Working with Dates & Times

• Make dates & datetimes with date(...), datetime(...), date.today() and datetime.now().
• Parse strings into datetimes with strptime.
• Format datetimes into strings with strftime.
• Do date arithmetic with timedelta — days, hours, "difference between two dates".

from datetime import date, datetime, timedelta

print(date.today())          # → 2026-05-28
print(datetime.now())        # → 2026-05-28 14:33:21.456789
print(datetime.now().year)   # → 2026

Three modules' worth of functionality lives in datetime:

• date — just a calendar date.
• datetime — date + time.
• timedelta — a length of time (days, hours, seconds).

The format codes

Both strptime and strftime use the same format codes. The five you'll use most:

%Y  → 4-digit year      2026
%m  → 2-digit month     05
%d  → 2-digit day       28
%H  → 2-digit hour (24) 14
%M  → 2-digit minute    33

String → datetime (strptime, "p" for parse)

from datetime import datetime

text = "2026-05-28"
d = datetime.strptime(text, "%Y-%m-%d")
print(d.year, d.month, d.day)        # → 2026 5 28

datetime → string (strftime, "f" for format)

from datetime import datetime
now = datetime.now()
print(now.strftime("%d/%m/%Y"))      # → 28/05/2026
print(now.strftime("%A, %d %B %Y"))  # → Thursday, 28 May 2026

You can mix any text you want with format codes; everything that isn't a %X is taken literally.

Arithmetic — timedelta

from datetime import date, timedelta

today    = date.today()
in_one_week  = today + timedelta(days=7)
last_friday  = today - timedelta(days=6)
diff = in_one_week - today
print(diff)            # → 7 days, 0:00:00
print(diff.days)       # → 7

Three rules:

• date ± timedelta → date
• date − date → timedelta
• timedelta(days=, hours=, minutes=, seconds=) — pick any combination.

The full pipeline

# A real CSV row → a calculation → back to a friendly string
from datetime import datetime

row = {"date": "2026-05-01", "qty": "3"}
when = datetime.strptime(row["date"], "%Y-%m-%d")

days_ago = (datetime.now() - when).days
print(f"That sale was {days_ago} days ago.")
print("Nice format:", when.strftime("%d %b %Y"))   # → 01 May 2026

Use the sales.csv from Lesson 3. Filter rows from the last 7 days, sum the revenue, print a friendly date range.

# last_7.py — date filtering + formatting
import csv
from datetime import datetime, date, timedelta

today      = date.today()
seven_ago  = today - timedelta(days=7)

with open("sales.csv", newline="") as f:
    rows = list(csv.DictReader(f))

# Parse dates into real objects
for r in rows:
    r["date_obj"] = datetime.strptime(r["date"], "%Y-%m-%d").date()
    r["revenue"]  = int(r["quantity"]) * float(r["price"])

# Filter
recent = [r for r in rows if seven_ago <= r["date_obj"] <= today]

# Report
total = sum(r["revenue"] for r in recent)
print(f"📅 {seven_ago.strftime('%d %b')} — {today.strftime('%d %b %Y')}")
print(f"   {len(recent)} transactions, total RM {total:.2f}")
print()
for r in recent:
    print(f"   {r['date_obj'].strftime('%a %d %b')}  "
          f"{r['product']:<8} RM {r['revenue']:.2f}")

Sample output (run on 2026-05-08)

📅 01 May — 08 May 2026
   6 transactions, total RM 41.50

   Fri 01 May  roti     RM 4.50
   Fri 01 May  milo     RM 6.00
   Sat 02 May  roti     RM 7.50
   Sat 02 May  nasi     RM 8.00
   Sun 03 May  milo     RM 12.00
   Sun 03 May  nasi     RM 16.00

Read the diff

The parse step (strptime → date()) happens once when data loads. After that, every comparison is between real date objects — clean, fast, no string parsing inside the loop. That's the "treat strings as enemies" rule in action.

Three classes: date, datetime, timedelta. The mantra: parse strings to objects on the way in (strptime), do every comparison and calculation on objects, format back to a string on the way out (strftime). timedelta handles all the "how long between" and "X days later" arithmetic without any maths from you.