Python processing pipeline

The parser/ package, loaded as a wheel in Pyodide. Related: Pyodide, Schedule generation.

parser/main.py:

• create_time_table(campus, year, time_table_json, subject_json, batch, electives_subject_codes)
• compare_timetables(tt1, tt2)
• create_and_compare_timetable([params1, params2])

flowchart TD
  CT[create_time_table]
  CT --> R{campus / year}
  R -->|62 1| F1[time_table_creator]
  R -->|62 2+| F2[time_table_creator_v2]
  R -->|128 1| F3[bando_year1]
  R -->|128 2+| F4[banado]
  R -->|BCA 1| F5[creator_year1]
  R -->|BCA 2+| F6[creator]

Campus modules:

• parser/modules/tt_parsers/sector_62/creator.py
• parser/modules/tt_parsers/sector_128/creator.py
• parser/modules/tt_parsers/BCA/creator.py

flowchart TD
  Days[for each day] --> Slots[for each timeslot]
  Slots --> Classes[for each activity string]
  Classes --> Code[subject_extractor]
  Classes --> Batches[batch_extractor + parse_batch_numbers]
  Batches --> Elective{is_elective?}
  Elective -->|core| Inc{is_batch_included?}
  Elective -->|elective year 1 heuristic| Inc
  Inc -->|yes| Row[append day, time, name, type, location]
  Row --> Out[fold into YourTietable]

Year 2+ uses is_enrolled_subject(code, electives_subject_codes) instead of the year-1 elective heuristic.

parser/utils/batch.py parse_batch_numbers:

BCA and 128 have extra extractors in their utils.py.

Example: TA4(CI121)-TS10/APR

flowchart LR
  S["TA4(CI121)-TS10/APR"] --> Type[T]
  S --> Batch[A4]
  S --> Code[CI121]
  S --> Loc[TS10]
  S --> Fac[APR]

subject_extractor takes the parenthesis (or dash fallback). location_extractor takes the room after -. Type is the leading L/T/P.

Year 1 is_elective treats some batch patterns as electives so they still attach when the batch matches. Year 2+ only keeps subjects whose code is in the form list.

subject_name_extractor matches Code or several slices of Full Code.

parser/utils/time.py:

• process_day: MON → Monday
• process_timeslot: 9-9.50 or 10:00-10:50 AM → 09:00, 10:00 (24h). Practicals can span two hours.

parser/modules/compare_tt/compare.py:

flowchart TD
  E[_expand_timetable_to_hourly]
  E --> Hours[08:00-17:00 each day]
  Hours --> Free[both empty to common_free_slots]
  Hours --> Same[same name type location to classes_together]

Multi-hour blocks split into 1-hour keys first.

flowchart LR
  toPy[pyodide.toPy] --> Fn[create_time_table]
  Fn --> toJs[result.toJs]

Home JSON-clones the result so Proxies do not survive into React state.

Per student this is O(days × slots × classes on that campus JSON). Fine in WASM. The expensive part is the first Pyodide download, not the loop.