byte-array

(byte-array size-or-seq) (byte-array size init)

Answers a byte array, Java's byte[]: size zeros, or the members of a seq, each stored as Number.byteValue stores it (an integer's low eight bits, so 300 is 44). With size and init, every element is the byte init, or the next member of the seq init (zeros past its end). An element reads back as a signed byte, -128 to 127. The array is mutable (aset, aset-byte); it is no collection (coll? and vector? are false, = is identity) but seqs its elements, so count, seq, vec, nth, get and the seq functions take it. It prints as the oracle's #object["[B" ...] without the identity hash. On the interpreter and the JVM it is a Java member's byte[], both ways (interop). As a value a function of one or two arguments.

.getBytes and (String. bytes ...) take the charset as a name ("UTF-8", "ISO-8859-1", "US-ASCII" or an alias) or as a java.nio.charset.Charset: one of the six StandardCharsets fields, or Charset/forName of a string literal. That value is the charset on every backend: .name, .displayName, str and .toString answer its name, =, .equals, .hashCode and .compareTo compare by name, and it is an instance? of Charset and Comparable. The decoding InputStreamReader, the encoding OutputStreamWriter and ByteArrayOutputStream.toString take it too, where :encoding of slurp and clojure.java.io takes a name only (a Charset is the oracle's ClassCastException). It crosses into a Java member as the host Charset (the interpreter and the JVM, interop). Charset/forName of a name the JDK rejects is the oracle's UnsupportedCharsetException, or IllegalCharsetNameException for a malformed one.