PandaGo 🐼

Pandoc, wrapped in Go.

Cloning this Project

It's better to use go get instead of git clone so that your repo ends up somewhere in the $GOPATH (which is relied on by govendor).

go get github.com/scripted/pandago
cd $GOPATH/src/github.com/scripted/pandago

PandaGo Locally

You can run the server directly.

go run main.go

Compile and Run a la Heroku

go install -v ./...
heroku local

Or run the app directly using pandago

go install -v ./...
pandago

Ping root to see if it's working

curl -i localhost:8080/ping
#
# HTTP/1.1 200 OK
# Content-Type: application/json; charset=utf-8
# Date: Tue, 11 Oct 2016 21:53:14 GMT
# Content-Length: 19
#
# {"message":"OK 🐼, Go!"}

Dependencies with govendor

PandaGo uses govendor to manage its dependencies.

go get -u github.com/kardianos/govendor

You can install all dependencies described in vendor/vendor.json into the vendor/ directory by running the following.

govendor sync

To update dependencies as described in the app, run the following.

govendor add +external
govendor remove +unused

For more info, checkout out Heroku's page on govendor.

Deploying to Heroku

Create the Heroku app as usual:

heroku create

Add the official heroku/go buildpack as well as the a buildpack for installing custom binaries so we can install Pandoc via .custom_binaries:

heroku buildpacks:add heroku/go
heroku buildpacks:add https://github.com/tonyta/heroku-buildpack-custom-binaries#v1.0.0

That's it. Then just push it to Heroku like normal.

git push heroku master
curl -i https://<YOUR_APP_NAME>.herokuapp.com/ping
#
# HTTP/1.1 200 OK
# Server: Cowboy
# Connection: keep-alive
# Content-Type: application/json; charset=utf-8
# Date: Tue, 11 Oct 2016 21:51:06 GMT
# Content-Length: 19
# Via: 1.1 vegur
#
# {"message":"OK 🐼, Go!"}

Compiling Pandoc

There might be a time we need to recompile Pandoc. If that day comes, here are some steps and tips.

Heroku Cedar-14 Environment

You can get a fresh Cedar-14 environment using Docker.

docker -it -v ~/local-shared-dir/:/container-shared-dir/ heroku/cedar:14 bash

You should probably go ahead and apt-get update

Pandoc Dependencies

Pandoc needs hsb2hs so we'll install Haskell Platform to get it using cabal. Then, append the Cabal binary path to $PATH so hsb2hs can be found.

apt-get update
apt-get install haskell-platform
cabal update
cabal install hsb2hs
export PATH=$PATH:/root/.cabal/bin/

We'll also install Haskell Stack, which we will use to actually build Pandoc.

curl -sSL https://get.haskellstack.org/ | sh
stack setup

Compiling Pandoc

Check the INSTALL instructions included in the Pandoc source.

We'll use Haskell Stack to install Pandoc, making sure to embed data-files so that the binary is relocatable. These include template files among others that Pandoc needs to create .docx files, for example.

wget https://hackage.haskell.org/package/pandoc-1.17.0.3/pandoc-1.17.0.3.tar.gz
tar xvzf pandoc-1.17.0.3.tar.gz
cd pandoc-1.17.0.3
stack install pandoc --flag pandoc:embed_data_files

Tarballing and Deploying

The Custom Binaries buildpack expects a tarball.

cp /root/.local/bin/pandoc /container-shared-dir/
cd /container-shared-dir/
tar -cvzf pandoc-embedded.tar.gz pandoc

Then we can just find pandoc-embedded.tar.gz at ~/local-shared-dir/ on our local machine.

From here, all we have to do is host it publicly (e.g, on AWS S3) and update our .custom_binaries manifest.