preface

from wikipedia: “Coffee cupping, or coffee tasting, is the practice of observing the tastes and aromas of brewed coffee.” this collection of scripts allows one to retrieve all of the test results in one’s jenkins jobs, arbitrarily inspect the builds with SQL, and generate a digest of test passage across repositories, branches and specific revisions in order to easily detect breakage and other trends.

explanation

fetch_tests.coffee

queries your jenkins for test results. can query the hwole server, a specific job, or even a particular build. can output to stdout or to a sqlite3 db.

display_tests.coffee

queries out.sqlite, which is written to by fetch_tests.coffee and displays an aggregated view of test passage

installation

termsql

sudo apt-get install sqlite3 cd /usr/src && sudo git clone https://github.com/tobimensch/termsql.git && cd termsql && sudo python setup.py install

nodejs

sudo apt-get install nodejs-legacy

nodejs deps

install #should install locally everything in package.json

coffeescript

sudo apt-get install npm npm install yarn yarn install

init an sqlite3 db, if not existing

cat schema.sql | sqlite3 out.sqlite

retrieval

retrieve a specific job and insert results into db

./fetch_tests.coffee –job=JOBNAME –toDB

retrieve all jobs, inserts results into db

./fetch_tests.coffee –toDB

retrieve a job, write to stdout (parseable by termsql)

./fetch_tests.coffee –job=JOBNAME

import into an sqlite database for heavy duty querying

(./fetch_tests.coffee ; cat passage/*.txt ) | termsql -1 -o out.sqlite

import a new job into an existing db without erasing old data

./fetch_tests.coffee –job=JOBNAME | termsql -a -1 -o out.sqlite

db querying