gcc/ada/scn.adb

   1 ------------------------------------------------------------------------------
   2 --                                                                          --
   3 --                         GNAT COMPILER COMPONENTS                         --
   4 --                                                                          --
   5 --                                  S C N                                   --
   6 --                                                                          --
   7 --                                 B o d y                                  --
   8 --                                                                          --
   9 --          Copyright (C) 1992-2010, Free Software Foundation, Inc.         --
  10 --                                                                          --
  11 -- GNAT is free software;  you can  redistribute it  and/or modify it under --
  12 -- terms of the  GNU General Public License as published  by the Free Soft- --
  13 -- ware  Foundation;  either version 3,  or (at your option) any later ver- --
  14 -- sion.  GNAT is distributed in the hope that it will be useful, but WITH- --
  15 -- OUT ANY WARRANTY;  without even the  implied warranty of MERCHANTABILITY --
  16 -- or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License --
  17 -- for  more details.  You should have  received  a copy of the GNU General --
  18 -- Public License  distributed with GNAT; see file COPYING3.  If not, go to --
  19 -- http://www.gnu.org/licenses for a complete copy of the license.          --
  20 --                                                                          --
  21 -- GNAT was originally developed  by the GNAT team at  New York University. --
  22 -- Extensive contributions were provided by Ada Core Technologies Inc.      --
  23 --                                                                          --
  24 ------------------------------------------------------------------------------
  25
  26 with Atree;    use Atree;
  27 with Csets;    use Csets;
  28 with Hostparm; use Hostparm;
  29 with Namet;    use Namet;
  30 with Opt;      use Opt;
  31 with Output;   use Output;
  32 with Restrict; use Restrict;
  33 with Rident;   use Rident;
  34 with Scans;    use Scans;
  35 with Sinfo;    use Sinfo;
  36 with Sinput;   use Sinput;
  37 with Uintp;    use Uintp;
  38
  39 with GNAT.Byte_Order_Mark; use GNAT.Byte_Order_Mark;
  40
  41 with System.WCh_Con; use System.WCh_Con;
  42
  43 package body Scn is
  44
  45    use ASCII;
  46
  47    Used_As_Identifier : array (Token_Type) of Boolean;
  48    --  Flags set True if a given keyword is used as an identifier (used to
  49    --  make sure that we only post an error message for incorrect use of a
  50    --  keyword as an identifier once for a given keyword).
  51
  52    procedure Check_End_Of_Line;
  53    --  Called when end of line encountered. Checks that line is not too long,
  54    --  and that other style checks for the end of line are met.
  55
  56    function Determine_License return License_Type;
  57    --  Scan header of file and check that it has an appropriate GNAT-style
  58    --  header with a proper license statement. Returns GPL, Unrestricted,
  59    --  or Modified_GPL depending on header. If none of these, returns Unknown.
  60
  61    procedure Error_Long_Line;
  62    --  Signal error of excessively long line
  63
  64    -----------------------
  65    -- Check_End_Of_Line --
  66    -----------------------
  67
  68    procedure Check_End_Of_Line is
  69       Len : constant Int := Int (Scan_Ptr) - Int (Current_Line_Start);
  70    begin
  71       if Style_Check then
  72          Style.Check_Line_Terminator (Len);
  73       elsif Len > Max_Line_Length then
  74          Error_Long_Line;
  75       end if;
  76    end Check_End_Of_Line;
  77
  78    -----------------------
  79    -- Determine_License --
  80    -----------------------
  81
  82    function Determine_License return License_Type is
  83       GPL_Found : Boolean := False;
  84       Result    : License_Type;
  85
  86       function Contains (S : String) return Boolean;
  87       --  See if current comment contains successive non-blank characters
  88       --  matching the contents of S. If so leave Scan_Ptr unchanged and
  89       --  return True, otherwise leave Scan_Ptr unchanged and return False.
  90
  91       procedure Skip_EOL;
  92       --  Skip to line terminator character
  93
  94       --------------
  95       -- Contains --
  96       --------------
  97
  98       function Contains (S : String) return Boolean is
  99          CP : Natural;
 100          SP : Source_Ptr;
 101          SS : Source_Ptr;
 102
 103       begin
 104          --  Loop to check characters. This loop is terminated by end of
 105          --  line, and also we need to check for the EOF case, to take
 106          --  care of files containing only comments.
 107
 108          SP := Scan_Ptr;
 109          while Source (SP) /= CR and then
 110                Source (SP) /= LF and then
 111                Source (SP) /= EOF
 112          loop
 113             if Source (SP) = S (S'First) then
 114                SS := SP;
 115                CP := S'First;
 116
 117                loop
 118                   SS := SS + 1;
 119                   CP := CP + 1;
 120
 121                   if CP > S'Last then
 122                      return True;
 123                   end if;
 124
 125                   while Source (SS) = ' ' loop
 126                      SS := SS + 1;
 127                   end loop;
 128
 129                   exit when Source (SS) /= S (CP);
 130                end loop;
 131             end if;
 132
 133             SP := SP + 1;
 134          end loop;
 135
 136          return False;
 137       end Contains;
 138
 139       --------------
 140       -- Skip_EOL --
 141       --------------
 142
 143       procedure Skip_EOL is
 144       begin
 145          while Source (Scan_Ptr) /= CR
 146            and then Source (Scan_Ptr) /= LF
 147            and then Source (Scan_Ptr) /= EOF
 148          loop
 149             Scan_Ptr := Scan_Ptr + 1;
 150          end loop;
 151       end Skip_EOL;
 152
 153    --  Start of processing for Determine_License
 154
 155    begin
 156       loop
 157          if Source (Scan_Ptr) /= '-'
 158            or else Source (Scan_Ptr + 1) /= '-'
 159          then
 160             if GPL_Found then
 161                Result := GPL;
 162                exit;
 163             else
 164                Result := Unknown;
 165                exit;
 166             end if;
 167
 168          elsif Contains ("Asaspecialexception") then
 169             if GPL_Found then
 170                Result := Modified_GPL;
 171                exit;
 172             end if;
 173
 174          elsif Contains ("GNUGeneralPublicLicense") then
 175             GPL_Found := True;
 176
 177          elsif
 178              Contains
 179                ("ThisspecificationisadaptedfromtheAdaSemanticInterface")
 180            or else
 181              Contains
 182               ("ThisspecificationisderivedfromtheAdaReferenceManual")
 183          then
 184             Result := Unrestricted;
 185             exit;
 186          end if;
 187
 188          Skip_EOL;
 189
 190          Check_End_Of_Line;
 191
 192          if Source (Scan_Ptr) /= EOF then
 193
 194             --  We have to take into account a degenerate case when the source
 195             --  file contains only comments and no Ada code.
 196
 197             declare
 198                Physical : Boolean;
 199
 200             begin
 201                Skip_Line_Terminators (Scan_Ptr, Physical);
 202
 203                --  If we are at start of physical line, update scan pointers
 204                --  to reflect the start of the new line.
 205
 206                if Physical then
 207                   Current_Line_Start       := Scan_Ptr;
 208                   Start_Column             := Scanner.Set_Start_Column;
 209                   First_Non_Blank_Location := Scan_Ptr;
 210                end if;
 211             end;
 212          end if;
 213       end loop;
 214
 215       return Result;
 216    end Determine_License;
 217
 218    ----------------------------
 219    -- Determine_Token_Casing --
 220    ----------------------------
 221
 222    function Determine_Token_Casing return Casing_Type is
 223    begin
 224       return Scanner.Determine_Token_Casing;
 225    end Determine_Token_Casing;
 226
 227    ---------------------
 228    -- Error_Long_Line --
 229    ---------------------
 230
 231    procedure Error_Long_Line is
 232    begin
 233       Error_Msg
 234         ("this line is too long",
 235          Current_Line_Start + Source_Ptr (Max_Line_Length));
 236    end Error_Long_Line;
 237
 238    ------------------------
 239    -- Initialize_Scanner --
 240    ------------------------
 241
 242    procedure Initialize_Scanner
 243      (Unit  : Unit_Number_Type;
 244       Index : Source_File_Index)
 245    is
 246       GNAT_Hedr : constant Text_Buffer (1 .. 78) := (others => '-');
 247
 248    begin
 249       Scanner.Initialize_Scanner (Index);
 250
 251       if Index /= Internal_Source_File then
 252          Set_Unit (Index, Unit);
 253       end if;
 254
 255       Current_Source_Unit := Unit;
 256
 257       --  Set default for Comes_From_Source (except if we are going to process
 258       --  an artificial string internally created within the compiler and
 259       --  placed into internal source duffer). All nodes built now until we
 260       --  reenter the analyzer will have Comes_From_Source set to True
 261
 262       if Index /= Internal_Source_File then
 263          Set_Comes_From_Source_Default (True);
 264       end if;
 265
 266       --  Check license if GNAT type header possibly present
 267
 268       if Source_Last (Index) - Scan_Ptr > 80
 269         and then Source (Scan_Ptr .. Scan_Ptr + 77) = GNAT_Hedr
 270       then
 271          Set_License (Current_Source_File, Determine_License);
 272       end if;
 273
 274       --  Check for BOM
 275
 276       declare
 277          BOM : BOM_Kind;
 278          Len : Natural;
 279          Tst : String (1 .. 5);
 280
 281       begin
 282          for J in 1 .. 5 loop
 283             Tst (J) := Source (Scan_Ptr + Source_Ptr (J) - 1);
 284          end loop;
 285
 286          Read_BOM (Tst, Len, BOM, False);
 287
 288          case BOM is
 289             when UTF8_All =>
 290                Scan_Ptr := Scan_Ptr + Source_Ptr (Len);
 291                Wide_Character_Encoding_Method := WCEM_UTF8;
 292                Upper_Half_Encoding := True;
 293
 294             when UTF16_LE | UTF16_BE =>
 295                Set_Standard_Error;
 296                Write_Line ("UTF-16 encoding format not recognized");
 297                Set_Standard_Output;
 298                raise Unrecoverable_Error;
 299
 300             when UTF32_LE | UTF32_BE =>
 301                Set_Standard_Error;
 302                Write_Line ("UTF-32 encoding format not recognized");
 303                Set_Standard_Output;
 304                raise Unrecoverable_Error;
 305
 306             when Unknown =>
 307                null;
 308
 309             when others =>
 310                raise Program_Error;
 311          end case;
 312       end;
 313
 314       --  Because of the License stuff above, Scng.Initialize_Scanner cannot
 315       --  call Scan. Scan initial token (note this initializes Prev_Token,
 316       --  Prev_Token_Ptr).
 317
 318       --  There are two reasons not to do the Scan step in case if we
 319       --  initialize the scanner for the internal source buffer:
 320
 321       --  - The artificial string may not be created by the compiler in this
 322       --    buffer when we call Initialize_Scanner
 323
 324       --  - For these artificial strings a special way of scanning is used, so
 325       --    the standard step of the scanner may just break the algorithm of
 326       --    processing these strings.
 327
 328       if Index /= Internal_Source_File then
 329          Scan;
 330       end if;
 331
 332       --  Clear flags for reserved words used as identifiers
 333
 334       for J in Token_Type loop
 335          Used_As_Identifier (J) := False;
 336       end loop;
 337    end Initialize_Scanner;
 338
 339    ---------------
 340    -- Post_Scan --
 341    ---------------
 342
 343    procedure Post_Scan is
 344       procedure Check_Obsolescent_Features_Restriction (S : Source_Ptr);
 345       --  This checks for Obsolescent_Features restriction being active, and
 346       --  if so, flags the restriction as occurring at the given scan location.
 347
 348       procedure Check_Obsolete_Base_Char;
 349       --  Check for numeric literal using ':' instead of '#' for based case
 350
 351       --------------------------------------------
 352       -- Check_Obsolescent_Features_Restriction --
 353       --------------------------------------------
 354
 355       procedure Check_Obsolescent_Features_Restriction (S : Source_Ptr) is
 356       begin
 357          --  Normally we have a node handy for posting restrictions. We don't
 358          --  have such a node here, so construct a dummy one with the right
 359          --  scan pointer. This is only used to get the Sloc value anyway.
 360
 361          Check_Restriction (No_Obsolescent_Features, New_Node (N_Empty, S));
 362       end Check_Obsolescent_Features_Restriction;
 363
 364       ------------------------------
 365       -- Check_Obsolete_Base_Char --
 366       ------------------------------
 367
 368       procedure Check_Obsolete_Base_Char is
 369          S : Source_Ptr;
 370
 371       begin
 372          if Based_Literal_Uses_Colon then
 373
 374             --  Find the : for the restriction or warning message
 375
 376             S := Token_Ptr;
 377             while Source (S) /= ':' loop
 378                S := S + 1;
 379             end loop;
 380
 381             Check_Obsolescent_Features_Restriction (S);
 382
 383             if Warn_On_Obsolescent_Feature then
 384                Error_Msg
 385                  ("use of "":"" is an obsolescent feature (RM J.2(3))?", S);
 386                Error_Msg
 387                  ("\use ""'#"" instead?", S);
 388             end if;
 389          end if;
 390       end Check_Obsolete_Base_Char;
 391
 392    --  Start of processing for Post_Scan
 393
 394    begin
 395       case Token is
 396          when Tok_Char_Literal =>
 397             Token_Node := New_Node (N_Character_Literal, Token_Ptr);
 398             Set_Char_Literal_Value (Token_Node, UI_From_CC (Character_Code));
 399             Set_Chars (Token_Node, Token_Name);
 400
 401          when Tok_Identifier =>
 402             Token_Node := New_Node (N_Identifier, Token_Ptr);
 403             Set_Chars (Token_Node, Token_Name);
 404
 405          when Tok_Real_Literal =>
 406             Token_Node := New_Node (N_Real_Literal, Token_Ptr);
 407             Set_Realval (Token_Node, Real_Literal_Value);
 408             Check_Obsolete_Base_Char;
 409
 410          when Tok_Integer_Literal =>
 411             Token_Node := New_Node (N_Integer_Literal, Token_Ptr);
 412             Set_Intval (Token_Node, Int_Literal_Value);
 413             Check_Obsolete_Base_Char;
 414
 415          when Tok_String_Literal =>
 416             Token_Node := New_Node (N_String_Literal, Token_Ptr);
 417             Set_Has_Wide_Character
 418               (Token_Node, Wide_Character_Found);
 419             Set_Has_Wide_Wide_Character
 420               (Token_Node, Wide_Wide_Character_Found);
 421             Set_Strval (Token_Node, String_Literal_Id);
 422
 423             if Source (Token_Ptr) = '%' then
 424                Check_Obsolescent_Features_Restriction (Token_Ptr);
 425
 426                if Warn_On_Obsolescent_Feature then
 427                   Error_Msg_SC
 428                     ("use of ""'%"" is an obsolescent feature (RM J.2(4))?");
 429                   Error_Msg_SC ("\use """""" instead?");
 430                end if;
 431             end if;
 432
 433          when Tok_Operator_Symbol =>
 434             Token_Node := New_Node (N_Operator_Symbol, Token_Ptr);
 435             Set_Chars (Token_Node, Token_Name);
 436             Set_Strval (Token_Node, String_Literal_Id);
 437
 438          when Tok_Vertical_Bar =>
 439             if Source (Token_Ptr) = '!' then
 440                Check_Obsolescent_Features_Restriction (Token_Ptr);
 441
 442                if Warn_On_Obsolescent_Feature then
 443                   Error_Msg_SC
 444                     ("use of ""'!"" is an obsolescent feature (RM J.2(2))?");
 445                   Error_Msg_SC ("\use ""'|"" instead?");
 446                end if;
 447             end if;
 448
 449          when others =>
 450             null;
 451       end case;
 452    end Post_Scan;
 453
 454    ------------------------------
 455    -- Scan_Reserved_Identifier --
 456    ------------------------------
 457
 458    procedure Scan_Reserved_Identifier (Force_Msg : Boolean) is
 459       Token_Chars : constant String := Token_Type'Image (Token);
 460
 461    begin
 462       --  We have in Token_Chars the image of the Token name, i.e. Tok_xxx.
 463       --  This code extracts the xxx and makes an identifier out of it.
 464
 465       Name_Len := 0;
 466
 467       for J in 5 .. Token_Chars'Length loop
 468          Name_Len := Name_Len + 1;
 469          Name_Buffer (Name_Len) := Fold_Lower (Token_Chars (J));
 470       end loop;
 471
 472       Token_Name := Name_Find;
 473
 474       if not Used_As_Identifier (Token) or else Force_Msg then
 475          Error_Msg_Name_1 := Token_Name;
 476          Error_Msg_SC ("reserved word* cannot be used as identifier!");
 477          Used_As_Identifier (Token) := True;
 478       end if;
 479
 480       Token := Tok_Identifier;
 481       Token_Node := New_Node (N_Identifier, Token_Ptr);
 482       Set_Chars (Token_Node, Token_Name);
 483    end Scan_Reserved_Identifier;
 484
 485 end Scn;