Teddy: Automatic Recommendation of Pythonic Idiom Usage For Pull-Based Software Projects

A repo for the Teddy tool.

Required Components

• Elasticsearch 2.2.0
• Apache Maven 3.6.3 or higher
• Java 8.x.x

Setting up Elasticsearch 2.2.0 and restore the python-patterns index

1. In ~/elasticsearch-2.2.0-teddyplus/config/elasticsearch.yml, specify the absolute path to the snapshot folder as follow path.repo: <CLONED_REPO_FOLDER>/elasticsearch-2.2.0-teddyplus/snapshot/backups
2. Save the config file and launch the Elastic server (.~/elasticsearch-2.2.0-teddyplus/bin/elasticsearch)
3. Bind the snapshot folder to the running Elastic via API: curl -XPUT 'localhost:9200/_snapshot/backups?pretty' -d '{"type":"fs","settings":{"location":"<SAME_PATH_AS_path.repo>"},"compress":true}}'
4. Restore the snapshot of python-patterns index to the running Elastic using API: curl -XPOST 'localhost:9200/_snapshot/backups/snapshot_1/_restore?pretty'
5. (Optional) Check that the python-patterns index has been restored in the Elastic using API: curl -XGET 'localhost:9200/_cat/indices?pretty'

DETECTION MODE USER INSTRUCTION (directly via Linux terminal)

0. Needed dependencies and library before start using:
   • Java JDK 1.8.X
   • Python 3.X.X
   • pandas library for Python
   • bokeh library for Python
   • git bash for Linux
1. Make sure that your local Elasticsearch is running and has the python-patterns index inside (refer to Setting up Elasticsearch 2.2.0 and restore the python-patterns index in the above section)
2. Make sure there is no any folder located in the directory ~/detection-mode/
3. Open another terminal and execute AutoRun.sh at ~/detection-mode/ with the following order of arguments:
   • .git URL of a GitHub repository
   • directory for the result CSV files generated by Siamese Engine to be saved, NOT within THE REPO (absolute path recommended)
   • Name of the repository you are analyzing (no whitespace allowed)
   • Example: $ ./AutoRun https://github.com/pallets/flask.git /home/flask-results flask
4. (Optional) You can also specify the interval of commits using -a and -b.
   • Example: If you would like to collect the commits after 2011/01/01: $ ./AutoRun https://github.com/pallets/flask.git /home/flask-results flask -a "2011/01/01"
   • Example: If you would like to collect the commits before 2011/01/01: $ ./AutoRun https://github.com/pallets/flask.git /home/flask-results flask -b "2011/01/01"
   • Example: If you would like to collect the commits between 2011/01/01 and 2020/01/01: $ ./AutoRun https://github.com/pallets/flask.git /home/flask-results flask -a "2011/01/01" -b "2020/01/01" If Teddy-plus cannot get any commits, it shows that message, then please check your specified format is correct or not. : There is no commits.
5. Finally, a file named myplot.html will be in the same folder as Siamese's output CSV folder. This is the final visualization of detection mode. Move it to wherever you like. Use a web browswer (Google Chrome is recommended) to view the file.
6. NOTE: Detection mode takes a considerable amount of time to complete depending on the repository's size and number of commits present.