This README
is basically the meta-documentation for the Pulsar website and documentation. You will find instructions on running the site locally.
Framework Docusaurus.
Ensure you have installed the latest version of Node. You can install Yarn as well.
You have to be on Node >= 8.x and Yarn >= 1.5.
To run the site locally:
cd website
yarn install
yarn start
Notes
- If you have installed
yarn
, you can skip theyarn install
command.- After you enter the
yarn start
command, you will be navigated to a local address, for example,http://localhost:3000
. ClickDocs
to see documentation for the latest release of Pulsar.- The
http://localhost:3000/en/versions
path shows the documentation for all versions. To view your local changes, clickDocumentation
in Latest Version, or enterhttp://localhost:3000/docs/en/next/standalone
in a browser.
The website is comprised of two parts: one is documentation, the other is website pages (including blog posts).
Documentation related pages are placed in the docs
directory. They are written in Markdown.
All documentation pages are versioned. For more details, refer to versioning.
Website pages are non-versioned. They are placed in the website
directory.
All the markdown files are placed in the docs
directory. It is a flat structure.
├── docs
│ ├── adaptors-kafka.md
│ ├── adaptors-spark.md
│ ├── adaptors-storm.md
│ ├── admin-api-brokers.md
│ ├── admin-api-clusters.md
│ ├── admin-api-namespaces.md
│ ├── admin-api-non-persistent-topics.md
│ ├── admin-api-overview.md
│ ├── admin-api-partitioned-topics.md
│ ├── admin-api-permissions.md
│ ├── admin-api-persistent-topics.md
│ ├── admin-api-tenants.md
│ ├── administration-dashboard.md
│ ├── administration-geo.md
│ ├── administration-load-distribution.md
│ ├── administration-proxy.md
...
All the files are named in the following convention:
<category>-<page-name>.md
<category>
is the category within the sidebar that this file belongs to, while <page-name>
is the string to name the file within this category.
There isn't any constraints on how files are named. It is just a naming convention for better maintenance.
All the documents are usual Markdown files. However you need to add some Docusaurus-specific fields in Markdown headers in order to link them correctly to the Sidebar and Navigation Bar.
id
: A unique document ID. If this field is not specified, the document ID defaults to its file name (without the extension).
title
: The title of the document. If this field is not specified, the document title defaults to its id.
hide_title
: Whether to hide the title at the top of the doc.
sidebar_label
: The text shown in the document sidebar for this document. If this field is not specified, the document sidebar_label
defaults to its title.
For example:
---
id: io-overview
title: Pulsar IO Overview
sidebar_label: Overview
---
To link to other documentation files, you can use relative URLs, which will be automatically converted to the corresponding HTML links when they are rendered.
Example:
[This links to another document](other-document.md)
The markdown file will be automatically converted into a link to /docs/other-document.html (or the appropriately translated/versioned link) once it is rendered.
This helps when you want to navigate through docs on GitHub since the links there are functional links to other documents (still on GitHub), and the documents have the correct HTML links when they are rendered.
We have a remarkable plugin to generate links to the javadoc for Pulsar classes. You can write them in the following syntax:
{@inject: javadoc:<Display Name>:<Relative-Path-To-Javadoc-Html-File>}
For example, the following line generates a hyperlink to the javadoc of PulsarAdmin
class.
{@inject: javadoc:PulsarAdmin:/admin/org/apache/pulsar/client/admin/PulsarAdmin.html}
We use the same remarkable plugin to generate links to files in Pulsar GitHub repository.
You can write it using similar syntax:
{@inject: github:<Display Text>:<Relative-Path-To-Files>}
For example, the following line generates a hyperlink to the dashboard Dockerfile.
{@inject: github:`Dockerfile`:/dashboard/Dockerfile}
For more details about markdown features, read here.
All the sidebars are defined in a sidebars.json
file in the website
directory. The documentation sidebar is named docs
in the JSON structure.
When you want to add a page to sidebar, you can add the document id
you used in the document header to the existing sidebar/category. In the example below,
docs
is the name of the sidebar, "Getting started" is a category within the sidebar, and "pulsar-2.0" is the id
of a document.
{
"docs": {
"Getting started": [
"pulsar-2.0",
"standalone",
"standalone-docker",
"client-libraries",
"concepts-architecture"
],
...
}
}
To add links to the top navigation bar, you can add entries to the headerLinks
of siteConfig.js
under website
directory.
To learn different types of links you can add to the top navigation bar, refer to Navigation and Sidebars.
Documentation versioning with Docusaurus becomes simpler. When done with a new release, just simply run following command:
yarn run version ${version}
This preserves all markdown files in the docs
directory and make them available as documentation for version ${version}
.
Versioned documents are placed into website/versioned_docs/version-${version}
, where ${version}
is the version number
you supplied in the command above.
Versioned sidebars are also copied into website/versioned_sidebars
and are named as version-${version}-sidebars.json
.
If you want to change the documentation for a previous version, you can access files for that respective version.
For more details about versioning, refer to Versioning.
Docusaurus makes it easy to use translation functionality using Crowdin.
All the markdown files are written in English. These markdown files are uploaded to Crowdin
for translation by users within a community. Top-level pages are also written in English.
The strings that are needed to be translated are wrapped in a <translate>
tag.
Pulsar Website Build pulls down and uploads translation for all the Pulsar website documentation files automatically. Once it pulls down translation from Crowdin, it will build the translation into the website.
Translation is stored and managed in the Pulsar Crowdin project. To contribute translation, you can simply create a Crowdin account, join the project and make contributions. Crowdin provides very good documentation for translators. You can read Crowdin Knowledge Base before contributing.
Translation you contribute is licensed under Apache License V2. Pulsar Committers will review translation. If your translation is not reviewed or approved by any committer, feel free to reach out via slack channel or mailing lists.
When you find display issues on the translated pages, you can download the translated docs from Crowdin, and follow the instructions below to debug and fix issues.
- Install Java (optional) If you have installed Java, skip this step. If you have not installed Java, install the latest version. If you are using Mac OS, you can use the following command to install Java:
brew cask install java
- Install Crowdin CLI
To download the translated markdown files, you need to install Crowdin CLI.
- Set environment variables
You need to set the following environment variables:
export CROWDIN_DOCUSAURUS_PROJECT_ID="apache-pulsar"
export CROWDIN_DOCUSAURUS_API_KEY=<crowdin-pulsar-api-key>
You can find the API Key of Pulsar Crowdin project here. Only PMC members and committers are able to retrieve the API key. If the API key is invalid, regenerate.
- Download the translated docs
Now you are ready to download the translated docs from Crowdin.
$ cd ${PULSAR_HOME}/site2/website
# download all translated docs
$ yarn crowdin-download
# download the translated docs for `zh-CN`
$ yarn crowdin-download -l zh-CN
The translated docs are downloaded to the site2/website/translated_docs
directory.
After download the translated documents, you can open the target markdown file, check issues and fix them. To verify if you have fixed the issues correctly, run the site locally.