path: root/jni/ruby/lib/rdoc/parser/c.rb

require 'tsort'

##
# RDoc::Parser::C attempts to parse C extension files.  It looks for
# the standard patterns that you find in extensions: <tt>rb_define_class,
# rb_define_method</tt> and so on.  It tries to find the corresponding
# C source for the methods and extract comments, but if we fail
# we don't worry too much.
#
# The comments associated with a Ruby method are extracted from the C
# comment block associated with the routine that _implements_ that
# method, that is to say the method whose name is given in the
# <tt>rb_define_method</tt> call. For example, you might write:
#
#   /*
#    * Returns a new array that is a one-dimensional flattening of this
#    * array (recursively). That is, for every element that is an array,
#    * extract its elements into the new array.
#    *
#    *    s = [ 1, 2, 3 ]           #=> [1, 2, 3]
#    *    t = [ 4, 5, 6, [7, 8] ]   #=> [4, 5, 6, [7, 8]]
#    *    a = [ s, t, 9, 10 ]       #=> [[1, 2, 3], [4, 5, 6, [7, 8]], 9, 10]
#    *    a.flatten                 #=> [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]
#    */
#    static VALUE
#    rb_ary_flatten(ary)
#        VALUE ary;
#    {
#        ary = rb_obj_dup(ary);
#        rb_ary_flatten_bang(ary);
#        return ary;
#    }
#
#    ...
#
#    void
#    Init_Array()
#    {
#      ...
#      rb_define_method(rb_cArray, "flatten", rb_ary_flatten, 0);
#
# Here RDoc will determine from the rb_define_method line that there's a
# method called "flatten" in class Array, and will look for the implementation
# in the method rb_ary_flatten. It will then use the comment from that
# method in the HTML output. This method must be in the same source file
# as the rb_define_method.
#
# The comment blocks may include special directives:
#
# [Document-class: +name+]
#   Documentation for the named class.
#
# [Document-module: +name+]
#   Documentation for the named module.
#
# [Document-const: +name+]
#   Documentation for the named +rb_define_const+.
#
#   Constant values can be supplied on the first line of the comment like so:
#
#     /* 300: The highest possible score in bowling */
#     rb_define_const(cFoo, "PERFECT", INT2FIX(300));
#
#   The value can contain internal colons so long as they are escaped with a \
#
# [Document-global: +name+]
#   Documentation for the named +rb_define_global_const+
#
# [Document-variable: +name+]
#   Documentation for the named +rb_define_variable+
#
# [Document-method: +method_name+]
#   Documentation for the named method.  Use this when the method name is
#   unambiguous.
#
# [Document-method: <tt>ClassName::method_name<tt>]
#   Documentation for a singleton method in the given class.  Use this when
#   the method name alone is ambiguous.
#
# [Document-method: <tt>ClassName#method_name<tt>]
#   Documentation for a instance method in the given class.  Use this when the
#   method name alone is ambiguous.
#
# [Document-attr: +name+]
#   Documentation for the named attribute.
#
# [call-seq:  <i>text up to an empty line</i>]
#   Because C source doesn't give descriptive names to Ruby-level parameters,
#   you need to document the calling sequence explicitly
#
# In addition, RDoc assumes by default that the C method implementing a
# Ruby function is in the same source file as the rb_define_method call.
# If this isn't the case, add the comment:
#
#   rb_define_method(....);  // in filename
#
# As an example, we might have an extension that defines multiple classes
# in its Init_xxx method. We could document them using
#
#   /*
#    * Document-class:  MyClass
#    *
#    * Encapsulate the writing and reading of the configuration
#    * file. ...
#    */
#
#   /*
#    * Document-method: read_value
#    *
#    * call-seq:
#    *   cfg.read_value(key)            -> value
#    *   cfg.read_value(key} { |key| }  -> value
#    *
#    * Return the value corresponding to +key+ from the configuration.
#    * In the second form, if the key isn't found, invoke the
#    * block and return its value.
#    */

class RDoc::Parser::C < RDoc::Parser

  parse_files_matching(/\.(?:([CcHh])\1?|c([+xp])\2|y)\z/)

  include RDoc::Text

  ##
  # Maps C variable names to names of Ruby classes or modules

  attr_reader :classes

  ##
  # C file the parser is parsing

  attr_accessor :content

  ##
  # Dependencies from a missing enclosing class to the classes in
  # missing_dependencies that depend upon it.

  attr_reader :enclosure_dependencies

  ##
  # Maps C variable names to names of Ruby classes (and singleton classes)

  attr_reader :known_classes

  ##
  # Classes found while parsing the C file that were not yet registered due to
  # a missing enclosing class.  These are processed by do_missing

  attr_reader :missing_dependencies

  ##
  # Maps C variable names to names of Ruby singleton classes

  attr_reader :singleton_classes

  ##
  # The TopLevel items in the parsed file belong to

  attr_reader :top_level

  ##
  # Prepares for parsing a C file.  See RDoc::Parser#initialize for details on
  # the arguments.

  def initialize top_level, file_name, content, options, stats
    super

    @known_classes = RDoc::KNOWN_CLASSES.dup
    @content = handle_tab_width handle_ifdefs_in @content
    @file_dir = File.dirname @file_name

    @classes           = load_variable_map :c_class_variables
    @singleton_classes = load_variable_map :c_singleton_class_variables

    # class_variable => { function => [method, ...] }
    @methods = Hash.new { |h, f| h[f] = Hash.new { |i, m| i[m] = [] } }

    # missing variable => [handle_class_module arguments]
    @missing_dependencies = {}

    # missing enclosure variable => [dependent handle_class_module arguments]
    @enclosure_dependencies = Hash.new { |h, k| h[k] = [] }
    @enclosure_dependencies.instance_variable_set :@missing_dependencies,
                                                  @missing_dependencies

    @enclosure_dependencies.extend TSort

    def @enclosure_dependencies.tsort_each_node &block
      each_key(&block)
    rescue TSort::Cyclic => e
      cycle_vars = e.message.scan(/"(.*?)"/).flatten

      cycle = cycle_vars.sort.map do |var_name|
        delete var_name

        var_name, type, mod_name, = @missing_dependencies[var_name]

        "#{type} #{mod_name} (#{var_name})"
      end.join ', '

      warn "Unable to create #{cycle} due to a cyclic class or module creation"

      retry
    end

    def @enclosure_dependencies.tsort_each_child node, &block
      fetch(node, []).each(&block)
    end
  end

  ##
  # Removes duplicate call-seq entries for methods using the same
  # implementation.

  def deduplicate_call_seq
    @methods.each do |var_name, functions|
      class_name = @known_classes[var_name]
      class_obj  = find_class var_name, class_name

      functions.each_value do |method_names|
        next if method_names.length == 1

        method_names.each do |method_name|
          deduplicate_method_name class_obj, method_name
        end
      end
    end
  end

  ##
  # If two ruby methods share a C implementation (and comment) this
  # deduplicates the examples in the call_seq for the method to reduce
  # confusion in the output.

  def deduplicate_method_name class_obj, method_name # :nodoc:
    return unless
      method = class_obj.method_list.find { |m| m.name == method_name }
    return unless call_seq = method.call_seq

    method_name = method_name[0, 1] if method_name =~ /\A\[/

    entries = call_seq.split "\n"

    matching = entries.select do |entry|
      entry =~ /^\w*\.?#{Regexp.escape method_name}/ or
        entry =~ /\s#{Regexp.escape method_name}\s/
    end

    method.call_seq = matching.join "\n"
  end

  ##
  # Scans #content for rb_define_alias

  def do_aliases
    @content.scan(/rb_define_alias\s*\(
                   \s*(\w+),
                   \s*"(.+?)",
                   \s*"(.+?)"
                   \s*\)/xm) do |var_name, new_name, old_name|
      class_name = @known_classes[var_name]

      unless class_name then
        @options.warn "Enclosing class or module %p for alias %s %s is not known" % [
          var_name, new_name, old_name]
        next
      end

      class_obj = find_class var_name, class_name

      al = RDoc::Alias.new '', old_name, new_name, ''
      al.singleton = @singleton_classes.key? var_name

      comment = find_alias_comment var_name, new_name, old_name

      comment.normalize

      al.comment = comment

      al.record_location @top_level

      class_obj.add_alias al
      @stats.add_alias al
    end
  end

  ##
  # Scans #content for rb_attr and rb_define_attr

  def do_attrs
    @content.scan(/rb_attr\s*\(
                   \s*(\w+),
                   \s*([\w"()]+),
                   \s*([01]),
                   \s*([01]),
                   \s*\w+\);/xm) do |var_name, attr_name, read, write|
      handle_attr var_name, attr_name, read, write
    end

    @content.scan(%r%rb_define_attr\(
                             \s*([\w\.]+),
                             \s*"([^"]+)",
                             \s*(\d+),
                             \s*(\d+)\s*\);
                %xm) do |var_name, attr_name, read, write|
      handle_attr var_name, attr_name, read, write
    end
  end

  ##
  # Scans #content for boot_defclass

  def do_boot_defclass
    @content.scan(/(\w+)\s*=\s*boot_defclass\s*\(\s*"(\w+?)",\s*(\w+?)\s*\)/) do
      |var_name, class_name, parent|
      parent = nil if parent == "0"
      handle_class_module(var_name, :class, class_name, parent, nil)
    end
  end

  ##
  # Scans #content for rb_define_class, boot_defclass, rb_define_class_under
  # and rb_singleton_class

  def do_classes
    do_boot_defclass
    do_define_class
    do_define_class_under
    do_singleton_class
    do_struct_define_without_accessor
  end

  ##
  # Scans #content for rb_define_variable, rb_define_readonly_variable,
  # rb_define_const and rb_define_global_const

  def do_constants
    @content.scan(%r%\Wrb_define_
                   ( variable          |
                     readonly_variable |
                     const             |
                     global_const        )
               \s*\(
                 (?:\s*(\w+),)?
                 \s*"(\w+)",
                 \s*(.*?)\s*\)\s*;
                 %xm) do |type, var_name, const_name, definition|
      var_name = "rb_cObject" if !var_name or var_name == "rb_mKernel"
      handle_constants type, var_name, const_name, definition
    end

    @content.scan(%r%
                  \Wrb_curses_define_const
                  \s*\(
                    \s*
                    (\w+)
                    \s*
                  \)
                  \s*;%xm) do |consts|
      const = consts.first

      handle_constants 'const', 'mCurses', const, "UINT2NUM(#{const})"
    end

    @content.scan(%r%
                  \Wrb_file_const
                  \s*\(
                    \s*
                    "([^"]+)",
                    \s*
                    (.*?)
                    \s*
                  \)
                  \s*;%xm) do |name, value|
      handle_constants 'const', 'rb_mFConst', name, value
    end
  end

  ##
  # Scans #content for rb_define_class

  def do_define_class
    # The '.' lets us handle SWIG-generated files
    @content.scan(/([\w\.]+)\s* = \s*rb_define_class\s*
              \(
                 \s*"(\w+)",
                 \s*(\w+)\s*
              \)/mx) do |var_name, class_name, parent|
      handle_class_module(var_name, :class, class_name, parent, nil)
    end
  end

  ##
  # Scans #content for rb_define_class_under

  def do_define_class_under
    @content.scan(/([\w\.]+)\s* =                  # var_name
                   \s*rb_define_class_under\s*
                   \(
                     \s* (\w+),                    # under
                     \s* "(\w+)",                  # class_name
                     \s*
                     (?:
                       ([\w\*\s\(\)\.\->]+) |      # parent_name
                       rb_path2class\("([\w:]+)"\) # path
                     )
                     \s*
                   \)
                  /mx) do |var_name, under, class_name, parent_name, path|
      parent = path || parent_name

      handle_class_module var_name, :class, class_name, parent, under
    end
  end

  ##
  # Scans #content for rb_define_module

  def do_define_module
    @content.scan(/(\w+)\s* = \s*rb_define_module\s*\(\s*"(\w+)"\s*\)/mx) do
      |var_name, class_name|
      handle_class_module(var_name, :module, class_name, nil, nil)
    end
  end

  ##
  # Scans #content for rb_define_module_under

  def do_define_module_under
    @content.scan(/(\w+)\s* = \s*rb_define_module_under\s*
              \(
                 \s*(\w+),
                 \s*"(\w+)"
              \s*\)/mx) do |var_name, in_module, class_name|
      handle_class_module(var_name, :module, class_name, nil, in_module)
    end
  end

  ##
  # Scans #content for rb_include_module

  def do_includes
    @content.scan(/rb_include_module\s*\(\s*(\w+?),\s*(\w+?)\s*\)/) do |c,m|
      next unless cls = @classes[c]
      m = @known_classes[m] || m

      comment = RDoc::Comment.new '', @top_level
      incl = cls.add_include RDoc::Include.new(m, comment)
      incl.record_location @top_level
    end
  end

  ##
  # Scans #content for rb_define_method, rb_define_singleton_method,
  # rb_define_module_function, rb_define_private_method,
  # rb_define_global_function and define_filetest_function

  def do_methods
    @content.scan(%r%rb_define_
                   (
                      singleton_method |
                      method           |
                      module_function  |
                      private_method
                   )
                   \s*\(\s*([\w\.]+),
                     \s*"([^"]+)",
                     \s*(?:RUBY_METHOD_FUNC\(|VALUEFUNC\(|\(METHOD\))?(\w+)\)?,
                     \s*(-?\w+)\s*\)
                   (?:;\s*/[*/]\s+in\s+(\w+?\.(?:cpp|c|y)))?
                 %xm) do |type, var_name, meth_name, function, param_count, source_file|

      # Ignore top-object and weird struct.c dynamic stuff
      next if var_name == "ruby_top_self"
      next if var_name == "nstr"

      var_name = "rb_cObject" if var_name == "rb_mKernel"
      handle_method(type, var_name, meth_name, function, param_count,
                    source_file)
    end

    @content.scan(%r%rb_define_global_function\s*\(
                             \s*"([^"]+)",
                             \s*(?:RUBY_METHOD_FUNC\(|VALUEFUNC\()?(\w+)\)?,
                             \s*(-?\w+)\s*\)
                (?:;\s*/[*/]\s+in\s+(\w+?\.[cy]))?
                %xm) do |meth_name, function, param_count, source_file|
      handle_method("method", "rb_mKernel", meth_name, function, param_count,
                    source_file)
    end

    @content.scan(/define_filetest_function\s*\(
                     \s*"([^"]+)",
                     \s*(?:RUBY_METHOD_FUNC\(|VALUEFUNC\()?(\w+)\)?,
                     \s*(-?\w+)\s*\)/xm) do |meth_name, function, param_count|

      handle_method("method", "rb_mFileTest", meth_name, function, param_count)
      handle_method("singleton_method", "rb_cFile", meth_name, function,
                    param_count)
    end
  end

  ##
  # Creates classes and module that were missing were defined due to the file
  # order being different than the declaration order.

  def do_missing
    return if @missing_dependencies.empty?

    @enclosure_dependencies.tsort.each do |in_module|
      arguments = @missing_dependencies.delete in_module

      next unless arguments # dependency on existing class

      handle_class_module(*arguments)
    end
  end

  ##
  # Scans #content for rb_define_module and rb_define_module_under

  def do_modules
    do_define_module
    do_define_module_under
  end

  ##
  # Scans #content for rb_singleton_class

  def do_singleton_class
    @content.scan(/([\w\.]+)\s* = \s*rb_singleton_class\s*
                  \(
                    \s*(\w+)
                  \s*\)/mx) do |sclass_var, class_var|
      handle_singleton sclass_var, class_var
    end
  end

  ##
  # Scans #content for struct_define_without_accessor

  def do_struct_define_without_accessor
    @content.scan(/([\w\.]+)\s* = \s*rb_struct_define_without_accessor\s*
              \(
                 \s*"(\w+)",  # Class name
                 \s*(\w+),    # Parent class
                 \s*\w+,      # Allocation function
                 (\s*