Defines an ordered map of signed integers with the specified bit-lengths. The sum of the bit-lengths must be divisible by 8.

(bit-map :a 3, :b 4, :c 1) <=> {:a 3, :b -15, :c false}

Defines a sequence of unsigned integers with the specified bit-lengths. The sum of the bit-lengths must be divisable by 8. Single bit values are treated specially, and will decode to simply ‘true’ or ‘false’.

(bit-seq 4 3 1) <=> [15 7 true]

If a number is larger than its bit-count can contain, it will be truncated during encoding.

Returns the number of bytes in the value. Compatible with any data structure that can be transformed into a sequence of ByteBuffers.

Takes a frame, and returns a codec. This function is idempotent; passing in a codec is a safe operation.

Functions that transform the values after they are decoded and before they are encoded can be specified, which allows the frame to only be an intermediate representation of the final Clojure data structure.

Defines a compiled frame.

Defines a private compiled frame.

Defines a frame which is just a byte sequence terminated by one or more delimiters.

If strip-delimiters? is true, the resulting byte sequences will not contain the delimiters.

Defines a frame which is terminated by delimiters.

Takes a list of enumerations, or a map of enumerations onto values, and returns a codec which associates each enumeration with a unique encoded value.

(enum :byte :a :b :c) (enum :int32 {:a 100, :b 200, :c 300})

Defines a byte sequence which is either of fixed length or has a prefix which describes its length.

Defines a frame which is either of finite length, or has a prefix which describes its length.

A header is a frame which describes the frame that follows. The decoded value from the header frame will be passed into ‘header->body,’ which will return the resulting codec for the body. When encoding, the value of the body will be passed into ‘body->header,’ which will return the resulting value for the header.

Creates a codec which consumes and emits standard Clojure hash-maps, but ensures that the values are encoded in the specified order. Useful for interop with C structs and network protocols.

A prefix is a specialized form of header, which only describes the length of the sequence that follows. It is only meant to be used in the context of ‘finite-frame’ or ‘repeated’. A prefix may be as simple as a primitive type:

(prefix :int32)

But may also be a more complex frame:

(prefix [(string :utf-8 :delimiters [“\0”]) :int64] second (fn [n] [“hello” n]))

For complex prefixes, ‘to-count’ must take the value of the header and return the length of the sequence that follows, and ‘from-count’ must take the length of the sequence and return the value of the prefix.

Describes a sequence of frames. By default, the sequence is prefixed with a 32-bit integer describing the length of the sequence. However, specifying a custom :prefix or :delimiters that terminate the sequence is also allowed.

Returns the number of bytes this codec will encode to, or nil if it is value-dependent.

Defines a frame which contains a string. The charset must be a keyword, such as :utf-8 or :ascii. Available options are :length, :delimiters, :suffix, and :char-sequence.

A string with :length specified is of finite byte length:

(string :ascii :length 3)

A string with :delimiters specified is terminated by one or more delimiters:

(string :utf-8 :delimiters [“\r\n” “\r”])

By specifying :value->delimiter you can selectively delimit a value when encoding:

(string :utf-8 :delimiters [“\r\n” “ ”] :value->delimiter (fn [v] (if (= v “END”) [“\r\n”] [“ ”])))

If a string is already bounded in length, but has a terminating sequence, use :suffix

(string :utf-8, :length 3, :suffix “\r\n”)

:suffix can be used in conjunction with both :length and :delimiters. The given :length is assumed to not include the suffix, and the delimiters are assumed to be followed by the suffix.

By default, this frame will return a string, but it can be more memory-efficient to have it return a java.lang.CharSequence instead. This frame also will encode CharSequences. To have the decoded result be a CharSequence, set :char-sequence to true:

(string :utf-8, :length 3, :char-sequence true)