Using Doctor on Azure DevOps to generate your documentation

GitHub Actions is hot these days, but still, Azure DevOps is being used in many companies, and I do as well. Doctor a tool for publishing your markdown documentation on SharePoint, works nicely on Azure DevOps as well.

The credentials

To start, you need to choose how you will publish your documentation to SharePoint. Doctor allows three types of authentication.

1. Certificate authentication
2. Username and password
3. Device code

The third option is only preferred when using it on your local machine. When you want to use certificate authentication, you need to do some steps before making use of it.

Using the credentials

On Azure DevOps, I recommend you make use of a variable group to connect to an Azure Key Vault. That way, all your secrets are secure and easy to be changed from one place.

Using a variable group is always a good idea; you do not need to manage your variables in your pipelines.

Show image

The pipeline

The pipeline itself is relatively straightforward — the azure-pipelines.yml file contents look like this:

1
trigger:
2
- main
3
- dev
4

5
variables:
6
  - group: doctor-sharepoint
7

8
stages:
9
- stage: 'publish'
10
  displayName: 'Publish to SharePoint'
11
  jobs:
12
  - job: 'mkdocs_build'
13
    displayName: 'Building Documentation'
14
    pool:
15
      vmImage: ubuntu-latest
16
    steps:
17
      - script: |
18
          debug=""
19
          if [[ $(System.Debug) = true ]]; then
20
            debug="--debug"
21
          fi
22

23
          cleanStart=""
24
          if [[ $(build.SourceBranchName) = "main" ]]; then
25
            cleanStart="--cleanStart --confirm"
26
          fi
27

28
          sudo npm i @estruyf/doctor@next -g
29
          doctor publish --auth certificate --certificateBase64Encoded $(CERTBASE64) --password $(CERTPASSWORD) --appId $(APPID) --tenant $(TENANT) --url $(URL) $cleanStart $debug
30
        displayName: 'Publish via Doctor'

Under the variables property, the variable group called doctor-sharepoint in my case gets retrieved, which includes all the required environment variables.

The pipeline will get triggered whenever there is a push to the main or dev branch and does a clean-up first on main. If you set the debug system variable to true. Doctor will automatically run in debug mode, which gives you extra logging.

Running the pipeline

Running the pipeline on Azure DevOps will give you the following result:

Show image