Notifier - Price & Availability Tracking Bot

Description

A bot used for scraping data from the websites of selected fashion retailers, tracking items and notifying by telegram messages upon changes in price or availability. The main features are:

• Per-user tracking of multiple variants (sizes, colors) of an item; adding, listing and removing items.
• Notification upon changes in price and availability status of tracked items. The bot can be queried to check for changes manually or instructed to do so at specified time intervals.
• Export of tracked items to a .json file, restoring the list of tracked items from such a file.
• Currently supported retailers (Dec. 24'): Zara & Uniqlo.

Demo:

notifier_demo.mp4

Instructions

• Notifier can be run (ideally on a server¹) by calling python3 notifier_main.py.
• In order to run, there needs to be a config folder in the project root directory with two files,bot_token.txt (see the telegram bot tutorial "From BotFather to 'Hello World' "), and allowed_users.txt, containing linewise the telegram handles of allowed users.
• The first user listed in allowed_users.txt can issue the command /shutdown to the bot which terminates the bot graciously and ensures that tracked items are stored persistently.
• For usage instructions, call /help from the bot.

Overview of the code

Core routines

file	description
notifier_main.py	Main routine, to be executed on server side.
notifier_bot.py	Core functions of the bot (bot-user interaction, querying, adding & removing products, tracking, updating).

Extraction / Scraping

file	description
src/zara_scraper.py	Class for scraping data from Zara
src/zalando_scraper.py	Class for scraping data from Zalando
src/hm_scraper.py	Class for scraping data from H&M
src/cos_scraper.py	Class for scraping data from COS
src/uniqlo_scraper.py	Class for scraping data from Uniqlo

Data Representation

file	description
src/product.py	Abstract base class for products from various retailers
src/zara_product.py	Class that encapsulates data about a product tracked from Zara
src/uniqlo_product.py	Class that encapsulates data about a product tracked from Uniqlo
src/zalando_product.py	Class that encapsulates data about a product tracked from Zalando
src/hm_product.py	Class that encapsulates data about a product tracked from H&M
src/cos_product.py	Class that encapsulates data about a product tracked from COS

Footnotes

1. Some sites (e.g. Uniqlo) require the Selenium framework, for which the server should not be too weak. We found a two-core VM.Standard.A1.Flex (Ampere/ARM) server from Oracle to be sufficient, while an VM.Standard.E5.Flex (AMD) instance proved too slow. To get Selenium running on an arm64 based architecture, see this answer on stackoverflow. ↩