Skip to content

Latest commit

 

History

History
227 lines (179 loc) · 9.55 KB

README.md

File metadata and controls

227 lines (179 loc) · 9.55 KB
Raw

logo

EO principles respected here DevOps By Rultor.com We recommend RubyMine

rake Availability at SixNines Webhook via ReHTTP PDD status Maintainability Test Coverage Hits-of-Code Codacy Badge

Read this blog post first: PDD in Action. TL;DR:

  1. Your boss tells you to fix issue #42
  2. You do it, but not completely (you have no time, you are lazy, etc.)
  3. You put TODO #42:30min bla-bla-bla into the code base (in a pull request)
  4. CI checks that you didn't break the format of the TODO (reuse our pdd.yml)
  5. You merge the pull request
  6. The bot picks up the TODO and creates issue #43 (new one)
  7. The boss asks your friend to fix #43
  8. The friend fixes it, and merges
  9. The TODO is gone from the code base
  10. The bot closes the issue #43

0pdd.com is a hosted service that finds new "puzzles" in your repository and posts them as GitHub issues. To start using it just create a Webhook in your repository just for push events with https://www.0pdd.com/hook/github payload URL and application/json content type.

Then, add @0pdd GitHub user as a collaborator to your repository, if it's private (you don't need this for a public repository). If your invitation is not accepted by @0pdd in 30mins, please visit this address https://0pdd.com/invitation?repo={REPO_FULL_NAME} - REPO_FULL_NAME is the full name of your repo e.g yegor256/0pdd

Then, add a @todo puzzle to the source code (format it right).

Then, git push to master branch something and see what happens. You should see a new issue created in your repository by @0pdd.

The dependency tree of all puzzles in your repository you can find here: https://www.0pdd.com/p?name=yegor256/0pdd (just replace the name of the repo in the URL).

Don't forget to add that cute little badge to your README.md, just like we did here in this repo (see above). The Markdown you need will look like this (replace yegor256/0pdd with GitHub coordinates of your own repository):

[![PDD status](https://www.0pdd.com/svg?name=yegor256/0pdd)](https://www.0pdd.com/p?name=yegor256/0pdd)

How to configure?

The only way to configure 0pdd is to add .0pdd.yml file to the root directory of your master branch (see this one as a live example). It has to be a YAML file with the following optional parameters inside:

threshold: 10
model: true
errors:
  - [email protected]
alerts:
  suppress:
    - on-found-puzzle
    - on-lost-puzzle
    - on-scope
  github:
    - yegor256
format:
  - short-title
  - title-length=100
tags:
  - pdd
  - bug

The element threshold allows you to limit the number of issues created from the puzzles in your code. In the example above, each time the appropriate push event is sent to your webhook up to 10 issues will be created regardless of the number of puzzles found in the code. If this limit is not set, threshold is assumed to be equal to 256.

Section errors allows you to specify a list of email addresses which will receive notifications when PDD processing fails for your repo. It's a very useful feature, since very often programmers make mistakes in PDD puzzle formatting. We would recommend you use this feature.

Section alerts allows you to specify users that will be notified when new PDD puzzles show up. By default we will just submit GitHub tickets and that's it. If you add github subsection there, you can list GitHub users who will be "notified": their GitHub nicknames will be added to each puzzle description and GitHub will notify them by email.

Subsection suppress lets you make 0pdd more quiet, where it's necessary:

  • on-found-puzzle: stay quiet when a new puzzle is discovered

  • on-lost-puzzle: stay quiet when a puzzle is gone

  • on-scope: stay quiet when child puzzles change statuses

--model

The model option used by 0pdd to opt-in to use ML model which prioritizes puzzles generated by pdd. If you would like to opt-in to puzzle prioritization, then add this option to your .0pdd.yml config file at the root of your project.

pdd is the tool that parses your source code files. You can configure its behavior by adding .pdd file to the root directory of the repository. Take this one, as an example.

The format section helps you instruct 0pdd about GitHub issues formatting. These options are supported:

  • short-title: issue title will not include file name and line numbers

  • title-length=...: you may configure the length of the title of GitHub issues we create. Minimim length is 30, maximum is 255. Any other values will be silently ignored. The default length is 60.

The tags section lists GitHub labels that will automatically be attached to all new issues we create. If you don't have that labels in your GitHub repository, they will automatically be created.

To exclude files from analysis, create a .pdd file with the following content:

--exclude=path/to/file.txt

See: pdd usage

What to expect?

Pay attention to the comments @0pdd posts to your commits. They will contain valuable information about its recent actions. If something goes wrong, you will receive exception messages there. Please, post them here as new issues.

Remember that GitHub triggers us only when you do git push. This means that if you make a number of commits, we will process them all together. Only the latest one will be commented. It may not be the one with new puzzles though.

After we create GitHub issues you can modify their titles and descriptions. You can work with them as with any other issues. We will touch them only one more time, when the puzzle disappears from the source code. At that moment we will try to close the issue. If it is already closed, nothing will happen. However, it's not a good practice to close them manually. You better remove the necessary puzzle from the source code and let us close the issue.

How to contribute?

It is a Ruby project. First, install Java SDK 8+, Maven 3.2+, Ruby 2.3+, Rubygems, and Bundler. Then:

$ bundle update
$ rake

The build has to be clean. If it's not, submit an issue.

Then, make your changes, make sure the build is still clean, and submit a pull request.

To run it locally:

$ rake run

If you want to run it on your own machine, you will need to add this config.yml file to the root directory of this repository:

s3:
  region: us-east-1
  bucket: xml.0pdd.com
  key: AKIAI..........UTSQA
  secret: Z2FbKB..........viCKaYo4H..........vva21
sentry: https://[email protected]/229223
dynamo:
  region: us-east-1
  key: AKIAI..........UTSQA
  secret: Z2FbKB..........viCKaYo4H..........vva21
github:
  client_id: b96a3b5..........87e
  client_secret: be61c471154e2..........66f434d33e0f63a5f
  encryption_secret: some-random-text
  login: 0pdd
  pwd: GitHub-Password
smtp:
  host: email-smtp.us-east-1.amazonaws.com
  port: 587
  key: AKIAI..........UTSQA
  secret: Z2FbKB..........viCKaYo4H..........vva21
id_rsa: |
  -----BEGIN RSA PRIVATE KEY-----
  MIIJKAIBAAKCAgEAoE94Xy8TGMbnoK5cKJXWccr9qLLDc/liKpMAMlnQEFDCgi0l
  ...
  NaaFpowFg8LKSiwc04ERduu72Imv5GJBCkhS8F7laURXFcZiYNqBnWYzY0U=
  -----END RSA PRIVATE KEY-----

We add this file to the repository while deploying to Heroku, see how it's done in .rultor.yml.

How to install in Heroku

Don't forget this:

heroku buildpacks:add --index 1 https://github.com/heroku/heroku-buildpack-apt