src/parser/xss/lexer.xss

   1 $ // This file is part of NIT ( http://www.nitlanguage.org ).
   2 $ //
   3 $ // Copyright 2008 Jean Privat <jean@pryen.org>
   4 $ // Based on algorithms developped for ( http://www.sablecc.org/ ).
   5 $ //
   6 $ // Licensed under the Apache License, Version 2.0 (the "License");
   7 $ // you may not use this file except in compliance with the License.
   8 $ // You may obtain a copy of the License at
   9 $ //
  10 $ //     http://www.apache.org/licenses/LICENSE-2.0
  11 $ //
  12 $ // Unless required by applicable law or agreed to in writing, software
  13 $ // distributed under the License is distributed on an "AS IS" BASIS,
  14 $ // WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  15 $ // See the License for the specific language governing permissions and
  16 $ // limitations under the License.
  17
  18 $ template make_lexer()
  19
  20 # The lexer extract NIT tokens from an input stream.
  21 # It is better user with the Parser
  22 class Lexer
  23         super TablesCapable
  24         # Last peeked token
  25         var _token: nullable Token
  26
  27         # Lexer current state
  28         var _state: Int = 0
  29
  30         # The source file
  31         readable var _file: SourceFile
  32
  33         # Current character in the stream
  34         var _stream_pos: Int = 0
  35
  36         # Current line number in the input stream
  37         var _line: Int = 0
  38
  39         # Current column in the input stream
  40         var _pos: Int = 0
  41
  42         # Was the last character a cariage-return?
  43         var _cr: Bool = false
  44
  45 $ foreach {lexer_data/state}
  46         # Constante state values
  47         private fun state_${translate(@name,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}: Int do return @id end
  48 $ end foreach
  49
  50         # Create a new lexer for a stream (and a name)
  51         init(file: SourceFile)
  52         do
  53                 _file = file
  54         end
  55
  56         # Give the next token (but do not consume it)
  57         fun peek: Token
  58         do
  59                 while _token == null do
  60                         _token = get_token
  61                 end
  62                 return _token.as(not null)
  63         end
  64
  65         # Give and consume the next token
  66         fun next: Token
  67         do
  68                 var result = peek
  69                 _token = null
  70                 return result
  71         end
  72
  73         # Primitive method to return a token, or return null if it is discarded
  74         # Is used to implement `peek` and `next`
  75         protected fun get_token: nullable Token
  76         do
  77                 var dfa_state = 0
  78
  79                 var sp = _stream_pos
  80                 var start_stream_pos = sp
  81                 var start_pos = _pos
  82                 var start_line = _line
  83                 var string = _file.string
  84                 var string_len = string.length
  85
  86                 var accept_state = -1
  87                 var accept_token = -1
  88                 var accept_length = -1
  89                 var accept_pos = -1
  90                 var accept_line = -1
  91
  92                 loop
  93                         if sp >= string_len then
  94                                 dfa_state = -1
  95                         else
  96                                 var c = string[sp].ascii
  97                                 sp += 1
  98
  99                                 var cr = _cr
 100                                 var line = _line
 101                                 var pos = _pos
 102                                 if c == 10 then
 103                                         if cr then
 104                                                 cr = false
 105                                                 _file.line_starts[line] = sp
 106                                         else
 107                                                 line = line + 1
 108                                                 pos = 0
 109                                                 _file.line_starts[line] = sp
 110                                         end
 111                                 else if c == 13 then
 112                                         line = line + 1
 113                                         pos = 0
 114                                         cr = true
 115                                         _file.line_starts[line] = sp
 116                                 else
 117                                         pos = pos + 1
 118                                         cr = false
 119                                 end
 120
 121                                 loop
 122                                         var old_state = dfa_state
 123                                         if dfa_state < -1 then
 124                                                 old_state = -2 - dfa_state
 125                                         end
 126
 127                                         dfa_state = -1
 128
 129                                         var low = 0
 130                                         var high = lexer_goto(old_state, 0) - 1
 131
 132                                         if high >= 0 then
 133                                                 while low <= high do
 134                                                         var middle = (low + high) / 2
 135                                                         var offset = middle * 3 + 1 # +1 because length is at 0
 136
 137                                                         if c < lexer_goto(old_state, offset) then
 138                                                                 high = middle - 1
 139                                                         else if c > lexer_goto(old_state, offset+1) then
 140                                                                 low = middle + 1
 141                                                         else
 142                                                                 dfa_state = lexer_goto(old_state, offset+2)
 143                                                                 break
 144                                                         end
 145                                                 end
 146                                         end
 147                                         if dfa_state > -2 then break
 148                                 end
 149
 150                                 _cr = cr
 151                                 _line = line
 152                                 _pos = pos
 153                         end
 154
 155                         if dfa_state >= 0 then
 156                                 var tok = lexer_accept(dfa_state)
 157                                 if tok != -1 then
 158                                         accept_state = dfa_state
 159                                         accept_token = tok
 160                                         accept_length = sp - start_stream_pos
 161                                         accept_pos = _pos
 162                                         accept_line = _line
 163                                 end
 164                         else
 165                                 if accept_state != -1 then
 166                                         var location = new Location(_file, start_line + 1, accept_line + 1, start_pos + 1, accept_pos)
 167                                         _pos = accept_pos
 168                                         _line = accept_line
 169                                         _stream_pos = start_stream_pos + accept_length
 170 $ foreach {//token}
 171                                         if accept_token == ${position()-1} then
 172 $    if {count(transition[@from!=@to])!=0}
 173                                                 var state_id = _state
 174 $        foreach transition in {transition[@from!=@to]}
 175                                                 if state_id == ${/parser/lexer_data/state[@name=$transition/@from]/@id} then
 176                                                         _state = state_${translate(@to,"ABCDEFGHIJKLMNOPQRSTUVWXYZ","abcdefghijklmnopqrstuvwxyz")}
 177                                                 end
 178 $        end
 179 $    end if
 180 $    if {@parser_index}
 181                                                 return new @ename.init_tk(location)
 182 $    else
 183                                                 return null
 184 $    end
 185                                         end
 186 $ end foreach
 187                                 else
 188                                         _stream_pos = sp
 189                                         var location = new Location(_file, start_line + 1, start_line + 1, start_pos + 1, start_pos + 1)
 190                                         if sp > start_stream_pos then
 191                                                 var text = string.substring(start_stream_pos, sp-start_stream_pos)
 192                                                 var token = new PLexerError.init_lexer_error("Syntax error: unknown token {text}.", location, text)
 193                                                 return token
 194                                         else
 195                                                 var token = new EOF.init_tk(location)
 196                                                 return token
 197                                         end
 198                                 end
 199                         end
 200                 end
 201         end
 202 end
 203
 204 $ end template
 205
 206
 207
 208 $ template make_lexer_table()
 209 $ foreach {lexer_data/goto_table/state}
 210 $     foreach {row}
 211 $         if {count(goto)!=0}
 212 static const int lexer_goto_row${position()}[] = {
 213         ${count(goto)},
 214 $             foreach {goto}
 215         @low, @high, @state[-sep ','-]
 216 $             end foreach
 217 };
 218 $         end
 219 $     end foreach
 220 static const int lexer_goto_row_null[] = {0};
 221 const int* const lexer_goto_table[] = {
 222 $     foreach {row}
 223 $         if {count(goto)!=0}
 224         lexer_goto_row${position()}[-sep ','-]
 225 $         else
 226         lexer_goto_row_null[-sep ','-]
 227 $         end
 228 $     end foreach
 229 };
 230 $ end foreach
 231
 232 $ foreach {lexer_data/accept_table/state}
 233 const int lexer_accept_table[] = {
 234         [-foreach {i}-]${.}[-sep ','-][-end foreach-]
 235 };
 236 $ end foreach
 237
 238 $ end template