Include classes that lack schedules + split off IAP/summer #100
Conversation
This still excludes classes that are not listed in the current course catalog, since those are filtered out by catalog.py.
We check that a class is offered in the fall/spring in fireroad.py, and check that it is offered during this academic year in catalog.py. This does have the effect of excluding IAP when the term is set to spring, until we properly split IAP and spring.
This allows capturing both IAP and spring, or summer and fall, in the same latestTerm.json file; the functionality is currently unchanged as only the "semester" term (fall/spring) is used.
Since no other fields of latestTerm.json are ever read in the frontend (the actual term file is used for start/end dates, Monday schedule, etc.).
The pre-semester (IAP or summer) is currently output to latestTermPreSemester, though (it needs to be renamed to i25, etc.). The scrapers also don't look at scheduleFall, scheduleIAP, scheduleSpring.
This does mean that the pre-semester file -- currently i25.json -- needs to be added to the .gitignore (and that this needs to be updated each semester), but it avoids special-casing the latest pre-semester in the code (i.e., saying that the penultimate term should be read from latestPreSemester.json rather than the expected file name).
psvenk
changed the title
| public/latest.json | ||
| public/i25.json |
There was a problem hiding this comment.
Is there no other way of making sure we don't commit presemester? I feel like this will start getting bloated after a few semesters
There was a problem hiding this comment.
The main other way that I can think of is to put the latest pre-semester in public/latestPreSemester.json and then to add a special case to src/lib/TermSwitcher.tsx so that the urlName corresponding to the latest pre-semester will read from the file public/latestPreSemester.json.
This would break the correspondence between urlNames and JSON file names that we have now; e.g., https://hydrant.mit.edu/?t=latest loads Spring 2025 while https://hydrant.mit.edu/?t=s25 does not, since the file is latest.json and not yet s25.json.
Alternatively, we could make public/i25.json a symlink to public/latestPreSemester.json, but then the maintenance task on each semester is changed from editing .gitignore to modifying a symlink (which is IMO marginally worse).
Bloat shouldn't become an issue here, since it's only the latest pre-semester that needs to be ignored; when the fall 2025 listings are out, public/i25.json should be committed (along with public/s25.json) and it will be public/m25.json that will need to be ignored.
|
Also, classes with pre-8am times show up in the class listing now but with an "unknown schedule," relevant for issue #86 |
|
This has caused some "not offered" classes to show up: #110 |
Fixes #27 and #7. In the scraper, replace the check that a class has a schedule with a check that a class is offered in the current term (with the fall/IAP/spring check done in
catalog.pywhile the academic year check is done infireroad.py), and add proper support for IAP and summer terms.Since this explicitly checks for classes to be offered in the current term, it excludes IAP classes when the current term is set to spring (and, in principle, summer classes when the current term is set to fall); this means that merging this PR in its current state would remove IAP classes from Hydrant. I plan on resolving #7 within this PR before marking it as ready for review.