reviewdog for github action, refs wemake-services#1023 (wemake-servic…

…es#1024)

.github/workflows/wps.yml

@@ -16,3 +16,7 @@ jobs:
     steps:
     - uses: actions/checkout@v1
     - uses: wemake-services/wemake-python-styleguide@master
+      with:
+        reporter: 'github-pr-review'
+      env:
+        GITHUB_TOKEN: ${{ secrets.github_token }}

CHANGELOG.md

@@ -3,6 +3,7 @@
 We follow Semantic Versions since the `0.1.0` release.
 We used to have incremental versioning before `0.1.0`.
 
+
 ## 0.13.1 WIP
 
 ### Bufixes
@@ -12,6 +13,10 @@ We used to have incremental versioning before `0.1.0`.
 - Fixes invalid docs for `BracketBlankLineViolation` #1020
 - Add more complex example to `ParametersIndentationViolation` #1021
 
+### Misc
+
+- Now our GitHub Action can be used to leave PR review comments
+
 
 ## 0.13.0 aka The Lintoberfest
 

Dockerfile

@@ -4,11 +4,14 @@
 # This is Github Action docker-based image.
 # It is not intended for local development!
 #
+# You can find docs about how to setup your own Github Action here:
+# https://wemake-python-stylegui.de/en/latest/pages/usage/integrations/github-actions.html
+#
 # It can still be used as a raw image for your own containers.
 # See `action.yml` in case you want to learn more about Github Actions.
 # See it live:
 # https://github.com/wemake-services/wemake-python-styleguide/actions
-
+#
 # This image is also available on Dockerhub:
 # https://hub.docker.com/r/wemakeservices/wemake-python-styleguide
 
@@ -19,9 +22,13 @@ LABEL vendor="wemake.services"
 
 ENV WPS_VERSION='0.13.0'
 
-RUN apk add --no-cache bash
-RUN pip install "wemake-python-styleguide==$WPS_VERSION"
+RUN apk add --no-cache bash git wget
+RUN pip install "wemake-python-styleguide==$WPS_VERSION" \
+  # Installing reviewdog to optionally comment on pull requests:
+  && wget -O - -q 'https://raw.githubusercontent.com/reviewdog/reviewdog/master/install.sh' \
+  | sh -s -- -b /usr/local/bin/ v0.9.13
 
+COPY ./scripts/action-config.cfg /
 COPY ./scripts/entrypoint.sh /
 RUN chmod +x /entrypoint.sh
 ENTRYPOINT ["/entrypoint.sh"]

README.md

@@ -7,6 +7,7 @@
 [![Github Action](https://github.com/wemake-services/wemake-python-styleguide/workflows/wps/badge.svg)](https://github.com/wemake-services/wemake-python-styleguide/actions)
 [![Python Version](https://img.shields.io/pypi/pyversions/wemake-python-styleguide.svg)](https://pypi.org/project/wemake-python-styleguide/)
 [![wemake-python-styleguide](https://img.shields.io/badge/style-wemake-000000.svg)](https://github.com/wemake-services/wemake-python-styleguide)
+
 ---
 
 Welcome to the strictest and most opinionated python linter ever.
@@ -52,6 +53,9 @@ And it won't change your existing workflow.
 See ["Usage" section](https://wemake-python-stylegui.de/en/latest/pages/usage/setup.html)
 in the docs for examples and integrations.
 
+We also support [Github Actions](https://wemake-python-stylegui.de/en/latest/pages/usage/integrations/github-actions.html) as first class-citizens.
+[Try it out](https://github.com/marketplace/actions/wemake-python-styleguide)!
+
 
 ## What we are about
 

action.yml

@@ -15,7 +15,11 @@ inputs:
     description: 'Path or space-separated list of paths to lint'
     required: false
     default: '.'
-
+  reporter:
+    description: 'How would you like the results to be displayed?'
+    required: false
+    default: 'terminal'
+
 outputs:
   output:
     description: 'The output of wemake-python-styleguide run'
@@ -24,3 +28,4 @@ runs:
   image: 'Dockerfile'
   args:
     - ${{ inputs.path }}
+    - ${{ inputs.reporter }}

docs/pages/usage/integrations/github-actions.rst

@@ -16,11 +16,31 @@ You can use it from the `Github Marketplace <https://github.com/marketplace/acti
 
 You can also specify any version
 starting from ``0.12.5`` instead of the ``latest`` tag.
+You can also specify to leave inline PR comments
+and PR summary review starting from ``0.13.1`` version.
 
 Inputs
 ~~~~~~
 
-We also support custom path to be specified:
+We support three reporting options:
+
+- ``terminal`` (default one) when we just dump the output into Action's logs.
+  Is the easiest one to setup, that's why we use it by default
+- ``github-pr-review`` (recommended) when we use `inline comments <https://github.com/reviewdog/reviewdog#reporter-github-pullrequest-review-comment--reportergithub-pr-review>`_ inside code reviews
+- ``github-pr-check`` when we use `Github Checks <https://github.com/reviewdog/reviewdog#reporter-github-checks--reportergithub-pr-check>`_ for the output
+
+That's how it can be setup:
+
+.. code:: yaml
+
+  - name: wemake-python-styleguide
+    uses: wemake-python-styleguide@latest
+    with:
+      reporter: 'github-pr-review'
+    env:
+      GITHUB_TOKEN: ${{ secrets.github_token }}
+
+We also support custom ``path`` to be specified:
 
 .. code:: yaml
 
@@ -34,15 +54,10 @@ Outputs
 
 We also support ``outputs`` from the spec, so you can later
 pass the output of ``wemake-python-styleguide`` to somewhere else.
-For example to `reviewdog <https://github.com/reviewdog/reviewdog>`_ app.
 
 .. code:: yaml
 
   - name: wemake-python-styleguide
     uses: wemake-python-styleguide@latest
-    with:
-      path: './your/custom/path'
-  - name: Custom reviewdog Action
-    runs: echo "{{ steps.wemake-python-styleguide.outputs.output }}" | reviewdog -f pep8
-
-Note, that Github Actions are currently in beta.
+  - name: Custom Action
+    runs: echo "{{ steps.wemake-python-styleguide.outputs.output }}"

scripts/action-config.cfg

@@ -0,0 +1,10 @@
+# We need to overload some default configuration,
+# when running inside Github Actions + ReviewDog.
+# Because we struggle to get the correct formatting.
+
+[flake8]
+# Base flake8 configuration:
+# https://flake8.pycqa.org/en/latest/user/configuration.html
+format = default
+show-source = False
+statistics = False

scripts/entrypoint.sh

@@ -1,13 +1,29 @@
 #!/bin/bash
 
 # Diagnostic output for the passed path:
-echo "Linting path: $1"
+flake8 --version
+echo "Linting path: $INPUT_PATH"
 echo '================================='
 echo
 
-# Runs flake8:
-output=$(flake8 "$1")
-status="$?"
+# Runs flake8, possibly with reviewdog:
+if [ "$INPUT_REPORTER" == 'terminal' ]; then
+  output=$(flake8 "$INPUT_PATH")
+  status="$?"
+elif [ "$INPUT_REPORTER" == 'github-pr-review' ] ||
+     [ "$INPUT_REPORTER" == 'github-pr-check' ]; then
+  # We will need this token for `reviewdog` to work:
+  export REVIEWDOG_GITHUB_API_TOKEN="$GITHUB_TOKEN"
+
+  # Running special version of `flake8` to mathc the `reviewdog` format:
+  output=$(flake8 "$INPUT_PATH" --append-config='/action-config.cfg')
+  echo "$output" | reviewdog -f=pep8 -reporter="$INPUT_REPORTER" -level=error
+  # `reviewdog` does not fail with any status code, so we have to get dirty:
+  status=$(test "$output" = '')
+else
+  output="Invalid reporter specified: $INPUT_REPORTER"
+  status=1
+fi
 
 # Sets the output variable for Github Action API:
 # See: https://help.github.com/en/articles/development-tools-for-github-action