rdoc 3.9.5 → 3.10.pre.1

data/lib/rdoc/test_case.rb

@@ -0,0 +1,59 @@
+require 'rubygems'
+require 'minitest/autorun'
+
+require 'fileutils'
+require 'pp'
+require 'tempfile'
+require 'tmpdir'
+require 'stringio'
+
+require 'rdoc'
+
+##
+# RDoc::TestCase is an abstract TestCase to provide common setup and teardown
+# across all RDoc tests.  The test case uses minitest, so all the assertions
+# of minitest may be used.
+#
+# The testcase provides the following:
+#
+# * A reset code-object tree
+# * A reset markup preprocessor (RDoc::Markup::PreProcess)
+# * The <code>@RM</code> alias of RDoc::Markup (for less typing)
+# * <code>@pwd</code> containing the current working directory
+# * FileUtils, pp, Tempfile, Dir.tmpdir and StringIO
+
+class RDoc::TestCase < MiniTest::Unit::TestCase
+
+  ##
+  # Abstract test-case setup
+
+  def setup
+    super
+
+    @top_level = nil
+
+    @RM = RDoc::Markup
+
+    RDoc::RDoc.reset
+    RDoc::Markup::PreProcess.reset
+
+    @pwd = Dir.pwd
+  end
+
+  ##
+  # Creates an RDoc::Comment with +text+ which was defined on +top_level+.
+  # By default the comment has the 'rdoc' format.
+
+  def comment text, top_level = @top_level
+    RDoc::Comment.new text, top_level
+  end
+
+end
+
+# This hack allows autoload to work when Dir.pwd is changed for Ruby 1.8 since
+# -I paths are not expanded.
+$LOAD_PATH.each do |load_path|
+  break if load_path[0] == ?/
+  load_path.replace File.expand_path load_path
+end if RUBY_VERSION < '1.9'
+

data/lib/rdoc/text.rb

@@ -5,11 +5,33 @@
 
 require 'strscan'
 
+##
+# For RDoc::Text#snippet
+
+begin
+  gem 'json'
+rescue Gem::LoadError
+end
+
+require 'json'
+
 ##
 # Methods for manipulating comment text
 
 module RDoc::Text
 
+  ##
+  # Maps markup formats to classes that can parse them.  If the format is
+  # unknown, "rdoc" format is used.
+
+  MARKUP_FORMAT = {
+    'rdoc'   => RDoc::Markup,
+    'rd'     => RDoc::RD,
+    'tomdoc' => RDoc::TomDoc,
+  }
+
+  MARKUP_FORMAT.default = RDoc::Markup
+
   ##
   # Maps an encoding to a Hash of characters properly transcoded for that
   # encoding.
@@ -80,9 +102,7 @@ module RDoc::Text
   # Requires the including class to implement #formatter
 
   def markup text
-    document = parse text
-
-    document.accept formatter
+    parse(text).accept formatter
   end
 
   ##
@@ -91,9 +111,10 @@ module RDoc::Text
   def normalize_comment text
     return text if text.empty?
 
-    text = strip_hashes text
-    text = expand_tabs text
-    text = flush_left text
+    text = strip_stars    text
+    text = strip_hashes   text
+    text = expand_tabs    text
+    text = flush_left     text
     text = strip_newlines text
     text
   end
@@ -101,35 +122,24 @@ module RDoc::Text
   ##
   # Normalizes +text+ then builds a RDoc::Markup::Document from it
 
-  def parse text
+  def parse text, format = 'rdoc'
     return text if RDoc::Markup::Document === text
+    return text.parse if RDoc::Comment === text
 
-    text = normalize_comment text
+    text = normalize_comment text # TODO remove, should not be necessary
 
     return RDoc::Markup::Document.new if text =~ /\A\n*\z/
 
-    RDoc::Markup::Parser.parse text
-  rescue RDoc::Markup::Parser::Error => e
-    $stderr.puts <<-EOF
-While parsing markup, RDoc encountered a #{e.class}:
-
-#{e}
-\tfrom #{e.backtrace.join "\n\tfrom "}
-
----8<---
-#{text}
----8<---
-
-RDoc #{RDoc::VERSION}
-
-Ruby #{RUBY_VERSION}-p#{RUBY_PATCHLEVEL} #{RUBY_RELEASE_DATE}
+    MARKUP_FORMAT[format].parse text
+  end
 
-Please file a bug report with the above information at:
+  ##
+  # The first +limit+ characters of +text+ as HTML
 
-https://github.com/rdoc/rdoc/issues
+  def snippet text, limit = 100
+    document = parse text
 
-    EOF
-    raise
+    RDoc::Markup::ToHtmlSnippet.new(limit).convert document
   end
 
   ##
@@ -155,6 +165,8 @@ https://github.com/rdoc/rdoc/issues
   # Strips /* */ style comments
 
   def strip_stars text
+    return text unless text =~ %r%/\*.*\*/%m
+
     encoding = text.encoding if Object.const_defined? :Encoding
 
     text = text.gsub %r%Document-method:\s+[\w:.#=!?]+%, ''
@@ -202,10 +214,10 @@ https://github.com/rdoc/rdoc/issues
 
     until s.eos? do
       case
-      when s.scan(/<tt>.*?<\/tt>/) then # skip contents of tt
+      when s.scan(/<(tt|code)>.*?<\/\1>/) then # skip contents of tt
         html << s.matched.gsub('\\\\', '\\')
-      when s.scan(/<tt>.*?/) then
-        warn 'mismatched <tt> tag' # TODO signal file/line
+      when s.scan(/<(tt|code)>.*?/) then
+        warn "mismatched <#{s[1]}> tag" # TODO signal file/line
         html << s.matched
       when s.scan(/<[^>]+\/?s*>/) then # skip HTML tags
         html << s.matched

data/lib/rdoc/token_stream.rb

@@ -7,6 +7,51 @@
 
 module RDoc::TokenStream
 
+  ##
+  # Converts +token_stream+ to HTML wrapping various tokens with
+  # <tt><span></tt> elements.  The following tokens types are wrapped in spans
+  # with the given class names:
+  #
+  # TkCONSTANT :: 'ruby-constant'
+  # TkKW       :: 'ruby-keyword'
+  # TkIVAR     :: 'ruby-ivar'
+  # TkOp       :: 'ruby-operator'
+  # TkId       :: 'ruby-identifier'
+  # TkNode     :: 'ruby-node'
+  # TkCOMMENT  :: 'ruby-comment'
+  # TkREGEXP   :: 'ruby-regexp'
+  # TkSTRING   :: 'ruby-string'
+  # TkVal      :: 'ruby-value'
+  #
+  # Other token types are not wrapped in spans.
+
+  def self.to_html token_stream
+    token_stream.map do |t|
+      next unless t
+
+      style = case t
+              when RDoc::RubyToken::TkCONSTANT then 'ruby-constant'
+              when RDoc::RubyToken::TkKW       then 'ruby-keyword'
+              when RDoc::RubyToken::TkIVAR     then 'ruby-ivar'
+              when RDoc::RubyToken::TkOp       then 'ruby-operator'
+              when RDoc::RubyToken::TkId       then 'ruby-identifier'
+              when RDoc::RubyToken::TkNode     then 'ruby-node'
+              when RDoc::RubyToken::TkCOMMENT  then 'ruby-comment'
+              when RDoc::RubyToken::TkREGEXP   then 'ruby-regexp'
+              when RDoc::RubyToken::TkSTRING   then 'ruby-string'
+              when RDoc::RubyToken::TkVal      then 'ruby-value'
+              end
+
+      text = CGI.escapeHTML t.text
+
+      if style then
+        "<span class=\"#{style}\">#{text}</span>"
+      else
+        text
+      end
+    end.join
+  end
+
   ##
   # Adds +tokens+ to the collected tokens
 

data/lib/rdoc/tom_doc.rb

@@ -0,0 +1,232 @@
+# :markup: tomdoc
+
+# A parser for TomDoc based on TomDoc 1.0.0-rc1 (02adef9b5a)
+#
+# The TomDoc specification can be found at:
+#
+# http://tomdoc.org
+#
+# The latest version of the TomDoc specification can be found at:
+#
+# https://github.com/mojombo/tomdoc/blob/master/tomdoc.md
+#
+# There are a few differences between this parser and the specification.  A
+# best-effort was made to follow the specification as closely as possible but
+# some choices to deviate were made.
+#
+# A future version of RDoc will warn when a MUST or MUST NOT is violated and
+# may warn when a SHOULD or SHOULD NOT is violated.  RDoc will always try
+# to emit documentation even if given invalid TomDoc.
+#
+# Here are some implementation choices this parser currently makes:
+#
+# This parser allows rdoc-style inline markup but you should not depended on
+# it.
+#
+# This parser allows a space between the comment and the method body.
+#
+# This parser does not require the default value to be described for an
+# optional argument.
+#
+# This parser does not examine the order of sections.  An Examples section may
+# precede the Arguments section.
+#
+# This class is documented in TomDoc format.  Since this is a subclass of the
+# RDoc markup parser there isn't much to see here, unfortunately.
+
+class RDoc::TomDoc < RDoc::Markup::Parser
+
+  # Internal: Token accessor
+
+  attr_reader :tokens
+
+  # Internal: Adds a post-processor which sets the RDoc section based on the
+  # comment's status.
+  #
+  # Returns nothing.
+
+  def self.add_post_processor # :nodoc:
+    RDoc::Markup::PreProcess.post_process do |comment, code_object|
+      next unless code_object and
+                  RDoc::Comment === comment and comment.format == 'tomdoc'
+
+      comment.text.gsub!(/(\A\s*# )(Public|Internal|Deprecated):\s+/) do
+        section = code_object.add_section $2
+        code_object.temporary_section = section
+
+        $1
+      end
+    end
+  end
+
+  add_post_processor
+
+  # Public: Parses TomDoc from text
+  #
+  # text - A String containing TomDoc-format text.
+  #
+  # Examples
+  #
+  #   RDoc::TomDoc.parse <<-TOMDOC
+  #   This method does some things
+  #
+  #   Returns nothing.
+  #   TOMDOC
+  #   # => #<RDoc::Markup::Document:0xXXX @parts=[...], @file=nil>
+  #
+  # Returns an RDoc::Markup::Document representing the TomDoc format.
+
+  def self.parse text
+    parser = new
+
+    parser.tokenize text
+    doc = RDoc::Markup::Document.new
+    parser.parse doc
+    doc
+  end
+
+  # Internal: Extracts the Signature section's method signature
+  #
+  # comment - An RDoc::Comment that will be parsed and have the signature
+  #           extracted
+  #
+  # Returns a String containing the signature and nil if not
+
+  def self.signature comment
+    return unless comment.tomdoc?
+
+    document = comment.parse
+
+    signature = nil
+    found_heading = false
+    found_signature = false
+
+    document.parts.delete_if do |part|
+      next false if found_signature
+
+      found_heading ||=
+        RDoc::Markup::Heading === part && part.text == 'Signature'
+
+      next false unless found_heading
+
+      next true if RDoc::Markup::BlankLine === part
+
+      if RDoc::Markup::Verbatim === part then
+        signature = part
+        found_signature = true
+      end
+    end
+
+    signature and signature.text
+  end
+
+  # Public: Creates a new TomDoc parser.  See also RDoc::Markup::parse
+
+  def initialize
+    super
+
+    @section = nil
+  end
+
+  # Internal: Builds a heading from the token stream
+  #
+  # level - The level of heading to create
+  #
+  # Returns an RDoc::Markup::Heading
+
+  def build_heading level
+    heading = super
+
+    @section = heading.text
+
+    heading
+  end
+
+  # Internal: Builds a verbatim from the token stream.  A verbatim in the
+  # Examples section will be marked as in ruby format.
+  #
+  # margin - The indentation from the margin for lines that belong to this
+  #          verbatim section.
+  #
+  # Returns an RDoc::Markup::Verbatim
+
+  def build_verbatim margin
+    verbatim = super
+
+    verbatim.format = :ruby if @section == 'Examples'
+
+    verbatim
+  end
+
+  # Internal: Builds a paragraph from the token stream
+  #
+  # margin - Unused
+  #
+  # Returns an RDoc::Markup::Paragraph.
+
+  def build_paragraph margin
+    p :paragraph_start => margin if @debug
+
+    paragraph = RDoc::Markup::Paragraph.new
+
+    until @tokens.empty? do
+      type, data, = get
+
+      if type == :TEXT then
+        paragraph << data
+        skip :NEWLINE
+      else
+        unget
+        break
+      end
+    end
+
+    p :paragraph_end => margin if @debug
+
+    paragraph
+  end
+
+  # Internal: Turns text into an Array of tokens
+  #
+  # text - A String containing TomDoc-format text.
+  #
+  # Returns self.
+
+  def tokenize text
+    text.sub!(/\A(Public|Internal|Deprecated):\s+/, '')
+
+    s = StringScanner.new text
+
+    @line = 0
+    @line_pos = 0
+
+    until s.eos? do
+      pos = s.pos
+
+      # leading spaces will be reflected by the column of the next token
+      # the only thing we loose are trailing spaces at the end of the file
+      next if s.scan(/ +/)
+
+      @tokens << case
+                 when s.scan(/\r?\n/) then
+                   token = [:NEWLINE, s.matched, *token_pos(pos)]
+                   @line_pos = s.pos
+                   @line += 1
+                   token
+                 when s.scan(/(Examples|Signature)$/) then
+                   @tokens << [:HEADER, 3, *token_pos(pos)]
+
+                   [:TEXT, s[1], *token_pos(pos)]
+                 when s.scan(/([:\w]\w*)[ ]+- /) then
+                   [:NOTE, s[1], *token_pos(pos)]
+                 else
+                   s.scan(/.*/)
+                   [:TEXT, s.matched.sub(/\r$/, ''), *token_pos(pos)]
+                 end
+    end
+
+    self
+  end
+
+end
+