abstract class IO

Overview

The IO class is the basis for all input and output in Crystal.

This class is inherited by types like File, Socket and IO::Memory and provides many useful methods for reading from and writing to an IO, like #print, #puts, #gets and #printf.

The only requirement for a type including the IO module is to define these two methods:

For example, this is a simple IO on top of a Bytes:

class SimpleSliceIO < IO
  def initialize(@slice : Bytes)
  end

  def read(slice : Bytes)
    slice.size.times { |i| slice[i] = @slice[i] }
    @slice += slice.size
    slice.size
  end

  def write(slice : Bytes)
    slice.size.times { |i| @slice[i] = slice[i] }
    @slice += slice.size
    nil
  end
end

slice = Slice.new(9) { |i| ('a'.ord + i).to_u8 }
String.new(slice) # => "abcdefghi"

io = SimpleSliceIO.new(slice)
io.gets(3) # => "abc"
io.print "xyz"
String.new(slice) # => "abcxyzghi"

Encoding

An IO can be set an encoding with the #set_encoding method. When this is set, all string operations (#gets, #gets_to_end, #read_char, #<<, #print, #puts #printf) will write in the given encoding, and read from the given encoding. Byte operations (#read, #write, #read_byte, #write_byte) never do encoding/decoding operations.

If an encoding is not set, the default one is UTF-8.

Mixing string and byte operations might not give correct results and should be avoided, as string operations might need to read extra bytes in order to get characters in the given encoding.

Direct Known Subclasses

Defined in:

io.cr
io/encoding.cr
io/error.cr

Class Method Summary

Instance Method Summary

Instance methods inherited from class Reference

==(other : self)
==(other : JSON::Any)
==(other : YAML::Any)
==(other)
==
, dup dup, hash(hasher) hash, inspect(io : IO) : Nil inspect, object_id : UInt64 object_id, pretty_print(pp) : Nil pretty_print, same?(other : Reference)
same?(other : Nil)
same?
, to_s(io : IO) : Nil to_s

Constructor methods inherited from class Reference

new new

Instance methods inherited from class Object

!=(other) !=, !~(other) !~, ==(other) ==, ===(other : JSON::Any)
===(other : YAML::Any)
===(other)
===
, =~(other) =~, class class, dup dup, hash(hasher)
hash
hash
, inspect(io : IO)
inspect
inspect
, itself itself, not_nil! not_nil!, pretty_inspect(width = 79, newline = "\n", indent = 0) : String pretty_inspect, pretty_print(pp : PrettyPrint) : Nil pretty_print, tap(&block) tap, to_json(io : IO)
to_json
to_json
, to_pretty_json(indent : String = " ")
to_pretty_json(io : IO, indent : String = " ")
to_pretty_json
, to_s
to_s(io : IO)
to_s
, to_yaml(io : IO)
to_yaml
to_yaml
, try(&block) try, unsafe_as(type : T.class) forall T unsafe_as

Constructor methods inherited from class Object

from_json(string_or_io, root : String) : self
from_json(string_or_io) : self
from_json
, from_yaml(string_or_io : String | IO) : self from_yaml

Class Method Detail

def self.copy(src, dst, limit : Int) #

Copy at most limit bytes from src to dst.

io = IO::Memory.new "hello"
io2 = IO::Memory.new

IO.copy io, io2, 3

io2.to_s # => "hel"

[View source]
def self.copy(src, dst) #

Copy all contents from src to dst.

io = IO::Memory.new "hello"
io2 = IO::Memory.new

IO.copy io, io2

io2.to_s # => "hello"

[View source]
def self.pipe(read_blocking = false, write_blocking = false) : Tuple(IO::FileDescriptor, IO::FileDescriptor) #

Creates a pair of pipe endpoints (connected to each other) and returns them as a two-element Tuple.

reader, writer = IO.pipe
writer.puts "hello"
writer.puts "world"
reader.gets # => "hello"
reader.gets # => "world"

[View source]
def self.pipe(read_blocking = false, write_blocking = false, &block) #

Creates a pair of pipe endpoints (connected to each other) and passes them to the given block. Both endpoints are closed after the block.

IO.pipe do |reader, writer|
  writer.puts "hello"
  writer.puts "world"
  reader.gets # => "hello"
  reader.gets # => "world"
end

[View source]

Instance Method Detail

def <<(obj) : self #

Writes the given object into this IO. This ends up calling to_s(io) on the object.

io = IO::Memory.new
io << 1
io << '-'
io << "Crystal"
io.to_s # => "1-Crystal"

[View source]
def close #

Closes this IO.

IO defines this is a no-op method, but including types may override.


[View source]
def closed? #

Returns true if this IO is closed.

IO defines returns false, but including types may override.


[View source]
def each_byte(&block) : Nil #

Invokes the given block with each byte (UInt8) in this IO.

io = IO::Memory.new("aあ")
io.each_byte do |byte|
  puts byte
end

Output:

97
227
129
130

[View source]
def each_byte #

Returns an Iterator for the bytes in this IO.

io = IO::Memory.new("aあ")
iter = io.each_byte
iter.next # => 97
iter.next # => 227
iter.next # => 129
iter.next # => 130

[View source]
def each_char(&block) : Nil #

Invokes the given block with each Char in this IO.

io = IO::Memory.new("あめ")
io.each_char do |char|
  puts char
end

Output:

あ
め

[View source]
def each_char #

Returns an Iterator for the chars in this IO.

io = IO::Memory.new("あめ")
iter = io.each_char
iter.next # => 'あ'
iter.next # => 'め'

[View source]
def each_line(*args, **options, &block) : Nil #

Invokes the given block with each line in this IO, where a line is defined by the arguments passed to this method, which can be the same ones as in the #gets methods.

io = IO::Memory.new("hello\nworld")
io.each_line do |line|
  puts line.chomp.reverse
end

Output:

olleh
dlrow

[View source]
def each_line(*args, **options) #

Returns an Iterator for the lines in this IO, where a line is defined by the arguments passed to this method, which can be the same ones as in the #gets methods.

io = IO::Memory.new("hello\nworld")
iter = io.each_line
iter.next # => "hello"
iter.next # => "world"

[View source]
def encoding : String #

Returns this IO's encoding. The default is UTF-8.


[View source]
def flush #

Flushes buffered data, if any.

IO defines this is a no-op method, but including types may override.


[View source]
def gets(delimiter : Char, limit : Int, chomp = false) : String? #

Reads until delimiter is found, limit bytes are read, or the end of the IO is reached. Returns nil if called at the end of this IO.

io = IO::Memory.new "hello\nworld"
io.gets('o', 3)  # => "hel"
io.gets('r', 10) # => "lo\nwor"
io.gets('z', 10) # => "ld"
io.gets('w', 10) # => nil

[View source]
def gets(delimiter : Char, chomp = false) : String? #

Reads until delimiter is found, or the end of the IO is reached. Returns nil if called at the end of this IO.

io = IO::Memory.new "hello\nworld"
io.gets('o') # => "hello"
io.gets('r') # => "\nwor"
io.gets('z') # => "ld"
io.gets('w') # => nil

[View source]
def gets(chomp = true) : String? #

Reads a line from this IO. A line is terminated by the \n character. Returns nil if called at the end of this IO.

By default the newline is removed from the returned string, unless chomp is false.

io = IO::Memory.new "hello\nworld\nfoo\n"
io.gets               # => "hello"
io.gets(chomp: false) # => "world\n"
io.gets               # => "foo"
io.gets               # => nil

[View source]
def gets(limit : Int, chomp = false) : String? #

Reads a line of at most limit bytes from this IO. A line is terminated by the \n character. Returns nil if called at the end of this IO.

io = IO::Memory.new "hello\nworld"
io.gets(3) # => "hel"
io.gets(3) # => "lo\n"
io.gets(3) # => "wor"
io.gets(3) # => "ld"
io.gets(3) # => nil

[View source]
def gets(delimiter : String, chomp = false) : String? #

Reads until delimiter is found or the end of the IO is reached. Returns nil if called at the end of this IO.

io = IO::Memory.new "hello\nworld"
io.gets("wo") # => "hello\nwo"
io.gets("wo") # => "rld"
io.gets("wo") # => nil

[View source]
def gets_to_end : String #

Reads the rest of this IO data as a String.

io = IO::Memory.new "hello world"
io.gets_to_end # => "hello world"

[View source]
def peek : Bytes? #

Peeks into this IO, if possible.

It returns:

  • nil if this IO isn't peekable
  • an empty slice if it is, but EOF was reached
  • a non-empty slice if some data can be peeked

The returned bytes are only valid data until a next call to any method that reads from this IO is invoked.

By default this method returns nil, but IO implementations that provide buffering or wrap other IOs should override this method.


[View source]
def pos #

Returns the current position (in bytes) in this IO.

The IO class raises on this method, but some subclasses, notable IO::FileDescriptor and IO::Memory implement it.

File.write("testfile", "hello")

file = File.new("testfile")
file.pos     # => 0
file.gets(2) # => "he"
file.pos     # => 2

[View source]
def pos=(value) #

Sets the current position (in bytes) in this IO.

The IO class raises on this method, but some subclasses, notable IO::FileDescriptor and IO::Memory implement it.

File.write("testfile", "hello")

file = File.new("testfile")
file.pos = 3
file.gets_to_end # => "lo"

[View source]
def print(*objects : _) : Nil #

Writes the given objects into this IO by invoking to_s(io) on each of the objects.

io = IO::Memory.new
io.print 1, '-', "Crystal"
io.to_s # => "1-Crystal"

[View source]
def print(obj) : Nil #

Same as #<<.

io = IO::Memory.new
io.print 1
io.print '-'
io.print "Crystal"
io.to_s # => "1-Crystal"

[View source]
def printf(format_string, *args) : Nil #

Writes a formatted string to this IO. For details on the format string, see Kernel::sprintf.


[View source]
def printf(format_string, args : Array | Tuple) : Nil #

Writes a formatted string to this IO. For details on the format string, see Kernel::sprintf.


[View source]
def puts(*objects : _) : Nil #

Writes the given objects, each followed by a newline character.

io = IO::Memory.new
io.puts 1, '-', "Crystal"
io.to_s # => "1\n-\nCrystal\n"

[View source]
def puts : Nil #

Writes a newline character.

io = IO::Memory.new
io.puts
io.to_s # => "\n"

[View source]
def puts(obj) : Nil #

Writes the given object to this IO followed by a newline character.

io = IO::Memory.new
io.puts 1
io.puts "Crystal"
io.to_s # => "1\nCrystal\n"

[View source]
def puts(string : String) : Nil #

Writes the given string to this IO followed by a newline character unless the string already ends with one.

io = IO::Memory.new
io.puts "hello\n"
io.puts "world"
io.to_s # => "hello\nworld\n"

[View source]
abstract def read(slice : Bytes) #

Reads at most slice.size bytes from this IO into slice. Returns the number of bytes read, which is 0 if and only if there is no more data to read (so checking for 0 is the way to detect end of file).

io = IO::Memory.new "hello"
slice = Bytes.new(4)
io.read(slice) # => 4
slice          # => Bytes[104, 101, 108, 108]
io.read(slice) # => 1
slice          # => Bytes[111, 101, 108, 108]
io.read(slice) # => 0

[View source]
def read_at(offset, bytesize, &block) #

Yields an IO to read a section inside this IO.

The IO class raises on this method, but some subclasses, notable File and IO::Memory implement it.

Multiple sections can be read concurrently.


[View source]
def read_byte : UInt8? #

Reads a single byte from this IO. Returns nil if there is no more data to read.

io = IO::Memory.new "a"
io.read_byte # => 97
io.read_byte # => nil

[View source]
def read_bytes(type, format : IO::ByteFormat = IO::ByteFormat::SystemEndian) #

Reads an instance of the given type from this IO using the specified format.

This ends up invoking type.from_io(self, format), so any type defining a from_io(io : IO, format : IO::ByteFormat = IO::ByteFormat::SystemEndian) method can be read in this way.

See Int.from_io and Float.from_io.

io = IO::Memory.new
io.puts "\u{4}\u{3}\u{2}\u{1}"
io.rewind
io.read_bytes(Int32, IO::ByteFormat::LittleEndian) # => 0x01020304

[View source]
def read_char : Char? #

Reads a single Char from this IO. Returns nil if there is no more data to read.

io = IO::Memory.new "あ"
io.read_char # => 'あ'
io.read_char # => nil

[View source]
def read_fully(slice : Bytes) #

Tries to read exactly slice.size bytes from this IO into slice. Raises EOFError if there aren't slice.size bytes of data.

io = IO::Memory.new "123451234"
slice = Bytes.new(5)
io.read_fully(slice) # => 5
slice                # => Bytes[49, 50, 51, 52, 53]
io.read_fully(slice) # raises IO::EOFError

[View source]
def read_fully?(slice : Bytes) #

Tries to read exactly slice.size bytes from this IO into slice. Returns nil if there aren't slice.size bytes of data, otherwise returns the number of bytes read.

io = IO::Memory.new "123451234"
slice = Bytes.new(5)
io.read_fully?(slice) # => 5
slice                 # => Bytes[49, 50, 51, 52, 53]
io.read_fully?(slice) # => nil

[View source]
def read_line(*args, **options) : String #

Same as #gets, but raises EOFError if called at the end of this IO.


[View source]
def read_string(bytesize : Int) : String #

Reads an UTF-8 encoded string of exactly bytesize bytes. Raises EOFError if there are not enough bytes to build the string.

io = IO::Memory.new("hello world")
io.read_string(5) # => "hello"
io.read_string(1) # => " "
io.read_string(6) # raises IO::EOFError

[View source]
def read_utf8(slice : Bytes) #

Reads UTF-8 decoded bytes into the given slice. Returns the number of UTF-8 bytes read.

If no encoding is set, this is the same as #read(slice).

bytes = "你".encode("GB2312") # => Bytes[196, 227]

io = IO::Memory.new(bytes)
io.set_encoding("GB2312")

buffer = uninitialized UInt8[1024]
bytes_read = io.read_utf8(buffer.to_slice) # => 3
buffer.to_slice[0, bytes_read]             # => Bytes[228, 189, 160]

"你".bytes # => [228, 189, 160]

[View source]
def read_utf8_byte #

Reads a single decoded UTF-8 byte from this IO. Returns nil if there is no more data to read.

If no encoding is set, this is the same as #read_byte.

bytes = "你".encode("GB2312") # => Bytes[196, 227]

io = IO::Memory.new(bytes)
io.set_encoding("GB2312")
io.read_utf8_byte # => 228
io.read_utf8_byte # => 189
io.read_utf8_byte # => 160
io.read_utf8_byte # => nil

"你".bytes # => [228, 189, 160]

[View source]
def rewind #

Rewinds this IO. By default this method raises, but including types may implement it.


[View source]
def seek(offset, whence : Seek = Seek::Set) #

Seeks to a given offset (in bytes) according to the whence argument.

The IO class raises on this method, but some subclasses, notable IO::FileDescriptor and IO::Memory implement it.

Returns self.

File.write("testfile", "abc")

file = File.new("testfile")
file.gets(3) # => "abc"
file.seek(1, IO::Seek::Set)
file.gets(2) # => "bc"
file.seek(-1, IO::Seek::Current)
file.gets(1) # => "c"

[View source]
def set_encoding(encoding : String, invalid : Symbol? = nil) #

Sets the encoding of this IO.

The invalid argument can be:

  • nil: an exception is raised on invalid byte sequences
  • :skip: invalid byte sequences are ignored

String operations (#gets, #gets_to_end, #read_char, #<<, #print, #puts #printf) will use this encoding.


[View source]
def skip(bytes_count : Int) : Nil #

Reads and discards exactly bytes_count bytes. Raises IO::EOFError if there aren't at least bytes_count bytes.

io = IO::Memory.new "hello world"
io.skip(6)
io.gets    # => "world"
io.skip(1) # raises IO::EOFError

[View source]
def skip_to_end : Nil #

Reads and discards bytes from self until there are no more bytes.


[View source]
def tell #

Same as #pos.


[View source]
def tty? : Bool #

Returns true if this IO is associated with a terminal device (tty), false otherwise.

IO returns false, but including types may override.

STDIN.tty?          # => true
IO::Memory.new.tty? # => false

[View source]
abstract def write(slice : Bytes) : Nil #

Writes the contents of slice into this IO.

io = IO::Memory.new
slice = Bytes.new(4) { |i| ('a'.ord + i).to_u8 }
io.write(slice)
io.to_s # => "abcd"

[View source]
def write_byte(byte : UInt8) #

Writes a single byte into this IO.

io = IO::Memory.new
io.write_byte 97_u8
io.to_s # => "a"

[View source]
def write_bytes(object, format : IO::ByteFormat = IO::ByteFormat::SystemEndian) #

Writes the given object to this IO using the specified format.

This ends up invoking object.to_io(self, format), so any object defining a to_io(io : IO, format : IO::ByteFormat = IO::ByteFormat::SystemEndian) method can be written in this way.

See Int#to_io and Float#to_io.

io = IO::Memory.new
io.write_bytes(0x01020304, IO::ByteFormat::LittleEndian)
io.rewind
io.gets(4) # => "\u{4}\u{3}\u{2}\u{1}"

[View source]
def write_utf8(slice : Bytes) #

Writes a slice of UTF-8 encoded bytes to this IO, using the current encoding.


[View source]