ebnf 0.3.5 → 0.3.6

data/lib/ebnf/parser.rb

@@ -17,9 +17,12 @@ module EBNF
           cur_lineno += s.count("\n")
           #debug("eachRule(ws)") { "[#{cur_lineno}] #{s.inspect}" }
         when s = scanner.scan(%r(/\*([^\*]|\*[^\/])*\*/)m)
-          # Eat comments
+          # Eat comments /* .. */
+          debug("eachRule(comment)") { "[#{cur_lineno}] #{s.inspect}" }
+        when s = scanner.scan(%r(\(\*([^\*]|\*[^\)])*\*\))m)
+          # Eat comments (* .. *)
           debug("eachRule(comment)") { "[#{cur_lineno}] #{s.inspect}" }
-        when s = scanner.scan(%r((#|//).*$))
+        when s = scanner.scan(%r((#(?!x)|//).*$))
           # Eat comments
           cur_lineno += s.count("\n")
           debug("eachRule(comment)") { "[#{cur_lineno}] #{s.inspect}" }
@@ -266,18 +269,18 @@ module EBNF
     def terminal(s)
       s = s.strip
       case m = s[0,1]
-      when '"', "'" # STRING1 or STRING2 Terminated by line-end or whitespace
-        l, s = s[1..-1].split(m.rstrip  , 2)
+      when '"', "'" # STRING1 or STRING2
+        l, s = s[1..-1].split(m.rstrip, 2)
         [LL1::Lexer.unescape_string(l), s]
-      when '[' # ENUM, RANGE, O_ENUM, or O_RANGE
+      when '[' # RANGE, O_RANGE
         l, s = s[1..-1].split(/(?<=[^\\])\]/, 2)
         [[:range, LL1::Lexer.unescape_string(l)], s]
       when '#' # HEX
-        s.match(/(#\w+)(.*)$/)
+        s.match(/(#x\h+)(.*)$/)
         l, s = $1, $2
         [[:hex, l], s]
       when /[\w\.]/ # SYMBOL
-        s.match(/(\w+)(.*)$/)
+        s.match(/([\w\.]+)(.*)$/)
         l, s = $1, $2
         [l.to_sym, s]
       when '@' # @pass or @terminals

data/lib/ebnf/rule.rb

@@ -161,11 +161,14 @@ module EBNF
     # Transform EBNF rule to BNF rules:
     #
     #   * Transform (a [n] rule (op1 (op2))) into two rules:
-    #     (a [n] rule (op1 a.2))
+    #     (a [n] rule (op1 _a_1))
     #     (_a_1 [n.1] rule (op2))
-    #   * Transform (a rule (opt b)) into (a rule (alt _empty "foo"))
+    #   * Transform (a rule (opt b)) into (a rule (alt _empty b))
     #   * Transform (a rule (star b)) into (a rule (alt _empty (seq b a)))
     #   * Transform (a rule (plus b)) into (a rule (seq b (star b)
+    #
+    # Transformation includes information used to re-construct non-transformed
+    # AST representation
     # @return [Array<Rule>]
     def to_bnf
       return [self] unless rule?

data/lib/ebnf/writer.rb

@@ -1,5 +1,6 @@
 # -*- encoding: utf-8 -*-
 require 'rdf'
+require 'strscan'    unless defined?(StringScanner)
 
 ##
 # Serialize ruleset back to EBNF
@@ -38,12 +39,27 @@ module EBNF
       Writer.new(rules, out: out)
     end
 
+    ##
+    # Write formatted rules to an IO like object as HTML
+    #
+    # @param  [Array<Rule>] rules
+    # @return [Object]
+    def self.html(*rules)
+      require 'stringio' unless defined?(StringIO)
+      buf = StringIO.new
+      Writer.new(rules, out: buf, html: true)
+      buf.string
+    end
+
     ##
     # @param [Array<Rule>] rules
     # @param [Hash{Symbol => Object}] options
     # @option options [Symbol] :format
     # @option options [#write] :out ($stdout)
+    # @option options [Boolean] :html (false)
+    #   Format as HTML
     def initialize(rules, options = {})
+      @options = options.dup
       out = options.fetch(:out, $stdio)
       #fmt = options.fetch(:format, :ebnf)
 
@@ -51,13 +67,24 @@ module EBNF
       max_id = rules.max_by {|r| r.id.to_s.length}.id.to_s.length
       max_sym = rules.max_by {|r| r.sym.to_s.length}.sym.to_s.length
       lhs_length = max_sym + 3
-      lhs_fmt = "%-#{max_sym}{sym} ::= "
+      lhs_fmt = "%<sym>-#{max_sym}s ::= "
       if max_id > 0
-        lhs_fmt = "%-#{max_id+2}{id} " + lhs_fmt
+        lhs_fmt = "%<id>-#{max_id+2}s " + lhs_fmt
         lhs_length += max_id + 3
       end
       rhs_length = LINE_LENGTH - lhs_length
 
+      if @options[:html]
+        # Output as formatted HTML
+        require 'haml'
+        html = Haml::Engine.new(HAML_DESC).render(self, rules: rules) do |rule|
+          formatted_expr = format(rule.expr)
+          formatted_expr.length > rhs_length ? format(rule.expr, "\n") : formatted_expr
+        end
+        out.write html
+        return
+      end
+
       # Format each rule, considering the available rhs size
       rules.each do |rule|
         buffer = if rule.pass?
@@ -78,48 +105,118 @@ module EBNF
     protected
     # Format the expression part of a rule
     def format(expr, sep = nil)
-      return expr.to_s if expr.is_a?(Symbol)
-      return %("#{escape(expr)}") if expr.is_a?(String)
+      return (@options[:html] ? %(<a href="#grammar-production-#{expr}">#{expr}</a>) : expr.to_s) if expr.is_a?(Symbol)
+      if expr.is_a?(String)
+        if expr.length == 1
+          return format_char(expr)
+        elsif expr =~ /\A#x\h+/
+          return (@options[:html] ? %(<code class="grammar-char-escape">#{expr}</code>) : expr)
+        elsif expr =~ /"/
+          return (@options[:html] ? %('<code class="grammar-literal">#{escape(expr, "'")}</code>') : %('#{escape(expr, "'")}'))
+        else
+          return (@options[:html] ? %("<code class="grammar-literal">#{escape(expr, '"')}</code>") : %("#{escape(expr, '"')}"))
+        end
+      end
+      parts = {
+        alt:    (@options[:html] ? "<code>|</code> " : "| "),
+        diff:   (@options[:html] ? "<code>-</code> " : "- "),
+        star:   (@options[:html] ? "<code>*</code> " : "*"),
+        plus:   (@options[:html] ? "<code>+</code> " : "+"),
+        opt:    (@options[:html] ? "<code>?</code> " : "?")
+      }
+      lparen = (@options[:html] ? "<code>(</code> " : "(")
+      rparen = (@options[:html] ? "<code>)</code> " : ")")
 
       case expr.first
       when :alt, :diff
-        this_sep = (sep ? sep : " ") + {alt: "| ", diff: "- "}[expr.first.to_sym]
+        this_sep = (sep ? sep : " ") + parts[expr.first.to_sym]
         expr[1..-1].map {|e| format(e)}.join(this_sep)
       when :star, :plus, :opt
         raise "Expected star expression to have a single operand" unless expr.length == 2
-        char = {star: "*", plus: "+", opt: "?"}[expr.first.to_sym]
+        char = parts[expr.first.to_sym]
         r = format(expr[1])
         (r.start_with?("(") || Array(expr[1]).length == 1) ? "#{r}#{char}" : "(#{r})#{char}"
+      when :hex
+        (@options[:html] ? %(<code class="grammar-char-escape">#{expr.last}</code>) : expr.last)
       when :range
-        parts = expr.last.split(/(?!\\)-/, 2)
-        "[" + parts.map {|e| format(e)[1..-2]}.join("-") + "]"
+        format_range(expr.last)
       when :seq
         this_sep = (sep ? sep : " ")
-        expr[1..-1].map {|e| r = format(e); Array(e).length > 2 ? "(#{r})" : r}.join(this_sep)
+        expr[1..-1].map {|e| r = format(e); Array(e).length > 2 ? "#{lparen}#{r}#{rparen}" : r}.join(this_sep)
       else
         raise "Unknown operator: #{expr.first}"
       end
     end
 
-    def escape(string)
+    # Format a single-character string, prefering hex for non-main ASCII
+    def format_char(c)
+      case c.ord
+      when 0x22         then (@options[:html] ? %('<code class="grammar-literal">"</code>') : %{'"'})
+      when (0x23..0x7e) then (@options[:html] ? %("<code class="grammar-literal">#{c}</code>") : %{"#{c}"})
+      else                   (@options[:html] ? %(<code class="grammar-char-escape">#{escape_hex(c)}</code>) : escape_hex(c))
+      end
+    end
+
+    # Format a range
+    def format_range(string)
+      lbrac  =  (@options[:html] ? "<code>[</code> " : "[")
+      rbrac  =  (@options[:html] ? "<code>]</code> " : "]")
+      dash   = (@options[:html] ? "<code>-</code> " : "-")
+
+      buffer = lbrac
+      s = StringScanner.new(string)
+      while !s.eos?
+        case
+        when s.scan(/\A[!"\u0024-\u007e]+/)
+          buffer << (@options[:html] ? %(<code class="grammar-literal">#{s.matched}</code>) : s.matched)
+        when s.scan(/\A#x\h+/)
+          buffer << (@options[:html] ? %(<code class="grammar-char-escape">#{s.matched}</code>) : s.matched)
+        when s.scan(/\A-/)
+          buffer << dash
+        else
+          buffer << (@options[:html] ? %(<code class="grammar-char-escape">#{escape_hex(s.getch)}</code>) : escape_hex(s.getch))
+        end
+      end
+      buffer + rbrac
+    end
+
+    # Escape a string, using as many UTF-8 characters as possible
+    def escape(string, quote = '"')
       buffer = ""
       string.each_char do |c|
-        buffer << case c.to_s
-        when "\t" then "\\t"
-        when "\n" then "\\n"
-        when "\r" then "\\r"
-        when "\\" then "\\\\"
-        #when "("  then "\\("
-        #when ")"  then "\\)"
-        #when "["  then "\\["
-        #when "]"  then "\\]"
-        #when "-"  then "\\\\-"
-        when "'"  then "\\'"
-        when '"'  then "\\\""
-        else c
+        buffer << case (u = c.ord)
+        when (0x00..0x1f) then "#x%02X" % u
+        when quote.ord    then "#x%02X" % u
+        else                   c
         end
       end
       buffer
     end
+
+    def escape_hex(u)
+      fmt = case u.ord
+      when 0x0000..0x00ff then "#x%02X"
+      when 0x0100..0xffff then "#x%04X"
+      else                     "#x%08X"
+      end
+      sprintf(fmt, u.ord)
+    end
+
+    HAML_DESC = %q(
+      %table.grammar
+        %tbody#grammar-productions
+          - rules.each do |rule|
+            %tr{id: "grammar-production-#{rule.sym}"}
+              - if rule.pass?
+                %td{colspan: 3}
+                  %code<="@pass"
+              - else
+                %td<= "[#{rule.id}]"
+                %td<
+                  %code<= rule.sym
+                %td<= "::="
+              %td
+                != yield rule
+    ).gsub(/^      /, '')
   end
 end

metadata

@@ -1,83 +1,103 @@
 --- !ruby/object:Gem::Specification
 name: ebnf
 version: !ruby/object:Gem::Version
-  version: 0.3.5
+  version: 0.3.6
 platform: ruby
 authors:
 - Gregg Kellogg
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-12-22 00:00:00.000000000 Z
+date: 2014-04-23 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: sxp
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '0.1'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.1.3
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '0.1'
+    - - ">="
+      - !ruby/object:Gem::Version
+        version: 0.1.3
 - !ruby/object:Gem::Dependency
   name: rdf
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '1.1'
   type: :runtime
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: '0'
+        version: '1.1'
+- !ruby/object:Gem::Dependency
+  name: haml
+  requirement: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '4.0'
+  type: :runtime
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    requirements:
+    - - "~>"
+      - !ruby/object:Gem::Version
+        version: '4.0'
 - !ruby/object:Gem::Dependency
   name: rspec
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: 2.12.0
+        version: '2.14'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: 2.12.0
+        version: '2.14'
 - !ruby/object:Gem::Dependency
   name: yard
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.8.3
+        version: '0.8'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - "~>"
       - !ruby/object:Gem::Version
-        version: 0.8.3
+        version: '0.8'
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '0'
 description: EBNF is a Ruby parser for W3C EBNF and a parser generator for compliant
@@ -93,19 +113,10 @@ files:
 - README.md
 - UNLICENSE
 - VERSION
-- lib/ebnf/base.rb
-- lib/ebnf/bnf.rb
-- lib/ebnf/ll1/lexer.rb
-- lib/ebnf/ll1/parser.rb
-- lib/ebnf/ll1/scanner.rb
-- lib/ebnf/ll1.rb
-- lib/ebnf/parser.rb
-- lib/ebnf/rule.rb
-- lib/ebnf/version.rb
-- lib/ebnf/writer.rb
-- lib/ebnf.rb
+- bin/ebnf
 - etc/doap.ttl
 - etc/ebnf.ebnf
+- etc/ebnf.html
 - etc/ebnf.ll1.sxp
 - etc/ebnf.rb
 - etc/ebnf.sxp
@@ -113,7 +124,17 @@ files:
 - etc/turtle.ll1.sxp
 - etc/turtle.rb
 - etc/turtle.sxp
-- bin/ebnf
+- lib/ebnf.rb
+- lib/ebnf/base.rb
+- lib/ebnf/bnf.rb
+- lib/ebnf/ll1.rb
+- lib/ebnf/ll1/lexer.rb
+- lib/ebnf/ll1/parser.rb
+- lib/ebnf/ll1/scanner.rb
+- lib/ebnf/parser.rb
+- lib/ebnf/rule.rb
+- lib/ebnf/version.rb
+- lib/ebnf/writer.rb
 homepage: http://github.com/gkellogg/ebnf
 licenses:
 - Public Domain
@@ -124,17 +145,17 @@ require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: 1.9.2
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
 rubyforge_project: 
-rubygems_version: 2.1.11
+rubygems_version: 2.2.2
 signing_key: 
 specification_version: 4
 summary: EBNF parser and parser generator.