Content parsers

λ-blog uses a very simple scheme of content management - you supply paths to files containing your content, and tell how to process them by specifying a parser.

A parser is simply a Clojure function that takes the text contents of a file and returns a data structure representing the file contents. The returned data structure can be arbitrary as long as your generation pipeline knows how to handle it. A sensible idea is to return a Clojure map containing processed file contents & some metadata.

Here's a simple definition of such a parser:

(defn hello-parser [contents]
  {:metadata {:length (count contents)}
   :contents (str "<h1>Hello world!</h1><p>" contents "</p>")})

Builtin parsers

Out of the box, λ-blog comes bundled with a Markdown parser based on markdown-clj. To use it, simply require its namespace and add it to your generation pipeline:

(require '[lambda-blog.parsers.md :as md])

;; Elswhere in the code:
(-> filename
    slurp
    md/parse)

The value returned is a Clojure map containing at least :contents & :metadata keys:

user=> (md/parse "Some: \"Metadata\"\n\n# Some\nMarkdown\n## Contents")
{:metadata {:some "Metadata"}
 :contents "<h1>Some</h1>Markdown<h2>Contents</h2>"}

Each of metadata value is assumed to be in Clojure EDN format and thus is parsed as such. In order to give a certain metadata key multiple values, simply use a Clojure vector:

user=> (md/parse "Some: [Multiple Values]\n\n# Some\nMarkdown\n## Contents" {})
{:metadata {:some [Multiple Values]}
 :contents "<h1>Some</h1>Markdown<h2>Contents</h2>"}

Additional arguments

The builtin Markdown parser accepts a number of extra, keyword arguments:

:footnotes? - toggles footnote syntax parsing; true by default,
:heading-anchors - toggles HTML anchors in headers; true by default,
previews? - toggles generation of entry prievews (snipping text at <!– more –> tag),
:reference-links? - toggles reference link syntax parsing; true by default.