As of August 2018, please open all Pull Requests DIRECTLY TO THE MASTER BRANCH ON THE PUBLIC VERSION OF THIS REPOSITORY. THERE IS NO MORE NEED FOR THE INTERNAL REPOSITORY
This site is maintained by the Product Communications and Experience team. Please contact [email protected] for issues, questions, and the such.
This repository generates LivePerson's Developer Center, which can be found at https://developers.liveperson.com. The site is generated using Jekyll. If you find an issue with the documentation, site structure, meta or anything else, please open an issue and we'll respond as soon as possible.
Table of Contents
- π‘ Updating the Documentation
- π¨ Building the Site Locally
- π Template
- π Licensing
- π» How to Hide files and still make them Accessible
All pages on the site correspond to a Markdown file (.md) which can be found inside pages/Documents
.
To update a file, please branch off of the master
branch, edit the file in question and create a Pull Request back to the master branch.
There's no need for the old Development branch, so please don't create pull requests to it.
Before making any commits, please make sure to read the Updating/Creating Headers section.
There is now a github precommit hook that makes sure you follow the rules on markdown file creation.
This hook will run on every commit and deny commits if they fail the test.
The errors will be outputed to ./_scripts/docOutputError.log
.
If you are adding new content, please make sure you are updating the content in documentsupdated.yaml file.
Our tests will use the documents updated yaml file as the source of truth. So make sure your header naming structure matches the documentsupdated.yaml.
File name rules:
- All markdown files must match the pagename that is provided in the headers
- Spaces in pagename must be replaced with a '-'
- All words are lowercased
- All special characters excluding periods and dashes need to be removed from the filename
- Periods are replaced with dashes.
Folder name rules:
- They must be TitleCase
- All files in the folder, must include a reference in its header to its folder name. This will be in either the documentname, categoryname, or subfoldername ( depending on the location of the folder)
- All special characters excluding periods, dashes, and & need to be removed from Foldername
- Periods are replaced with dashes.
E.G pagename : Customizing the Conversational Cloud!
should be filename customizing-the-conversational-cloud.md
E.G documentname: Add Agent Widgets
should be a folder with name AddAgentWidgets
- Category name will always be the top most folder in the sidebar
Example Normal Layout:
- categoryname: Agent Experience
image: agent-experience-new
documents:
- documentname: Add Agent Widgets
pages:
- pagename: Add Your Own Widgets to the Agent Workspace
- documentname: Agent Workspace Widget SDK
pages:
- pagename: Overview
- pagename: Limitations
- pagename: Methods
- pagename: Public Model Structure
- pagename: Public Properties
- pagename: Best Practices and Troubleshooting
- pagename: Release Notes
- documentname: Chat Agent API
basedomain: https://{domain}/api/account/{accountId}/agentSession
pages:
- pagename: Overview
The Top layer 0 in this structure is the category name Agent Experience. Its folder name isAgentExperience
.
Add Agent Widgets is a folder in layer 1 with path AgentExperience/AddAgentWidgets
The Add Agent Widgets folder only contains one page in it. AgentExperience/AddAgentWidgets/add-your-own-widgets-to-the-agent-workspace.md
the permalink for the file must include all parent folders excluding the layer 0 categoryname. For above example the permalink is:
add-agent-widgets-add-your-own-widgets-to-the-agent-workspace.html
If you look at the next document name Agent Workspace Widget SDK
it is still on level 1 with path AgentExperience/AgentWorkspaceWidgetSDK/
The files in the folder are all listed below pages until you get to last pagename.
IE documentname Chat Agent API is not a folder located inside Agent Workspace Widget SDK
Example Subfolder Layout:
- categoryname: Conversational AI
image: agent-experience
documents:
- documentname: Conversational AI
pages:
- pagename: Overview
- documentname: Conversation Builder
pages:
- pagename: Tutorials & Guides
subpages:
- subpagename: Getting Started
- subpagename: Using Meta Intents with Conversation Builder
- subpagename: Implementing a Web View Integration
- subpagename: Using LivePerson Functions with a Bot
- pagename
Tutorial & guides
because it has subpages is actually a folder not a file - in the markdown file for
Using Meta Intents with Conversation Builder
its pagename must match the subpagename. - the file must include
Tutorial & Guides
as asubfoldername
header since it is at level 2 - the file must include
Conversation Builder
as adocumentname
in the header. - the permalink must be
conversation-builder-tutorials-guides-using-meta-intents-with-conversation-builder.html
- notice how the
&
in the subfoldername is replaced with a dash in the permalink.
Example Documentname is pagename:
- categoryname: Getting Started
image: getting-started
documents:
- documentname: Getting Started with your Free Trial Account
- documentname: Do More with the Conversational Cloud
- documentname: Customizing the Conversational Cloud
- documentname: Starting with your Concierge Bot
- documentname: API Guidelines
pages:
- pagename: Accessing LivePerson APIs
- pagename: Create API keys
- pagename: Domain API
- pagename: Data APIs
- pagename: API Data Metrics
- pagename: Engagement Attributes
- pagename: Analytics Builder Data Metrics
- pagename: Retry Policy Recommendations
- Documentname name is
Getting Started with your Free Trial Account
is not a folder name becauase it does not contain a pages key below it - Since this above file only contains one parent, there should not be a document name in the file
getting-started-with-your-free-trial-account.md
- The pagename for
getting-started-with-your-free-trial-account.md
must match the documentname in the yaml fileGetting Started with your Free Trial Account
To updated production and staging enviornments, create a pull request for master or Staging. When pull request is merged a automated release cycle will start and publish those changes in around five minutes.
-
Production (built from the
master
branch): https://developers.liveperson.com/ -
Staging (built from the
Staging
branch): https://developers.liveperson.com.staging-vdt2zeq-jlynhjefjcpgg.us.platform.sh/
Jekyll uses a front-matter to arrange and define the various documents in the site. This is the text which appears in between the "---" at the top of each document. It's technically a YAML snippet, so all YAML formatting and rules apply to it. Our headers are usually comprised of the following key/value pairs:
-
pagename
: This is the name of the page that will appear at the top of the document. -
keywords
: This replaces the keywords found in the<meta>
tag of the page. Leave it unpopulated. -
sitesection
: This key accepts eitherDocuments
orSolutions
. This designates which part of the site the document is under. -
categoryname
: This is the category to which the document's API belongs (for example, the "Create Users" method belongs to the Users API which is under Contact Center Management). -
documentname
: This is the API to which the document belongs. -
subfoldername
: This is a sub-folder to which the document belongs, if there is one. -
permalink
: this key defines the link at which the document can be found. The format of this value MUST BE as follows. Any other value format will cause the sidebar to malfunction:-
If the page has a
subfoldername
value: documentname - subfoldername - pagename. For example:mobile-app-messaging-sdk-for-android-advanced-features-audio-messages.html
-
If the page does not have a
subfoldername
value: documentname - pagename. For example: users-api-overview.html
-
-
indicator
: this key sets the Chat or Messaging indicator (or both) on a document. It acceptschat
,messaging
orboth
as its value. -
layout
: set the value tohidden-layout
if you want this file to be ignored in search engines
- Add the Files to the Hidden->Hidden folder.
- Do not include them to documentsupdated.yaml file.
- Make sure Layout header is set to
hidden-layout
(This specifically makes it so search engines can't find it) - Make sure that the headers are set correctly follow the file header structure of
blank.md
in the Hidden->Hidden folder
Any other parameters which are not documented here which you might find in the front-matter are deprecated and are only present for backwards compatibility purposes. These should not be used.
Once you've created a new document, you'll need to have it manually added. We chose a manual process for the sidebar for a few reasons. First, it reduces the fragility of the sidebar (the extra, manual step gives us another layer of QA). Secondly, it increases the flexibility of the sidebar (we write code once and then maintain a YAML file, making it easier to add options). Lastly, it decreases site build times (since the forloops
needed to dynamically build a sidebar in a site of our size and complexity are time and resource consuming).
The sidebar's YAML file can be found in the _data
folder. It's called documentsupdated.yaml
. You must make sure the name of the file and the pagename in the sidebar correspond; the link the sidebar sends to is auto-generated and must match the permalink
in the file's header (see above). Please make sure the markdown file created contains its pagename, documentname, categoryname, and permalink in its header. The markdown file might need more information depending on where it will need to be in the sidebar.
The max width for an image in this repo is 800px.
If you have not already done so, make sure your computer has Ruby installed. Here's a helpful guide on how best do that on Mac (you can stop once Ruby is installed, you don't need Rails) and on any other system.
Once you have installed Ruby, clone this repository to your machine. Once done, navigate to it using Terminal or your preferred command line interface. Follow the steps below to run the site from your machine. If you're on Windows, don't forget to run your CLI as an admin.
First time install
- Run
npm install
- Run
npm run serve
- Navigate to http://localhost:4000/ (or the port you chose) and you'll see the site.
OSX Installation
- We recommend installing a standalone Ruby Installation and RVM
- See this for an example: Stack Overflow
Serving the site after the first install
You have two options to run the site locally after the first install:
-
Using gulp.js. Gulp is a toolkit for automating painful or time-consuming tasks. By simply typing in
gulp
in your command line, it takes care of all the build commands needed to serve the site. It also watches the root directory and will automatically refresh your browser once any changes were built. Gulp and its dependencies are installed locally in the project, so there's no further installation needed from your end. -
Using Jekyll's standard commands. All you need to run in consequent builds of the site is
bundle exec jekyll serve
. You can add the suffix--incremental
to enable incremental building of the site. This saves build times since the regeneration feature is enabled by default (the site rebuilds every time you hit "save"). When--incremental
is used, Jekyll won't rebuild the entire site on every save, only the affected sections. If you'd like the project to automatically open in a new tab, you can add the-o
flag to the end of the above command.
Note: changes that alter site navigation or other changes that change the site as a whole might not show up when using --incremental
. If that occurs, simply "kill" the build and run bundle exec jekyll serve
without the suffix. This is also true for gulp: you will need to kill your gulp instance and then run the direct Jekyll command.
See the _template
folder above for a complete template of a simple REST API. Other templates will follow in the future. However, if you have a unique API to document or need further assistance, please reach out to Product Communications before starting to write your document so that we can advise on its structure.
Aloglia is the tool we use for the search bar. It generates a list of searchable items by indexing it in their dashboard which is then pulled into the search bar within the project
To get the latest data to be added into algolia simply do a pull request on the production branch (master)
All usage of the contents, documentation or code found in this repository is subject to the LivePerson API Terms of Use. Please use the link above to read them carefully before utilizing the site.