A play time recording and calendar heatmap visualization tool in Python 3, based on the data provided by Steam, deployed on Windows 10 and Linux platform.
-
An official Steam client. (Win 10)
-
Python 3.6 (3.2+ required), and some additional packages by running
pip3 install pandas requests numpy calplot(Win 10 & Ubuntu 18.04)
Fill out the form here to obtain a Steam API key for your account. The domain names seem to be used for generating keys only.
Ways to obtain your 64-bit steamid.
Fill in the "key" and "steamid" entries as strings in config.json as above mentioned.
To query Steam, open PowerShell or equivalent terminal in the script's directory, run
python.exe steam_stat.py
You can check the log file ./log/steam_record.log to see if the whole thing works.
It is recommended to deploy auto querying on a PC that you use most frequently or runs 24/7, if a server is not available.
The idea is to use Windows 10 built-in task scheduler to run steam_stat.py periodically.
A scheduled task steam_rec.xml is provided. To use it,
- edit line 77, 78 within the
<Actions>tag correctly in the.xmlfile.pythonw.exeis recommended as it won't prompt a terminal window. - Type and choose your own Windows system account in the security options block, when interactively importing to the task scheduler.
You can adjust query frequency in the scheduler, up to once per minute.
Note: Since Steam does not provide play time on a daily basis, querying the total playtime periodically is inevitable if you rely on its data. One can use batch scripts or 3rd party tools to trigger accurate querying, beyond the simple methods here.
It may be a good idea to record already played time before starting steam client. To do this,
- Edit the
steam_colauncher.pyfile to ensure the path toSteam.exeis correct. - Edit the batch file
run_steam_launcher.batto ensure the paths to python executable and scriptsteam_colauncher.pyare correct. - Create a shortcut from
run_steam_launcher.batto the desktop, and you can change its icon if you like, a.icofile is provided in the./assetsfolder. - Launch Steam from the shortcut from now on.
The pre-installed task scheduler cron is used. After installing python packages and cron,
- Edit
query_steam.shfor the correct path to the scriptsteam_stat.py - Add permission
chmod +x query_steam.sh crontab -eto edit user schedules, e.g.1 0 * * * /path/to/job.sh(runs 00:01 every day). check cron syntax.
Suppose the default database file <steamid>-<suffix>.db is generated automatically, and at least 2 queries finish successfully, run
python.exe steam_recorder_plot.py y1 m1 d1 y2 m2 d2
where parameters are the year, month, date of the start and end of the querying. Game-id pairs will be printed to the console.
Or you can specify a database path for plotting:
python.exe steam_recorder_plot.py y1 m1 d1 y2 m2 d2 path/to/data.db
Plotting playtime record of one game is also supported, as long as you know the integer game id:
python.exe steam_recorder_plot.py y1 m1 d1 y2 m2 d2 [path/to/data.db] -g <game-id>
Put several <steamid>-<suffix>.db files to merge in the ./data/<steamid>/ folder, then run
python.exe steam_stat.py -m
will produce a merged database with name <steamid>-<merged_suffix>.db in the above folder.
MIT License.
- TODO: cannot rely on db filenames! Use userinfo table instead.
- Mask steamkey and steamid in the log string for privacy.
- Display heatmap in readme when I played enough.
- Database optimization for high frequency queries.
- sparsify to hours for active apps, remove continuous blanks for inactive apps.
- transaction for large insertion(merging). Does not have autocommit by default ref, meaning it operates modification in a transaction by default. So separate conn to the outside the function.
- index for history database, say per month.
- Add user info and game info to the database and heatmap.
- Heatmap without calplot.
- Directly calling Anaconda's
python.exeorpythonw.exefrom the batch file or by the task scheduler may cause "SSL module is not available" failure.