steamreviews 0.9.6

An API to download Steam reviews

Download Steam Reviews

This repository contains Python code to download every Steam review for the games of your choice.

Requirements

• Install the latest version of Python 3.X (at least version 3.11).

Installation

The code is packaged for PyPI, so that the installation consists in running:

pip install steamreviews

Usage

The Steam API is rate-limited so you should be able to download about 10 reviews per second.

NB: If you do not know the appID of a game, look for it on the Steam store. The appID is a unique number in the URL.

For instance, for SpyParty, the appID is 329070.

Process a batch of appIDs

import steamreviews

app_ids = [329070, 573170]
steamreviews.download_reviews_for_app_id_batch(app_ids)

Process a batch of appIDs, written down in a text file

• For every game of interest, write down its appID in a text file named idlist.txt. There should be an appID per line.
• Then proceed as follows:

import steamreviews

steamreviews.download_reviews_for_app_id_batch()

Load reviews for one appID

import steamreviews

app_id = 329070
review_dict = steamreviews.load_review_dict(app_id)

Download reviews for one appID

import steamreviews

app_id = 573170
review_dict, query_count = steamreviews.download_reviews_for_app_id(app_id)

Download reviews for one appID, with specific request parameters (language, sentiment, store)

Caveat: the following parameters do not appear in the output filename, so make sure that you start the download from scratch (instead of updating existing JSON review data) if you ever decide to change them, e.g the value of the review_type (set to all, positive or negative).

Caveat²: if review_type is set to positive (or negative), then the value of total_reviews can be misleading. It is indeed arbitrarily set to total_positive (respectively total_negative). In this case, if you need the total number of reviews, compute it as the sum of total_positive and total_negative.

import steamreviews

request_params = dict()
# Reference: https://partner.steamgames.com/doc/store/localization#supported_languages
request_params['language'] = 'english'
# Reference: https://partner.steamgames.com/doc/store/getreviews
request_params['review_type'] = 'positive'
request_params['purchase_type'] = 'steam'

app_id = 573170
review_dict, query_count = steamreviews.download_reviews_for_app_id(app_id,
                                                                    chosen_request_params=request_params)

Download a few of the most helpful reviews for one appID, which were created in a time-window

Caveat: with filter set to all, you will only be able to download a few reviews within the specified time-window.

import steamreviews

request_params = dict()
# Reference: https://partner.steamgames.com/doc/store/getreviews
request_params['filter'] = 'all'  # reviews are sorted by helpfulness instead of chronology
request_params['day_range'] = '28'  # focus on reviews which were published during the past four weeks

app_id = 573170
review_dict, query_count = steamreviews.download_reviews_for_app_id(app_id,
                                                                    chosen_request_params=request_params)

Download reviews for one appID, which were created within a specific time-window

import steamreviews

request_params = dict()
request_params['filter'] = 'recent'
request_params['day_range'] = '28'

app_id = 573170
review_dict, query_count = steamreviews.download_reviews_for_app_id(app_id,
                                                                    chosen_request_params=request_params)

Download reviews for one appID, which were updated within a specific time-window

import steamreviews

request_params = dict()
request_params['filter'] = 'updated'
request_params['day_range'] = '28'

app_id = 573170
review_dict, query_count = steamreviews.download_reviews_for_app_id(app_id,
                                                                    chosen_request_params=request_params)

References

• my original Steam-Reviews repository

• a snapshot of Steam-Reviews data for hidden gems