Skip to content

christophM/LolCrawler

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

87 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

LolCrawler

Crawls League of Legends summoner match lists and match data and stores the data

What it does

  1. Start with the seed summoner as defined in config.py
  2. Download match list of the chosen summoner and store it in MongoDB
  3. Choose random match from the match list
  4. Download match via API and store it in MongoDB
  5. Choose random player from the downloaded match and continue with step 2

Usage

Place your Riot API key in config.py. LolCrawler needs a region and a summoner id to start crawling. The config.py file provides a summoner seed for Europe West (EUW). You need to have MongoDB installed for storing the match and match list JSON.

Start crawling:

python crawl.py

If you want to specifically crawl yesterdays challenger and master matches in euw, na and kr, do:

python crawl.py top

Run the script in the background by using nohup:

nohup python -u crawl.py &

Check if the script is still running with:

ps ax | grep crawl.py

And kill LolCrawler by checking if the script is still running, checking the process id and then type:

kill <process-id>

If something goes wrong, have a look at the log file named crawler.log.

Run tests with

py.test test/

Champion statistics

You can create a MongoDB table "champStats" by running the following script:

python aggregate.py update

The script sequentially processes the matches. It creates statistic about champions: number of matches, victories and defeats. These numbers are split by patch, lane, role and so on. The script sets a flag for processed matches. So if you run update again, it will only count the new matches. It should be no problem to run the crawler at the same time as the aggregation. If you want to remove the flags and delete the aggregations, run:

python aggregate.py reset

And to first reset and then update:

python aggregate.py reprocess

Changelog

v0.8

  • Now there are two crawler versions: One that does a random walk through the summoners, the other systematically crawls challenger and master matches
  • Included the 'extract'-module that enriches the matches with additional information

v0.7

  • Added script that creates a MongodB table with champion statistics by processing the matches. The statistics are count of matches, victories and defeats, grouped by patch, champion, role, lane and so on
  • Instead of nesting extracted information (like the patch numbers) in the raw data, a field "extractions" was added on top level of the document structure and additional information will be nested in "extractions"

v0.6

  • Adding patch fields to match, extracted from the field "matchVersion"
    • "patch": The patch version as a string, for example "5.24"
    • "patchMajorNumeric": The major patch version as integer, resembling the season, for example 5
    • "patchMinorNumeric": The minor patch version as integer, for example 24

v0.5

  • Be smarter about match and summoner sampling
  • By default, don't fetch the timeline
  • Sample only from the latest matches of a summoner
  • Introduce logging

v0.4

  • If there are already matchlists inserted in MongoDB, start with latest summoner id as seed summoner, instead of seed summoner id from config file
  • Move files in own module

v0.3

  • Change from Riot Api's matchhistory endpoint to matchlist, since matchhistory endpoint will be deprecated starting 2015/09/22
  • Be more intelligent about error handling and make crawler more robust

v0.2

  • Write to MongoDB, not to disk

v0.1

  • Initial release

Disclaimer

LolCrawler isn’t endorsed by Riot Games and doesn’t reflect the views or opinions of Riot Games or anyone officially involved in producing or managing League of Legends. League of Legends and Riot Games are trademarks or registered trademarks of Riot Games, Inc. League of Legends © Riot Games, Inc.

About

Crawls League of Legends matches

Resources

Stars

Watchers

Forks

Packages

No packages published

Languages