sanger-pathogens/sanger-pathogens.github.io

Summary of Sanger Pathogen's Repos

Organisation Summary

Creates a summary of an organisation's Github repos and publishes it to github.io.

Contents

• Introduction
• Usage (for sanger-pathogens)
• Automated usage (for sanger-pathogens)
• Things to change (not for sanger-pathogens)
• License
• Feedback/Issues

Introduction

Github Pages serves static content from the master branch of <your-org-name>/<your-org-name>.github.io on https://<your-org-name>.github.io. In our case, this is https://sanger-pathogens.github.io.

Usage (for sanger-pathogens)

Because master is served to the world, development in this repo is done on the code branch.

update_repo_data.py queries the Github API to get the latest data and 'score' each repo; this data is stored in json.
update_pages.py parses this data and uses a template to create a static html page.
serve_local.sh can be used to serve a copy locally on port 8080.
deploy.sh takes your local changes in the site/ directory, copies the contents into a temporary copy of the master branch and then pushes them to the master branch on Github.

You need to provide a username and api token to get the data from Github. This is entered in config.yml. If you're automating daily updates (via cron) then you probably also want to use an SSH key for Github. It's probably worth setting up a repo specific deploy key to reduce the impact if your key is compromised.

• clone this repo
• create a personal access token
• update config.yml
• Run update_repo_data.py to download the latest data
• Run update_pages.py to update the html
• Run serve_local.py to check your updates (see http://localhost:8080)
• Run deploy.sh to deploy your changes to github.io

Automated usage (for sanger-pathogens)

• clone the repo using SSH onto a server
• create a personal access token
• update config.yml
• create a deploy key and install it on the server (without a password)
• create a cron job to run cronrun.sh

Because we're automatically making and pushing commits, these scripts will only delete and modify files already known to Github to reduce the risk of a sensitive file being distributed to the world. If you want to start serving a new file, you will need to manually push it to the master branch on the first occasion.

cronrun.sh is a script we need to use because of potential file permissions incompatibility between the user we run the cron job with and the location our software is installed in. Other teams probably just need to use blind_update.sh instead.

Things to change (not for sanger-pathogens)

• fork the repo
• clone and checkout the code branch git clone --branch code <URL>
• Update the following files:
  • the config
  • default config
  • the template
  • the logo
  • the favicon
• delete the master branch git branch -D master
• create a new master branch git subtree split --prefix site -b master
• force push to master (danger!) git push -f origin master
• rename the repo to <your-org-name-here>.github.io

NB Your pages won't be available after you fork; you need to actually push some changes to master before they're browsable.

License

• The scripts, config and templates are licenced under GPLv3.
• All rights are reserved for the images including the Wellcome Trust Sanger Institute logo. Please don't use them unless otherwise authorised to do so.
• The distinct CSS and Javascript libraries are licenced under their respective licences. Please review the source for details.

Feedback/Issues

Please report any issues to path-help@sanger.ac.uk.