Class: EBNF::Writer

Inherits:
Object
  • Object
show all
Defined in:
lib/ebnf/writer.rb

Constant Summary collapse

LINE_LENGTH = 
80
LINE_LENGTH_HTML = 
200
UNICODE_ESCAPE_NAMES =

UNICODE escape names From en.wikipedia.org/wiki/List_of_Unicode_characters

{
  0x00 => 'null',
  0x01 => 'start of heading',
  0x02 => 'start of text',
  0x03 => 'end of text',
  0x04 => 'end of transmission',
  0x05 => 'enquiry',
  0x06 => 'acknowledge',
  0x07 => 'bell',
  0x08 => 'backspace',
  0x09 => 'horizontal tab',
  0x0A => 'new line',
  0x0B => 'vertical tab',
  0x0C => 'form feed',
  0x0D => 'carriage return',
  0x0E => 'shift out',
  0x0F => 'shift in',
  0x10 => 'data link escape',
  0x11 => 'device control 1',
  0x12 => 'device control 2',
  0x13 => 'device control 3',
  0x14 => 'device control 4',
  0x15 => 'negative acknowledge',
  0x16 => 'synchronous idle',
  0x17 => 'end of trans. block',
  0x18 => 'cancel',
  0x19 => 'end of medium',
  0x1A => 'substitute',
  0x1B => 'escape',
  0x1C => 'file separator',
  0x1D => 'group separator',
  0x1E => 'record separator',
  0x1F => 'unit separator',
  0x20 => 'space',
  0x22 => 'dquote',
  0x27 => 'apos',
  0x2F => 'slash',
  0x5C => 'backslash',
  0x60 => 'grave',
  0x7F => 'delete',
  0x80 => 'padding character',
  0x81 => 'high octet preset',
  0x82 => 'break permitted here',
  0x83 => 'no break here',
  0x84 => 'index',
  0x85 => 'next line',
  0x86 => 'start of selected area',
  0x87 => 'end of selected area',
  0x88 => 'character tabulation set',
  0x89 => 'character tabulation with justification',
  0x8A => 'line tabulation set',
  0x8B => 'partial line forward',
  0x8C => 'partial line backward',
  0x8D => 'reverse line feed',
  0x8E => 'single-shift two',
  0x8F => 'single-shift three',
  0x90 => 'device control string',
  0x91 => 'private use 1',
  0x92 => 'private use 2',
  0x93 => 'set transmit state',
  0x94 => 'cancel character',
  0x95 => 'message waiting',
  0x96 => 'start of protected area',
  0x97 => 'end of protected area',
  0x98 => 'start of string',
  0x99 => 'single graphic character introducer',
  0x9A => 'single character intro introducer',
  0x9B => 'control sequence introducer',
  0x9C => 'string terminator',
  0x9D => 'operating system command',
  0x9E => 'private message',
  0x9F => 'application program command',
}
ERB_DESC = 
%(<!-- Generated with ebnf version #{EBNF::VERSION}. See https://github.com/dryruby/ebnf. -->\n) +
  %q(<table class="grammar">
    <tbody id="grammar-productions" class="<%= @format %>">
      <% for rule in @rules %>
      <tr<%= %{ id="grammar-#{rule[:class]}-#{rule.sym}"} unless %w(=/ |).include?(rule.assign) || rule.sym.nil?%>>
        <% if rule.id %>
        <td<%= " colspan=2" unless rule.sym %>><%= rule.id %></td>
        <% end %>
        <% if rule.sym %>
        <td><code><%== (rule.sym unless rule.class == :declaration) %></code></td>
        <% end %>
        <td><%= rule.assign %></td>
        <td><%= rule.formatted %></td>
      </tr>
      <% end %>
    </tbody>
  </table>
).gsub(/^      /, '')

Class Method Summary collapse

Instance Method Summary collapse

Constructor Details

#initialize(rules, out: $stdout, html: false, format: :ebnf, validate: false, **options) ⇒ Writer

Returns a new instance of Writer.

Parameters:

  • rules (Array<Rule>)
  • format (:abnf, :ebnf, :isoebnf) (defaults to: :ebnf) —

    (:ebnf)

  • html (Boolean) (defaults to: false) —

    (false) generate HTML output

  • validate (Boolean) (defaults to: false) —

    (false) validate generated HTML.

  • options (Hash{Symbol => Object})
  • out (#write) (defaults to: $stdout) —

    ($stdout)



145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
 
# File 'lib/ebnf/writer.rb', line 145

def initialize(rules, out: $stdout, html: false, format: :ebnf, validate: false, **options)
  @options = options.merge(html: html)
  return if rules.empty?

  # Determine max LHS length
  format_meth = "format_#{format}".to_sym
  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 + 1
  lhs_fmt = case format
  when :abnf    then "%<sym>-#{max_sym}s = "
  when :ebnf    then "%<sym>-#{max_sym}s ::= "
  when :isoebnf then "%<sym>-#{max_sym}s = "
  end
  if format == :ebnf && max_id > 0
    lhs_fmt = "%<id>-#{max_id+2}s " + lhs_fmt
    lhs_length += max_id + 3
  end
  rhs_length = (html ? LINE_LENGTH_HTML : LINE_LENGTH) - lhs_length

  if html
    # Output as formatted HTML
    begin
      require 'erubis'
      require 'htmlentities'
      @coder = HTMLEntities.new
      eruby = Erubis::Eruby.new(ERB_DESC)
      formatted_rules = rules.map do |rule|
        if rule.kind == :terminals || rule.kind == :pass
          OpenStruct.new(id: ("@#{rule.kind}"),
                         class: :declaration,
                         sym: rule.kind,
                         assign: nil,
                         formatted: (
                           rule.kind == :terminals ?
                             "<strong># Productions for terminals</strong>" :
                             self.send(format_meth, rule.expr)))
        else
          formatted_expr = self.send(format_meth, rule.expr)
          # Measure text without markup
          formatted_expr_text = formatted_expr.gsub(%r{</?\w+[^>]*>}, '')
          if formatted_expr_text.length > rhs_length && (format != :abnf || rule.alt?)
            lines = []
            # Can only reasonably split apart alts
            self.send(format_meth, rule.expr, sep: "--rule-extensions--").
            split(/\s*--rule-extensions--\s*/).each_with_index do |formatted, ndx|
              assign = case format
              when :ebnf
                formatted.sub!(%r{\s*<code[^>]*>\|</code>\s*}, '')
                (ndx > 0 ? (rule.alt? ? '<code class="grammar-alt">|</code>' : '') : '::=')
              when :abnf
                formatted.sub!(%r{\s*<code[^>]>/</code>\s*}, '')
                (ndx > 0 ? '<code class="grammar-alt">=/</code>' : '=')
              else
                formatted.sub!(%r{\s*<code[^>]>\|</code>\s*}, '')
                (ndx > 0 ? (rule.alt? ? '<code class="grammar-alt">|</code>' : '') : '=')
              end
              lines << OpenStruct.new(id: ((ndx == 0 ? "[#{rule.id}]" : "") if rule.id),
                                      sym: (rule.sym if ndx == 0 || format == :abnf),
                                      class: :production,
                                      assign: assign,
                                      formatted: formatted)
            end
            if format == :isoebnf
              lines << OpenStruct.new(assign: ';')
            end
            lines
          else
            OpenStruct.new(id: ("[#{rule.id}]" if rule.id),
                           class: :production,
                           sym: rule.sym,
                           assign: (format == :ebnf ? '::=' : '='),
                           formatted: (formatted_expr + (format == :isoebnf ? ' ;' : '')))
          end
        end
      end.flatten

      html_result = eruby.evaluate(format: format, rules: formatted_rules)

      if validate
        begin
          require 'nokogiri'
          # Validate the output HTML
          doc = ::Nokogiri::HTML5("<!DOCTYPE html>" + html_result, max_errors: 10)
          raise EncodingError, "Errors found in generated HTML:\n  " +
            doc.errors.map(&:to_s).join("\n  ") unless doc.errors.empty?
        rescue LoadError, NoMethodError
          # Skip
        end
      end

      out.write html_result
      return
    rescue LoadError
      $stderr.puts "Generating HTML requires erubis and htmlentities gems to be loaded"
    end
  end

  # Format each rule, considering the available rhs size
  rules.each do |rule|
    buffer = if rule.pass?
      "\n%-#{lhs_length-2}s      " % '@pass'
    elsif rule.kind == :terminals
      "\n%-#{lhs_length-2}s" % '@terminals'
    else
      lhs_fmt % {id: "[#{rule.id}]", sym: rule.sym}
    end
    formatted_expr = self.send(format_meth, rule.expr)
    if formatted_expr.length > rhs_length && (format != :abnf || rule.alt?)
      if format == :abnf
        # No whitespace, use =/
        self.send(format_meth, rule.expr, sep: "--rule-extensions--").
        split(/\s*--rule-extensions--\s*/).each_with_index do |formatted, ndx|
          if ndx > 0
            buffer << "\n" + lhs_fmt.sub('= ', '=/') % {id: "[#{rule.id}]", sym: rule.sym}
          end
          buffer << formatted.sub(/\s*\/\s*/, '')
        end
      else
        # Space out past "= "
        buffer << self.send(format_meth, rule.expr, sep: ("\n" + " " * (lhs_length + (rule.alt? ? 2 : 4) - (format == :ebnf ? 0 : 2))))
        buffer << ("\n" + " " * (lhs_length) + ';') if format == :isoebnf
      end
    else
      buffer << formatted_expr + (format == :isoebnf ? ' ;'  : '')
    end
    buffer << "\n\n" if [:terminals, :pass].include?(rule.kind)
    out.puts(buffer)
  end
end

Class Method Details

.html(*rules, format: :ebnf, validate: false) ⇒ Object

Write formatted rules to an IO like object as HTML

Parameters:

  • rules (Array<Rule>)
  • format (:abnf, :ebnf, :isoebnf) (defaults to: :ebnf) —

    (:ebnf)

  • validate (Boolean) (defaults to: false) —

    (false) validate generated HTML.

Returns:

  • (Object) 


131
132
133
134
135
136
 
# File 'lib/ebnf/writer.rb', line 131

def self.html(*rules, format: :ebnf, validate: false)
  require 'stringio' unless defined?(StringIO)
  buf = StringIO.new
  Writer.new(rules, out: buf, html: true, format: format, validate: validate)
  buf.string
end

.print(*rules, format: :ebnf) ⇒ Object

Format rules to $stdout

Parameters:

  • rules (Array<Rule>)
  • format (:abnf, :ebnf, :isoebnf) (defaults to: :ebnf) —

    (:ebnf)

Returns:

  • (Object) 


109
110
111
 
# File 'lib/ebnf/writer.rb', line 109

def self.print(*rules, format: :ebnf)
  write($stdout, *rules, format: format)
end

.string(*rules, format: :ebnf) ⇒ Object

Format rules to a String

Parameters:

  • rules (Array<Rule>)
  • format (:abnf, :ebnf, :isoebnf) (defaults to: :ebnf) —

    (:ebnf)

Returns:

  • (Object) 


96
97
98
99
100
101
 
# File 'lib/ebnf/writer.rb', line 96

def self.string(*rules, format: :ebnf)
  require 'stringio' unless defined?(StringIO)
  buf = StringIO.new
  write(buf, *rules, format: format)
  buf.string
end

.write(out, *rules, format: :ebnf) ⇒ Object

Write formatted rules to an IO like object

Parameters:

  • out (Object)
  • rules (Array<Rule>)
  • format (:abnf, :ebnf, :isoebnf) (defaults to: :ebnf) —

    (:ebnf)

Returns:

  • (Object) 


120
121
122
 
# File 'lib/ebnf/writer.rb', line 120

def self.write(out, *rules, format: :ebnf)
  Writer.new(rules, out: out, format: format)
end

Instance Method Details

#escape_abnf_hex(u) ⇒ Object (protected) 



589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
 
# File 'lib/ebnf/writer.rb', line 589

def escape_abnf_hex(u)
  fmt = case u.ord
  when 0x0000..0x00ff then "%02X"
  when 0x0100..0xffff then "%04X"
  else                     "%08X"
  end
  char =  "%x" + (fmt % u.ord).upcase
  if @options[:html]
    char = if UNICODE_ESCAPE_NAMES.include?(u.ord)
      %(<abbr title="#{UNICODE_ESCAPE_NAMES[u.ord]}">#{char}</abbr>)
    elsif ([::Unicode::Types.of(u)] - %w(Control Private-use Surrogate Noncharacter Reserved)).empty?
      %(<abbr title="unicode '#{@coder.encode u}'">#{char}</abbr>)
    else
      uni_esc = "U+%04X" % u.ord
      %(<abbr title="unicode #{uni_esc}">#{char}</abbr>)
    end
    %(<code class="grammar-char-escape">#{char}</code>)
  else
    char
  end
end

#escape_ebnf_hex(u) ⇒ Object (protected) 



402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
 
# File 'lib/ebnf/writer.rb', line 402

def escape_ebnf_hex(u)
  fmt = case u.ord
  when 0x0000..0x00ff then "#x%02X"
  when 0x0100..0xffff then "#x%04X"
  else                     "#x%08X"
  end
  char = fmt % u.ord
  if @options[:html]
    char = if UNICODE_ESCAPE_NAMES.include?(u.ord)
      %(<abbr title="#{UNICODE_ESCAPE_NAMES[u.ord]}">#{char}</abbr>)
    elsif ([::Unicode::Types.of(u)] - %w(Control Private-use Surrogate Noncharacter Reserved)).empty?
      %(<abbr title="unicode '#{@coder.encode u}'">#{char}</abbr>)
    else
      uni_esc = "U+%04X" % u.ord
      %(<abbr title="unicode #{uni_esc}">#{char}</abbr>)
    end
    %(<code class="grammar-char-escape">#{char}</code>)
  else
    char
  end
end

#format_abnf(expr, sep: nil, embedded: false, sensitive: true) ⇒ Object (protected)

Format the expression part of a rule



429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
 
# File 'lib/ebnf/writer.rb', line 429

def format_abnf(expr, sep: nil, embedded: false, sensitive: true)
  return (@options[:html] ? %(<a href="#grammar-production-#{@coder.encode expr}">#{@coder.encode expr}</a>) : expr.to_s) if expr.is_a?(Symbol)
  if expr.is_a?(String)
    if expr.length == 1
      return format_abnf_char(expr)
    elsif expr.start_with?('%')
      # Already encoded
      return expr
    elsif expr =~ /"/
      # Split into segments
      segments = expr.split('"')

      return format_abnf_char(expr) if segments.empty?

      seq = segments.inject([]) {|memo, s| memo.concat([[:hex, "#x22"], s])}[1..-1]
      seq.unshift(:seq)
      return format_abnf(seq, sep: nil, embedded: false)
    else
      return (@options[:html] ? %("<code class="grammar-literal">#{@coder.encode expr}</code>") : %("#{expr}"))
    end
  end
  parts = {
    alt:    (@options[:html] ? %(<code class="grammar-alt">/</code>) : "/ "),
    star:   (@options[:html] ? %(<code class="grammar-star">*</code>) : "*"),
    plus:   (@options[:html] ? %(<code class="grammar-plus">+</code>) : "1*"),
    opt:    (@options[:html] ? %(<code class="grammar-opt">?</code>) : "?")
  }
  lbrac = (@options[:html] ? %(<code class="grammar-brac">[</code>) : "[")
  rbrac = (@options[:html] ? %(<code class="grammar-brac">]</code>) : "]")
  lparen = (@options[:html] ? %[<code class="grammar-paren">(</code>] : "(")
  rparen = (@options[:html] ? %[<code class="grammar-paren">)</code>] : ")")

  case expr.first
  when :istr
    # FIXME: if string part is segmented, need to do something different
    format_abnf(expr.last, embedded: true, sensitive: false)
  when :alt
    this_sep = (sep ? sep : " ") + parts[expr.first.to_sym]
    res = expr[1..-1].map {|e| format_abnf(e, embedded: true)}.join(this_sep)
    embedded ? (lparen + res + rparen) : res
  when :diff
    raise RangeError, "ABNF does not support the diff operator"
  when :opt
    char = parts[expr.first.to_sym]
    r = format_abnf(expr[1], embedded: true)
    "#{lbrac}#{r}#{rbrac}"
  when :plus, :star
    char = parts[expr.first.to_sym]
    r = format_abnf(expr[1], embedded: true)
    "#{char}#{r}"
  when :hex
    escape_abnf_hex(expr.last[2..-1].hex.chr)
  when :range
    # Returns an [:alt] or [:not [:alt]] if composed of multiple sequences
    # Note: ABNF does not support the `not` operator
    res = format_abnf_range(expr.last)
    res.is_a?(Array) ?
      format_abnf(res, embedded: true) :
      res
  when :seq
    this_sep = (sep ? sep : " ")
    res = expr[1..-1].map do |e|
      format_abnf(e, embedded: true)
    end.join(this_sep)
    embedded ? (lparen + res + rparen) : res
  when :rept
    # Expand repetition
    min, max, value = expr[1..-1]
    r = format_abnf(value, embedded: true)
    if min == max
      "#{min}#{r}"
    elsif min == 0 && max == '*'
      "#{parts[:star]}#{r}"
    elsif min > 0 && max == '*'
      "#{min}#{parts[:star]}#{r}"
    elsif min == 0
      "#{parts[:star]}#{max}#{r}"
    else
      "#{min}#{parts[:star]}#{max}#{r}"
    end
  else
    raise "Unknown operator: #{expr.first}"
  end
end

#format_abnf_char(c) ⇒ Object (protected)

Format a single-character string, prefering hex for non-main ASCII



515
516
517
518
519
520
521
 
# File 'lib/ebnf/writer.rb', line 515

def format_abnf_char(c)
  if /[\x20-\x21\x23-\x7E]/.match?(c)
    @options[:html] ? %("<code class="grammar-literal">#{@coder.encode c}</code>") : c.inspect
  else
    escape_abnf_hex(c)
  end
end

#format_abnf_range(string) ⇒ Object (protected)

Format a range

Presumes range has already been validated

Raises:

  • (RangeError) 


526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
 
# File 'lib/ebnf/writer.rb', line 526

def format_abnf_range(string)
  alt, o_dash = [:alt], false

  raise RangeError, "cannot format #{string.inspect} an ABNF range" if string.start_with?('^')

  if string.end_with?('-')
    o_dash = true
    string = string[0..-2]
  end

  scanner = StringScanner.new(string)
  hexes, deces = [], []
  in_range = false
  # Build op (alt) from different ranges/enums
  while !scanner.eos?
    if hex = scanner.scan(Terminals::HEX)
      # Append any decimal values
      alt << "%d" + deces.join(".") unless deces.empty?
      deces = []
      hex = hex.upcase

      if in_range
        # Add "." sequences for any previous hexes
        alt << "%x" + hexes[0..-2].join(".") if hexes.length > 1
        alt << "%x#{hexes.last}-#{hex[2..-1]}"
        in_range, hexes = false, []
      else
        hexes << hex[2..-1]
      end
    elsif dec = scanner.scan(Terminals::R_CHAR)
      # Append any hexadecimal values
      alt << "%x" + hexes.join(".") unless hexes.empty?
      hexes = []

      if in_range
        # Add "." sequences for any previous hexes
        alt << "%d" + deces[0..-2].join(".") if deces.length > 1
        alt << "%d#{deces.last}-#{dec.codepoints.first}"
        in_range, deces = false, []
      else
        deces << dec.codepoints.first.to_s
      end
    end

    in_range = true if scanner.scan(/\-/)
  end

  deces << '45' if o_dash

  # Append hexes and deces as "." sequences (should be only one)
  alt << "%d" + deces.join(".") unless deces.empty?
  alt << "%x" + hexes.join(".") unless hexes.empty?

  # FIXME: HTML abbreviations?
  if alt.length == 2
    # Just return the range or enum
    alt.last
  else
    # Return the alt, which will be further formatted
    alt
  end
end

#format_ebnf(expr, sep: nil, embedded: false) ⇒ Object (protected)

Format the expression part of a rule



283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
 
# File 'lib/ebnf/writer.rb', line 283

def format_ebnf(expr, sep: nil, embedded: false)
  return (@options[:html] ? %(<a href="#grammar-production-#{@coder.encode expr}">#{@coder.encode expr}</a>) : expr.to_s) if expr.is_a?(Symbol)
  if expr.is_a?(String)
    return expr.length == 1 ?
      format_ebnf_char(expr) :
      format_ebnf_string(expr)
  end
  parts = {
    alt:    (@options[:html] ? %(<code class="grammar-alt">|</code> ) : "| "),
    diff:   (@options[:html] ? %(<code class="grammar-diff">-</code> ) : "- "),
    star:   (@options[:html] ? %(<code class="grammar-star">*</code>) : "*"),
    plus:   (@options[:html] ? %(<code class="grammar-plus">+</code>) : "+"),
    opt:    (@options[:html] ? %(<code class="grammar-opt">?</code>) : "?")
  }
  lparen = (@options[:html] ? %[<code class="grammar-paren">(</code>] : "(")
  rparen = (@options[:html] ? %[<code class="grammar-paren">)</code>] : ")")

  case expr.first
  when :istr
    # Looses fidelity, but, oh well ...
    format_ebnf(expr.last, embedded: true)
  when :alt, :diff
    this_sep = (sep ? sep : " ") + parts[expr.first.to_sym]
    res = expr[1..-1].map {|e| format_ebnf(e, embedded: true)}.join(this_sep)
    embedded ? (lparen + res + rparen) : res
  when :star, :plus, :opt
    char = parts[expr.first.to_sym]
    r = format_ebnf(expr[1], embedded: true)
    "#{r}#{char}"
  when :hex
    escape_ebnf_hex(expr.last[2..-1].hex.chr(Encoding::UTF_8))
  when :range
    format_ebnf_range(expr.last)
  when :seq
    this_sep = (sep ? sep : " ")
    res = expr[1..-1].map do |e|
      format_ebnf(e, embedded: true)
    end.join(this_sep)
    embedded ? (lparen + res + rparen) : res
  when :rept
    # Expand repetition
    min, max, value = expr[1..-1]
    if min == 0 && max == 1
      format_ebnf([:opt, value], sep: sep, embedded: embedded)
    elsif min == 0 && max == '*'
      format_ebnf([:star, value], sep: sep, embedded: embedded)
    elsif min == 1 && max == '*'
      format_ebnf([:plus, value], sep: sep, embedded: embedded)
    else
      val2 = [:seq]
      while min > 0
        val2 << value
        min -= 1
        max -= 1 unless max == '*'
      end
      if max == '*'
        val2 << [:star, value]
      else
        opt = nil
        while max > 0
          opt = [:opt, opt ? [:seq, value, opt] : value]
          max -= 1
        end
        val2 << opt if opt
      end
      format_ebnf(val2, sep: sep, embedded: embedded)
    end
  else
    raise "Unknown operator: #{expr.first}"
  end
end

#format_ebnf_char(c) ⇒ Object (protected)

Format a single-character string, prefering hex for non-main ASCII



356
357
358
359
360
361
362
363
364
365
 
# File 'lib/ebnf/writer.rb', line 356

def format_ebnf_char(c)
  quote = c.as_dquote? ? '"' : "'"
  case c.ord
  when 0x21           then (@options[:html] ? %("<code class="grammar-literal">#{@coder.encode c}</code>") : %{"#{c}"})
  when 0x22           then (@options[:html] ? %('<code class="grammar-literal">&quot;</code>') : %{'"'})
  when (0x23..0x7e)   then (@options[:html] ? %(#{quote}<code class="grammar-literal">#{@coder.encode c}</code>#{quote}) : %{#{quote}#{c}#{quote}})
  when (0x80..0xFFFD) then (@options[:html] ? %(#{quote}<code class="grammar-literal">#{@coder.encode c}</code>#{quote}) : %{#{quote}#{c}#{quote}})
  else                     escape_ebnf_hex(c)
  end
end

#format_ebnf_range(string) ⇒ Object (protected)

Format a range



368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
 
# File 'lib/ebnf/writer.rb', line 368

def format_ebnf_range(string)
  lbrac  =  (@options[:html] ? %(<code class="grammar-brac">[</code>) : "[")
  rbrac  =  (@options[:html] ? %(<code class="grammar-brac">]</code>) : "]")

  buffer = lbrac
  s = StringScanner.new(string)
  while !s.eos?
    case
    when s.scan(/\A[!"\u0024-\u007e]+/)
      buffer << (@options[:html] ? %(<code class="grammar-literal">#{@coder.encode s.matched}</code>) : s.matched)
    when s.scan(/\A#x\h+/)
      buffer << escape_ebnf_hex(s.matched[2..-1].hex.chr(Encoding::UTF_8))
    else
      buffer << escape_ebnf_hex(s.getch)
    end
  end
  buffer + rbrac
end

#format_ebnf_string(string) ⇒ Object (protected)

Escape a string, using as many UTF-8 characters as possible



388
389
390
391
392
393
394
395
396
397
398
399
400
 
# File 'lib/ebnf/writer.rb', line 388

def format_ebnf_string(string)
  quote = string.as_dquote? ? '"' : "'"
  string.each_char do |c|
    case c.ord
    when 0x00..0x19, quote.ord
      raise RangeError, "cannot format #{string.inspect} as an EBNF String: #{c.inspect} is out of range" unless
        ISOEBNF::TERMINAL_CHARACTER.match?(c)
    end
  end

  res = @options[:html] ? %(<code class="grammar-literal">#{@coder.encode(string)}</code>) : string
  res = "#{quote}#{res}#{quote}"
end

#format_isoebnf(expr, sep: nil, embedded: false) ⇒ Object (protected)

Format the expression part of a rule



616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
 
# File 'lib/ebnf/writer.rb', line 616

def format_isoebnf(expr, sep: nil, embedded: false)
  return (@options[:html] ? %(<a href="#grammar-production-#{@coder.encode expr}">#{@coder.encode expr}</a>) : expr.to_s) if expr.is_a?(Symbol)
  if expr.is_a?(String)
    expr = expr[2..-1].hex.chr if expr =~ /\A#x\h+/
    expr.chars.each do |c|
      raise RangeError, "cannot format #{expr.inspect} as an ISO EBNF String: #{c.inspect} is out of range" unless
        ISOEBNF::TERMINAL_CHARACTER.match?(c)
    end
    if expr =~ /"/
      return (@options[:html] ? %('<code class="grammar-literal">#{@coder.encode expr}</code>') : %('#{expr}'))
    else
      return (@options[:html] ? %("<code class="grammar-literal">#{@coder.encode expr}</code>") : %("#{expr}"))
    end
  end
  parts = {
    alt:    (@options[:html] ? %(<code class="grammar-alt">|</code> ) : "| "),
    diff:   (@options[:html] ? %(<code class="grammar-diff">-</code> ) : "- "),
  }
  lparen = (@options[:html] ? %[<code class="grammar-paren">(</code>] : "(")
  rparen = (@options[:html] ? %[<code class="grammar-paren">)</code>] : ")")

  case expr.first
  when :istr
    # Looses fidelity, but, oh well ...
    format_isoebnf(expr.last, embedded: true)
  when :alt, :diff
    this_sep = (sep ? sep : " ") + parts[expr.first.to_sym]
    res = expr[1..-1].map {|e| format_isoebnf(e, embedded: true)}.join(this_sep)
    embedded ? (lparen + res + rparen) : res
  when :opt
    r = format_isoebnf(expr[1], embedded: true)
    "[#{r}]"
  when :star
    r = format_isoebnf(expr[1], embedded: true)
    "{#{r}}"
  when :plus
    r = format_isoebnf(expr[1], embedded: true)
    "#{r}, {#{r}}"
  when :hex
    format_isoebnf(expr[1], embedded: true)
  when :range
    res = format_isoebnf_range(expr.last)
    res.is_a?(Array) ?
      format_isoebnf(res, embedded: true) :
      res
  when :seq
    this_sep = "," + (sep ? sep : " ")
    res = expr[1..-1].map do |e|
      format_isoebnf(e, embedded: true)
    end.join(this_sep)
    embedded ? (lparen + res + rparen) : res
  when :rept
    # Expand repetition
    min, max, value = expr[1..-1]
    if min == 0 && max == 1
      format_isoebnf([:opt, value], sep: sep, embedded: embedded)
    elsif min == 0 && max == '*'
      format_isoebnf([:star, value], sep: sep, embedded: embedded)
    elsif min == 1 && max == '*'
      format_isoebnf([:plus, value], sep: sep, embedded: embedded)
    else
      val2 = [:seq]
      while min > 0
        val2 << value
        min -= 1
        max -= 1 unless max == '*'
      end
      if max == '*'
        val2 << [:star, value]
      else
        opt = nil
        while max > 0
          opt = [:opt, opt ? [:seq, value, opt] : value]
          max -= 1
        end
        val2 << opt if opt
      end
      format_isoebnf(val2, sep: sep, embedded: embedded)
    end
  else
    raise "Unknown operator: #{expr.first}"
  end
end

#format_isoebnf_range(string) ⇒ Object (protected)

Format a range Range is formatted as a aliteration of characters

Raises:

  • (RangeError) 


702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
 
# File 'lib/ebnf/writer.rb', line 702

def format_isoebnf_range(string)
  chars = []
  o_dash = false

  raise RangeError, "cannot format #{string.inspect} an ABNF range" if string.start_with?('^')

  if string.end_with?('-')
    o_dash = true
    string = string[0..-2]
  end

  scanner = StringScanner.new(string)
  in_range = false
  # Build chars from different ranges/enums
  while !scanner.eos?
    char = if hex = scanner.scan(Terminals::HEX)
      hex[2..-1].hex.ord.char(Encoding::UTF_8)
    else scanner.scan(Terminals::R_CHAR)
    end
    raise RangeError, "cannot format #{string.inspect} as an ISO EBNF Aliteration: #{char.inspect} is out of range" unless
      char && ISOEBNF::TERMINAL_CHARACTER.match?(char)

    if in_range
      # calculate characters from chars.last to this char
      raise RangeError, "cannot format #{string.inspect} as an ISO EBNF Aliteration" unless chars.last < char
      chars.concat (chars.last..char).to_a[1..-1]
      in_range = false
    else
      chars << char
    end

    in_range = true if scanner.scan(/\-/)
  end

  chars << '-' if o_dash

  # Possibly only a single character (no character?)
  chars.length == 1 ? chars.last.inspect : chars.unshift(:alt)
end