|
info = block_structure.get_transformer_block_field(
|
|
key, ContentLibraryTransformer, 'block_analytics_summary'
|
|
)
|
|
json_result.append(info)
|
|
return json_result
|
|
|
|
def publish_event(event_name, result, **kwargs):
|
|
"""
|
|
Helper function to publish an event for analytics purposes
|
|
"""
|
|
event_data = {
|
|
"location": six.text_type(location),
|
|
"previous_count": previous_count,
|
|
"result": result,
|
|
"max_count": max_count,
|
|
}
|
|
event_data.update(kwargs)
|
|
context = contexts.course_context_from_course_id(location.course_key)
|
|
if user_id:
|
|
context['user_id'] = user_id
|
|
full_event_name = "edx.librarycontentblock.content.{}".format(event_name)
|
|
with tracker.get_tracker().context(full_event_name, context):
|
|
tracker.emit(full_event_name, event_data)
|
|
|
|
LibraryContentModule.publish_selected_children_events(
|
|
block_keys,
|
|
format_block_keys,
|
|
publish_event,
|
|
)
|
|
|
|
|
|
class ContentLibraryOrderTransformer(BlockStructureTransformer):
|
|
"""
|
|
A transformer that manipulates the block structure by modifying the order of the
|
|
selected blocks within a library_content module to match the order of the selections
|
|
made by the ContentLibraryTransformer or the corresponding XBlock. So this transformer
|
|
requires the selections for the randomized content block to be already
|
|
made either by the ContentLibraryTransformer or the XBlock.
|
|
|
|
Staff users are *not* exempted from library content pathways.
|
|
"""
|
|
WRITE_VERSION = 1
|
|
READ_VERSION = 1
|
|
|
|
@classmethod
|
|
def name(cls):
|
|
"""
|
|
Unique identifier for the transformer's class;
|
|
same identifier used in setup.py
|
|
"""
|
|
return "library_content_randomize"
|
|
|
|
@classmethod
|
|
def collect(cls, block_structure):
|
|
"""
|
|
Collects any information that's necessary to execute this
|
|
transformer's transform method.
|
|
"""
|
|
# There is nothing to collect
|
|
pass # pylint:disable=unnecessary-pass
|
|
|
|
def transform(self, usage_info, block_structure):
|
|
"""
|
|
Transforms the order of the children of the randomized content block
|
|
to match the order of the selections made and stored in the XBlock 'selected' field.
|
|
"""
|
|
for block_key in block_structure:
|
|
if block_key.block_type != 'library_content':
|
|
continue
|
|
|
|
library_children = block_structure.get_children(block_key)
|
|
|
|
if library_children:
|
|
state_dict = get_student_module_as_dict(usage_info.user, usage_info.course_key, block_key)
|
|
current_children_blocks = set(block.block_id for block in library_children)
|
|
current_selected_blocks = set(item[1] for item in state_dict['selected'])
|
|
|
|
# As the selections should have already been made by the ContentLibraryTransformer,
|
|
# the current children of the library_content block should be the same as the stored
|
|
# selections. If they aren't, some other transformer that ran before this transformer
|
|
# has modified those blocks (for example, content gating may have affected this). So do not
|
|
# transform the order in that case.
|
|
if current_children_blocks != current_selected_blocks:
|
|
logger.info(
|
|
u'Mismatch between the children of %s in the stored state and the actual children for user %s. '
|
|
'Continuing without order transformation.',
|
|
str(block_key),
|
|
usage_info.user.username
|
|
)
|
|
else:
|
|
ordering_data = {block[1]: position for position, block in enumerate(state_dict['selected'])}
|
|
library_children.sort(key=lambda block, data=ordering_data: data[block.block_id])
|