Skip to content

Latest commit

 

History

History
944 lines (849 loc) · 31.3 KB

README.md

File metadata and controls

944 lines (849 loc) · 31.3 KB

🧱 Core

Global configuration and options

Supported features
→ Full specification
👤 Users 👥 Organizations 📓 Repositories
🗝️ token 🗝️ committer_token

➡️ Jump to all available options

🌐 Configure used timezone

By default, dates use Greenwich meridian (GMT/UTC).

Configure config_timezone (see supported timezone) to avoid time offsets.

Example: configuring timezone

- uses: lowlighter/metrics@latest
  with:
    config_timezone: Europe/Paris

📦 Ordering content

Content can be manually ordered using config_order option.

Example: display base.header, isocalendar, languages and stars in this specific order

- uses: lowlighter/metrics@latest
  with:
    base: header
    plugin_isocalendar: yes
    plugin_languages: yes
    plugin_stars: yes
    config_order: base.header, isocalendar, languages, stars

💡 Omitted sections will be appended at the end using default order

🪛 Using presets

🚧 This feature is an early implementation and may change before official release

It is possible to reuse the same configuration across different repositories and workflows using configuration presets. A preset override the default values of inputs, and multiple presets can be provided at once through URLs or file paths.

Options resolution is done in the following order:

  • default values
  • presets, from first to last
  • user values

Example: using a configuration preset from an url

- uses: lowlighter/metrics@latest
  with:
    config_presets: https://raw.githubusercontent.com/lowlighter/metrics/presets/lunar-red/preset.yaml

Some presets are hosted on this repository on the @presets branch and can be used directly by using using their identifier prefixed by an arobase (@).

Example: using a pre-defined configuration preset

- uses: lowlighter/metrics@latest
  with:
    config_presets: "@lunar-red"

⚠️ 🔐 Tokens and options marked with ⏯️ Cannot be preset, as they suggest, cannot be preset and thus requires to be explicitely defined to be set.

ℹ️ Presets configurations use schemas to ensure compatibility between format changes

🎨 Custom CSS styling

Additional CSS can be injected using extras_css option.

Example: changing the color of h2

- uses: lowlighter/metrics@latest
  with:
    base: header
    extras_css: |
      h2 {
        color: red;
      }

💡 metrics does not use !important keyword, so use it when having trouble when styling is not applied

💡 If you make an heavy use of this option, creating a community templates may be a better alternative

⚠️ CSS styles may slightly change between releases, backward compatibility is not guaranteed!

🗳️ Custom JavaScript scripting

Additional JavaScript can be injected using extras_js option.

Example: removing all h2

- uses: lowlighter/metrics@latest
  with:
    base: header
    extras_js: |
      document.querySelectorAll("h2")?.forEach(h2 => h2.remove())

ℹ️ JavaScript is executed in puppeteer context during the rendering phase, not in metrics context. It will be possible to access document and all other features accessibles like if the SVG was opened in a browser page

💡 If you make an heavy use of this option, creating a community templates may be a better alternative

⚠️ HTML elements may slightly change between releases, backward compatibility is not guaranteed!

🔲 Adjusting padding

SVG rendering is dependent on operating system, browser and fonts combination and may look different across different platforms.

It may not look like it, but computing the height of a SVG is not trivial. metrics spawns an headless browser and try to do its best to resize the result, but it may sometimes ends up in either cropped or oversized images.

Tweak config_padding option to manually adjust padding and solve this issue.

This settings supports the following format:

  • 1 value for both width and height
  • 2 values for width first and height second, separated by a comma (,)

💡 Both negative and positive values are allowed

Each value need to respect the following format:

  • {number}
  • {number} + {number}%
  • {number}%

💡 Percentage based values are relative to the height computed by puppeteer

Example: add 10px padding for both width and height

- uses: lowlighter/metrics@latest
  with:
    config_padding: 10

Example: add 10px padding to height and increase it by 8%

- uses: lowlighter/metrics@latest
  with:
    config_padding: 0, 10 + 8%

Example: remove 10% from height

- uses: lowlighter/metrics@latest
  with:
    config_padding: 0, -10%

↔️ Controlling display size

Some templates may support different output display size.

A regular display size will render a medium-sized image suitable for both desktop and mobile displays, while a large one will be more suitable only for desktop and some plugins (like 📌 topics or 🏅 contributors)

The columns display will render a full-width image with automatic resizing: two columns for desktop and a single one column for mobiles.

Example: output a PNG image

- uses: lowlighter/metrics@latest
  with:
    config_display: large

💱 Configuring output format

Use config_output to change output format.

Example: output a PNG image

- uses: lowlighter/metrics@latest
  with:
    config_output: png

A JSON output can be used to retrieved collected data and use it elsewhere.

Example: output a JSON data dump

- uses: lowlighter/metrics@latest
  with:
    config_output: json

When using a PDF output, it is advised to set config_base64: yes to encode embed images in base64 in order to make self-contained documents.

Example: output a self-contained PDF document

- uses: lowlighter/metrics@latest
  with:
    markdown: TEMPLATE.md
    config_output: markdown-pdf
    config_base64: yes

✨ Render Metrics insights statically

It is possible to generate a self-contained HTML file containing ✨ Metrics insights output by using config_output: insights.

💡 Note that like ✨ Metrics insights content is not configurable, thus any other plugin option will actually be ignored

Example: output ✨ Metrics insights report

- uses: lowlighter/metrics@latest
  with:
    config_output: insights

🧶 Configuring output action

Using commits (default)

Use config_output: commit to make the action directly push changes to committer_branch with a commit. A custom commit message can be used through committer_message.

💡 metrics will automatically ignore push events with a commit message containing [Skip GitHub Action] or Auto-generated metrics for run # to avoid infinite loops. Note that by default, GitHub already ignore events pushed by ${{ github.token }} or containing [skip ci] in commit message

Example: push output to metrics-renders branch rather than the default branch

- uses: lowlighter/metrics@latest
  with:
    output_action: commit
    committer_branch: metrics-renders
    committer_message: "chore: update metrics"

Using pull requests

Use config_output: pull-request to make the action open a new pull request and push changes from the same run on it.

The last step should use either pull-request-merge, pull-request-squash or pull-request-rebase to merge changes to committer_branch.

💡 When using pull-request output action, do not forget to change filename too or previous output will be overwritten!

Example: push two outputs using a merge pull request

- uses: lowlighter/metrics@latest
  with:
    filename: my-metrics-0.svg
    output_action: pull-request

- uses: lowlighter/metrics@latest
  with:
    filename: my-metrics-1.svg
    output_action: pull-request-merge

Using gists

Use config_output: gist to push output to a GitHub gist instead. It is required to provide a gist id to committer_gist option to make it work.

💡 This feature will use token instead of committer_token to push changes, so gists scope must be granted to the original token first

Example: push output to a gist

- uses: lowlighter/metrics@latest
  with:
    output_action: gist
    committer_gist: xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Manual handling

Use config_ouput: none to perform custom processing with outputs. They will be available under /metrics_renders/{filename} in the runner.

Example: generate outputs and manually push them

- name: Checkout repository
  uses: actions/checkout@v2
    with:
      fetch-depth: 0

- uses: lowlighter/metrics@latest
  with:
    output_action: none

- uses: lowlighter/metrics@latest
  run: |
    set +e
    git checkout metrics-renders
    git config user.name github-actions[bot]
    git config user.email 41898282+github-actions[bot]@users.noreply.github.com
    sudo mv /metrics_renders/* ./
    git add --all
    git commit -m "chore: push metrics"
    git push

♻️ Retrying automatically failed rendering and output action

Rendering is subject to external factors and can fail ocassionaly. Use retries and retries_delay options to automatically retry rendering.

Example: retry render up to 3 times (wait 5 minutes between each fail)

- uses: lowlighter/metrics@latest
  with:
    retries: 3
    retries_delay: 300

Output action is also subject to GitHub API rate-limiting and overall health status and can fail ocassionaly. Use retries_output_action and retries_delay_output_action options to automatically retry output action.

💡 As output action is a separate step from rendering, render step won't be called again

Example: retry output action up to 5 times (wait 2 minutes between each fail)

- uses: lowlighter/metrics@latest
  with:
    retries_output_action: 5
    retries_delay_output_action: 120

🗜️ Optimize SVG output

To reduce filesize and decrease loading time, metrics offers several optimization options, such as purging unused CSS and style minification, XML pretty-pretting (which also reduce diffs between changes) and general SVG optimation (still experimental).

💡 This option is enabled by default!

Example: optimize CSS and XML

- uses: lowlighter/metrics@latest
  with:
    optimize: css, xml

Example: optimize SVG (experimental)

- uses: lowlighter/metrics@latest
  with:
    optimize: svg
    experimental_features: --optimize-svg

🐳 Faster execution with prebuilt docker images

When using lowlighter/metrics official releases as a GitHub Action, a prebuilt docker container image will be pulled from GitHub Container Registry. It allows to significantly reduce workflow execution time.

💡 This option is enabled by default!

On forks, this feature is disable to take into account any changes you made on it.

Example: using prebuilt docker image

- uses: lowlighter/metrics@latest
  with:
    use_prebuilt_image: yes

➡️ Available options

TypeDescription
token

GitHub Personal Access Token

No scopes are required by default, though some plugins and features may require additional scopes

When using a configuration which does not requires a GitHub PAT, you may pass NOT_NEEDED instead

✔️ Required
🔐 Token
type: token
user

GitHub username

Defaults to token owner username.

⏯️ Cannot be preset
type: string
repo

GitHub repository

This option is revevalant only for repositories templates

⏯️ Cannot be preset
type: string
committer_token

GitHub Token used to commit metrics

Leave this to ${{ github.token }} or ${{ secrets.GITHUB_TOKEN }}, which is a special auto-generated token restricted to current repository scope.

💡 When using output_action: gist, it will use token instead, since gists are outside of scope

🔐 Token
type: token
default: ${{ github.token }}
committer_branch

Target branch

Default value is set to your repository default branch

type: string
committer_message

Commit message

Use ${filename} to display filename

type: string
default: Update ${filename} - [Skip GitHub Action]
committer_gist

Gist id

Specify an existing gist id (can be retrieved from its URL) when using output_action: gist.

⏯️ Cannot be preset
type: string
filename

Output path

When using an asterisk (*), correct extension will automatically be applied according to config_output value

type: string
default: github-metrics.*
markdown

Markdown template path

It can be either a local path or a full link (e.g. https://raw.githubusercontent.com)

type: string
default: TEMPLATE.md
markdown_cache

Markdown file cache

type: string
default: .cache
output_action

Output action

  • none: just create file in /metrics_renders directory of action runner
  • commit: push output to committer_branch
  • pull-request: push output to a new branch and open a pull request to committer_branch
  • pull-request-merge: same as pull-request and additionaly merge pull request
  • pull-request-squash: same as pull-request and additionaly squash and merge pull request
  • pull-request-rebase: same as pull-request and additionaly rebase and merge pull request
  • gist: push output to committer_gist

💡 When using pull-request, you will need to set the last job with a pull-request-* action instead, else it won't be merged

type: string
default: commit
allowed values:
  • none
  • commit
  • pull-request
  • pull-request-merge
  • pull-request-squash
  • pull-request-rebase
  • gist
output_condition

Output condition

  • always: always try to push changes
  • data-changed: skip changes if no data changed (e.g. like when only metadata changed)
type: undefined
optimize

Optimization features

  • css: purge and minify CSS styles
  • xml: pretty-print XML (useful to reduce diff)
  • svg: optimization with SVGO (experimental, require --optimize-svg experimental flag)

Some templates may not support all options

type: array (comma-separated)
default: css, xml
allowed values:
  • css
  • xml
  • svg
setup_community_templates

Community templates to setup

See community templates guide for more informations

🌐 Web instances must configure settings.json
type: array (comma-separated)
template

Template

Community templates must be prefixed by at sign (@) See list of supported templates

type: string
default: classic
query

Query parameters

Pass additional parameters to templates. This is mostly useful for custom templates.

⚠️ Do not use this option to pass other existing parameters, they will be overwritten

type: json
default: {}
extras_css

Extra CSS

Custom CSS that will be injected in used template. Useful to avoid creating a new template just to tweak some styling

🌐 Web instances must configure settings.json
type: string
extras_js

Extra JavaScript

Custom JavaScript that will be executed during puppeteer rendering. Useful to avoid creating a new template just to tweak some content.

Note that is it executed within puppeteer context and not metrics context. It is run after transformations and optimizations, but just before resizing.

✨ On master/main
🌐 Web instances must configure settings.json
type: string
config_timezone

Timezone for dates

See list of supported timezone

⏭️ Global option
type: string
config_order

Plugin order

By default, templates use partials/_.json ordering. You can override the content order by using this setting.

If some partials are omitted, they will be appended at the end with default ordering

⏭️ Global option
type: array (comma-separated)
config_twemoji

Use twemojis

Replace emojis by twemojis to have a consistent render across all platforms May increase filesize.

⏭️ Global option
type: boolean
default: no
config_gemoji

Use GitHub custom emojis

GitHub supports additional emojis which are not registered in Unicode standard (:octocat:, :shipit:, :trollface:, ...) See full list at https://api.github.com/emojis.

May increase filesize

⏭️ Global option
type: boolean
default: yes
config_display

Display width (for image output formats)

  • regular: 480px width
  • large: 960px width (may not be supported by all templates)
  • columns: Full width with auto-sizing (two columns for desktops, and one column for mobile)
⏭️ Global option
type: string
default: regular
allowed values:
  • regular
  • large
  • columns
config_animations

Use CSS animations

⏭️ Global option
type: boolean
default: yes
config_base64

Base64-encoded images

Enable this option to make self-contained ouput (i.e. with no external links)

⏭️ Global option
type: boolean
default: yes
config_padding

Output padding

Although metrics try to auto-guess resulting height, rendering is still dependent on OS and browser settings. It can result in cropped or oversized outputs.

This settings let you manually adjust padding with the following format:

  • 1 value for both width and height
  • 2 values for width fist and height second, separated by a comma (,)

Each value need to respect the following format:

  • {number}
  • {number} + {number}%
  • {number}%

Percentage are relative to computed dimensions

type: string
default: 0, 8 + 11%
config_output

Output format

  • auto: Template default (usually svg or markdown)
  • svg: SVG image
  • png: PNG image (animations not supported)
  • jpeg: JPEG image (animations and transparency not supported)
  • json: JSON data dump
  • markdown: Markdown rendered file
  • markdown-pdf: PDF from markdown rendered file
  • insights: Metrics Insights self-contained HTML file (not configurable)
type: string
default: auto
allowed values:
  • auto
  • svg
  • png
  • jpeg
  • json
  • markdown
  • markdown-pdf
  • insights
config_presets

Configuration presets

⏯️ Cannot be preset
✨ On master/main
type: array (comma-separated)
retries

Retries in case of failures (for rendering)

type: number (1 ≤ 𝑥 ≤ 10)
default: 3
retries_delay

Delay between each retry (in seconds, for rendering)

type: number (0 ≤ 𝑥 ≤ 3600)
default: 300
retries_output_action

Retries in case of failures (for output action)

type: number (1 ≤ 𝑥 ≤ 10)
default: 5
retries_delay_output_action

Delay between each retry (in seconds, for output action)

type: number (0 ≤ 𝑥 ≤ 3600)
default: 120
delay

Job delay

This can be used to avoid triggering GitHub abuse mechanics on large workflows

type: number (0 ≤ 𝑥 ≤ 3600)
default: 0
use_prebuilt_image

Use pre-built docker image from GitHub container registry

It allows to save build time and make job significantly faster, and there is almost no reason to disable this settings. This option has no effects on forks (images will always be rebuilt from Dockerfile)

⏯️ Cannot be preset
🔧 For development
type: boolean
default: yes
plugins_errors_fatal

Fatal plugin errors

When enabled, the job will fail in case of plugin errors, else it will be handled gracefully in output with an error message

⏯️ Cannot be preset
🔧 For development
type: boolean
default: no
debug

Debug mode

This setting is automatically enable if a job fail (useful with plugins_errors_fatal: yes)

⏯️ Cannot be preset
🔧 For development
type: boolean
default: no
verify

SVG validity check

⏯️ Cannot be preset
🔧 For development
type: boolean
default: no
debug_flags

Debug flags

  • --cakeday: simulate registration anniversary
  • --hireable: simulate "Available for hire" account setting
  • --halloween: enable halloween colors
  • --error: force render error
⏯️ Cannot be preset
🔧 For development
type: array (space-separated)
allowed values:
  • --cakeday
  • --hireable
  • --halloween
  • --error
dryrun

Dry-run

Contrary to output_action: none, output file won't be available in /metrics_renders directory

⏯️ Cannot be preset
🔧 For development
type: boolean
default: no
experimental_features

Experimental features

No backward compatibility is guaranteed for these features

⏯️ Cannot be preset
🔧 For development
type: array (space-separated)
allowed values:
  • --optimize-svg
use_mocked_data

Use mocked data instead of live APIs

⏯️ Cannot be preset
🔧 For development
type: boolean
default: no