TODO

Disclaimer: this list of notes/TODOs belongs in an issue tracker if
there are ever >2 people working on this project.

[Critical features]
[1] Parse questions to enums.
[2] Aggregated search API/page. (depends: [1])
[3] Teacher model
[4] Create minified bundle of JS resources with un-minified (or source-mapped) debug versions
[6] APIs for single course/class/teacher
[7] UIs for single course/class/teacher [7]
[8] Logging/Tracing wrapper(s) for requests. Make sure they don't block actual responses.

[Possible performance improvements] (don't build any before [8])
* Relational-DB solution for searching
* Add a PushStrategy to Jetty for HTTP/2 push (depends: [4])


[Open bugs]
* Questions not parsed to enums
* Term strings not parsed to term codes (less important for now)
* Invalid pages (not logged in, load failures, etc.) shouldn't be saved/cached.


[Other notes]
* Need to parse TceClass page contents into meaningful data, including:
** Normalization of different questions
** Normalization of different answer ranges, thanks to 2011-12

* May need to pull in a solution for indexing/searching.
** Primary source of truth is HTML files on disk. Parsed format should be
   regenerated on the fly at server start time.
** Data is highly relational. Relational DB might be best.