bse_history.py — Trading Calendar
BSEWorkingDays() / get_next_bseworkingdate() mirror
nseeod/utils.py's NSEWorkingDays() / get_next_nseworkingdate(),
reading holidays/specialdays with exch=3 instead of exch=1.
Calendar Seeding — exch=3, Shared Tables
holidays/specialdays already have an exch column (NSE=1, MCX=2 — see
MCXEOD's calendar notes).
BSE slots in as exch=3 with zero schema change — no migration needed,
unlike every other piece of BSE's schema.
India's BSE and NSE equity trading calendars are effectively identical in
practice, so seed_bse_holidays_from_nse() seeds BSE's exch=3 rows by
copying NSE's exch=1 rows, rather than scraping a separate source —
the same approach MCX's calendar took as a starting point (derived from
NSE's, with MCX-specific closures layered on top).
cd backend
python -m bseeod.bse_history # dry run - reports what would be inserted
python -m bseeod.bse_history --apply # actually insert
Confirmed live 2026-10-11: 465 holidays + 59 special (Muhurat) days copied
into exch=3 from NSE's existing calendar. Safe to re-run — only inserts
dates not already present under exch=3.
This is a reasonable default, not a substitute for checking BSE's own published trading-holiday circular before relying on it for a specific historical backfill range — the two exchanges' calendars are usually identical but not contractually guaranteed to be.
BSEWorkingDays(fromdate, todate)
Returns a DataFrame indexed by date, built from a
pandas.tseries.offsets.CustomBusinessDay over holidays (exch=3) plus
any specialdays (exch=3) rows in range — same construction as
NSEWorkingDays.
get_next_bseworkingdate(fromdate)
The next BSE trading day strictly after fromdate, walking a 20-day window
forward (mirrors get_next_nseworkingdate's approach).
Related
- bse_backfill.py — the main consumer, via
BSEWorkingDays()to compute pending dates - bse_downloader.py —
BseEodEqUpdater.update()and friends useBSEWorkingDays()to walk missing dates in a daily sync