fix overzealous turndown filter on quiz rule

[gumaerc]

Pre-Flight checklist

• [ ] Testing
  • [ ] Code is tested
  • [ ] Changes have been manually tested

What are the relevant tickets?

Closes https://github.com/mitodl/ocw-to-hugo/issues/493

What's this PR do?

In https://github.com/mitodl/ocw-to-hugo/pull/418 we added support for parsing multiple choice type quizzes embedded in legacy OCW courses. One of the turndown rules that does this, multiple_choice_question, was filtering on:

        node.getAttribute("class") === "problem_question" &&
        node.getElementsByClassName("problem_radio_input")

The issue with this is that even if getElementsByClassName doesn't find anything, it still returns an object. This means that HTML that shouldn't have been matching that rule was being matched and processed as if it were an multiple choice question. I corrected this by changing it to:

        node.getAttribute("class") === "problem_question" &&
        node.getElementsByClassName("problem_radio_input").length > 0

A try / catch statement with Winston file logging was also added to the html2markdown function so any errors converting html to markdown are logged.

How should this be manually tested?

• Read the readme if you've never used ocw-to-hugo before and make sure you're set up to download courses from S3
• Place the following in private/courses.json:

  {
  "courses": [
      "6-042j-mathematics-for-computer-science-spring-2015",
      "16-90-computational-methods-in-aerospace-engineering-spring-2014",
      "15-071-the-analytics-edge-spring-2017",
      "14-01sc-principles-of-microeconomics-fall-2011"
  ]
  }

• Run node . -i private/input -o private/output -c private/courses.json --download --rm
• Inspect the following pages and make sure there is no raw HTML:

• 6-042j-mathematics-for-computer-science-spring-2015/content/pages/proofs/tp2-2/vertical-a28e46f96fa1.md
• 6-042j-mathematics-for-computer-science-spring-2015/content/pages/proofs/tp4-3/set-theory-axioms-optional-0.md
• 16-90-computational-methods-in-aerospace-engineering-spring-2014/content/pages/numerical-integration-of-ordinary-differential-equations/discretizing-odes/_index.md
• 15-071-the-analytics-edge-spring-2017/content/pages/logistic-regression/modeling-the-expert-an-introduction-to-logistic-regression/quick-question-152.md
• 14-01sc-principles-of-microeconomics-fall-2011/content/pages/unit-6-topics-in-intermediate-microeconomics/international-trade.md

[codecov-commenter]

Codecov Report

Merging #497 (d0148c4) into master (0e394ac) will decrease coverage by 0.13%. The diff coverage is 80.00%.

@@            Coverage Diff             @@
##           master     #497      +/-   ##
==========================================
- Coverage   96.47%   96.33%   -0.14%     
==========================================
  Files          18       18              
  Lines        2126     2130       +4     
==========================================
+ Hits         2051     2052       +1     
- Misses         75       78       +3     

Impacted Files	Coverage Δ
src/turndown.js	92.30% <80.00%> (-0.25%)	:arrow_down:
src/markdown_generators.js	97.22% <0.00%> (-1.39%)	:arrow_down:

---

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 0e394ac...d0148c4. Read the comment docs.