shafqatfarhan/problem-builder

An XBlock for posing questions to students

Problem Builder and Step Builder

This repository provides two XBlocks: Problem Builder and Step Builder.

Both blocks allow to create questions of various types. They can be used to simulate the workflow of real-life mentoring, within an edX course.

Supported features include:

• Free-form answers (textarea) which can be shared accross different XBlock instances (for example, to allow a student to review and edit an answer they gave before).
• Self-assessment MCQs (multiple choice questions), to display predetermined feedback to a student based on his choices in the self-assessment. Supports rating scales and arbitrary answers.
• MRQs (Multiple Response Questions), a type of multiple choice question that allows the student to select more than one choice.
• Answer recaps that display a read-only summary of a user's answer to a free-form question asked earlier in the course.
• Progression tracking, to require that the student has completed a particular step before allowing them to complete the next step. Provides a link to the next step to the student.
• Tables, which allow to present answers from the student to free-form answers in a concise way. Supports custom headers.
• Dashboards, for displaying a summary of the student's answers to multiple choice questions. Details

The following screenshot shows an example of a Problem Builder block containing a free-form question, two MCQs and one MRQ:

Installation

You can install Problem Builder from PyPI using this command:

pip install xblock-problem-builder

For full details, see "Open edX Installation", below.

Usage

See Usage Instructions

Workbench installation and settings

For developers, you can install this XBlock into the XBlock SDK workbench's virtualenv by running the following command from the problem builder repo root:

pip install -r requirements.txt

In the main XBlock repository, create the following configuration file in workbench/settings_pb.py in the XBlock repository:

from settings import *

INSTALLED_APPS += ('problem_builder',)
DATABASES['default']['NAME'] = 'workbench.sqlite'

Because this XBlock uses a Django model, you need to sync the database before starting the workbench. Run this from the XBlock repository root:

$ ./manage.py syncdb --settings=workbench.settings_pb

Running the workbench

$ ./manage.py runserver 8000 --settings=workbench.settings_pb

Access it at http://localhost:8000/.

Running tests

First, make sure the XBlock SDK (Workbench) is installed in the same virtual environment as xblock-problem-builder.

From the xblock-problem-builder repository root, run the tests with the following command:

$ ./run_tests.py

If you want to run only the integration or the unit tests, append the directory to the command. You can also run separate modules in this manner.

$ ./run_tests.py problem_builder/tests/unit

Extracting Translatable Strings

To extract/update strings for translation, you will need i18n_tools:

pip install git+https://github.com/edx/i18n-tools.git#egg=i18n_tools

To extract strings, use i18n_tool. We added two dummy translations for testing. Easily set your openedx instance to use one of them (eo and rtl) and check.
If you want to add a new language:

1. Clone en directory to problem_builder/locale/<lang_code>/ for example: problem_builder/locale/fa_IR/
2. Make neccessary changes to translation files headers. Make sure you have proper Language and Plural-Forms lines.
3. Edit the contents of .po files located in problem_builder/locale/<lang_code>/LC_MESSAGES as you wish
4. When you finished your modification process, re-compile the translation messages manually by executing the following command in the root of xblock:

i18n_tool generate

Adding custom scenarios to the workbench

Within the xblock-problem-builder repository, create the templates/xml and add XML scenarios to it - all files with the *.xml extension will be automatically loaded by the workbench:

$ mkdir templates/xml
$ cat > templates/xml/my_pb_scenario.xml

Restart the workbench to take the new scenarios into account.

Upgrading from Version 1

To upgrade a course from xblock-mentoring ("v1") to xblock-problem-builder ("v2"), run the following command on a system with edx-platform, xblock-mentoring, and xblock-problem-builder installed:

$ SERVICE_VARIANT=cms DJANGO_SETTINGS_MODULE="cms.envs.devstack" python -m problem_builder.v1.upgrade "Org/Course/Run"

Where "Org/Course/Run" is replaced with the ID of the course to upgrade.

Open edX Installation

Problem Builder releases are tagged with a version number, e.g. v2.6.0, v2.6.5. We recommend installing the most recently tagged version, with the exception of the following compatibility issues:

• edx-platform version open-release/eucalyptus.2 and earlier must use ≤v2.6.0. See PR 128 for details.
• edx-platform version named-release/dogwood.3 and earlier must use v2.0.0.

The edx-platform master branch will generally always be compatible with the most recent Problem Builder tag. See the EDXAPP_PRIVATE_REQUIREMENTS setting for the version currently installed on edx.org.

To install new versions of Problem Builder (v3.1.3+), use pip install xblock-problem-builder or specify a version using e.g. pip install xblock-problem-builder==3.1.3. To do this on Open edX could look like:

$ sudo -Hu edxapp bash
edxapp $ cd && . edxapp_env  && . ./venvs/edxapp/bin/activate && cd edx-platform/
edxapp $ pip install xblock-problem-builder
edxapp $ ./manage.py lms migrate --settings=aws  # or openstack, as appropriate

Then, restart the edxapp services:

$ sudo /edx/bin/supervisorctl restart edxapp:
$ sudo /edx/bin/supervisorctl restart edxapp_workers:

To install old verions of Problem Builder (< v3.1.3) on an Open edX installation, choose the tag you wish to install, follow the above instructions but instead of the pip install xblock-problem-builder command, use:

TAG='v2.6.5' pip install "git+https://github.com/open-craft/problem-builder.git@$TAG#egg=xblock-problem-builder==$TAG"

Note that Problem Builder requires xblock-utils. If you are installing it into a virtualenv used by edx-platform, xblock-utils should already be installed. But if you are installing it into another virtualenv, you may need to first install xblock-utils manually (recent versions of it are not available on PyPI so will not be automatically installed).

See Usage Instructions for how to enable in Studio.

Publishing to PyPI

Whenever we tag a new version, e.g. v3.1.3 and push it to GitHub, CircleCI will build it and deploy it to PyPI automatically. For details on how this works, see this pull request.

License

This XBlock is available under the GNU Affero General Public License (AGPLv3).