# -*- mode: ruby; ruby-indent-level: 4; tab-width: 4 -*- vim: sw=4 ts=4
# $Id: yaml.rb 16084 2008-04-19 11:45:39Z knu $
# = yaml.rb: top-level module with methods for loading and parsing YAML documents
# Author:: why the lucky stiff
# YAML(tm) (rhymes with 'camel') is a
# straightforward machine parsable data serialization format designed for
# human readability and interaction with scripting languages such as Perl
# and Python. YAML is optimized for data serialization, formatted
# dumping, configuration files, log files, Internet messaging and
# filtering. This specification describes the YAML information model and
# serialization format. Together with the Unicode standard for characters, it
# provides all the information necessary to understand YAML Version 1.0
# and construct computer programs to process it.
# See http://yaml.org/ for more information. For a quick tutorial, please
# visit YAML In Five Minutes (http://yaml.kwiki.org/?YamlInFiveMinutes).
# The YAML 1.0 specification outlines four stages of YAML loading and dumping.
# This library honors all four of those stages, although data is really only
# available to you in three stages.
# The four stages are: native, representation, serialization, and presentation.
# The native stage refers to data which has been loaded completely into Ruby's
# own types. (See +YAML::load+.)
# The representation stage means data which has been composed into
# +YAML::BaseNode+ objects. In this stage, the document is available as a
# tree of node objects. You can perform YPath queries and transformations
# at this level. (See +YAML::parse+.)
# The serialization stage happens inside the parser. The YAML parser used in
# Ruby is called Syck. Serialized nodes are available in the extension as
# The presentation stage is the YAML document itself. This is accessible
# to you as a string. (See +YAML::dump+.)
# For more information about the various information models, see Chapter
# 3 of the YAML 1.0 Specification (http://yaml.org/spec/#id2491269).
# The YAML module provides quick access to the most common loading (YAML::load)
# and dumping (YAML::dump) tasks. This module also provides an API for registering
# global types (YAML::add_domain_type).
# A simple round-trip (load and dump) of an object.
# test_obj = ["dogs", "cats", "badgers"]
# yaml_obj = YAML::dump( test_obj )
# ruby_obj = YAML::load( yaml_obj )
# # => ["dogs", "cats", "badgers"]
# To register your custom types with the global resolver, use +add_domain_type+.
# YAML::add_domain_type( "your-site.com,2004", "widget" ) do |type, val|
Resolver = YAML::Syck::Resolver
DefaultResolver = YAML::Syck::DefaultResolver
DefaultResolver.use_types_at( @@tagged_classes )
GenericResolver = YAML::Syck::GenericResolver
Parser = YAML::Syck::Parser
Emitter = YAML::Syck::Emitter
# Returns a new default parser
def YAML.parser; Parser.new.set_resolver( YAML.resolver ); end
# Returns a new generic parser
def YAML.generic_parser; Parser.new.set_resolver( GenericResolver ); end
# Returns the default resolver
def YAML.resolver; DefaultResolver; end
# Returns a new default emitter
def YAML.emitter; Emitter.new.set_resolver( YAML.resolver ); end
# Converts _obj_ to YAML and writes the YAML result to _io_.
# File.open( 'animals.yaml', 'w' ) do |out|
# YAML.dump( ['badger', 'elephant', 'tiger'], out )
# If no _io_ is provided, a string containing the dumped YAML
def YAML.dump( obj, io = nil )
obj.to_yaml( io || io2 = StringIO.new )
io || ( io2.rewind; io2.read )
# Load a document from the current _io_ stream.
# File.open( 'animals.yaml' ) { |yf| YAML::load( yf ) }
# #=> ['badger', 'elephant', 'tiger']
# Can also load from a string.
# YAML.load( "--- :locked" )
# Load a document from the file located at _filepath_.
# YAML.load_file( 'animals.yaml' )
# #=> ['badger', 'elephant', 'tiger']
def YAML.load_file( filepath )
File.open( filepath ) do |f|
# Parse the first document from the current _io_ stream
# File.open( 'animals.yaml' ) { |yf| YAML::load( yf ) }
# #=> #<YAML::Syck::Node:0x82ccce0
# [#<YAML::Syck::Node:0x82ccd94
# #<YAML::Syck::Node:0x82ccd58
# #<YAML::Syck::Node:0x82ccd1c
# Can also load from a string.
# YAML.parse( "--- :locked" )
# #=> #<YAML::Syck::Node:0x82edddc
# @type_id="tag:ruby.yaml.org,2002:sym",
# @value=":locked", @kind=:scalar>
yp = generic_parser.load( io )
# Parse a document from the file located at _filepath_.
# YAML.parse_file( 'animals.yaml' )
# #=> #<YAML::Syck::Node:0x82ccce0
# [#<YAML::Syck::Node:0x82ccd94
# #<YAML::Syck::Node:0x82ccd58
# #<YAML::Syck::Node:0x82ccd1c
def YAML.parse_file( filepath )
File.open( filepath ) do |f|
# Calls _block_ with each consecutive document in the YAML
# stream contained in _io_.
# File.open( 'many-docs.yaml' ) do |yf|
# YAML.each_document( yf ) do |ydoc|
# ## ydoc contains the single object
# ## from the YAML document
def YAML.each_document( io, &block )
yp = parser.load_documents( io, &block )
# Calls _block_ with each consecutive document in the YAML
# stream contained in _io_.
# File.open( 'many-docs.yaml' ) do |yf|
# YAML.load_documents( yf ) do |ydoc|
# ## ydoc contains the single object
# ## from the YAML document
def YAML.load_documents( io, &doc_proc )
YAML.each_document( io, &doc_proc )
# Calls _block_ with a tree of +YAML::BaseNodes+, one tree for
# each consecutive document in the YAML stream contained in _io_.
# File.open( 'many-docs.yaml' ) do |yf|
# YAML.each_node( yf ) do |ydoc|
# ## ydoc contains a tree of nodes
# ## from the YAML document
def YAML.each_node( io, &doc_proc )
yp = generic_parser.load_documents( io, &doc_proc )
# Calls _block_ with a tree of +YAML::BaseNodes+, one tree for
# each consecutive document in the YAML stream contained in _io_.
# File.open( 'many-docs.yaml' ) do |yf|
# YAML.parse_documents( yf ) do |ydoc|
# ## ydoc contains a tree of nodes
# ## from the YAML document
def YAML.parse_documents( io, &doc_proc )
YAML.each_node( io, &doc_proc )
# Loads all documents from the current _io_ stream,
# returning a +YAML::Stream+ object containing all
def YAML.load_stream( io )
parser.load_documents( io ) do |doc|
d = YAML::Stream.new if not d
# Returns a YAML stream containing each of the items in +objs+,
# each having their own document.
# YAML.dump_stream( 0, [], {} )
def YAML.dump_stream( *objs )
# Add a global handler for a YAML domain type.
def YAML.add_domain_type( domain, type_tag, &transfer_proc )
resolver.add_type( "tag:#{ domain }:#{ type_tag }", transfer_proc )
# Add a transfer method for a builtin type
def YAML.add_builtin_type( type_tag, &transfer_proc )
resolver.add_type( "tag:yaml.org,2002:#{ type_tag }", transfer_proc )
# Add a transfer method for a builtin type
def YAML.add_ruby_type( type_tag, &transfer_proc )
resolver.add_type( "tag:ruby.yaml.org,2002:#{ type_tag }", transfer_proc )
# Add a private document type
def YAML.add_private_type( type_re, &transfer_proc )
resolver.add_type( "x-private:" + type_re, transfer_proc )
# Detect typing of a string
def YAML.detect_implicit( val )
resolver.detect_implicit( val )
# Convert a type_id to a taguri
# Apply a transfer method to a Ruby object
def YAML.transfer( type_id, obj )
resolver.transfer( YAML.tagurize( type_id ), obj )
# Apply any implicit a node may qualify for
def YAML.try_implicit( obj )
YAML.transfer( YAML.detect_implicit( obj ), obj )
# Method to extract colon-seperated type and class, returning
# the type and the constant of the class
def YAML.read_type_class( type, obj_class )
scheme, domain, type, tclass = type.split( ':', 4 )
tclass.split( "::" ).each { |c| obj_class = obj_class.const_get( c ) } if tclass
return [ type, obj_class ]
def YAML.object_maker( obj_class, val )
o.instance_variable_set("@#{k}", v)
raise YAML::Error, "Invalid object explicitly tagged !ruby/Object: " + val.inspect
# Allocate an Emitter if needed
def YAML.quick_emit( oid, opts = {}, &e )
if opts.is_a? YAML::Emitter
case oid when Fixnum, NilClass; oid
else oid = "#{oid.object_id}-#{oid.hash}"
# ryan:: You know how Kernel.p is a really convenient way to dump ruby
# structures? The only downside is that it's not as legible as
# ryan:: I know you don't want to urinate all over your users' namespaces.
# But, on the other hand, convenience of dumping for debugging is,
# IMO, a big YAML use case.
# _why:: Go nuts! Have a pony parade!
# ryan:: Either way, I certainly will have a pony parade.
# Prints any supplied _objects_ out in YAML. Intended as
# a variation on +Kernel::p+.
# S = Struct.new(:name, :state)
def y( object, *objects )
puts( if objects.length == 1
YAML::dump_stream( *objects )
It is recommended that you Edit text format, this type of Fix handles quite a lot in one request
