OLX Cleaner

This library aims to perform two functions:

• Parse the XML code for an edX course, loading it into python objects
• Validate the objects for errors

Based on this, two scripts are provided that leverage the library:

• edx-cleaner constructs an error report, course tree and course statistics
• edx-reporter constructs a LaTeX file representation of the course structure

Version 0.1.3

Copyright (C) 2018-2019 Jolyon Bloomfield

Links

• Error Listing
• Wishlist
• Vision
• Changelog
• License

Installation

This package may be installed from PYPI using pip install olxcleaner. It requires python 3.6 or later.

Repository Installation (advanced)

Clone this repository, and set up a virtual environment for python 3.6 or later. Run pip install -r requirements.txt to install the libraries, followed by pytest to ensure that all tests are passing as expected.

edx-cleaner Usage

Used to validate OLX (edX XML) code. This is a very light wrapper around the olxcleaner library, but exposes all of the functionality thereof.

Basic usage: run edx-cleaner in the directory of the course you want to validate.

Command-line options:

edx-cleaner [-h] 
            [-c COURSE]
            [-p {1,2,3,4,5,6,7,8}] 
            [-t TREE] [-l {0,1,2,3,4}]
            [-q] [-e] [-s] [-S]
            [-f {0,1,2,3,4}]
            [-i IGNORE [IGNORE ...]]

• -h: Display help.
• -c: Specify the course file to analyze. If not specified, looks for course.xml in the current directory. If given a directory, looks for course.xml in that directory.
• -p: Specify the validation level you wish analyze the course at:
  • 1: Load the course
  • 2: Load the policy and grading policy
  • 3: Validate url_names
  • 4: Merge policy data with course, ensuring that all references are valid
  • 5: Validate the grading policy
  • 6: Have every object validate itself
  • 7: Parse the course for global errors
  • 8: Parse the course for detailed global errors (default)
• -t TREE: Specify a file to output the tree structure to.
• -l: Specify the depth level to output the tree structure to. Only used if the -t option is set. 0 = Course, 1 = Chapter, 2 = Sequential, 3 = Vertical, 4 = Content.
• -q: Quiet mode. Does not output anything to the screen.
• -e: Suppress error listing. Implied by -q.
• -s: Suppress summary of errors. Implied by -q.
• -S: Display course statistics (off by default). Overridden by -q.
• -f: Select the error level at which to exit with an error code. 0 = DEBUG, 1 = INFO, 2 = WARNING, 3 = ERROR (default), 4 = NEVER. Exit code is set to 1 if an error at the specified level or higher is present.
• -i: Specify a space-separated list of error names to ignore. See Error Listing.

edx-reporter Usage

The olxcleaner library includes modules that parse a course into python objects. This can be useful if you want to scan a course to generate a report. We exploit this in edx-reporter to generate a LaTeX report of course structure.

Basic usage: run edx-reporter in the directory of the course you want to generate a report about.

Command-line options:

edx-reporter.py [-h] 
                [-c COURSE]
                [-u]
                [> latexfile.tex]

• -h: Display help.
• -c: Specify the course file to analyze. If not specified, looks for course.xml in the current directory. If given a directory, looks for course.xml in that directory.
• -u: Include url_names for verticals.
• > latexfile.tex: Output the report to a file.

If you get an error like Character cannot be encoded into LaTeX: U+FEFF - `', then you have some bad unicode in your display_name entries. Look through the LaTeX output for {\bfseries ?}, which is what that character is converted into.

Once you have generated a latex file, you can compile it into a PDF file by running pdflatex latexfile.tex. Note that the latex file can be modified with any text editor; its format should be self-explanatory.

Library usage

The workhorse of the library is olxcleaner.validate, which validates a course in a number of steps.

olxcleaner.validate(filename, steps=8, ignore=None)

• filename: Pass in either the course directory or the path of course.xml for the course you wish to validate.
• steps: Choose how many validation steps you wish to perform:
  • 1: Load the course
  • 2: Load the policy and grading policy
  • 3: Validate url_names
  • 4: Merge policy data with course, ensuring that all references are valid
  • 5: Validate the grading policy
  • 6: Have every object validate itself
  • 7: Parse the course for global errors
  • 8: Parse the course for global errors that may be time-consuming to detect
• ignore: A list of error names to ignore

Returns EdxCourse, ErrorStore, url_names (dictionary {'url_name': EdxObject}, or None if steps < 3)

See examples of how to use olxcleaner.validate and the objects it returns in olxcleaner.entries.