No site would be complete without content. In Liara, content is provided in a content directory. Documents within that directory get processed through templates and routed. For example, a document placed in
content-root/foo/bar.md will be routed to
/foo/bar/index.html. The content root can be set in the Configuration.
In addition to documents static files can be also placed in the content directory. Those are just copied to the output without processing. Static files within the content root can use metadata.
If you have static files without any metadata, they should be placed into the static directory. In this case they get directly copied into the output without any extra processing.
Static files get always symlinked to the output if possible, no matter if they’re in the content or static directory tree. The difference is how metadata is handled.
An example content layout could be:
. ├── content │ ├── about │ │ ├── _index.md // Document │ │ ├── about1.mp4 // Static file │ │ ├── about1.meta // Metadata file │ │ ├── about2.mp4 // Static file │ │ └── about2.meta // Metadata file │ └── blog │ ├── first-blog-post.md // Document │ └── second-blog-post.md // Document └── static └── background-image.jpg // Static file
In the example above, the about movie files in
/content/about/ are binaries with metadata attached to them, making them discoverable (for instance, one could store a caption in the metadata and reference it in a template.) The background image on the other hand would get copied over without any processing.
The bulk of the content are document nodes – Markdown or Html files which get processed by Liara to Html and which get templates applied. Liara supports some common Markdown extensions to handle tables and code snippets.
By default, every file is mapped directly to an output URL. For example, a file named
/about/index.md will produce a page at
/about/index. Sometimes it’s desirable to create a page at the parent location, for instance, when you want to move some content around. In this case, an
_index file can be used. The index file behaves as-if a file with the same name as the parent folder had been placed inside the parent folder. The following two directory trees are thus equivalent:
. └── about └── _index.md // maps to /about
. └── about.md // maps to /about
Every document in Liara must have metadata associated with it, which contains at least the document title. There are two ways to add metadata to a document: Embedding it inside the document or by using a separate
.meta file. For non-document files, using a separate
.meta file is the only way to associate metadata with a file.
Metadata has to be either embedded or placed in a
.meta file, but not both. If a
.meta file is found the whole document content will be used and no search for metadata within it will be performed.
When embedding it inside the document, the metadata must be placed at the beginning of the file. The metadata can be provided as YAML or TOML. For YAML, use
--- as the delimiter, for TOML, use
+++. Documents can be empty as long as the metadata is present, so this is a valid document with YAML metadata:
--- title: "This is an example" ---
You cannot mix the delimiters, i.e. using
--- to start and
+++ to end will result in a failure. Using more characters is also not supported.
Alternatively, the metadata can be stored in a
.meta file next to the document. This method also works for static files like images, videos, and other binaries. In this case, the
.meta file must contain plain YAML.
.meta file name must be the same as the original file name, with the last suffix changed to
.meta. For instance, for a file named
blog-post.md, the metadata file would be
blog-post.meta. If you have a file with multiple suffixes like
blog-post.new.md, then the metadata file has to be named
Some metadata fields in Liara are processed by a
date expects a timestamp, for example:
--- title: "My blog post" date: 2096-11-22 19:30:56+01:00 ---
Documents with a date that lies in the future relative to the time the build is invoked will get filtered by the
status can be used to hide content by setting it to
private – which in turn will make the
StatusFilter filter out the page. The filters can be set up in the Configuration.
Besides normal content, Liara allows storing arbitrary structured data in data files. This data is available via
site.data. This is very similar to the global metadata, but data nodes can be placed anywhere inside the content tree.