Building a Price Tracker Bot

Architecture

1. Fetch product page (or API)
2. Parse price
3. Save price history (CSV/SQLite)
4. Compare with threshold
5. Notify (email/telegram)

Example: parse a price from HTML

from bs4 import BeautifulSoup
 
html = "<span class='price'>$199.99</span>"
soup = BeautifulSoup(html, "html.parser")
price_text = soup.select_one(".price").get_text(strip=True)
print(price_text)

Persist history (CSV)

import csv
from datetime import datetime
 
 
def append_price(csv_path: str, price: str):
    with open(csv_path, "a", newline="", encoding="utf-8") as f:
        w = csv.writer(f)
        w.writerow([datetime.now().isoformat(), price])

Notes

Real sites often block scraping.

Prefer APIs, and respect site policies.

Visualize it

A tracker bot loops through fetch, parse, and store steps, retrying politely when a request fails.

  flowchart TD
    A["Fetch the page"] --> B{"Request successful?"}
    B -- "No" --> E["Wait/retry (respect rate limits)"]
    E --> A
    B -- "Yes" --> C["Parse the HTML"]
    C --> D["Extract the price"]
    D --> F["Store it"]

How the tracker fits together

A price tracker is a small loop that runs on a schedule. Each run does the same five things, and each is a separate function, so you can test the parts that do not need the network.

The parts that are pure Python are the most valuable to get right, and the price text is the trickiest. A web page shows $1,199.99 or 1.199,99 €, and you need a number. Parse it in one function, and handle failure:

import csv
from datetime import datetime
from pathlib import Path
 
 
def parse_price(text: str) -> float:
    cleaned = text.replace("$", "").replace(",", "").strip()
    return float(cleaned)          # raises ValueError if the layout changed
 
 
def last_price(csv_path: Path) -> float | None:
    if not csv_path.exists():
        return None
    with csv_path.open(newline="", encoding="utf-8") as f:
        rows = list(csv.reader(f))
    return float(rows[-1][1]) if rows else None
 
 
def check(csv_path: Path, price_text: str, threshold: float) -> None:
    price = parse_price(price_text)
    previous = last_price(csv_path)
 
    with csv_path.open("a", newline="", encoding="utf-8") as f:
        csv.writer(f).writerow([datetime.now().isoformat(), price])
 
    if price <= threshold and (previous is None or previous > threshold):
        print(f"ALERT: price fell to {price} (threshold {threshold})")
 
 
check(Path("history.csv"), "$199.99", threshold=200)

The alert fires only when the price crosses the threshold, not on every run while it stays below. Without that check you would get an email every hour for as long as the sale lasts.

Why store history

Storing every price, not just the latest, lets you see trends, spot a fake "discount", and debug. CSV is fine for one product. For many products, use SQLite (sqlite3 is in the standard library) with a table of product, time, and price.

The fetch step, done politely

Fetch with requests and a timeout, send a User-Agent that names your bot, and run at most a few times a day. Check robots.txt and the terms first. Many shops forbid scraping. If the store has an official API or a price feed, use that. Scrape only what the site allows, and stop if you get 403 or 429 responses.

Common mistakes

• Trusting a selector forever. Sites change their layout. When select_one returns None, log it and alert yourself instead of crashing silently.
• Parsing without handling locale. 1.299,00 and 1,299.00 mean the same thing but parse differently.
• Alerting on every run. Track the previous state.
• Running every minute. That is rude to the site and gets you blocked. Hourly or daily is plenty.
• Storing the price as text. Store numbers so you can compare them.

Adapting it

Wire the alert into the email, Telegram, or Slack lessons, and run the script with cron, Task Scheduler, or the schedule library.

🧪 Try It Yourself

Exercise 1 – Parse a Price and Compare

Press Run to see what your code prints.

Check your understanding