#!/usr/bin/env ruby
#--
# Copyright 2004 by Jim Weirich (jim@weirichhouse.org).
# All rights reserved.
# Permission is granted for use, copying, modification, distribution,
# and distribution of modified versions of this work as long as the
# above copyright notice is included.
#++
# Provide a flexible and easy to use Builder for creating XML markup.
# See XmlBuilder for usage details.
require 'builder/xmlbase'
module Builder
# Create XML markup easily. All (well, almost all) methods sent to
# an XmlMarkup object will be translated to the equivalent XML
# markup. Any method with a block will be treated as an XML markup
# tag with nested markup in the block.
#
# Examples will demonstrate this easier than words. In the
# following, +xm+ is an +XmlMarkup+ object.
#
# xm.em("emphasized") # => <em>emphasized</em>
# xm.em { xmm.b("emp & bold") } # => <em><b>emph & bold</b></em>
# xm.a("A Link", "href"=>"http://onestepback.org")
# # => <a href="http://onestepback.org">A Link</a>
# xm.div { br } # => <div><br/></div>
# xm.target("name"=>"compile", "option"=>"fast")
# # => <target option="fast" name="compile"\>
# # NOTE: order of attributes is not specified.
#
# xm.instruct! # <?xml version="1.0" encoding="UTF-8"?>
# xm.html { # <html>
# xm.head { # <head>
# xm.title("History") # <title>History</title>
# } # </head>
# xm.body { # <body>
# xm.comment! "HI" # <!-- HI -->
# xm.h1("Header") # <h1>Header</h1>
# xm.p("paragraph") # <p>paragraph</p>
# } # </body>
# } # </html>
#
# == Notes:
#
# * The order that attributes are inserted in markup tags is
# undefined.
#
# * Sometimes you wish to insert text without enclosing tags. Use
# the <tt>text!</tt> method to accomplish this.
#
# Example:
#
# xm.div { # <div>
# xm.text! "line"; xm.br # line<br/>
# xm.text! "another line"; xmbr # another line<br/>
# } # </div>
#
# * The special XML characters <, >, and & are converted to <,
# > and & automatically. Use the <tt><<</tt> operation to
# insert text without modification.
#
# * Sometimes tags use special characters not allowed in ruby
# identifiers. Use the <tt>tag!</tt> method to handle these
# cases.
#
# Example:
#
# xml.tag!("SOAP:Envelope") { ... }
#
# will produce ...
#
# <SOAP:Envelope> ... </SOAP:Envelope>"
#
# <tt>tag!</tt> will also take text and attribute arguments (after
# the tag name) like normal markup methods. (But see the next
# bullet item for a better way to handle XML namespaces).
#
# * Direct support for XML namespaces is now available. If the
# first argument to a tag call is a symbol, it will be joined to
# the tag to produce a namespace:tag combination. It is easier to
# show this than describe it.
#
# xml.SOAP :Envelope do ... end
#
# Just put a space before the colon in a namespace to produce the
# right form for builder (e.g. "<tt>SOAP:Envelope</tt>" =>
# "<tt>xml.SOAP :Envelope</tt>")
#
# * XmlMarkup builds the markup in any object (called a _target_)
# that accepts the <tt><<</tt> method. If no target is given,
# then XmlMarkup defaults to a string target.
#
# Examples:
#
# xm = Builder::XmlMarkup.new
# result = xm.title("yada")
# # result is a string containing the markup.
#
# buffer = ""
# xm = Builder::XmlMarkup.new(buffer)
# # The markup is appended to buffer (using <<)
#
# xm = Builder::XmlMarkup.new(STDOUT)
# # The markup is written to STDOUT (using <<)
#
# xm = Builder::XmlMarkup.new
# x2 = Builder::XmlMarkup.new(:target=>xm)
# # Markup written to +x2+ will be send to +xm+.
#
# * Indentation is enabled by providing the number of spaces to
# indent for each level as a second argument to XmlBuilder.new.
# Initial indentation may be specified using a third parameter.
#
# Example:
#
# xm = Builder.new(:ident=>2)
# # xm will produce nicely formatted and indented XML.
#
# xm = Builder.new(:indent=>2, :margin=>4)
# # xm will produce nicely formatted and indented XML with 2
# # spaces per indent and an over all indentation level of 4.
#
# builder = Builder::XmlMarkup.new(:target=>$stdout, :indent=>2)
# builder.name { |b| b.first("Jim"); b.last("Weirich) }
# # prints:
# # <name>
# # <first>Jim</first>
# # <last>Weirich</last>
# # </name>
#
# * The instance_eval implementation which forces self to refer to
# the message receiver as self is now obsolete. We now use normal
# block calls to execute the markup block. This means that all
# markup methods must now be explicitly send to the xml builder.
# For instance, instead of
#
# xml.div { strong("text") }
#
# you need to write:
#
# xml.div { xml.strong("text") }
#
# Although more verbose, the subtle change in semantics within the
# block was found to be prone to error. To make this change a
# little less cumbersome, the markup block now gets the markup
# object sent as an argument, allowing you to use a shorter alias
# within the block.
#
# For example:
#
# xml_builder = Builder::XmlMarkup.new
# xml_builder.div { |xml|
# xml.stong("text")
# }
#
class XmlMarkup < XmlBase
# Create an XML markup builder. Parameters are specified by an
# option hash.
#
# :target=><em>target_object</em>::
# Object receiving the markup. +out+ must respond to the
# <tt><<</tt> operator. The default is a plain string target.
# :indent=><em>indentation</em>::
# Number of spaces used for indentation. The default is no
# indentation and no line breaks.
# :margin=><em>initial_indentation_level</em>::
# Amount of initial indentation (specified in levels, not
# spaces).
#
def initialize(options={})
indent = options[:indent] || 0
margin = options[:margin] || 0
super(indent, margin)
@target = options[:target] || ""
end
# Return the target of the builder.
def target!
@target
end
def comment!(comment_text)
_ensure_no_block block_given?
_special("<!-- ", " -->", comment_text, nil)
end
# Insert an XML declaration into the XML markup.
#
# For example:
#
# xml.declare! :ELEMENT, :blah, "yada"
# # => <!ELEMENT blah "yada">
def declare!(inst, *args, &block)
_indent
@target << "<!#{inst}"
args.each do |arg|
case arg
when String
@target << %{ "#{arg}"}
when Symbol
@target << " #{arg}"
end
end
if block_given?
@target << " ["
_newline
_nested_structures(block)
@target << "]"
end
@target << ">"
_newline
end
# Insert a processing instruction into the XML markup. E.g.
#
# For example:
#
# xml.instruct!
# #=> <?xml version="1.0" encoding="UTF-8"?>
# xml.instruct! :aaa, :bbb=>"ccc"
# #=> <?aaa bbb="ccc"?>
#
def instruct!(directive_tag=:xml, attrs={})
_ensure_no_block block_given?
if directive_tag == :xml
a = { :version=>"1.0", :encoding=>"UTF-8" }
attrs = a.merge attrs
end
_special(
"<?#{directive_tag}",
"?>",
nil,
attrs,
[:version, :encoding, :standalone])
end
# Surrounds the given text with a CDATA tag
#
# For example:
#
# xml.cdata! "blah blah blah"
# # => <![CDATA[blah blah blah]]>
def cdata!(text)
_ensure_no_block block_given?
_special("<![CDATA[", "]]>", text, nil)
end
private
# NOTE: All private methods of a builder object are prefixed when
# a "_" character to avoid possible conflict with XML tag names.
# Insert text directly in to the builder's target.
def _text(text)
@target << text
end
# Insert special instruction.
def _special(open, close, data=nil, attrs=nil, order=[])
_indent
@target << open
@target << data if data
_insert_attributes(attrs, order) if attrs
@target << close
_newline
end
# Start an XML tag. If <tt>end_too</tt> is true, then the start
# tag is also the end tag (e.g. <br/>
def _start_tag(sym, attrs, end_too=false)
@target << "<#{sym}"
_insert_attributes(attrs)
@target << "/" if end_too
@target << ">"
end
# Insert an ending tag.
def _end_tag(sym)
@target << "</#{sym}>"
end
# Insert the attributes (given in the hash).
def _insert_attributes(attrs, order=[])
return if attrs.nil?
order.each do |k|
v = attrs[k]
@target << %{ #{k}="#{v}"} if v
end
attrs.each do |k, v|
@target << %{ #{k}="#{v}"} unless order.member?(k)
end
end
def _ensure_no_block(got_block)
if got_block
fail IllegalBlockError,
"Blocks are not allowed on XML instructions"
end
end
end
end
