**These docs are meant to be viewed on [docs.oceanprotocol.com](https://docs.oceanprotocol.com). You can still browse them here but links or images might not work in some places.**
That content is organized into subfolders corresponding to the sections mentioned above. The subfolder names, along with the filenames, define the final URL of a given piece of content.
Additionally, some content files live in other repositories and are maintained there. They are pulled into Gatsby's data layer at build time, and pages are created automatically for them as defined in [gatsby-node.js](gatsby-node.js).
For including a document from any of the above repositories, 4 values are required in a document's YAML frontmatter. If found, a page will be generated automatically, accessible under the defined `slug`.
This will NOT include this page in the doc's sidebar navigation, this needs to be done manually in the docs repo in one of the sidebar files. This is so you can check out everything before exposing it to visitors, and to ensure editorial workflow of the categorization.
TODO: In a later stage, we will include more documentation in this site, e.g.:
2. Don't include the page title or description in the Markdown section. That is, don't begin the Markdown content with `# This is the Title in Title Case`. Just write as if that were already there.
3. start your heading levels with `h2`, so `## My heading`
4. Internal links to other docs pages should be:
- to a absolute URL without the host, that looks like `/concepts/terminology/` with slashes on the beginning and end, and with no `.md` or `.html` at the end (before the last slash).
5. no TOC please, this will be generated automatically from all headings
6. for images and media, you can keep them in the original repo. Images will be automatically grabbed by the docs site on querying. When doing that, docs site will generate all sorts of image sizes to handle proper responsive images, so no need to keep an eye on image dimensions or file sizes
**Have a look at [docs.oceanprotocol.com/test/](https://docs.oceanprotocol.com/test/) to see what content elements can be used in all Markdown files included in docs site.**
Every article on [docs.oceanprotocol.com](https://docs.oceanprotocol.com) ends with an _Edit this page on GitHub_ link. Clicking that will put you to the right place in the repository from where you can hit the _Edit_ pencil icon on GitHub.
GitHub will automatically fork the repository if you are not part of the `oceanprotocol` organisation on GitHub. Members can simply push to a new branch on the original repo.
On the front page it is used to show an overview of all our key repositories. This repositories list is sourced from the [`/data/repositories.yml`](data/repositories.yml) file, defining the grouping, the display order, which repos to include, and what additional links to show for every repository.
Including a repo on the front page requires only the `name` key and value, and it needs to be exactly the same as the repo name on GitHub:
The GitHub link is automatically added for every repository and will always be displayed.
#### Release Versions
The displayed version number is based on the tag name of the latest release for a given repository. That means only GitHub releases will trigger a version number update, creating a new Git tag alone is not sufficient.
This Git repo has [Git submodules](https://git-scm.com/book/en/v2/Git-Tools-Submodules). They are the subdirectories of `external/`. Each one is basically another Git repo, i.e. an external Git repo that just _looks_ like it's part of this Git repo. It's very easy to get Git into a confusing state when there are submodules. If you follow the following tips, you should probably be okay:
1. Don't edit anything in the submodules (i.e. in `external/`). Instead, edit it over in the other Git repository and merge your changes over there.
1. Get all the submodules up-to-date using:
```bash
git submodule update --remote --recursive
```
That will get each submodule up-to-date with the HEAD commit of some branch in a remote repository. Which branch? That's set in the `.gitmodules` file. You can check the current commit hashes of all the submodules using `git submodule status`
1. Before doing any `git checkout ...` or other normal Git operations, do this:
```bash
git config --global submodule.recurse true
```
That's like adding the `--recurse-submodules` option to all those Git commands (except for `git clone`) so you don't have to.
### Using npm
As a prerequisite, you'll need the following on your machine:
To enforce a consistent code style, linting is setup for pretty much every file. Linting is part of the test suite, meaning builds on Travis will fail in case of linting errors.
In this repo the following tools are setup for that:
The GitHub GraphQL API integration is done through [gatsby-source-graphql](https://www.gatsbyjs.org/packages/gatsby-source-graphql/) and requires authorization.
An environment variable `GITHUB_TOKEN` needs to present, filled with a [personal access token](https://github.com/settings/tokens) with the scope `public_repo`.
When running the site locally, you can use the GraphiQL client running under [localhost:8000/\_\_\_graphql](http://localhost:8000/___graphql) to explore the whole GraphQL layer of the site (not just the GitHub infos).
This query should get you started to explore what information you can get from GitHub. All that is described in [GitHub GraphQL API](https://developer.github.com/v4/) can be used :
The GitHub GraphQL API is only queried on build time, further GitHub updates on client side need to be done through additional fetch API calls. At the moment this is done for the [repositories](#repositories) component, where the stars and forks numbers are updated on client-side.
We use [github-projects](https://github.com/oceanprotocol/github-projects) for all communications with the [GitHub REST API v3](https://developer.github.com/v3/), deployed on [Now](https://zeit.co/now).
This microservice should be used for all client-side integrations for performance and security reasons, required changes in data structure should be done over there. This service does data refetching automatically, caches results for 15min, and it has access to a secret GitHub token for making authorized API calls.
As a next step, using the REST API could be made obsolete by using some GraphQL client like [Apollo](https://www.apollographql.com) to query GitHub's GraphQL API on run time too.
The deploy command simply calls the [`scripts/deploy.sh`](scripts/deploy.sh) script, syncing the contents of the `public/` folder to S3:
```bash
npm run deploy
```
Requires authorization against AWS with [one of the various ways](https://docs.aws.amazon.com/sdk-for-javascript/v2/developer-guide/setting-credentials-node.html), on Travis this is done with those environment variables:
