Grammar-based Clojure reader.
Parcera can safely read any Clojure file without any code evaluation.
Parcera uses the wonderful Antlr4 as its parsing engine and focuses on the grammar definition instead.
Add [org.antlr/antlr4-runtime "4.7.1"]
to your dependencies in addition to parcera. Parcera assumes that
this dependency will be in the classpath to avoid collisions.
(ns example.core
(:require [parcera.core :as parcera]))
;;parse clojure code from a string
(parcera/ast (str '(ns parcera.core
(:require [foo.bar :as bar]
[clojure.data :as data]
[clojure.string :as str]))))
;; => returns a data structure with the result from the parser
(:code
(:list
(:symbol "ns")
(:whitespace " ")
(:symbol "parcera.core")
(:whitespace " ")
(:list
(:simple_keyword "require")
(:whitespace " ")
(:vector
(:symbol "foo.bar")
(:whitespace " ")
(:simple_keyword "as")
(:whitespace " ")
(:symbol "bar"))
(:whitespace " ")
(:vector (:symbol "clojure.data") (:whitespace " ") (:simple_keyword "as") (:whitespace " ") (:symbol "data"))
(:whitespace " ")
(:vector (:symbol "clojure.string") (:whitespace " ") (:simple_keyword "as") (:whitespace " ") (:symbol "str")))))
;; get meta data from the parsed code
(meta (second (parcera/ast (str :hello))))
#:parcera.core{:start {:row 1, :column 0}, :end {:row 1, :column 6}}
;; convert an AST back into a string
(parcera/code [:symbol "ns"])
;; "ns"
note: parcera is a bit permissive for symbols and keywords, you should be able to parse any valid Clojure file however, I cannot guarantee that an invalid symbol/keyword will yield a failure. This is because Clojure's reader is very stateful and embedding all those rules into the grammar would make it prohibitively complex.
Can you improve this documentation?Edit on GitHub
cljdoc is a website building & hosting documentation for Clojure/Script libraries
× close