JIIT Time Table Website

A Nextjs and Pyodide website to simplify JIIT Timetables

Source 15 Live site

timetable data format reference

On-disk JSON for classes, calendars, and exams. Related: Python pipeline, Types.

Directories#

JSON lives in website/data/ (API also checks ../data). It is not under public/data/.

Path Contents
website/data/time-table/{year}/{SEM}/{campus}.json Year-keyed timetable + subjects
website/data/calender/{yy1yy2}.json Academic events
website/data/exam/{year}/{SEM}/T*.json Exam rows
website/public/parser/*.whl Parser wheel (not JSON)

Campus files: 62, 128, BCA. ODD26 currently has 62 and BCA; EVEN26 has all three. Default semester constant: ODD26.

HTTP:

http
GET /api/time-table
GET /api/time-table/{semester}/{batch}
GET /api/academic-calendar
GET /api/academic-calendar/{year}
GET /api/exam-schedule
GET /api/exam-schedule/{semester}
GET /api/mess-menu

batch here is the campus file stem (62), not a student batch like A6. GET /api/time-table/EVEN26/62 returns the whole campus file below.

Timetable file#

One JSON object per campus. Top-level keys are year numbers as strings. B.Tech files use "1" through "4". BCA stops at "3". Each year has two siblings: timetable (the grid) and subjects (the catalog for that year).

EVEN26 Sector 62 year 1, trimmed to one morning. This is the real shape, not a made-up schema:

json
{
  "1": {
    "timetable": {
      "MON": {
        "9-9.50": [
          "LB3,B4(CI121)-CR325/ANP",
          "TA4(CI121)-TS10/APR",
          "TG2(CI121)-F7/NIY"
        ],
        "10-10.50": [
          "LA1,A2(PH211)-G1/ANU",
          "LA3,A4(PH211)-FF1/RKD"
        ],
        "12-12.50": ["LUNCH"]
      },
      "TUES": {},
      "WED": {},
      "THUR": {},
      "FRI": {},
      "SAT": {}
    },
    "subjects": [
      {
        "Code": "CI111",
        "Full Code": "",
        "Subject": "Software Development Fundamentals-"
      },
      {
        "Code": "PH211",
        "Full Code": "",
        "Subject": "Physics-2"
      },
      {
        "Code": "PH271",
        "Full Code": "",
        "Subject": "Physics Lab-2"
      }
    ]
  },
  "2": {
    "timetable": {},
    "subjects": [
      {
        "Code": "MA223",
        "Full Code": "25B12MA223",
        "Subject": "Time Series Analysis and Forecasting"
      }
    ]
  }
}

Days in the file are MON, TUES, WED, THUR, FRI, SAT. Not every year has every day. Year 3 EVEN26/62 has no TUES. Year 4 has no SAT.

Slot keys are the Excel headings, written the way the creator dumped them: 9-9.50, 10-10.50, …, 4-4.50. Year 2+ sometimes adds a 5pm slot. A slot value is always an array of strings. One string is one overlapping class in that hour. LUNCH is a real entry, not a hole in the object.

Year 1 EVEN26/62 has 12 subjects. Year 2 has 38. Year 1 Full Code is often empty; the parser then matches on Code or a slice of whatever sits in the activity string. Year 2 fills Full Code with the catalog id (25B12MA223).

TypeScript for the file (not exported as one type; this is the on-disk contract):

ts
type CampusFile = {
  [year: string]: {
    timetable: {
      [day: string]: {
        [slot: string]: string[];
      };
    };
    subjects: Array<{
      Code: string;
      "Full Code": string;
      Subject: string;
    }>;
  };
};

Activity strings#

Each class cell is a packed token, not a JSON object. From EVEN26/62 Monday 9am:

text
TA4(CI121)-TS10/APR
LB3,B4(CI121)-CR325/ANP
LA1,A2(PH211)-G1/ANU
PA5(PH271)-PL1/MKC/MTR
LUNCH

TA4(CI121)-TS10/APR reads left to right:

Chunk Value Meaning
leading letter T class type: L lecture, T tutorial, P practical
batch A4 student batch. Commas and ranges pack several: LB3,B4, LA1,A2
(…) CI121 subject code the parser looks up
after - TS10 room
after / APR faculty initials. Labs can list two: MKC/MTR

The parser does not store faculty. It keeps name, type, and room. subject_extractor takes the parenthesis (dash fallback if there is none). location_extractor takes the room. Type is that first L/T/P. Batch matching is the pipeline page.

Generated timetable#

Raw slots stay strings until Pyodide. create_time_table(campus, year, time_table_json, subject_json, batch, electives) walks every day, every slot, every string, and keeps the rows whose batch (and electives) match. That object is the timetable the grid and timeline render. It is also what HomeContent writes to localStorage as cachedSchedule.

Days become full names (MON → Monday). Slot keys become 24h ranges (9-9.50 → 09:00-10:00). Each cell is one class, not an array. LUNCH is dropped. Practicals (P) get an extra hour on the end, so 10-10.50 lab is 10:00-12:00. If two matching classes land on the same hour, the later one overwrites. Type C is not in this output. EditEventDialog writes it onto editedSchedule.

Subject lookup uses Code, then slices of Full Code. Year 1 EVEN26/62 often has empty Full Code, so a token like CI121 or MA211 can stay the code if nothing in subjects matches.

Batch A4, year 1, EVEN26 Sector 62, no electives. This is the week time_table_creator builds from the campus file above:

json
{
  "Monday": {
    "09:00-10:00": {
      "subject_name": "CI121",
      "type": "T",
      "location": "TS10"
    },
    "10:00-11:00": {
      "subject_name": "Physics-2",
      "type": "L",
      "location": "FF1"
    },
    "11:00-12:00": {
      "subject_name": "MA211",
      "type": "L",
      "location": "FF1"
    }
  },
  "Tuesday": {
    "10:00-11:00": {
      "subject_name": "HS111",
      "type": "L",
      "location": "FF1"
    },
    "11:00-12:00": {
      "subject_name": "CI121",
      "type": "L",
      "location": "FF1"
    }
  },
  "Wednesday": {
    "09:00-10:00": {
      "subject_name": "CI121",
      "type": "L",
      "location": "CR325"
    },
    "10:00-11:00": {
      "subject_name": "Workshop",
      "type": "T",
      "location": "TS17"
    },
    "11:00-12:00": {
      "subject_name": "HS111",
      "type": "L",
      "location": "FF4"
    },
    "13:00-15:00": {
      "subject_name": "Software Development Fundamentals Lab-",
      "type": "P",
      "location": "CL02"
    },
    "16:00-17:00": {
      "subject_name": "MA211",
      "type": "L",
      "location": "FF1"
    }
  },
  "Thursday": {
    "09:00-10:00": {
      "subject_name": "Physics-2",
      "type": "T",
      "location": "TS6"
    },
    "10:00-12:00": {
      "subject_name": "Physics Lab-2",
      "type": "P",
      "location": "PL2"
    },
    "13:00-14:00": {
      "subject_name": "Physics-2",
      "type": "L",
      "location": "G1"
    },
    "14:00-15:00": {
      "subject_name": "MA211",
      "type": "L",
      "location": "FF1"
    },
    "15:00-16:00": {
      "subject_name": "MA211",
      "type": "T",
      "location": "TS7"
    },
    "16:00-17:00": {
      "subject_name": "CI121",
      "type": "L",
      "location": "G8"
    }
  },
  "Friday": {
    "09:00-11:00": {
      "subject_name": "Workshop",
      "type": "P",
      "location": "EW2"
    },
    "13:00-14:00": {
      "subject_name": "HS111",
      "type": "T",
      "location": "TS6"
    }
  },
  "Saturday": {
    "10:00-11:00": {
      "subject_name": "Physics-2",
      "type": "L",
      "location": "G1"
    },
    "11:00-13:00": {
      "subject_name": "HS111",
      "type": "P",
      "location": "LL1"
    }
  }
}

TypeScript (website/types/index.ts):

ts
interface YourTietable {
  [day: string]: {
    [timeSlot: string]: {
      subject_name: string;
      type: "L" | "T" | "P" | "C";
      location: string;
    };
  };
}

WeekSchedule in types/schedule.ts is the same nested shape. After toJs(), Home JSON-clones it so Pyodide Proxies do not land in React state. Display reads editedSchedule || schedule. A new generate sets editedSchedule to null.

cachedSchedule in localStorage is that week object. Next to it:

json
{
  "year": "1",
  "batch": "A4",
  "campus": "62",
  "selectedSubjects": []
}

That params blob is cachedScheduleParams. classConfigs stores named form presets, not this grid. The share URL is year/batch/campus/electives, not the generated cells.

Calendar file#

website/data/calender/{yy1yy2}.json is a JSON array. The folder name is a typo that stuck (calender). 2526.json is academic year 2025-26 (146 events). Google-calendar shaped: summary plus start.date / end.date as ISO dates. Same-day events repeat the date. Ranges use a later end.

json
[
  {
    "summary": "Registration of 1st Semester (B.Tech, Intgt M.Tech, All UG and Diploma)",
    "start": { "date": "2025-07-10" },
    "end": { "date": "2025-07-10" }
  },
  {
    "summary": "T1 Examination & Results - Examination Schedule",
    "start": { "date": "2025-08-29" },
    "end": { "date": "2025-09-06" }
  },
  {
    "summary": "Holiday - Semester Break - Diwali (Odd)",
    "start": { "date": "2025-10-19" },
    "end": { "date": "2025-10-26" }
  }
]

The UI treats a title that starts with Holiday - as a holiday. Everything else is an academic event.

Exam file#

website/data/exam/{year}/{SEM}/T*.json is a JSON array. EVEN26 T3 has 316 rows, all subject_type: "L" in that dump.

json
[
  {
    "subject_code": "19M21BT117",
    "subject_type": "L",
    "subject_name": "ENZYMES AND BIOPROCESS TECHNOLOGY",
    "exam_date": "11-05-2026",
    "exam_time": "10.00 AM",
    "exam_day": "Monday",
    "semester": "2"
  },
  {
    "subject_code": "24B11CS243",
    "subject_type": "L",
    "subject_name": "Data Science & Data Analytics: Theory & Practice",
    "exam_date": "11-05-2026",
    "exam_time": "10.00 AM",
    "exam_day": "Monday",
    "semester": "2, 4"
  }
]

semester on disk is a string: "2", "8", or a list like "2, 4" / "1, 2, 3". The UI splits that into semesters: number[]. ExamContent searches with Fuse and can filter to “my exams” using the current timetable subjects. Dates stay DD-MM-YYYY. Times stay 10.00 AM.

How files get there#

creator/ Streamlit + Typer (Gemini for PDF notices) writes this JSON. Copy into website/data/ (and repo-root data/ if you use the parent fallback). The website does not parse Excel at runtime.