JKAN

A lightweight, backend-free open data portal, powered by Jekyll

Open-source data portals can be really hard to install and maintain. But their basic purpose of providing links to download data really isn't that complicated. JKAN is a proof-of-concept that allows a small, resource-strapped government agency to stand-up an open data portal by simply clicking the fork button.

Please note this is still a work in progress! Check out the issues to help out or give feedback.

Demo site

Installation Options

1. See Get Started on jkan.io for an installation wizard,
2. follow the manual installation instructions yourself.
3. Do a manual fork and do not install gatekeeper at all (login bits won't work, but Heroku is no longer free)

For configuration details, see the wiki

Development

The recommended to build the site for development or making changes is docker compose. Just run docker compose up

$ docker compose up
[+] Running 1/0
 ⠿ Container jkan-jekyll-1  Created                                                                                                                                                                                               0.0s
Attaching to jkan-jekyll-1
jkan-jekyll-1  | ruby 2.6.3p62 (2019-04-16 revision 67580) [x86_64-linux-musl]
jkan-jekyll-1  | Configuration file: /srv/jekyll/_config.yml
jkan-jekyll-1  |             Source: /srv/jekyll
jkan-jekyll-1  |        Destination: /srv/jekyll/_site
jkan-jekyll-1  |  Incremental build: enabled
jkan-jekyll-1  |       Generating... 
jkan-jekyll-1  |                     done in 0.025 seconds.
jkan-jekyll-1  |  Auto-regeneration: enabled for '/srv/jekyll'
jkan-jekyll-1  |     Server address: http://0.0.0.0:4000/jkan/
jkan-jekyll-1  |   Server running... press ctrl-c to stop.

Then connect to http://0.0.0.0:4000/jkan/ via a web browser.

Read more about the Architecture on the Wiki.

Metadata generation instruction

Custom scripts are located in the resources folder.

Mugi's version is contained in the resources/metadata_process_scripts directory.
The main workflow involves generating markdown files from an Excel file.
The directory has the following structure:

├── md_old                                              # Roy's version of markdowns
│   ├── 2018-SU2-Trento-Accelerometer Event.md
│   ├── 2018-SU2-Trento-Activities Per Time.md
│   ├── ...
├── sources                                             # Input/Output files
│   ├── 2024-LivePeople_Metadata_Description-v2.xlsx    # Used for data completion
│   ├── 2024_LivePeople PROJECT Metadata.xlsx           # Used for data completion
│   ├── catalog.xlsx                                    # Metadata catalog
│   └── list_of_datasets.csv                            # Dataset list
├── generate_excel.py                                   # Generates Excel from old markdowns
├── generate_markdowns.py                               # Generates markdowns from Excel
├── generate_md_desc.py                                 # Generates metadata descriptions for the website
├── get_dataset_list.py                                 # Generates dataset list for distribution
├── modeling.py                                         # Maps old markdown structure to the new one

How to Use

To generate metadata for future use, follow these steps:

1. Add the new catalog to catalog.xlsx.

2. Run generate_markdowns.py to generate the markdown files (they will be saved in the _datasets folder).

3. Run get_dataset_list.py to generate the list_of_datasets.csv file, which is used for data download requests.

Note:
If you encounter any issues, the older version of the markdowns and scripts is still available in the repository.