v1.0.0-beta.3

Hopefully, the last beta

We are very close to our final and stable release for v1. Why we didn't ship
this version already? Some people like to move fast & break things. But we
wanted to have a correct design for the v1 so future majors version won't break
too much things.

What we did for v0.x was an experiment and we learned a lot of things by running
this in production on moderate websites. We have spent a lot of brain energy to
make a scalable & flexible solution with a small API surface.

We have designed an architecture that follow React way of thinking, by offering
a centralized lifecycle. With that in
mind we have adjusted our goal and make React & Webpack, 2 core pieces of v0.x,
first class plugins in our v1.x. This opens up lot of nice things to support
React or Webpack alternatives.

Documentation is a work in progress 🎉

First thing to celebrate, we started to work on our website! You can now find
documentation for the core and the
react preset tutorial! More things will
follow!

2 small breaking changes

Below you will find 2 small breaking changes, very easy to adjust (that will
probably consist of adding content in front of all your current existing query
to our content api or to adjust your configuration).

🚨 @phenomic/core: content options now supports multiples entries and globs
(+
fixup)
by @MoOx

All your previous queries won't work with this breaking change but don't
worry, it's very easy to adjust your code.

We are sorry to add a breaking change now, but it's for the better.

Solution 1 (recommended): You will have to add "content/" in front of all
your path in your queries. Why? It's for being more explicit and also be able
to support multiples sources (multiples folders) to avoid duplicate names
without a weird configuration. The upgrade is pretty easy.

For example

const HomeContainer = withPhenomicApi(Home, props => ({
  posts: query({
    path: "posts"
  })
}));

Will become

const HomeContainer = withPhenomicApi(Home, props => ({
  posts: query({
    path: "content/posts"
  })
}));

Solution 2 (for lazy people): You can keep the previous behavior very easily
by using this value for content option: "": {root: "content", globs: ["**/*"]}. You can throw that in your package.json in a phenomic like this:

  "phenomic": {
    "content": {
      "": {root: "content", globs: ["**/*"]}
    },
    "presets": ["@phenomic/preset-react-app"]
  }

🚨 @phenomic/plugin-rss-feed: now lookup in content/posts by default.
by @MoOx

To get previous behavior back, just use this configuration (adjust to match
yours):

  "phenomic": {
    "presets": ["@phenomic/preset-react-app"],
    "plugins": [
      [
        "@phenomic/plugin-rss-feed",
        {
          "feeds": {
            "feed.xml": {
              "query": {
                "path": "content/posts"
              }
            }
          }
        }
      ]
    ]
  },

Bugfixes

🐛 @phenomic/plugin-renderer-react: Handle scroll like it should (to top or to the hash if any)
by @MoOx

🐛phenomic/plugin-collector-files: correctly clean extensions from filename for id (not only md|json but all extensions by the plugin loaded at runtime)
by @MoOx

Tons of small improvements

As we started to actively working on our documentation, we have added tons of
small improvements.

✨ @phenomic/plugin-renderer-react: add render + callback option to createApp/renderApp for custom rendering
by @MoOx

This commit allows you to use a custom render() function. This way to render the
app is required by some solutions that needs a custom wrapper for styles
pre-rendering (eg: react-native-web).

🎉 examples/react-native-web-app
by @MoOx

✨ @phenomic/core: add socketPort option (and switch to 3334 since 1415 can be used by dbstar)
by @MoOx

✨@phenomic/core: Improve error shown from database access
by @MoOx

✨ @phenomic/core: Improve error message from database not found entry
by @MoOx

🔥 Add a new helper to extract meta from body nodes (title + headings list)
(+
fixup)
by @MoOx

This helper is used in @phenomic/plugin-transform-markdown and
@phenomic/plugin-transform-asciidoc to automatically add a title meta
even if you don't have frontmatter. Pretty cool for files that you don't
control. This helper also returns a list of headings in your content, which is
pretty convenient to make a table of content at the top of you page or in a
sidebar.

✨ @phenomic/plugin-transform-markdown: add title fallback extracted from markdown body if no title is in the frontmatter + add a list of markdown headings
(+
fixup)
by @MoOx

✨ @phenomic/plugin-transform-asciidoc: add title fallback extracted from markdown body if no title is in the frontmatter + add a list of markdown headings
(+
fixup)
by @MoOx

✨ @phenomic/plugin-transform-json: add title fallback
by @MoOx

✨ @phenomic/core: better display in case of error during start/build
by @MoOx

✨ @phenomic/core: add options to sort db results
by @MoOx

This feature is pretty awesome and allows you to add custom order to your query.
As a good example, we wanted to sort our showcase with the following algorithm:

• Sites with a flag curated first
• Sites with many tags
• Sites with tags "blog" at the end.

In order to do so, we have introduced this option and we hope you will like it.
You can check out the
documentation for db.sortFunctions
and take a look to the
one we use for the showcase.

✨ @phenomic/cli: add preview command
by @MoOx

✨ @phenomic/core: add preview command to build & serve results
by @MoOx

Now you can run

phenomic preview

This will build and serve the website, to mimic production (as you will have the
pre-rendered files and the client side reconciliation).