edn
extensible data notation [eed-n]
Rationale
edn is an extensible data notation. A superset of edn is used by Clojure to represent programs, and it is used by Datomic and other applications as a data transfer format. This spec describes edn in isolation from those and other specific use cases, to help facilitate implementation of readers and writers in other languages, and for other uses.
edn supports a rich set of built-in elements, and the definition of extension elements in terms of the others. Users of data formats without such facilities must rely on either convention or context to convey elements not included in the base set. This greatly complicates application logic, betraying the apparent simplicity of the format. edn is simple, yet powerful enough to meet the demands of applications without convention or complex context-sensitive logic.
edn is a system for the conveyance of values. It is not a type system, and has no schemas. Nor is it a system for representing objects - there are no reference types, nor should a consumer have an expectation that two equivalent elements in some body of edn will yield distinct object identities when read, unless a reader implementation goes out of its way to make such a promise. Thus the resulting values should be considered immutable, and a reader implementation should yield values that ensure this, to the extent possible.
edn is a set of definitions for acceptable elements. A use of edn might be a stream or file containing elements, but it could be as small as the conveyance of a single element in e.g. an HTTP query param.
There is no enclosing element at the top level. Thus edn is suitable for streaming and interactive applications.
The base set of elements in edn is meant to cover the basic set of data structures common to most programming languages. While edn specifies how those elements are formatted in text, it does not dictate the representation that results on the consumer side. A well behaved reader library should endeavor to map the elements to programming language types with similar semantics.
Spec
Currently this specification is casual, as we gather feedback from implementors. A more rigorous e.g. BNF will follow.
General considerations
edn elements, streams and files should be encoded using UTF-8.
Elements are generally separated by whitespace. Whitespace, other than within strings, is not
otherwise significant, nor need redundant whitespace be preserved during transmissions. Commas ,
are also considered whitespace, other than within strings.
The delimiters { } ( ) [ ]
need not be separated from adjacent elements by whitespace.
# dispatch character
Tokens beginning with #
are reserved. The character following #
determines the behavior. The
dispatches #{
(sets), #_
(discard), #alphabetic-char (tag) are defined below. #
is not a
delimiter.
Built-in elements
nil
nil
represents nil, null or nothing. It should be read as an object with similar meaning on the
target platform.
booleans
true
and false
should be mapped to booleans.
If a platform has canonic values for true and false, it is a further semantic of booleans that all
instances of true
yield that (identical) value, and similarly for false
.
strings
Strings are enclosed in "double quotes"
. May span multiple lines. Standard C/Java escape
characters \t, \r, \n, \\ and \"
are supported.
characters
Characters are preceded by a backslash: \c
, \newline
, \return
, \space
and \tab
yield the
corresponding characters. Unicode characters are represented with \uNNNN as in Java. Backslash cannot be
followed by whitespace.
symbols
Symbols are used to represent identifiers, and should map to something other than strings, if possible.
Symbols begin with a non-numeric character and can contain alphanumeric characters and . * + ! - _ ? $ % & = < >
. If -
, +
or .
are the first character, the second character (if any) must be
non-numeric. Additionally, : #
are allowed as constituent characters in symbols other than as the
first character.
/
has special meaning in symbols. It can be used once only in the middle of a symbol to separate
the prefix (often a namespace) from the name, e.g. my-namespace/foo
. /
by itself is a legal
symbol, but otherwise neither the prefix nor the name part can be empty when the symbol
contains /
.
If a symbol has a prefix and /
, the following name component should follow the
first-character restrictions for symbols as a whole. This is to avoid ambiguity in reading contexts
where prefixes might be presumed as implicitly included namespaces and elided thereafter.
keywords
Keywords are identifiers that typically designate themselves. They are semantically akin to
enumeration values. Keywords follow the rules of symbols, except they can (and must) begin with :
, e.g. :fred
or :my/fred
. If the target platform does not have a keyword type distinct
from a symbol type, the same type can be used without conflict, since the mandatory leading :
of
keywords is disallowed for symbols. Per the symbol rules above, :/ and :/anything are not legal keywords.
A keyword cannot begin with ::
If the target platform supports some notion of interning, it is a further semantic of keywords that all instances of the same keyword yield the identical object.
integers
Integers consist of the digits 0
- 9
, optionally prefixed by -
to indicate a negative number, or
(redundantly) by +
. No integer other than 0 may begin with 0. 64-bit (signed integer) precision is
expected. An integer can have the suffix N
to indicate that arbitrary precision is desired. -0 is a
valid integer not distinct from 0.
integer
int
int N
digit
0-9
int
digit
1-9 digits
+ digit
+ 1-9 digits
- digit
- 1-9 digits
floating point numbers
64-bit (double) precision is expected.
floating-point-number
int M
int frac
int exp
int frac exp
digit
0-9
int
digit
1-9 digits
+ digit
+ 1-9 digits
- digit
- 1-9 digits
frac
. digits
exp
ex digits
digits
digit
digit digits
ex
e
e+
e-
E
E+
E-
In addition, a floating-point number may have the suffix M
to indicate that exact precision is
desired.
lists
A list is a sequence of values. Lists are represented by zero or more elements enclosed in
parentheses ()
. Note that lists can be heterogeneous.
(a b 42)
vectors
A vector is a sequence of values that supports random access. Vectors are represented by zero or
more elements enclosed in square brackets []
. Note that vectors can be heterogeneous.
[a b 42]
maps
A map is a collection of associations between keys and values. Maps are represented by zero or more
key and value pairs enclosed in curly braces {}
. Each key should appear at most once. No
semantics should be associated with the order in which the pairs appear.
{:a 1, "foo" :bar, [1 2 3] four}
Note that keys and values can be elements of any type. The use of commas above is optional, as they are parsed as whitespace.
sets
A set is a collection of unique values. Sets are represented by zero or more elements enclosed in
curly braces preceded by #
#{}
. No semantics should be associated with the order in which the
elements appear. Note that sets can be heterogeneous.
#{a b [1 2 3]}
tagged elements
edn supports extensibility through a simple mechanism. #
followed immediately by a symbol
starting with an alphabetic character indicates that that symbol is a tag. A tag indicates
the semantic interpretation of the following element. It is envisioned that a reader
implementation will allow clients to register handlers for specific tags. Upon encountering a tag,
the reader will first read the next element (which may itself be or comprise other tagged elements),
then pass the result to the corresponding handler for further interpretation, and the result of the
handler will be the data value yielded by the tag + tagged element, i.e. reading a tag and tagged
element yields one value. This value is the value to be returned to the program and is not further
interpreted as edn data by the reader.
This process will bottom out on elements either understood or built-in.
Thus you can build new distinct readable elements out of (and only out of) other readable elements, keeping extenders and extension consumers out of the text business.
The semantics of a tag, and the type and interpretation of the tagged element are defined by the steward of the tag.
#myapp/Person {:first "Fred" :last "Mertz"}
If a reader encounters a tag for which no handler is registered, the implementation can either report an error, call a designated 'unknown element' handler, or create a well-known generic representation that contains both the tag and the tagged element, as it sees fit. Note that the non-error strategies allow for readers which are capable of reading any and all edn, in spite of being unaware of the details of any extensions present.
rules for tags
Tag symbols without a prefix are reserved by edn for built-ins defined using the tag system.
User tags must contain a prefix component, which must be owned by the user (e.g. trademark or domain) or known unique in the communication context.
A tag may specify more than one format for the tagged element, e.g. both a string and a vector representation.
Tags themselves are not elements. It is an error to have a tag without a corresponding tagged element.
built-in tagged elements
#inst "rfc-3339-format"
An instant in time. The tagged element is a string in RFC-3339 format.
#inst "1985-04-12T23:20:50.52Z"
#uuid "f81d4fae-7dec-11d0-a765-00a0c91e6bf6"
A UUID. The tagged element is a canonical UUID string representation.
comments
If a ;
character is encountered outside of a string, that character and all subsequent characters
to the next newline should be ignored.
discard
#
followed immediately by _
is the discard sequence, indicating that the next element (whether
separated from #_
by whitespace or not) should be read and discarded. Note that the next element
must still be a readable element. A reader should not call user-supplied tag handlers during the
processing of the element to be discarded.
[a b #_foo 42] => [a b 42]
The discard sequence is not an element. It is an error to have a discard sequence without a following element.
equality
Sets and maps have requirements that their elements and keys respectively be unique, which requires a mechanism for determining when 2 values are not unique (i.e. are equal).
nil, booleans, strings, characters, and symbols are equal to values of the same type with the same edn representation.
integers and floating point numbers should be considered equal to values only of the same magnitude, type, and precision. Comingling numeric types and precision in map/set key/elements, or constituents therein, is not advised.
sequences (lists and vectors) are equal to other sequences whose count of elements is the same, and for which each corresponding pair of elements (by ordinal) is equal.
sets are equal if they have the same count of elements and, for every element in one set, an equal element is in the other.
maps are equal if they have the same number of entries, and for every key/value entry in one map an equal key is present and mapped to an equal value in the other.
tagged elements must define their own equality semantics. #uuid elements are equal if their canonic representations are equal. #inst elements are equal if their representation strings designate the same timestamp per RFC-3339.