The compiler for our training material. Sensei is a replacement for zenika-formation-framework.
-
Pull the image with
docker image pull zenika/sensei
-
Create an alias
sensei
alias sensei='docker container run \ --pull always \ --interactive \ --tty \ --rm \ --volume $(pwd):/$(basename $(pwd)) \ --workdir /$(basename $(pwd)) \ --publish ${SENSEI_PORT:-8080}:${SENSEI_PORT:-8080} \ --env SENSEI_PORT \ --env SENSEI_WATCH_POLL \ --cap-add=SYS_ADMIN \ zenika/sensei'
โ When running sensei inside a Docker container, the
--material
is limited to descendants of the working directory.
โ If sensei fails to recompile on changes, try setting
SENSEI_WATCH_POLL
totrue
.
โ To change
SENSEI_PORT
when using this alias, use the following syntax:export SENSEI_PORT=9000; sensei
. See here.
โน๏ธ The following also applies to Colima on macOS.
When bind-mounting files in Docker for Windows with WSL2,
the recommendation
is to store files in the Linux filesystem, i.e. inside WSL2.
The Linux filesystem can then be accessed from Windows through the path \\wsl$\
(or \\wsl.localhost\
).
When using VSCode, you can also use the WSL extension
to edit files on the Linux filesystem.
Therefore it is recommended to clone the training repository in the Linux filesystem then to run the alias from WSL2.
โ If you use the Windows filesystem, hot reload when changing training content won't work out-of-the-box, but you can set
SENSEI_WATCH_POLL
totrue
to enable it.
โ If you use the Windows filesystem and expect to use the alias within Git Bash for Windows, prepend the
--volume
and--workdir
options with an additional slash (i.e.--volume /$(pwd):/$(basename $(pwd)) --workdir //$(basename $(pwd)
). See known issues of Git for Windows. This avoids theC:/Program Files/Git/...: no such file or directory
kind of errors.
- Clone this repo and
cd
into the created folder - Build the image with
docker image build --tag zenika/sensei ./
- Create the same alias as for the published Docker image but without the
--pull always
โ The same warnings as for the published Docker image apply.
- Install with
npm install --global https://github.com/Zenika/sensei
โ You may use Yarn, however it's been known to have cache issues when installing packages from GitHub, resulting in failures to update sensei correctly.
npm install
, otherwise you will have an unused version of chromium installed in ~/.cache/puppeteer
.
# Install chromium without Mac OS Quarantine to allow puppeteer to launch it
brew install --cask chromium --no-quarantine
# Put these variables export in your .bashrc or .zshrc to make sure puppeteer use the right chromium
export PUPPETEER_SKIP_CHROMIUM_DOWNLOAD=true
export PUPPETEER_EXECUTABLE_PATH=`which chromium`
Run sensei --help
for available commands and options.
cd
into a training material folder (must haveSlides/slides.json
andWorkbook/workbook.json
)- Run
sensei pdf
- PDFs are generated inside
pdf
folder
cd
into a training material folder (must haveSlides/slides.json
andWorkbook/workbook.json
)- Run
sensei serve
- Navigate to
http://localhost:8080/
Write slides as you would usual Reveal.js slides. Refer to Reveal's documentation for features. Note that the following plugins are enabled:
- Markdown
- Highlight
- Zoom
- Notes
- Math
Write the workbook as you would a usual markdown document.
โน You may use
<!-- toc -->
to have a table of content inserted at that place.
Install dependencies (npm i
) then use npm start --
to run the CLI
(e.g. npm start -- serve --material=./training-material
where ./training-material
points to directory with training material in it).
You may alternatively use npm run dev --
instead to enable restart on change
(e.g. npm run dev -- serve --material=./training-material
).
Note that npm run dev
requires Node.js 18.11 or later.
Tests are written using the built-in test module, which requires Node.js 18.3 or later.
- Run all tests:
npm test
- Run one test file:
npm test path/to/test/file.test.js
src/app
: source code for the web app that embeds training materialsrc/build
: source code that builds the previously mentioned web app, including the training materialsrc/pdf
: source code that builds PDFs from the training materialsrc/cli
: source code for the CLI
This project uses Prettier. Don't forget to format before committing! You may
use npm run prettier:write
to do that.