1------------------------------------------------------------------------------
2--                                                                          --
3--                         GNAT COMPILER COMPONENTS                         --
4--                                                                          --
5--                                 S C N G                                  --
6--                                                                          --
7--                                 B o d y                                  --
8--                                                                          --
9--          Copyright (C) 1992-2015, 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
26with Atree;    use Atree;
27with Csets;    use Csets;
28with Hostparm; use Hostparm;
29with Namet;    use Namet;
30with Opt;      use Opt;
31with Restrict; use Restrict;
32with Rident;   use Rident;
33with Scans;    use Scans;
34with Sinput;   use Sinput;
35with Snames;   use Snames;
36with Stringt;  use Stringt;
37with Stylesw;  use Stylesw;
38with Uintp;    use Uintp;
39with Urealp;   use Urealp;
40with Widechar; use Widechar;
41
42pragma Warnings (Off);
43--  This package is used also by gnatcoll
44with System.CRC32;
45with System.UTF_32;  use System.UTF_32;
46with System.WCh_Con; use System.WCh_Con;
47pragma Warnings (On);
48
49package body Scng is
50
51   use ASCII;
52   --  Make control characters visible
53
54   Special_Characters : array (Character) of Boolean := (others => False);
55   --  For characters that are Special token, the value is True
56
57   Comment_Is_Token : Boolean := False;
58   --  True if comments are tokens
59
60   End_Of_Line_Is_Token : Boolean := False;
61   --  True if End_Of_Line is a token
62
63   -----------------------
64   -- Local Subprograms --
65   -----------------------
66
67   procedure Accumulate_Token_Checksum;
68   pragma Inline (Accumulate_Token_Checksum);
69   --  Called after each numeric literal and identifier/keyword. For keywords,
70   --  the token used is Tok_Identifier. This allows detection of additional
71   --  spaces added in sources when using the builder switch -m.
72
73   procedure Accumulate_Token_Checksum_GNAT_6_3;
74   --  Used in place of Accumulate_Token_Checksum for GNAT versions 5.04 to
75   --  6.3, when Tok_Some was not included in Token_Type and the actual
76   --  Token_Type was used for keywords. This procedure is never used in the
77   --  compiler or gnatmake, only in gprbuild.
78
79   procedure Accumulate_Token_Checksum_GNAT_5_03;
80   --  Used in place of Accumulate_Token_Checksum for GNAT version 5.03, when
81   --  Tok_Interface, Tok_Some, Tok_Synchronized and Tok_Overriding were not
82   --  included in Token_Type and the actual Token_Type was used for keywords.
83   --  This procedure is never used in the compiler or gnatmake, only in
84   --  gprbuild.
85
86   procedure Accumulate_Checksum (C : Character);
87   pragma Inline (Accumulate_Checksum);
88   --  This routine accumulates the checksum given character C. During the
89   --  scanning of a source file, this routine is called with every character
90   --  in the source, excluding blanks, and all control characters (except
91   --  that ESC is included in the checksum). Upper case letters not in string
92   --  literals are folded by the caller. See Sinput spec for the documentation
93   --  of the checksum algorithm. Note: checksum values are only used if we
94   --  generate code, so it is not necessary to worry about making the right
95   --  sequence of calls in any error situation.
96
97   procedure Accumulate_Checksum (C : Char_Code);
98   pragma Inline (Accumulate_Checksum);
99   --  This version is identical, except that the argument, C, is a character
100   --  code value instead of a character. This is used when wide characters
101   --  are scanned. We use the character code rather than the ASCII characters
102   --  so that the checksum is independent of wide character encoding method.
103
104   procedure Initialize_Checksum;
105   pragma Inline (Initialize_Checksum);
106   --  Initialize checksum value
107
108   -------------------------
109   -- Accumulate_Checksum --
110   -------------------------
111
112   procedure Accumulate_Checksum (C : Character) is
113   begin
114      System.CRC32.Update (System.CRC32.CRC32 (Checksum), C);
115   end Accumulate_Checksum;
116
117   procedure Accumulate_Checksum (C : Char_Code) is
118   begin
119      if C > 16#FFFF# then
120         Accumulate_Checksum (Character'Val (C / 2 ** 24));
121         Accumulate_Checksum (Character'Val ((C / 2 ** 16) mod 256));
122         Accumulate_Checksum (Character'Val ((C / 256) mod 256));
123      else
124         Accumulate_Checksum (Character'Val (C / 256));
125      end if;
126
127      Accumulate_Checksum (Character'Val (C mod 256));
128   end Accumulate_Checksum;
129
130   -------------------------------
131   -- Accumulate_Token_Checksum --
132   -------------------------------
133
134   procedure Accumulate_Token_Checksum is
135   begin
136      System.CRC32.Update
137        (System.CRC32.CRC32 (Checksum),
138         Character'Val (Token_Type'Pos (Token)));
139   end Accumulate_Token_Checksum;
140
141   ----------------------------------------
142   -- Accumulate_Token_Checksum_GNAT_6_3 --
143   ----------------------------------------
144
145   procedure Accumulate_Token_Checksum_GNAT_6_3 is
146   begin
147      --  Individual values of Token_Type are used, instead of subranges, so
148      --  that additions or suppressions of enumerated values in type
149      --  Token_Type are detected by the compiler.
150
151      case Token is
152         when Tok_Integer_Literal | Tok_Real_Literal | Tok_String_Literal |
153              Tok_Char_Literal | Tok_Operator_Symbol | Tok_Identifier |
154              Tok_Double_Asterisk | Tok_Ampersand | Tok_Minus | Tok_Plus |
155              Tok_Asterisk | Tok_Mod | Tok_Rem | Tok_Slash | Tok_New |
156              Tok_Abs | Tok_Others | Tok_Null | Tok_Dot | Tok_Apostrophe |
157              Tok_Left_Paren | Tok_Delta | Tok_Digits | Tok_Range |
158              Tok_Right_Paren | Tok_Comma | Tok_And | Tok_Or | Tok_Xor |
159              Tok_Less | Tok_Equal | Tok_Greater | Tok_Not_Equal |
160              Tok_Greater_Equal | Tok_Less_Equal | Tok_In | Tok_Not |
161              Tok_Box | Tok_Colon_Equal | Tok_Colon | Tok_Greater_Greater |
162              Tok_Abstract | Tok_Access | Tok_Aliased | Tok_All | Tok_Array |
163              Tok_At | Tok_Body | Tok_Constant | Tok_Do | Tok_Is |
164              Tok_Interface | Tok_Limited | Tok_Of | Tok_Out | Tok_Record |
165              Tok_Renames | Tok_Reverse =>
166
167            System.CRC32.Update
168              (System.CRC32.CRC32 (Checksum),
169               Character'Val (Token_Type'Pos (Token)));
170
171         when Tok_Some =>
172
173            System.CRC32.Update
174              (System.CRC32.CRC32 (Checksum),
175               Character'Val (Token_Type'Pos (Tok_Identifier)));
176
177         when Tok_Tagged | Tok_Then | Tok_Less_Less | Tok_Abort | Tok_Accept |
178              Tok_Case | Tok_Delay | Tok_Else | Tok_Elsif | Tok_End |
179              Tok_Exception | Tok_Exit | Tok_Goto | Tok_If | Tok_Pragma |
180              Tok_Raise | Tok_Requeue | Tok_Return | Tok_Select |
181              Tok_Terminate | Tok_Until | Tok_When | Tok_Begin | Tok_Declare |
182              Tok_For | Tok_Loop | Tok_While | Tok_Entry | Tok_Protected |
183              Tok_Task | Tok_Type | Tok_Subtype | Tok_Overriding |
184              Tok_Synchronized | Tok_Use | Tok_Function | Tok_Generic |
185              Tok_Package | Tok_Procedure | Tok_Private | Tok_With |
186              Tok_Separate | Tok_EOF | Tok_Semicolon | Tok_Arrow |
187              Tok_Vertical_Bar | Tok_Dot_Dot | Tok_Project | Tok_Extends |
188              Tok_External | Tok_External_As_List | Tok_Comment |
189              Tok_End_Of_Line | Tok_Special | Tok_SPARK_Hide | No_Token =>
190
191            System.CRC32.Update
192              (System.CRC32.CRC32 (Checksum),
193               Character'Val (Token_Type'Pos (Token_Type'Pred (Token))));
194      end case;
195   end Accumulate_Token_Checksum_GNAT_6_3;
196
197   -----------------------------------------
198   -- Accumulate_Token_Checksum_GNAT_5_03 --
199   -----------------------------------------
200
201   procedure Accumulate_Token_Checksum_GNAT_5_03 is
202   begin
203      --  Individual values of Token_Type are used, instead of subranges, so
204      --  that additions or suppressions of enumerated values in type
205      --  Token_Type are detected by the compiler.
206
207      case Token is
208         when Tok_Integer_Literal | Tok_Real_Literal | Tok_String_Literal |
209              Tok_Char_Literal | Tok_Operator_Symbol | Tok_Identifier |
210              Tok_Double_Asterisk | Tok_Ampersand | Tok_Minus | Tok_Plus |
211              Tok_Asterisk | Tok_Mod | Tok_Rem | Tok_Slash | Tok_New |
212              Tok_Abs | Tok_Others | Tok_Null | Tok_Dot | Tok_Apostrophe |
213              Tok_Left_Paren | Tok_Delta | Tok_Digits | Tok_Range |
214              Tok_Right_Paren | Tok_Comma | Tok_And | Tok_Or | Tok_Xor |
215              Tok_Less | Tok_Equal | Tok_Greater | Tok_Not_Equal |
216              Tok_Greater_Equal | Tok_Less_Equal | Tok_In | Tok_Not |
217              Tok_Box | Tok_Colon_Equal | Tok_Colon | Tok_Greater_Greater |
218              Tok_Abstract | Tok_Access | Tok_Aliased | Tok_All | Tok_Array |
219              Tok_At | Tok_Body | Tok_Constant | Tok_Do | Tok_Is =>
220
221            System.CRC32.Update
222              (System.CRC32.CRC32 (Checksum),
223               Character'Val (Token_Type'Pos (Token)));
224
225         when Tok_Interface | Tok_Some | Tok_Overriding | Tok_Synchronized =>
226            System.CRC32.Update
227              (System.CRC32.CRC32 (Checksum),
228               Character'Val (Token_Type'Pos (Tok_Identifier)));
229
230         when Tok_Limited | Tok_Of | Tok_Out | Tok_Record |
231              Tok_Renames | Tok_Reverse =>
232
233            System.CRC32.Update
234              (System.CRC32.CRC32 (Checksum),
235               Character'Val (Token_Type'Pos (Token) - 1));
236
237         when Tok_Tagged | Tok_Then | Tok_Less_Less | Tok_Abort | Tok_Accept |
238              Tok_Case | Tok_Delay | Tok_Else | Tok_Elsif | Tok_End |
239              Tok_Exception | Tok_Exit | Tok_Goto | Tok_If | Tok_Pragma |
240              Tok_Raise | Tok_Requeue | Tok_Return | Tok_Select |
241              Tok_Terminate | Tok_Until | Tok_When | Tok_Begin | Tok_Declare |
242              Tok_For | Tok_Loop | Tok_While | Tok_Entry | Tok_Protected |
243              Tok_Task | Tok_Type | Tok_Subtype =>
244
245            System.CRC32.Update
246              (System.CRC32.CRC32 (Checksum),
247               Character'Val (Token_Type'Pos (Token) - 2));
248
249         when Tok_Use | Tok_Function | Tok_Generic |
250              Tok_Package | Tok_Procedure | Tok_Private | Tok_With |
251              Tok_Separate | Tok_EOF | Tok_Semicolon | Tok_Arrow |
252              Tok_Vertical_Bar | Tok_Dot_Dot | Tok_Project | Tok_Extends |
253              Tok_External | Tok_External_As_List | Tok_Comment |
254              Tok_End_Of_Line | Tok_Special | Tok_SPARK_Hide | No_Token =>
255
256            System.CRC32.Update
257              (System.CRC32.CRC32 (Checksum),
258               Character'Val (Token_Type'Pos (Token) - 4));
259      end case;
260   end Accumulate_Token_Checksum_GNAT_5_03;
261
262   -----------------------
263   -- Check_End_Of_Line --
264   -----------------------
265
266   procedure Check_End_Of_Line is
267      Len : constant Int :=
268              Int (Scan_Ptr) -
269                Int (Current_Line_Start) -
270                  Wide_Char_Byte_Count;
271
272   --  Start of processing for Check_End_Of_Line
273
274   begin
275      if Style_Check then
276         Style.Check_Line_Terminator (Len);
277      end if;
278
279      --  Deal with checking maximum line length
280
281      if Style_Check and Style_Check_Max_Line_Length then
282         Style.Check_Line_Max_Length (Len);
283
284         --  If style checking is inactive, check maximum line length against
285         --  standard value.
286
287      elsif Len > Max_Line_Length then
288         Error_Msg
289           ("this line is too long",
290            Current_Line_Start + Source_Ptr (Max_Line_Length));
291      end if;
292
293      --  Now one more checking circuit. Normally we are only enforcing a limit
294      --  of physical characters, with tabs counting as one character. But if
295      --  after tab expansion we would have a total line length that exceeded
296      --  32766, that would really cause trouble, because column positions
297      --  would exceed the maximum we allow for a column count. Note: the limit
298      --  is 32766 rather than 32767, since we use a value of 32767 for special
299      --  purposes (see Sinput). Now we really do not want to go messing with
300      --  tabs in the normal case, so what we do is to check for a line that
301      --  has more than 4096 physical characters. Any shorter line could not
302      --  be a problem, even if it was all tabs.
303
304      if Len >= 4096 then
305         declare
306            Col : Natural;
307            Ptr : Source_Ptr;
308
309         begin
310            Col := 1;
311            Ptr := Current_Line_Start;
312            loop
313               exit when Ptr = Scan_Ptr;
314
315               if Source (Ptr) = ASCII.HT then
316                  Col := (Col - 1 + 8) / 8 * 8 + 1;
317               else
318                  Col := Col + 1;
319               end if;
320
321               if Col > 32766 then
322                  Error_Msg
323                    ("this line is longer than 32766 characters",
324                     Current_Line_Start);
325                  raise Unrecoverable_Error;
326               end if;
327
328               Ptr := Ptr + 1;
329            end loop;
330         end;
331      end if;
332
333      --  Reset wide character byte count for next line
334
335      Wide_Char_Byte_Count := 0;
336   end Check_End_Of_Line;
337
338   ----------------------------
339   -- Determine_Token_Casing --
340   ----------------------------
341
342   function Determine_Token_Casing return Casing_Type is
343   begin
344      return Determine_Casing (Source (Token_Ptr .. Scan_Ptr - 1));
345   end Determine_Token_Casing;
346
347   -------------------------
348   -- Initialize_Checksum --
349   -------------------------
350
351   procedure Initialize_Checksum is
352   begin
353      System.CRC32.Initialize (System.CRC32.CRC32 (Checksum));
354   end Initialize_Checksum;
355
356   ------------------------
357   -- Initialize_Scanner --
358   ------------------------
359
360   procedure Initialize_Scanner (Index : Source_File_Index) is
361   begin
362      --  Establish reserved words
363
364      Scans.Initialize_Ada_Keywords;
365
366      --  Initialize scan control variables
367
368      Current_Source_File       := Index;
369      Source                    := Source_Text (Current_Source_File);
370      Scan_Ptr                  := Source_First (Current_Source_File);
371      Token                     := No_Token;
372      Token_Ptr                 := Scan_Ptr;
373      Current_Line_Start        := Scan_Ptr;
374      Token_Node                := Empty;
375      Token_Name                := No_Name;
376      Start_Column              := Set_Start_Column;
377      First_Non_Blank_Location  := Scan_Ptr;
378
379      Initialize_Checksum;
380      Wide_Char_Byte_Count := 0;
381
382      --  Do not call Scan, otherwise the License stuff does not work in Scn
383
384   end Initialize_Scanner;
385
386   ------------------------------
387   -- Reset_Special_Characters --
388   ------------------------------
389
390   procedure Reset_Special_Characters is
391   begin
392      Special_Characters := (others => False);
393   end Reset_Special_Characters;
394
395   ----------
396   -- Scan --
397   ----------
398
399   procedure Scan is
400
401      Start_Of_Comment : Source_Ptr;
402      --  Record start of comment position
403
404      Underline_Found : Boolean;
405      --  During scanning of an identifier, set to True if last character
406      --  scanned was an underline or other punctuation character. This
407      --  is used to flag the error of two underlines/punctuations in a
408      --  row or ending an identifier with a underline/punctuation. Here
409      --  punctuation means any UTF_32 character in the Unicode category
410      --  Punctuation,Connector.
411
412      Wptr : Source_Ptr;
413      --  Used to remember start of last wide character scanned
414
415      function Double_Char_Token (C : Character) return Boolean;
416      --  This function is used for double character tokens like := or <>. It
417      --  checks if the character following Source (Scan_Ptr) is C, and if so
418      --  bumps Scan_Ptr past the pair of characters and returns True. A space
419      --  between the two characters is also recognized with an appropriate
420      --  error message being issued. If C is not present, False is returned.
421      --  Note that Double_Char_Token can only be used for tokens defined in
422      --  the Ada syntax (it's use for error cases like && is not appropriate
423      --  since we do not want a junk message for a case like &-space-&).
424
425      procedure Error_Illegal_Character;
426      --  Give illegal character error, Scan_Ptr points to character. On
427      --  return, Scan_Ptr is bumped past the illegal character.
428
429      procedure Error_Illegal_Wide_Character;
430      --  Give illegal wide character message. On return, Scan_Ptr is bumped
431      --  past the illegal character, which may still leave us pointing to
432      --  junk, not much we can do if the escape sequence is messed up.
433
434      procedure Error_No_Double_Underline;
435      --  Signal error of two underline or punctuation characters in a row.
436      --  Called with Scan_Ptr pointing to second underline/punctuation char.
437
438      procedure Nlit;
439      --  This is the procedure for scanning out numeric literals. On entry,
440      --  Scan_Ptr points to the digit that starts the numeric literal (the
441      --  checksum for this character has not been accumulated yet). On return
442      --  Scan_Ptr points past the last character of the numeric literal, Token
443      --  and Token_Node are set appropriately, and the checksum is updated.
444
445      procedure Slit;
446      --  This is the procedure for scanning out string literals. On entry,
447      --  Scan_Ptr points to the opening string quote (the checksum for this
448      --  character has not been accumulated yet). On return Scan_Ptr points
449      --  past the closing quote of the string literal, Token and Token_Node
450      --  are set appropriately, and the checksum is updated.
451
452      procedure Skip_Other_Format_Characters;
453      --  Skips past any "other format" category characters at the current
454      --  cursor location (does not skip past spaces or any other characters).
455
456      function Start_Of_Wide_Character return Boolean;
457      --  Returns True if the scan pointer is pointing to the start of a wide
458      --  character sequence, does not modify the scan pointer in any case.
459
460      -----------------------
461      -- Double_Char_Token --
462      -----------------------
463
464      function Double_Char_Token (C : Character) return Boolean is
465      begin
466         if Source (Scan_Ptr + 1) = C then
467            Accumulate_Checksum (C);
468            Scan_Ptr := Scan_Ptr + 2;
469            return True;
470
471         elsif Source (Scan_Ptr + 1) = ' '
472           and then Source (Scan_Ptr + 2) = C
473         then
474            Scan_Ptr := Scan_Ptr + 1;
475            Error_Msg_S -- CODEFIX
476              ("no space allowed here");
477            Scan_Ptr := Scan_Ptr + 2;
478            return True;
479
480         else
481            return False;
482         end if;
483      end Double_Char_Token;
484
485      -----------------------------
486      -- Error_Illegal_Character --
487      -----------------------------
488
489      procedure Error_Illegal_Character is
490      begin
491         Error_Msg_S ("illegal character");
492         Scan_Ptr := Scan_Ptr + 1;
493      end Error_Illegal_Character;
494
495      ----------------------------------
496      -- Error_Illegal_Wide_Character --
497      ----------------------------------
498
499      procedure Error_Illegal_Wide_Character is
500      begin
501         Scan_Ptr := Scan_Ptr + 1;
502         Error_Msg ("illegal wide character", Wptr);
503      end Error_Illegal_Wide_Character;
504
505      -------------------------------
506      -- Error_No_Double_Underline --
507      -------------------------------
508
509      procedure Error_No_Double_Underline is
510      begin
511         Underline_Found := False;
512
513         --  There are four cases, and we special case the messages
514
515         if Source (Scan_Ptr) = '_' then
516            if Source (Scan_Ptr - 1) = '_' then
517               Error_Msg_S -- CODEFIX
518                 ("two consecutive underlines not permitted");
519            else
520               Error_Msg_S ("underline cannot follow punctuation character");
521            end if;
522
523         else
524            if Source (Scan_Ptr - 1) = '_' then
525               Error_Msg_S ("punctuation character cannot follow underline");
526            else
527               Error_Msg_S
528                 ("two consecutive punctuation characters not permitted");
529            end if;
530         end if;
531      end Error_No_Double_Underline;
532
533      ----------
534      -- Nlit --
535      ----------
536
537      procedure Nlit is
538
539         C : Character;
540         --  Current source program character
541
542         Base_Char : Character;
543         --  Either # or : (character at start of based number)
544
545         Base : Int;
546         --  Value of base
547
548         UI_Base : Uint;
549         --  Value of base in Uint format
550
551         UI_Int_Value : Uint;
552         --  Value of integer scanned by Scan_Integer in Uint format
553
554         UI_Num_Value : Uint;
555         --  Value of integer in numeric value being scanned
556
557         Scale : Int;
558         --  Scale value for real literal
559
560         UI_Scale : Uint;
561         --  Scale in Uint format
562
563         Exponent_Is_Negative : Boolean;
564         --  Set true for negative exponent
565
566         Extended_Digit_Value : Int;
567         --  Extended digit value
568
569         Point_Scanned : Boolean;
570         --  Flag for decimal point scanned in numeric literal
571
572         -----------------------
573         -- Local Subprograms --
574         -----------------------
575
576         procedure Error_Digit_Expected;
577         --  Signal error of bad digit, Scan_Ptr points to the location at
578         --  which the digit was expected on input, and is unchanged on return.
579
580         procedure Scan_Integer;
581         --  Scan integer literal. On entry, Scan_Ptr points to a digit, on
582         --  exit Scan_Ptr points past the last character of the integer.
583         --
584         --  For each digit encountered, UI_Int_Value is multiplied by 10, and
585         --  the value of the digit added to the result. In addition, the value
586         --  in Scale is decremented by one for each actual digit scanned.
587
588         --------------------------
589         -- Error_Digit_Expected --
590         --------------------------
591
592         procedure Error_Digit_Expected is
593         begin
594            Error_Msg_S ("digit expected");
595         end Error_Digit_Expected;
596
597         ------------------
598         -- Scan_Integer --
599         ------------------
600
601         procedure Scan_Integer is
602            C : Character;
603            --  Next character scanned
604
605         begin
606            C := Source (Scan_Ptr);
607
608            --  Loop through digits (allowing underlines)
609
610            loop
611               Accumulate_Checksum (C);
612               UI_Int_Value :=
613                 UI_Int_Value * 10 + (Character'Pos (C) - Character'Pos ('0'));
614               Scan_Ptr := Scan_Ptr + 1;
615               Scale := Scale - 1;
616               C := Source (Scan_Ptr);
617
618               --  Case of underline encountered
619
620               if C = '_' then
621
622                  --  We do not accumulate the '_' in the checksum, so that
623                  --  1_234 is equivalent to 1234, and does not trigger
624                  --  compilation for "minimal recompilation" (gnatmake -m).
625
626                  loop
627                     Scan_Ptr := Scan_Ptr + 1;
628                     C := Source (Scan_Ptr);
629                     exit when C /= '_';
630                     Error_No_Double_Underline;
631                  end loop;
632
633                  if C not in '0' .. '9' then
634                     Error_Digit_Expected;
635                     exit;
636                  end if;
637
638               else
639                  exit when C not in '0' .. '9';
640               end if;
641            end loop;
642         end Scan_Integer;
643
644      --  Start of processing for Nlit
645
646      begin
647         Base := 10;
648         UI_Base := Uint_10;
649         UI_Int_Value := Uint_0;
650         Based_Literal_Uses_Colon := False;
651         Scale := 0;
652         Scan_Integer;
653         Point_Scanned := False;
654         UI_Num_Value := UI_Int_Value;
655
656         --  Various possibilities now for continuing the literal are period,
657         --  E/e (for exponent), or :/# (for based literal).
658
659         Scale := 0;
660         C := Source (Scan_Ptr);
661
662         if C = '.' then
663
664            --  Scan out point, but do not scan past .. which is a range
665            --  sequence, and must not be eaten up scanning a numeric literal.
666
667            while C = '.' and then Source (Scan_Ptr + 1) /= '.' loop
668               Accumulate_Checksum ('.');
669
670               if Point_Scanned then
671                  Error_Msg_S ("duplicate point ignored");
672               end if;
673
674               Point_Scanned := True;
675               Scan_Ptr := Scan_Ptr + 1;
676               C := Source (Scan_Ptr);
677
678               if C not in '0' .. '9' then
679                  Error_Msg
680                    ("real literal cannot end with point", Scan_Ptr - 1);
681               else
682                  Scan_Integer;
683                  UI_Num_Value := UI_Int_Value;
684               end if;
685            end loop;
686
687         --  Based literal case. The base is the value we already scanned.
688         --  In the case of colon, we insist that the following character
689         --  is indeed an extended digit or a period. This catches a number
690         --  of common errors, as well as catching the well known tricky
691         --  bug otherwise arising from "x : integer range 1 .. 10:= 6;"
692
693         elsif C = '#'
694           or else (C = ':' and then
695                      (Source (Scan_Ptr + 1) = '.'
696                         or else
697                       Source (Scan_Ptr + 1) in '0' .. '9'
698                         or else
699                       Source (Scan_Ptr + 1) in 'A' .. 'Z'
700                         or else
701                       Source (Scan_Ptr + 1) in 'a' .. 'z'))
702         then
703            Accumulate_Checksum (C);
704            Base_Char := C;
705            UI_Base := UI_Int_Value;
706
707            if Base_Char = ':' then
708               Based_Literal_Uses_Colon := True;
709            end if;
710
711            if UI_Base < 2 or else UI_Base > 16 then
712               Error_Msg_SC ("base not 2-16");
713               UI_Base := Uint_16;
714            end if;
715
716            Base := UI_To_Int (UI_Base);
717            Scan_Ptr := Scan_Ptr + 1;
718
719            --  Scan out extended integer [. integer]
720
721            C := Source (Scan_Ptr);
722            UI_Int_Value := Uint_0;
723            Scale := 0;
724
725            loop
726               if C in '0' .. '9' then
727                  Accumulate_Checksum (C);
728                  Extended_Digit_Value :=
729                    Int'(Character'Pos (C)) - Int'(Character'Pos ('0'));
730
731               elsif C in 'A' .. 'F' then
732                  Accumulate_Checksum (Character'Val (Character'Pos (C) + 32));
733                  Extended_Digit_Value :=
734                    Int'(Character'Pos (C)) - Int'(Character'Pos ('A')) + 10;
735
736               elsif C in 'a' .. 'f' then
737                  Accumulate_Checksum (C);
738                  Extended_Digit_Value :=
739                    Int'(Character'Pos (C)) - Int'(Character'Pos ('a')) + 10;
740
741               else
742                  Error_Msg_S ("extended digit expected");
743                  exit;
744               end if;
745
746               if Extended_Digit_Value >= Base then
747                  Error_Msg_S ("digit '>= base");
748               end if;
749
750               UI_Int_Value := UI_Int_Value * UI_Base + Extended_Digit_Value;
751               Scale := Scale - 1;
752               Scan_Ptr := Scan_Ptr + 1;
753               C := Source (Scan_Ptr);
754
755               if C = '_' then
756                  loop
757                     Accumulate_Checksum ('_');
758                     Scan_Ptr := Scan_Ptr + 1;
759                     C := Source (Scan_Ptr);
760                     exit when C /= '_';
761                     Error_No_Double_Underline;
762                  end loop;
763
764               elsif C = '.' then
765                  Accumulate_Checksum ('.');
766
767                  if Point_Scanned then
768                     Error_Msg_S ("duplicate point ignored");
769                  end if;
770
771                  Scan_Ptr := Scan_Ptr + 1;
772                  C := Source (Scan_Ptr);
773                  Point_Scanned := True;
774                  Scale := 0;
775
776               elsif C = Base_Char then
777                  Accumulate_Checksum (C);
778                  Scan_Ptr := Scan_Ptr + 1;
779                  exit;
780
781               elsif C = '#' or else C = ':' then
782                  Error_Msg_S ("based number delimiters must match");
783                  Scan_Ptr := Scan_Ptr + 1;
784                  exit;
785
786               elsif not Identifier_Char (C) then
787                  if Base_Char = '#' then
788                     Error_Msg_S -- CODEFIX
789                       ("missing '#");
790                  else
791                     Error_Msg_S -- CODEFIX
792                       ("missing ':");
793                  end if;
794
795                  exit;
796               end if;
797
798            end loop;
799
800            UI_Num_Value := UI_Int_Value;
801         end if;
802
803         --  Scan out exponent
804
805         if not Point_Scanned then
806            Scale := 0;
807            UI_Scale := Uint_0;
808         else
809            UI_Scale := UI_From_Int (Scale);
810         end if;
811
812         if Source (Scan_Ptr) = 'e' or else Source (Scan_Ptr) = 'E' then
813            Accumulate_Checksum ('e');
814            Scan_Ptr := Scan_Ptr + 1;
815            Exponent_Is_Negative := False;
816
817            if Source (Scan_Ptr) = '+' then
818               Accumulate_Checksum ('+');
819               Scan_Ptr := Scan_Ptr + 1;
820
821            elsif Source (Scan_Ptr) = '-' then
822               Accumulate_Checksum ('-');
823
824               if not Point_Scanned then
825                  Error_Msg_S
826                    ("negative exponent not allowed for integer literal");
827               else
828                  Exponent_Is_Negative := True;
829               end if;
830
831               Scan_Ptr := Scan_Ptr + 1;
832            end if;
833
834            UI_Int_Value := Uint_0;
835
836            if Source (Scan_Ptr) in '0' .. '9' then
837               Scan_Integer;
838            else
839               Error_Digit_Expected;
840            end if;
841
842            if Exponent_Is_Negative then
843               UI_Scale := UI_Scale - UI_Int_Value;
844            else
845               UI_Scale := UI_Scale + UI_Int_Value;
846            end if;
847         end if;
848
849         --  Case of real literal to be returned
850
851         if Point_Scanned then
852            Token := Tok_Real_Literal;
853            Real_Literal_Value :=
854              UR_From_Components (
855                                  Num   => UI_Num_Value,
856                                  Den   => -UI_Scale,
857                                  Rbase => Base);
858
859         --  Case of integer literal to be returned
860
861         else
862            Token := Tok_Integer_Literal;
863
864            if UI_Scale = 0 then
865               Int_Literal_Value := UI_Num_Value;
866
867            --  Avoid doing possibly expensive calculations in cases like
868            --  parsing 163E800_000# when semantics will not be done anyway.
869            --  This is especially useful when parsing garbled input.
870
871            elsif Operating_Mode /= Check_Syntax
872              and then (Serious_Errors_Detected = 0 or else Try_Semantics)
873            then
874               Int_Literal_Value := UI_Num_Value * UI_Base ** UI_Scale;
875
876            else
877               Int_Literal_Value := No_Uint;
878            end if;
879         end if;
880
881         if Checksum_Accumulate_Token_Checksum then
882            Accumulate_Token_Checksum;
883         end if;
884
885         return;
886      end Nlit;
887
888      ----------
889      -- Slit --
890      ----------
891
892      procedure Slit is
893
894         Delimiter : Character;
895         --  Delimiter (first character of string)
896
897         C : Character;
898         --  Current source program character
899
900         Code : Char_Code;
901         --  Current character code value
902
903         Err : Boolean;
904         --  Error flag for Scan_Wide call
905
906         String_Start : Source_Ptr;
907         --  Point to first character of string
908
909         procedure Error_Bad_String_Char;
910         --  Signal bad character in string/character literal. On entry
911         --  Scan_Ptr points to the improper character encountered during the
912         --  scan. Scan_Ptr is not modified, so it still points to the bad
913         --  character on return.
914
915         procedure Error_Unterminated_String;
916         --  Procedure called if a line terminator character is encountered
917         --  during scanning a string, meaning that the string is not properly
918         --  terminated.
919
920         procedure Set_String;
921         --  Procedure used to distinguish between string and operator symbol.
922         --  On entry the string has been scanned out, and its characters start
923         --  at Token_Ptr and end one character before Scan_Ptr. On exit Token
924         --  is set to Tok_String_Literal/Tok_Operator_Symbol as appropriate,
925         --  and Token_Node is appropriately initialized. In addition, in the
926         --  operator symbol case, Token_Name is appropriately set, and the
927         --  flags [Wide_]Wide_Character_Found are set appropriately.
928
929         ---------------------------
930         -- Error_Bad_String_Char --
931         ---------------------------
932
933         procedure Error_Bad_String_Char is
934            C : constant Character := Source (Scan_Ptr);
935
936         begin
937            if C = HT then
938               Error_Msg_S ("horizontal tab not allowed in string");
939
940            elsif C = VT or else C = FF then
941               Error_Msg_S ("format effector not allowed in string");
942
943            elsif C in Upper_Half_Character then
944               Error_Msg_S ("(Ada 83) upper half character not allowed");
945
946            else
947               Error_Msg_S ("control character not allowed in string");
948            end if;
949         end Error_Bad_String_Char;
950
951         -------------------------------
952         -- Error_Unterminated_String --
953         -------------------------------
954
955         procedure Error_Unterminated_String is
956            S : Source_Ptr;
957
958         begin
959            --  An interesting little refinement. Consider the following
960            --  examples:
961
962            --     A := "this is an unterminated string;
963            --     A := "this is an unterminated string &
964            --     P(A, "this is a parameter that didn't get terminated);
965            --     P("this is a parameter that didn't get terminated, A);
966
967            --  We fiddle a little to do slightly better placement in these
968            --  cases also if there is white space at the end of the line we
969            --  place the flag at the start of this white space, not at the
970            --  end. Note that we only have to test for blanks, since tabs
971            --  aren't allowed in strings in the first place and would have
972            --  caused an error message.
973
974            --  Two more cases that we treat specially are:
975
976            --     A := "this string uses the wrong terminator'
977            --     A := "this string uses the wrong terminator' &
978
979            --  In these cases we give a different error message as well
980
981            --  We actually reposition the scan pointer to the point where we
982            --  place the flag in these cases, since it seems a better bet on
983            --  the original intention.
984
985            while Source (Scan_Ptr - 1) = ' '
986              or else Source (Scan_Ptr - 1) = '&'
987            loop
988               Scan_Ptr := Scan_Ptr - 1;
989               Unstore_String_Char;
990            end loop;
991
992            --  Check for case of incorrect string terminator, but single quote
993            --  is not considered incorrect if the opening terminator misused
994            --  a single quote (error message already given).
995
996            if Delimiter /= '''
997              and then Source (Scan_Ptr - 1) = '''
998            then
999               Unstore_String_Char;
1000               Error_Msg
1001                 ("incorrect string terminator character", Scan_Ptr - 1);
1002               return;
1003            end if;
1004
1005            --  Backup over semicolon or right-paren/semicolon sequence
1006
1007            if Source (Scan_Ptr - 1) = ';' then
1008               Scan_Ptr := Scan_Ptr - 1;
1009               Unstore_String_Char;
1010
1011               if Source (Scan_Ptr - 1) = ')' then
1012                  Scan_Ptr := Scan_Ptr - 1;
1013                  Unstore_String_Char;
1014               end if;
1015            end if;
1016
1017            --  See if there is a comma in the string, if so, guess that
1018            --  the first comma terminates the string.
1019
1020            S := String_Start;
1021            while S < Scan_Ptr loop
1022               if Source (S) = ',' then
1023                  while Scan_Ptr > S loop
1024                     Scan_Ptr := Scan_Ptr - 1;
1025                     Unstore_String_Char;
1026                  end loop;
1027
1028                  exit;
1029               end if;
1030
1031               S := S + 1;
1032            end loop;
1033
1034            --  Now we have adjusted the scan pointer, give message
1035
1036            Error_Msg_S -- CODEFIX
1037              ("missing string quote");
1038         end Error_Unterminated_String;
1039
1040         ----------------
1041         -- Set_String --
1042         ----------------
1043
1044         procedure Set_String is
1045            Slen : constant Int := Int (Scan_Ptr - Token_Ptr - 2);
1046            C1   : Character;
1047            C2   : Character;
1048            C3   : Character;
1049
1050         begin
1051            --  Token_Name is currently set to Error_Name. The following
1052            --  section of code resets Token_Name to the proper Name_Op_xx
1053            --  value if the string is a valid operator symbol, otherwise it is
1054            --  left set to Error_Name.
1055
1056            if Slen = 1 then
1057               C1 := Source (Token_Ptr + 1);
1058
1059               case C1 is
1060                  when '=' =>
1061                     Token_Name := Name_Op_Eq;
1062
1063                  when '>' =>
1064                     Token_Name := Name_Op_Gt;
1065
1066                  when '<' =>
1067                     Token_Name := Name_Op_Lt;
1068
1069                  when '+' =>
1070                     Token_Name := Name_Op_Add;
1071
1072                  when '-' =>
1073                     Token_Name := Name_Op_Subtract;
1074
1075                  when '&' =>
1076                     Token_Name := Name_Op_Concat;
1077
1078                  when '*' =>
1079                     Token_Name := Name_Op_Multiply;
1080
1081                  when '/' =>
1082                     Token_Name := Name_Op_Divide;
1083
1084                  when others =>
1085                     null;
1086               end case;
1087
1088            elsif Slen = 2 then
1089               C1 := Source (Token_Ptr + 1);
1090               C2 := Source (Token_Ptr + 2);
1091
1092               if C1 = '*' and then C2 = '*' then
1093                  Token_Name := Name_Op_Expon;
1094
1095               elsif C2 = '=' then
1096
1097                  if C1 = '/' then
1098                     Token_Name := Name_Op_Ne;
1099                  elsif C1 = '<' then
1100                     Token_Name := Name_Op_Le;
1101                  elsif C1 = '>' then
1102                     Token_Name := Name_Op_Ge;
1103                  end if;
1104
1105               elsif (C1 = 'O' or else C1 = 'o') and then    -- OR
1106                 (C2 = 'R' or else C2 = 'r')
1107               then
1108                  Token_Name := Name_Op_Or;
1109               end if;
1110
1111            elsif Slen = 3 then
1112               C1 := Source (Token_Ptr + 1);
1113               C2 := Source (Token_Ptr + 2);
1114               C3 := Source (Token_Ptr + 3);
1115
1116               if (C1 = 'A' or else C1 = 'a') and then       -- AND
1117                 (C2 = 'N' or else C2 = 'n') and then
1118                 (C3 = 'D' or else C3 = 'd')
1119               then
1120                  Token_Name := Name_Op_And;
1121
1122               elsif (C1 = 'A' or else C1 = 'a') and then    -- ABS
1123                 (C2 = 'B' or else C2 = 'b') and then
1124                 (C3 = 'S' or else C3 = 's')
1125               then
1126                  Token_Name := Name_Op_Abs;
1127
1128               elsif (C1 = 'M' or else C1 = 'm') and then    -- MOD
1129                 (C2 = 'O' or else C2 = 'o') and then
1130                 (C3 = 'D' or else C3 = 'd')
1131               then
1132                  Token_Name := Name_Op_Mod;
1133
1134               elsif (C1 = 'N' or else C1 = 'n') and then    -- NOT
1135                 (C2 = 'O' or else C2 = 'o') and then
1136                 (C3 = 'T' or else C3 = 't')
1137               then
1138                  Token_Name := Name_Op_Not;
1139
1140               elsif (C1 = 'R' or else C1 = 'r') and then    -- REM
1141                 (C2 = 'E' or else C2 = 'e') and then
1142                 (C3 = 'M' or else C3 = 'm')
1143               then
1144                  Token_Name := Name_Op_Rem;
1145
1146               elsif (C1 = 'X' or else C1 = 'x') and then    -- XOR
1147                 (C2 = 'O' or else C2 = 'o') and then
1148                 (C3 = 'R' or else C3 = 'r')
1149               then
1150                  Token_Name := Name_Op_Xor;
1151               end if;
1152
1153            end if;
1154
1155            --  If it is an operator symbol, then Token_Name is set. If it is
1156            --  some other string value, then Token_Name still contains
1157            --  Error_Name.
1158
1159            if Token_Name = Error_Name then
1160               Token := Tok_String_Literal;
1161
1162            else
1163               Token := Tok_Operator_Symbol;
1164            end if;
1165         end Set_String;
1166
1167      --  Start of processing for Slit
1168
1169      begin
1170         --  On entry, Scan_Ptr points to the opening character of the string
1171         --  which is either a percent, double quote, or apostrophe (single
1172         --  quote). The latter case is an error detected by the character
1173         --  literal circuit.
1174
1175         String_Start := Scan_Ptr;
1176
1177         Delimiter := Source (Scan_Ptr);
1178         Accumulate_Checksum (Delimiter);
1179
1180         Start_String;
1181         Wide_Character_Found      := False;
1182         Wide_Wide_Character_Found := False;
1183         Scan_Ptr := Scan_Ptr + 1;
1184
1185         --  Loop to scan out characters of string literal
1186
1187         loop
1188            C := Source (Scan_Ptr);
1189
1190            if C = Delimiter then
1191               Accumulate_Checksum (C);
1192               Scan_Ptr := Scan_Ptr + 1;
1193               exit when Source (Scan_Ptr) /= Delimiter;
1194               Code := Get_Char_Code (C);
1195               Accumulate_Checksum (C);
1196               Scan_Ptr := Scan_Ptr + 1;
1197
1198            else
1199               if C = '"' and then Delimiter = '%' then
1200                  Error_Msg_S
1201                    ("quote not allowed in percent delimited string");
1202                  Code := Get_Char_Code (C);
1203                  Scan_Ptr := Scan_Ptr + 1;
1204
1205               elsif Start_Of_Wide_Character then
1206                  Wptr := Scan_Ptr;
1207                  Scan_Wide (Source, Scan_Ptr, Code, Err);
1208
1209                  if Err then
1210                     Error_Illegal_Wide_Character;
1211                     Code := Get_Char_Code (' ');
1212                  end if;
1213
1214                  Accumulate_Checksum (Code);
1215
1216                  --  In Ada 95 mode we allow any wide characters in a string
1217                  --  but in Ada 2005, the set of characters allowed has been
1218                  --  restricted to graphic characters.
1219
1220                  if Ada_Version >= Ada_2005
1221                    and then Is_UTF_32_Non_Graphic (UTF_32 (Code))
1222                  then
1223                     Error_Msg
1224                       ("(Ada 2005) non-graphic character not permitted " &
1225                        "in string literal", Wptr);
1226                  end if;
1227
1228               else
1229                  Accumulate_Checksum (C);
1230
1231                  if C not in Graphic_Character then
1232                     if C in Line_Terminator then
1233                        Error_Unterminated_String;
1234                        exit;
1235
1236                     elsif C in Upper_Half_Character then
1237                        if Ada_Version = Ada_83 then
1238                           Error_Bad_String_Char;
1239                        end if;
1240
1241                     else
1242                        Error_Bad_String_Char;
1243                     end if;
1244                  end if;
1245
1246                  Code := Get_Char_Code (C);
1247                  Scan_Ptr := Scan_Ptr + 1;
1248               end if;
1249            end if;
1250
1251            Store_String_Char (Code);
1252
1253            if not In_Character_Range (Code) then
1254               if In_Wide_Character_Range (Code) then
1255                  Wide_Character_Found := True;
1256               else
1257                  Wide_Wide_Character_Found := True;
1258               end if;
1259            end if;
1260         end loop;
1261
1262         String_Literal_Id := End_String;
1263         Set_String;
1264         return;
1265      end Slit;
1266
1267      ----------------------------------
1268      -- Skip_Other_Format_Characters --
1269      ----------------------------------
1270
1271      procedure Skip_Other_Format_Characters is
1272         P    : Source_Ptr;
1273         Code : Char_Code;
1274         Err  : Boolean;
1275
1276      begin
1277         while Start_Of_Wide_Character loop
1278            P := Scan_Ptr;
1279            Scan_Wide (Source, Scan_Ptr, Code, Err);
1280
1281            if not Is_UTF_32_Other (UTF_32 (Code)) then
1282               Scan_Ptr := P;
1283               return;
1284            end if;
1285         end loop;
1286      end Skip_Other_Format_Characters;
1287
1288      -----------------------------
1289      -- Start_Of_Wide_Character --
1290      -----------------------------
1291
1292      function Start_Of_Wide_Character return Boolean is
1293         C : constant Character := Source (Scan_Ptr);
1294
1295      begin
1296         --  ESC encoding method with ESC present
1297
1298         if C = ESC
1299           and then Wide_Character_Encoding_Method in WC_ESC_Encoding_Method
1300         then
1301            return True;
1302
1303         --  Upper half character with upper half encoding
1304
1305         elsif C in Upper_Half_Character and then Upper_Half_Encoding then
1306            return True;
1307
1308         --  Brackets encoding
1309
1310         elsif C = '['
1311           and then Source (Scan_Ptr + 1) = '"'
1312           and then Identifier_Char (Source (Scan_Ptr + 2))
1313         then
1314            return True;
1315
1316         --  Not the start of a wide character
1317
1318         else
1319            return False;
1320         end if;
1321      end Start_Of_Wide_Character;
1322
1323   --  Start of processing for Scan
1324
1325   begin
1326      Prev_Token := Token;
1327      Prev_Token_Ptr := Token_Ptr;
1328      Token_Name := Error_Name;
1329
1330      --  The following loop runs more than once only if a format effector
1331      --  (tab, vertical tab, form  feed, line feed, carriage return) is
1332      --  encountered and skipped, or some error situation, such as an
1333      --  illegal character, is encountered.
1334
1335      <<Scan_Next_Character>>
1336
1337      loop
1338         --  Skip past blanks, loop is opened up for speed
1339
1340         while Source (Scan_Ptr) = ' ' loop
1341            if Source (Scan_Ptr + 1) /= ' ' then
1342               Scan_Ptr := Scan_Ptr + 1;
1343               exit;
1344            end if;
1345
1346            if Source (Scan_Ptr + 2) /= ' ' then
1347               Scan_Ptr := Scan_Ptr + 2;
1348               exit;
1349            end if;
1350
1351            if Source (Scan_Ptr + 3) /= ' ' then
1352               Scan_Ptr := Scan_Ptr + 3;
1353               exit;
1354            end if;
1355
1356            if Source (Scan_Ptr + 4) /= ' ' then
1357               Scan_Ptr := Scan_Ptr + 4;
1358               exit;
1359            end if;
1360
1361            if Source (Scan_Ptr + 5) /= ' ' then
1362               Scan_Ptr := Scan_Ptr + 5;
1363               exit;
1364            end if;
1365
1366            if Source (Scan_Ptr + 6) /= ' ' then
1367               Scan_Ptr := Scan_Ptr + 6;
1368               exit;
1369            end if;
1370
1371            if Source (Scan_Ptr + 7) /= ' ' then
1372               Scan_Ptr := Scan_Ptr + 7;
1373               exit;
1374            end if;
1375
1376            Scan_Ptr := Scan_Ptr + 8;
1377         end loop;
1378
1379         --  We are now at a non-blank character, which is the first character
1380         --  of the token we will scan, and hence the value of Token_Ptr.
1381
1382         Token_Ptr := Scan_Ptr;
1383
1384         --  Here begins the main case statement which transfers control on the
1385         --  basis of the non-blank character we have encountered.
1386
1387         case Source (Scan_Ptr) is
1388
1389         --  Line terminator characters
1390
1391         when CR | LF | FF | VT =>
1392            goto Scan_Line_Terminator;
1393
1394         --  Horizontal tab, just skip past it
1395
1396         when HT =>
1397            if Style_Check then
1398               Style.Check_HT;
1399            end if;
1400
1401            Scan_Ptr := Scan_Ptr + 1;
1402
1403         --  End of file character, treated as an end of file only if it is
1404         --  the last character in the buffer, otherwise it is ignored.
1405
1406         when EOF =>
1407            if Scan_Ptr = Source_Last (Current_Source_File) then
1408               Check_End_Of_Line;
1409
1410               if Style_Check then
1411                  Style.Check_EOF;
1412               end if;
1413
1414               Token := Tok_EOF;
1415               return;
1416            else
1417               Scan_Ptr := Scan_Ptr + 1;
1418            end if;
1419
1420         --  Ampersand
1421
1422         when '&' =>
1423            Accumulate_Checksum ('&');
1424
1425            if Source (Scan_Ptr + 1) = '&' then
1426               Error_Msg_S -- CODEFIX
1427                 ("'&'& should be `AND THEN`");
1428               Scan_Ptr := Scan_Ptr + 2;
1429               Token := Tok_And;
1430               return;
1431
1432            else
1433               Scan_Ptr := Scan_Ptr + 1;
1434               Token := Tok_Ampersand;
1435               return;
1436            end if;
1437
1438         --  Asterisk (can be multiplication operator or double asterisk which
1439         --  is the exponentiation compound delimiter).
1440
1441         when '*' =>
1442            Accumulate_Checksum ('*');
1443
1444            if Source (Scan_Ptr + 1) = '*' then
1445               Accumulate_Checksum ('*');
1446               Scan_Ptr := Scan_Ptr + 2;
1447               Token := Tok_Double_Asterisk;
1448               return;
1449
1450            else
1451               Scan_Ptr := Scan_Ptr + 1;
1452               Token := Tok_Asterisk;
1453               return;
1454            end if;
1455
1456         --  Colon, which can either be an isolated colon, or part of an
1457         --  assignment compound delimiter.
1458
1459         when ':' =>
1460            Accumulate_Checksum (':');
1461
1462            if Double_Char_Token ('=') then
1463               Token := Tok_Colon_Equal;
1464
1465               if Style_Check then
1466                  Style.Check_Colon_Equal;
1467               end if;
1468
1469               return;
1470
1471            elsif Source (Scan_Ptr + 1) = '-'
1472              and then Source (Scan_Ptr + 2) /= '-'
1473            then
1474               Token := Tok_Colon_Equal;
1475               Error_Msg -- CODEFIX
1476                 (":- should be :=", Scan_Ptr);
1477               Scan_Ptr := Scan_Ptr + 2;
1478               return;
1479
1480            else
1481               Scan_Ptr := Scan_Ptr + 1;
1482               Token := Tok_Colon;
1483
1484               if Style_Check then
1485                  Style.Check_Colon;
1486               end if;
1487
1488               return;
1489            end if;
1490
1491         --  Left parenthesis
1492
1493         when '(' =>
1494            Accumulate_Checksum ('(');
1495            Scan_Ptr := Scan_Ptr + 1;
1496            Token := Tok_Left_Paren;
1497
1498            if Style_Check then
1499               Style.Check_Left_Paren;
1500            end if;
1501
1502            return;
1503
1504         --  Left bracket
1505
1506         when '[' =>
1507            if Source (Scan_Ptr + 1) = '"' then
1508               goto Scan_Wide_Character;
1509
1510            else
1511               Error_Msg_S ("illegal character, replaced by ""(""");
1512               Scan_Ptr := Scan_Ptr + 1;
1513               Token := Tok_Left_Paren;
1514               return;
1515            end if;
1516
1517         --  Left brace
1518
1519         when '{' =>
1520            Error_Msg_S ("illegal character, replaced by ""(""");
1521            Scan_Ptr := Scan_Ptr + 1;
1522            Token := Tok_Left_Paren;
1523            return;
1524
1525         --  Comma
1526
1527         when ',' =>
1528            Accumulate_Checksum (',');
1529            Scan_Ptr := Scan_Ptr + 1;
1530            Token := Tok_Comma;
1531
1532            if Style_Check then
1533               Style.Check_Comma;
1534            end if;
1535
1536            return;
1537
1538         --  Dot, which is either an isolated period, or part of a double dot
1539         --  compound delimiter sequence. We also check for the case of a
1540         --  digit following the period, to give a better error message.
1541
1542         when '.' =>
1543            Accumulate_Checksum ('.');
1544
1545            if Double_Char_Token ('.') then
1546               Token := Tok_Dot_Dot;
1547
1548               if Style_Check then
1549                  Style.Check_Dot_Dot;
1550               end if;
1551
1552               return;
1553
1554            elsif Source (Scan_Ptr + 1) in '0' .. '9' then
1555               Error_Msg_S ("numeric literal cannot start with point");
1556               Scan_Ptr := Scan_Ptr + 1;
1557
1558            else
1559               Scan_Ptr := Scan_Ptr + 1;
1560               Token := Tok_Dot;
1561               return;
1562            end if;
1563
1564         --  Equal, which can either be an equality operator, or part of the
1565         --  arrow (=>) compound delimiter.
1566
1567         when '=' =>
1568            Accumulate_Checksum ('=');
1569
1570            if Double_Char_Token ('>') then
1571               Token := Tok_Arrow;
1572
1573               if Style_Check then
1574                  Style.Check_Arrow (Inside_Depends);
1575               end if;
1576
1577               return;
1578
1579            elsif Source (Scan_Ptr + 1) = '=' then
1580               Error_Msg_S -- CODEFIX
1581                 ("== should be =");
1582               Scan_Ptr := Scan_Ptr + 1;
1583            end if;
1584
1585            Scan_Ptr := Scan_Ptr + 1;
1586            Token := Tok_Equal;
1587            return;
1588
1589         --  Greater than, which can be a greater than operator, greater than
1590         --  or equal operator, or first character of a right label bracket.
1591
1592         when '>' =>
1593            Accumulate_Checksum ('>');
1594
1595            if Double_Char_Token ('=') then
1596               Token := Tok_Greater_Equal;
1597               return;
1598
1599            elsif Double_Char_Token ('>') then
1600               Token := Tok_Greater_Greater;
1601               return;
1602
1603            else
1604               Scan_Ptr := Scan_Ptr + 1;
1605               Token := Tok_Greater;
1606               return;
1607            end if;
1608
1609         --  Less than, which can be a less than operator, less than or equal
1610         --  operator, or the first character of a left label bracket, or the
1611         --  first character of a box (<>) compound delimiter.
1612
1613         when '<' =>
1614            Accumulate_Checksum ('<');
1615
1616            if Double_Char_Token ('=') then
1617               Token := Tok_Less_Equal;
1618               return;
1619
1620            elsif Double_Char_Token ('>') then
1621               Token := Tok_Box;
1622
1623               if Style_Check then
1624                  Style.Check_Box;
1625               end if;
1626
1627               return;
1628
1629            elsif Double_Char_Token ('<') then
1630               Token := Tok_Less_Less;
1631               return;
1632
1633            else
1634               Scan_Ptr := Scan_Ptr + 1;
1635               Token := Tok_Less;
1636               return;
1637            end if;
1638
1639         --  Minus, which is either a subtraction operator, or the first
1640         --  character of double minus starting a comment
1641
1642         when '-' => Minus_Case : begin
1643            if Source (Scan_Ptr + 1) = '>' then
1644               Error_Msg_S ("invalid token");
1645               Scan_Ptr := Scan_Ptr + 2;
1646               Token := Tok_Arrow;
1647               return;
1648
1649            elsif Source (Scan_Ptr + 1) /= '-' then
1650               Accumulate_Checksum ('-');
1651               Scan_Ptr := Scan_Ptr + 1;
1652               Token := Tok_Minus;
1653               return;
1654
1655            --  Comment
1656
1657            else -- Source (Scan_Ptr + 1) = '-' then
1658               if Style_Check then
1659                  Style.Check_Comment;
1660               end if;
1661
1662               Scan_Ptr := Scan_Ptr + 2;
1663
1664               --  If we are in preprocessor mode with Replace_In_Comments set,
1665               --  then we return the "--" as a token on its own.
1666
1667               if Replace_In_Comments then
1668                  Token := Tok_Comment;
1669                  return;
1670               end if;
1671
1672               --  Otherwise scan out the comment
1673
1674               Start_Of_Comment := Scan_Ptr;
1675
1676               --  Loop to scan comment (this loop runs more than once only if
1677               --  a horizontal tab or other non-graphic character is scanned)
1678
1679               loop
1680                  --  Scan to non graphic character (opened up for speed)
1681
1682                  --  Note that we just eat left brackets, which means that
1683                  --  bracket notation cannot be used for end of line
1684                  --  characters in comments. This seems a reasonable choice,
1685                  --  since no one would ever use brackets notation in a real
1686                  --  program in this situation, and if we allow brackets
1687                  --  notation, we forbid some valid comments which contain a
1688                  --  brackets sequence that happens to match an end of line
1689                  --  character.
1690
1691                  loop
1692                     exit when Source (Scan_Ptr) not in Graphic_Character;
1693                     Scan_Ptr := Scan_Ptr + 1;
1694                     exit when Source (Scan_Ptr) not in Graphic_Character;
1695                     Scan_Ptr := Scan_Ptr + 1;
1696                     exit when Source (Scan_Ptr) not in Graphic_Character;
1697                     Scan_Ptr := Scan_Ptr + 1;
1698                     exit when Source (Scan_Ptr) not in Graphic_Character;
1699                     Scan_Ptr := Scan_Ptr + 1;
1700                     exit when Source (Scan_Ptr) not in Graphic_Character;
1701                     Scan_Ptr := Scan_Ptr + 1;
1702                  end loop;
1703
1704                  --  Keep going if horizontal tab
1705
1706                  if Source (Scan_Ptr) = HT then
1707                     if Style_Check then
1708                        Style.Check_HT;
1709                     end if;
1710
1711                     Scan_Ptr := Scan_Ptr + 1;
1712
1713                  --  Terminate scan of comment if line terminator
1714
1715                  elsif Source (Scan_Ptr) in Line_Terminator then
1716                     exit;
1717
1718                  --  Terminate scan of comment if end of file encountered
1719                  --  (embedded EOF character or real last character in file)
1720
1721                  elsif Source (Scan_Ptr) = EOF then
1722                     exit;
1723
1724                  --  If we have a wide character, we have to scan it out,
1725                  --  because it might be a legitimate line terminator
1726
1727                  elsif Start_Of_Wide_Character then
1728                     declare
1729                        Wptr : constant Source_Ptr := Scan_Ptr;
1730                        Code : Char_Code;
1731                        Err  : Boolean;
1732
1733                     begin
1734                        Scan_Wide (Source, Scan_Ptr, Code, Err);
1735
1736                        --  If not well formed wide character, then just skip
1737                        --  past it and ignore it.
1738
1739                        if Err then
1740                           Scan_Ptr := Wptr + 1;
1741
1742                        --  If UTF_32 terminator, terminate comment scan
1743
1744                        elsif Is_UTF_32_Line_Terminator (UTF_32 (Code)) then
1745                           Scan_Ptr := Wptr;
1746                           exit;
1747                        end if;
1748                     end;
1749
1750                  --  Keep going if character in 80-FF range, or is ESC. These
1751                  --  characters are allowed in comments by RM-2.1(1), 2.7(2).
1752                  --  They are allowed even in Ada 83 mode according to the
1753                  --  approved AI. ESC was added to the AI in June 93.
1754
1755                  elsif Source (Scan_Ptr) in Upper_Half_Character
1756                     or else Source (Scan_Ptr) = ESC
1757                  then
1758                     Scan_Ptr := Scan_Ptr + 1;
1759
1760                  --  Otherwise we have an illegal comment character, ignore
1761                  --  this error in relaxed semantics mode.
1762
1763                  else
1764                     if Relaxed_RM_Semantics then
1765                        Scan_Ptr := Scan_Ptr + 1;
1766                     else
1767                        Error_Illegal_Character;
1768                     end if;
1769                  end if;
1770               end loop;
1771
1772               --  Note that, except when comments are tokens, we do NOT
1773               --  execute a return here, instead we fall through to reexecute
1774               --  the scan loop to look for a token.
1775
1776               if Comment_Is_Token then
1777                  Name_Len := Integer (Scan_Ptr - Start_Of_Comment);
1778                  Name_Buffer (1 .. Name_Len) :=
1779                    String (Source (Start_Of_Comment .. Scan_Ptr - 1));
1780                  Comment_Id := Name_Find;
1781                  Token := Tok_Comment;
1782                  return;
1783               end if;
1784
1785               --  If the SPARK restriction is set for this unit, then generate
1786               --  a token Tok_SPARK_Hide for a SPARK HIDE directive.
1787
1788               if Restriction_Check_Required (SPARK_05)
1789                 and then Source (Start_Of_Comment) = '#'
1790               then
1791                  declare
1792                     Scan_SPARK_Ptr : Source_Ptr;
1793
1794                  begin
1795                     Scan_SPARK_Ptr := Start_Of_Comment + 1;
1796
1797                     --  Scan out blanks
1798
1799                     while Source (Scan_SPARK_Ptr) = ' '
1800                       or else Source (Scan_SPARK_Ptr) = HT
1801                     loop
1802                        Scan_SPARK_Ptr := Scan_SPARK_Ptr + 1;
1803                     end loop;
1804
1805                     --  Recognize HIDE directive. SPARK input cannot be
1806                     --  encoded as wide characters, so only deal with
1807                     --  lower/upper case.
1808
1809                     if (Source (Scan_SPARK_Ptr) = 'h'
1810                          or else Source (Scan_SPARK_Ptr) = 'H')
1811                       and then (Source (Scan_SPARK_Ptr + 1) = 'i'
1812                                  or else Source (Scan_SPARK_Ptr + 1) = 'I')
1813                       and then (Source (Scan_SPARK_Ptr + 2) = 'd'
1814                                  or else Source (Scan_SPARK_Ptr + 2) = 'D')
1815                       and then (Source (Scan_SPARK_Ptr + 3) = 'e'
1816                                  or else Source (Scan_SPARK_Ptr + 3) = 'E')
1817                       and then (Source (Scan_SPARK_Ptr + 4) = ' '
1818                                  or else Source (Scan_SPARK_Ptr + 4) = HT)
1819                     then
1820                        Token := Tok_SPARK_Hide;
1821                        return;
1822                     end if;
1823                  end;
1824               end if;
1825            end if;
1826         end Minus_Case;
1827
1828         --  Double quote or percent starting a string literal
1829
1830         when '"' | '%' =>
1831            Slit;
1832            Post_Scan;
1833            return;
1834
1835         --  Apostrophe. This can either be the start of a character literal,
1836         --  or an isolated apostrophe used in a qualified expression or an
1837         --  attribute. In the following:
1838
1839         --    A := CHARACTER'('A');
1840
1841         --  the first apostrophe is treated as an isolated apostrophe, and the
1842         --  second one is treated as the start of the character literal 'A'.
1843         --  Note that RM-2.2(7) does not require a separator between "'" and
1844         --  "(" in the above, so we cannot use lookahead to distinguish the
1845         --  cases; we use look-back instead. Analysis of the grammar shows
1846         --  that some tokens can be followed by an apostrophe, and some by a
1847         --  character literal, but none by both. Some cannot be followed by
1848         --  either, so it doesn't matter what we do in those cases, except to
1849         --  get good error behavior.
1850
1851         when ''' => Char_Literal_Case : declare
1852            Code : Char_Code;
1853            Err  : Boolean;
1854
1855         begin
1856            Accumulate_Checksum (''');
1857            Scan_Ptr := Scan_Ptr + 1;
1858
1859            --  Distinguish between apostrophe and character literal. It's an
1860            --  apostrophe if the previous token is one of the following.
1861            --  Reserved words are included for things like A.all'Address and
1862            --  T'Digits'Img. Strings literals are included for things like
1863            --  "abs"'Address. Other literals are included to give better error
1864            --  behavior for illegal cases like 123'Img.
1865
1866            if Prev_Token = Tok_Identifier
1867               or else Prev_Token = Tok_Right_Paren
1868               or else Prev_Token = Tok_All
1869               or else Prev_Token = Tok_Delta
1870               or else Prev_Token = Tok_Digits
1871               or else Prev_Token = Tok_Project
1872               or else Prev_Token in Token_Class_Literal
1873            then
1874               Token := Tok_Apostrophe;
1875
1876               if Style_Check then
1877                  Style.Check_Apostrophe;
1878               end if;
1879
1880               return;
1881
1882            --  Otherwise the apostrophe starts a character literal
1883
1884            else
1885               --  Case of wide character literal
1886
1887               if Start_Of_Wide_Character then
1888                  Wptr := Scan_Ptr;
1889                  Scan_Wide (Source, Scan_Ptr, Code, Err);
1890                  Accumulate_Checksum (Code);
1891
1892                  if Err then
1893                     Error_Illegal_Wide_Character;
1894                     Code := Character'Pos (' ');
1895
1896                  --  In Ada 95 mode we allow any wide character in a character
1897                  --  literal, but in Ada 2005, the set of characters allowed
1898                  --  is restricted to graphic characters.
1899
1900                  elsif Ada_Version >= Ada_2005
1901                    and then Is_UTF_32_Non_Graphic (UTF_32 (Code))
1902                  then
1903                     Error_Msg -- CODEFIX????
1904                       ("(Ada 2005) non-graphic character not permitted " &
1905                        "in character literal", Wptr);
1906                  end if;
1907
1908                  if Source (Scan_Ptr) /= ''' then
1909                        Error_Msg_S ("missing apostrophe");
1910                  else
1911                     Scan_Ptr := Scan_Ptr + 1;
1912                  end if;
1913
1914               --  If we do not find a closing quote in the expected place then
1915               --  assume that we have a misguided attempt at a string literal.
1916
1917               --  However, if previous token is RANGE, then we return an
1918               --  apostrophe instead since this gives better error recovery
1919
1920               elsif Source (Scan_Ptr + 1) /= ''' then
1921                  if Prev_Token = Tok_Range then
1922                     Token := Tok_Apostrophe;
1923                     return;
1924
1925                  else
1926                     Scan_Ptr := Scan_Ptr - 1;
1927                     Error_Msg_S
1928                       ("strings are delimited by double quote character");
1929                     Slit;
1930                     Post_Scan;
1931                     return;
1932                  end if;
1933
1934               --  Otherwise we have a (non-wide) character literal
1935
1936               else
1937                  Accumulate_Checksum (Source (Scan_Ptr));
1938
1939                  if Source (Scan_Ptr) not in Graphic_Character then
1940                     if Source (Scan_Ptr) in Upper_Half_Character then
1941                        if Ada_Version = Ada_83 then
1942                           Error_Illegal_Character;
1943                        end if;
1944
1945                     else
1946                        Error_Illegal_Character;
1947                     end if;
1948                  end if;
1949
1950                  Code := Get_Char_Code (Source (Scan_Ptr));
1951                  Scan_Ptr := Scan_Ptr + 2;
1952               end if;
1953
1954               --  Fall through here with Scan_Ptr updated past the closing
1955               --  quote, and Code set to the Char_Code value for the literal
1956
1957               Accumulate_Checksum (''');
1958               Token := Tok_Char_Literal;
1959               Set_Character_Literal_Name (Code);
1960               Token_Name := Name_Find;
1961               Character_Code := Code;
1962               Post_Scan;
1963               return;
1964            end if;
1965         end Char_Literal_Case;
1966
1967         --  Right parenthesis
1968
1969         when ')' =>
1970            Accumulate_Checksum (')');
1971            Scan_Ptr := Scan_Ptr + 1;
1972            Token := Tok_Right_Paren;
1973
1974            if Style_Check then
1975               Style.Check_Right_Paren;
1976            end if;
1977
1978            return;
1979
1980         --  Right bracket or right brace, treated as right paren
1981
1982         when ']' | '}' =>
1983            Error_Msg_S ("illegal character, replaced by "")""");
1984            Scan_Ptr := Scan_Ptr + 1;
1985            Token := Tok_Right_Paren;
1986            return;
1987
1988         --  Slash (can be division operator or first character of not equal)
1989
1990         when '/' =>
1991            Accumulate_Checksum ('/');
1992
1993            if Double_Char_Token ('=') then
1994               Token := Tok_Not_Equal;
1995               return;
1996            else
1997               Scan_Ptr := Scan_Ptr + 1;
1998               Token := Tok_Slash;
1999               return;
2000            end if;
2001
2002         --  Semicolon
2003
2004         when ';' =>
2005            Accumulate_Checksum (';');
2006            Scan_Ptr := Scan_Ptr + 1;
2007            Token := Tok_Semicolon;
2008
2009            if Style_Check then
2010               Style.Check_Semicolon;
2011            end if;
2012
2013            return;
2014
2015         --  Vertical bar
2016
2017         when '|' => Vertical_Bar_Case : begin
2018            Accumulate_Checksum ('|');
2019
2020            --  Special check for || to give nice message
2021
2022            if Source (Scan_Ptr + 1) = '|' then
2023               Error_Msg_S -- CODEFIX
2024                 ("""'|'|"" should be `OR ELSE`");
2025               Scan_Ptr := Scan_Ptr + 2;
2026               Token := Tok_Or;
2027               return;
2028
2029            else
2030               Scan_Ptr := Scan_Ptr + 1;
2031               Token := Tok_Vertical_Bar;
2032
2033               if Style_Check then
2034                  Style.Check_Vertical_Bar;
2035               end if;
2036
2037               Post_Scan;
2038               return;
2039            end if;
2040         end Vertical_Bar_Case;
2041
2042         --  Exclamation, replacement character for vertical bar
2043
2044         when '!' => Exclamation_Case : begin
2045            Accumulate_Checksum ('!');
2046
2047            if Source (Scan_Ptr + 1) = '=' then
2048               Error_Msg_S -- CODEFIX
2049                 ("'!= should be /=");
2050               Scan_Ptr := Scan_Ptr + 2;
2051               Token := Tok_Not_Equal;
2052               return;
2053
2054            else
2055               Scan_Ptr := Scan_Ptr + 1;
2056               Token := Tok_Vertical_Bar;
2057               Post_Scan;
2058               return;
2059            end if;
2060         end Exclamation_Case;
2061
2062         --  Plus
2063
2064         when '+' => Plus_Case : begin
2065            Accumulate_Checksum ('+');
2066            Scan_Ptr := Scan_Ptr + 1;
2067            Token := Tok_Plus;
2068            return;
2069         end Plus_Case;
2070
2071         --  Digits starting a numeric literal
2072
2073         when '0' .. '9' =>
2074
2075            --  First a bit of a scan ahead to see if we have a case of an
2076            --  identifier starting with a digit (remembering exponent case).
2077
2078            declare
2079               C : constant Character := Source (Scan_Ptr + 1);
2080
2081            begin
2082               --  OK literal if digit followed by digit or underscore
2083
2084               if C in '0' .. '9' or else C = '_' then
2085                  null;
2086
2087               --  OK literal if digit not followed by identifier char
2088
2089               elsif not Identifier_Char (C) then
2090                  null;
2091
2092               --  OK literal if digit followed by e/E followed by digit/sign.
2093               --  We also allow underscore after the E, which is an error, but
2094               --  better handled by Nlit than deciding this is an identifier.
2095
2096               elsif (C = 'e' or else C = 'E')
2097                 and then (Source (Scan_Ptr + 2) in '0' .. '9'
2098                             or else Source (Scan_Ptr + 2) = '+'
2099                             or else Source (Scan_Ptr + 2) = '-'
2100                             or else Source (Scan_Ptr + 2) = '_')
2101               then
2102                  null;
2103
2104               --  Here we have what really looks like an identifier that
2105               --  starts with a digit, so give error msg.
2106
2107               else
2108                  Error_Msg_S ("identifier may not start with digit");
2109                  Name_Len := 1;
2110                  Underline_Found := False;
2111                  Name_Buffer (1) := Source (Scan_Ptr);
2112                  Accumulate_Checksum (Name_Buffer (1));
2113                  Scan_Ptr := Scan_Ptr + 1;
2114                  goto Scan_Identifier;
2115               end if;
2116            end;
2117
2118            --  Here we have an OK integer literal
2119
2120            Nlit;
2121
2122            --  Check for proper delimiter, ignoring other format characters
2123
2124            Skip_Other_Format_Characters;
2125
2126            if Identifier_Char (Source (Scan_Ptr)) then
2127               Error_Msg_S
2128                 ("delimiter required between literal and identifier");
2129            end if;
2130
2131            Post_Scan;
2132            return;
2133
2134         --  Lower case letters
2135
2136         when 'a' .. 'z' =>
2137            Name_Len := 1;
2138            Underline_Found := False;
2139            Name_Buffer (1) := Source (Scan_Ptr);
2140            Accumulate_Checksum (Name_Buffer (1));
2141            Scan_Ptr := Scan_Ptr + 1;
2142            goto Scan_Identifier;
2143
2144         --  Upper case letters
2145
2146         when 'A' .. 'Z' =>
2147            Name_Len := 1;
2148            Underline_Found := False;
2149            Name_Buffer (1) :=
2150              Character'Val (Character'Pos (Source (Scan_Ptr)) + 32);
2151            Accumulate_Checksum (Name_Buffer (1));
2152            Scan_Ptr := Scan_Ptr + 1;
2153            goto Scan_Identifier;
2154
2155         --  Underline character
2156
2157         when '_' =>
2158            if Special_Characters ('_') then
2159               Token_Ptr := Scan_Ptr;
2160               Scan_Ptr := Scan_Ptr + 1;
2161               Token := Tok_Special;
2162               Special_Character := '_';
2163               return;
2164            end if;
2165
2166            Error_Msg_S ("identifier cannot start with underline");
2167            Name_Len := 1;
2168            Name_Buffer (1) := '_';
2169            Scan_Ptr := Scan_Ptr + 1;
2170            Underline_Found := False;
2171            goto Scan_Identifier;
2172
2173         --  Space (not possible, because we scanned past blanks)
2174
2175         when ' ' =>
2176            raise Program_Error;
2177
2178         --  Characters in top half of ASCII 8-bit chart
2179
2180         when Upper_Half_Character =>
2181
2182            --  Wide character case
2183
2184            if Upper_Half_Encoding then
2185               goto Scan_Wide_Character;
2186
2187            --  Otherwise we have OK Latin-1 character
2188
2189            else
2190               --  Upper half characters may possibly be identifier letters
2191               --  but can never be digits, so Identifier_Char can be used to
2192               --  test for a valid start of identifier character.
2193
2194               if Identifier_Char (Source (Scan_Ptr)) then
2195                  Name_Len := 0;
2196                  Underline_Found := False;
2197                  goto Scan_Identifier;
2198               else
2199                  Error_Illegal_Character;
2200               end if;
2201            end if;
2202
2203         when ESC =>
2204
2205            --  ESC character, possible start of identifier if wide characters
2206            --  using ESC encoding are allowed in identifiers, which we can
2207            --  tell by looking at the Identifier_Char flag for ESC, which is
2208            --  only true if these conditions are met. In Ada 2005 mode, may
2209            --  also be valid UTF_32 space or line terminator character.
2210
2211            if Identifier_Char (ESC) then
2212               Name_Len := 0;
2213               goto Scan_Wide_Character;
2214            else
2215               Error_Illegal_Character;
2216            end if;
2217
2218         --  Invalid control characters
2219
2220         when NUL | SOH | STX | ETX | EOT | ENQ | ACK | BEL | BS  | ASCII.SO |
2221              SI  | DLE | DC1 | DC2 | DC3 | DC4 | NAK | SYN | ETB | CAN |
2222              EM  | FS  | GS  | RS  | US  | DEL
2223         =>
2224            Error_Illegal_Character;
2225
2226         --  Invalid graphic characters
2227
2228         when '#' | '$' | '?' | '@' | '`' | '\' | '^' | '~' =>
2229
2230            --  If Set_Special_Character has been called for this character,
2231            --  set Scans.Special_Character and return a Special token.
2232
2233            if Special_Characters (Source (Scan_Ptr)) then
2234               Token_Ptr := Scan_Ptr;
2235               Token := Tok_Special;
2236               Special_Character := Source (Scan_Ptr);
2237               Scan_Ptr := Scan_Ptr + 1;
2238               return;
2239
2240            --  Check for something looking like a preprocessor directive
2241
2242            elsif Source (Scan_Ptr) = '#'
2243              and then (Source (Scan_Ptr + 1 .. Scan_Ptr + 2) = "if"
2244                          or else
2245                        Source (Scan_Ptr + 1 .. Scan_Ptr + 5) = "elsif"
2246                          or else
2247                        Source (Scan_Ptr + 1 .. Scan_Ptr + 4) = "else"
2248                          or else
2249                        Source (Scan_Ptr + 1 .. Scan_Ptr + 3) = "end")
2250            then
2251               Error_Msg_S
2252                 ("preprocessor directive ignored, preprocessor not active");
2253
2254               --  Skip to end of line
2255
2256               loop
2257                  if Source (Scan_Ptr) in Graphic_Character
2258                       or else
2259                     Source (Scan_Ptr) = HT
2260                  then
2261                     Scan_Ptr := Scan_Ptr + 1;
2262
2263                  --  Done if line terminator or EOF
2264
2265                  elsif Source (Scan_Ptr) in Line_Terminator
2266                          or else
2267                        Source (Scan_Ptr) = EOF
2268                  then
2269                     exit;
2270
2271                  --  If we have a wide character, we have to scan it out,
2272                  --  because it might be a legitimate line terminator
2273
2274                  elsif Start_Of_Wide_Character then
2275                     declare
2276                        Wptr : constant Source_Ptr := Scan_Ptr;
2277                        Code : Char_Code;
2278                        Err  : Boolean;
2279
2280                     begin
2281                        Scan_Wide (Source, Scan_Ptr, Code, Err);
2282
2283                        --  If not well formed wide character, then just skip
2284                        --  past it and ignore it.
2285
2286                        if Err then
2287                           Scan_Ptr := Wptr + 1;
2288
2289                        --  If UTF_32 terminator, terminate comment scan
2290
2291                        elsif Is_UTF_32_Line_Terminator (UTF_32 (Code)) then
2292                           Scan_Ptr := Wptr;
2293                           exit;
2294                        end if;
2295                     end;
2296
2297                  --  Else keep going (don't worry about bad comment chars
2298                  --  in this context, we just want to find the end of line.
2299
2300                  else
2301                     Scan_Ptr := Scan_Ptr + 1;
2302                  end if;
2303               end loop;
2304
2305            --  Otherwise, this is an illegal character
2306
2307            else
2308               Error_Illegal_Character;
2309            end if;
2310
2311         --  End switch on non-blank character
2312
2313         end case;
2314
2315      --  End loop past format effectors. The exit from this loop is by
2316      --  executing a return statement following completion of token scan
2317      --  (control never falls out of this loop to the code which follows)
2318
2319      end loop;
2320
2321      --  Wide_Character scanning routine. On entry we have encountered the
2322      --  initial character of a wide character sequence.
2323
2324      <<Scan_Wide_Character>>
2325
2326         declare
2327            Code : Char_Code;
2328            Cat  : Category;
2329            Err  : Boolean;
2330
2331         begin
2332            Wptr := Scan_Ptr;
2333            Scan_Wide (Source, Scan_Ptr, Code, Err);
2334
2335            --  If bad wide character, signal error and continue scan
2336
2337            if Err then
2338               Error_Illegal_Wide_Character;
2339               goto Scan_Next_Character;
2340            end if;
2341
2342            Cat := Get_Category (UTF_32 (Code));
2343
2344            --  If OK letter, reset scan ptr and go scan identifier
2345
2346            if Is_UTF_32_Letter (Cat) then
2347               Scan_Ptr := Wptr;
2348               Name_Len := 0;
2349               Underline_Found := False;
2350               goto Scan_Identifier;
2351
2352            --  If OK wide space, ignore and keep scanning (we do not include
2353            --  any ignored spaces in checksum)
2354
2355            elsif Is_UTF_32_Space (Cat) then
2356               goto Scan_Next_Character;
2357
2358            --  If other format character, ignore and keep scanning (again we
2359            --  do not include in the checksum) (this is for AI-0079).
2360
2361            elsif Is_UTF_32_Other (Cat) then
2362               goto Scan_Next_Character;
2363
2364            --  If OK wide line terminator, terminate current line
2365
2366            elsif Is_UTF_32_Line_Terminator (UTF_32 (Code)) then
2367               Scan_Ptr := Wptr;
2368               goto Scan_Line_Terminator;
2369
2370            --  Punctuation is an error (at start of identifier)
2371
2372            elsif Is_UTF_32_Punctuation (Cat) then
2373               Error_Msg ("identifier cannot start with punctuation", Wptr);
2374               Scan_Ptr := Wptr;
2375               Name_Len := 0;
2376               Underline_Found := False;
2377               goto Scan_Identifier;
2378
2379            --  Mark character is an error (at start of identifier)
2380
2381            elsif Is_UTF_32_Mark (Cat) then
2382               Error_Msg ("identifier cannot start with mark character", Wptr);
2383               Scan_Ptr := Wptr;
2384               Name_Len := 0;
2385               Underline_Found := False;
2386               goto Scan_Identifier;
2387
2388            --  Extended digit character is an error. Could be bad start of
2389            --  identifier or bad literal. Not worth doing too much to try to
2390            --  distinguish these cases, but we will do a little bit.
2391
2392            elsif Is_UTF_32_Digit (Cat) then
2393               Error_Msg
2394                 ("identifier cannot start with digit character", Wptr);
2395               Scan_Ptr := Wptr;
2396               Name_Len := 0;
2397               Underline_Found := False;
2398               goto Scan_Identifier;
2399
2400            --  All other wide characters are illegal here
2401
2402            else
2403               Error_Illegal_Wide_Character;
2404               goto Scan_Next_Character;
2405            end if;
2406         end;
2407
2408      --  Routine to scan line terminator. On entry Scan_Ptr points to a
2409      --  character which is one of FF,LR,CR,VT, or one of the wide characters
2410      --  that is treated as a line terminator.
2411
2412      <<Scan_Line_Terminator>>
2413
2414         --  Check line too long
2415
2416         Check_End_Of_Line;
2417
2418         --  Set Token_Ptr, if End_Of_Line is a token, for the case when it is
2419         --  a physical line.
2420
2421         if End_Of_Line_Is_Token then
2422            Token_Ptr := Scan_Ptr;
2423         end if;
2424
2425         declare
2426            Physical : Boolean;
2427
2428         begin
2429            Skip_Line_Terminators (Scan_Ptr, Physical);
2430
2431            --  If we are at start of physical line, update scan pointers to
2432            --  reflect the start of the new line.
2433
2434            if Physical then
2435               Current_Line_Start       := Scan_Ptr;
2436               Start_Column             := Set_Start_Column;
2437               First_Non_Blank_Location := Scan_Ptr;
2438
2439               --  If End_Of_Line is a token, we return it as it is a
2440               --  physical line.
2441
2442               if End_Of_Line_Is_Token then
2443                  Token := Tok_End_Of_Line;
2444                  return;
2445               end if;
2446            end if;
2447         end;
2448
2449         goto Scan_Next_Character;
2450
2451      --  Identifier scanning routine. On entry, some initial characters of
2452      --  the identifier may have already been stored in Name_Buffer. If so,
2453      --  Name_Len has the number of characters stored, otherwise Name_Len is
2454      --  set to zero on entry. Underline_Found is also set False on entry.
2455
2456      <<Scan_Identifier>>
2457
2458         --  This loop scans as fast as possible past lower half letters and
2459         --  digits, which we expect to be the most common characters.
2460
2461         loop
2462            if Source (Scan_Ptr) in 'a' .. 'z'
2463              or else Source (Scan_Ptr) in '0' .. '9'
2464            then
2465               Name_Buffer (Name_Len + 1) := Source (Scan_Ptr);
2466               Accumulate_Checksum (Source (Scan_Ptr));
2467
2468            elsif Source (Scan_Ptr) in 'A' .. 'Z' then
2469               Name_Buffer (Name_Len + 1) :=
2470                 Character'Val (Character'Pos (Source (Scan_Ptr)) + 32);
2471               Accumulate_Checksum (Name_Buffer (Name_Len + 1));
2472
2473            else
2474               exit;
2475            end if;
2476
2477            Underline_Found := False;
2478            Scan_Ptr := Scan_Ptr + 1;
2479            Name_Len := Name_Len + 1;
2480         end loop;
2481
2482         --  If we fall through, then we have encountered either an underline
2483         --  character, or an extended identifier character (i.e. one from the
2484         --  upper half), or a wide character, or an identifier terminator. The
2485         --  initial test speeds us up in the most common case where we have
2486         --  an identifier terminator. Note that ESC is an identifier character
2487         --  only if a wide character encoding method that uses ESC encoding
2488         --  is active, so if we find an ESC character we know that we have a
2489         --  wide character.
2490
2491         if Identifier_Char (Source (Scan_Ptr))
2492           or else (Source (Scan_Ptr) in Upper_Half_Character
2493                     and then Upper_Half_Encoding)
2494         then
2495            --  Case of underline
2496
2497            if Source (Scan_Ptr) = '_' then
2498               Accumulate_Checksum ('_');
2499
2500               if Underline_Found then
2501                  Error_No_Double_Underline;
2502               else
2503                  Underline_Found := True;
2504                  Name_Len := Name_Len + 1;
2505                  Name_Buffer (Name_Len) := '_';
2506               end if;
2507
2508               Scan_Ptr := Scan_Ptr + 1;
2509               goto Scan_Identifier;
2510
2511            --  Upper half character
2512
2513            elsif Source (Scan_Ptr) in Upper_Half_Character
2514              and then not Upper_Half_Encoding
2515            then
2516               Accumulate_Checksum (Source (Scan_Ptr));
2517               Store_Encoded_Character
2518                 (Get_Char_Code (Fold_Lower (Source (Scan_Ptr))));
2519               Scan_Ptr := Scan_Ptr + 1;
2520               Underline_Found := False;
2521               goto Scan_Identifier;
2522
2523            --  Left bracket not followed by a quote terminates an identifier.
2524            --  This is an error, but we don't want to give a junk error msg
2525            --  about wide characters in this case.
2526
2527            elsif Source (Scan_Ptr) = '['
2528              and then Source (Scan_Ptr + 1) /= '"'
2529            then
2530               null;
2531
2532            --  We know we have a wide character encoding here (the current
2533            --  character is either ESC, left bracket, or an upper half
2534            --  character depending on the encoding method).
2535
2536            else
2537               --  Scan out the wide character and insert the appropriate
2538               --  encoding into the name table entry for the identifier.
2539
2540               declare
2541                  Code : Char_Code;
2542                  Err  : Boolean;
2543                  Chr  : Character;
2544                  Cat  : Category;
2545
2546               begin
2547                  Wptr := Scan_Ptr;
2548                  Scan_Wide (Source, Scan_Ptr, Code, Err);
2549
2550                  --  If error, signal error
2551
2552                  if Err then
2553                     Error_Illegal_Wide_Character;
2554
2555                  --  If the character scanned is a normal identifier
2556                  --  character, then we treat it that way.
2557
2558                  elsif In_Character_Range (Code)
2559                    and then Identifier_Char (Get_Character (Code))
2560                  then
2561                     Chr := Get_Character (Code);
2562                     Accumulate_Checksum (Chr);
2563                     Store_Encoded_Character
2564                       (Get_Char_Code (Fold_Lower (Chr)));
2565                     Underline_Found := False;
2566
2567                  --  Here if not a normal identifier character
2568
2569                  else
2570                     Cat := Get_Category (UTF_32 (Code));
2571
2572                     --  Wide character in Unicode category "Other, Format"
2573                     --  is not accepted in an identifier. This is because it
2574                     --  it is considered a security risk (AI-0091).
2575
2576                     --  However, it is OK for such a character to appear at
2577                     --  the end of an identifier.
2578
2579                     if Is_UTF_32_Other (Cat) then
2580                        if not Identifier_Char (Source (Scan_Ptr)) then
2581                           goto Scan_Identifier_Complete;
2582                        else
2583                           Error_Msg
2584                             ("identifier cannot contain other_format "
2585                              & "character", Wptr);
2586                           goto Scan_Identifier;
2587                        end if;
2588
2589                     --  Wide character in category Separator,Space terminates
2590
2591                     elsif Is_UTF_32_Space (Cat) then
2592                        goto Scan_Identifier_Complete;
2593                     end if;
2594
2595                     --  Here if wide character is part of the identifier
2596
2597                     --  Make sure we are allowing wide characters in
2598                     --  identifiers. Note that we allow wide character
2599                     --  notation for an OK identifier character. This in
2600                     --  particular allows bracket or other notation to be
2601                     --  used for upper half letters.
2602
2603                     --  Wide characters are always allowed in Ada 2005
2604
2605                     if Identifier_Character_Set /= 'w'
2606                       and then Ada_Version < Ada_2005
2607                     then
2608                        Error_Msg
2609                          ("wide character not allowed in identifier", Wptr);
2610                     end if;
2611
2612                     --  If OK letter, store it folding to upper case. Note
2613                     --  that we include the folded letter in the checksum.
2614
2615                     if Is_UTF_32_Letter (Cat) then
2616                        Code :=
2617                          Char_Code (UTF_32_To_Upper_Case (UTF_32 (Code)));
2618                        Accumulate_Checksum (Code);
2619                        Store_Encoded_Character (Code);
2620                        Underline_Found := False;
2621
2622                     --  If OK extended digit or mark, then store it
2623
2624                     elsif Is_UTF_32_Digit (Cat)
2625                       or else Is_UTF_32_Mark (Cat)
2626                     then
2627                        Accumulate_Checksum (Code);
2628                        Store_Encoded_Character (Code);
2629                        Underline_Found := False;
2630
2631                     --  Wide punctuation is also stored, but counts as an
2632                     --  underline character for error checking purposes.
2633
2634                     elsif Is_UTF_32_Punctuation (Cat) then
2635                        Accumulate_Checksum (Code);
2636
2637                        if Underline_Found then
2638                           declare
2639                              Cend : constant Source_Ptr := Scan_Ptr;
2640                           begin
2641                              Scan_Ptr := Wptr;
2642                              Error_No_Double_Underline;
2643                              Scan_Ptr := Cend;
2644                           end;
2645
2646                        else
2647                           Store_Encoded_Character (Code);
2648                           Underline_Found := True;
2649                        end if;
2650
2651                     --  Any other wide character is not acceptable
2652
2653                     else
2654                        Error_Msg
2655                          ("invalid wide character in identifier", Wptr);
2656                     end if;
2657                  end if;
2658
2659                  goto Scan_Identifier;
2660               end;
2661            end if;
2662         end if;
2663
2664      --  Scan of identifier is complete. The identifier is stored in
2665      --  Name_Buffer, and Scan_Ptr points past the last character.
2666
2667      <<Scan_Identifier_Complete>>
2668         Token_Name := Name_Find;
2669
2670         --  Check for identifier ending with underline or punctuation char
2671
2672         if Underline_Found then
2673            Underline_Found := False;
2674
2675            if Source (Scan_Ptr - 1) = '_' then
2676               Error_Msg
2677                 ("identifier cannot end with underline", Scan_Ptr - 1);
2678            else
2679               Error_Msg
2680                 ("identifier cannot end with punctuation character", Wptr);
2681            end if;
2682         end if;
2683
2684         --  We will assume it is an identifier, not a keyword, so that the
2685         --  checksum is independent of the Ada version.
2686
2687         Token := Tok_Identifier;
2688
2689         --  Here is where we check if it was a keyword
2690
2691         if Is_Keyword_Name (Token_Name) then
2692            if Opt.Checksum_GNAT_6_3 then
2693               Token := Token_Type'Val (Get_Name_Table_Byte (Token_Name));
2694
2695               if Checksum_Accumulate_Token_Checksum then
2696                  if Checksum_GNAT_5_03 then
2697                     Accumulate_Token_Checksum_GNAT_5_03;
2698                  else
2699                     Accumulate_Token_Checksum_GNAT_6_3;
2700                  end if;
2701               end if;
2702
2703            else
2704               Accumulate_Token_Checksum;
2705               Token := Token_Type'Val (Get_Name_Table_Byte (Token_Name));
2706            end if;
2707
2708            --  Keyword style checks
2709
2710            if Style_Check then
2711
2712               --  Deal with possible style check for non-lower case keyword,
2713               --  but we don't treat ACCESS, DELTA, DIGITS, RANGE as keywords
2714               --  for this purpose if they appear as attribute designators.
2715               --  Actually we only check the first character for speed.
2716
2717               --  Ada 2005 (AI-284): Do not apply the style check in case of
2718               --  "pragma Interface"
2719
2720               --  Ada 2005 (AI-340): Do not apply the style check in case of
2721               --  MOD attribute.
2722
2723               if Source (Token_Ptr) <= 'Z'
2724                 and then (Prev_Token /= Tok_Apostrophe
2725                           or else
2726                             (Token /= Tok_Access and then
2727                              Token /= Tok_Delta  and then
2728                              Token /= Tok_Digits and then
2729                              Token /= Tok_Mod    and then
2730                              Token /= Tok_Range))
2731                       and then (Token /= Tok_Interface
2732                                  or else
2733                                    (Token = Tok_Interface
2734                                      and then Prev_Token /= Tok_Pragma))
2735               then
2736                  Style.Non_Lower_Case_Keyword;
2737               end if;
2738
2739               --  Check THEN/ELSE style rules. These do not apply to AND THEN
2740               --  or OR ELSE, and do not apply in if expressions.
2741
2742               if (Token = Tok_Then and then Prev_Token /= Tok_And)
2743                    or else
2744                  (Token = Tok_Else and then Prev_Token /= Tok_Or)
2745               then
2746                  if Inside_If_Expression = 0 then
2747                     Style.Check_Separate_Stmt_Lines;
2748                  end if;
2749               end if;
2750            end if;
2751
2752            --  We must reset Token_Name since this is not an identifier and
2753            --  if we leave Token_Name set, the parser gets confused because
2754            --  it thinks it is dealing with an identifier instead of the
2755            --  corresponding keyword.
2756
2757            Token_Name := No_Name;
2758            return;
2759
2760         --  It is an identifier after all
2761
2762         else
2763            if Checksum_Accumulate_Token_Checksum then
2764               Accumulate_Token_Checksum;
2765            end if;
2766
2767            Post_Scan;
2768            return;
2769         end if;
2770   end Scan;
2771
2772   --------------------------
2773   -- Set_Comment_As_Token --
2774   --------------------------
2775
2776   procedure Set_Comment_As_Token (Value : Boolean) is
2777   begin
2778      Comment_Is_Token := Value;
2779   end Set_Comment_As_Token;
2780
2781   ------------------------------
2782   -- Set_End_Of_Line_As_Token --
2783   ------------------------------
2784
2785   procedure Set_End_Of_Line_As_Token (Value : Boolean) is
2786   begin
2787      End_Of_Line_Is_Token := Value;
2788   end Set_End_Of_Line_As_Token;
2789
2790   ---------------------------
2791   -- Set_Special_Character --
2792   ---------------------------
2793
2794   procedure Set_Special_Character (C : Character) is
2795   begin
2796      case C is
2797         when '#' | '$' | '_' | '?' | '@' | '`' | '\' | '^' | '~' =>
2798            Special_Characters (C) := True;
2799
2800         when others =>
2801            null;
2802      end case;
2803   end Set_Special_Character;
2804
2805   ----------------------
2806   -- Set_Start_Column --
2807   ----------------------
2808
2809   --  Note: it seems at first glance a little expensive to compute this value
2810   --  for every source line (since it is certainly not used for all source
2811   --  lines). On the other hand, it doesn't take much more work to skip past
2812   --  the initial white space on the line counting the columns than it would
2813   --  to scan past the white space using the standard scanning circuits.
2814
2815   function Set_Start_Column return Column_Number is
2816      Start_Column : Column_Number := 0;
2817
2818   begin
2819      --  Outer loop scans past horizontal tab characters
2820
2821      Tabs_Loop : loop
2822
2823         --  Inner loop scans past blanks as fast as possible, bumping Scan_Ptr
2824         --  past the blanks and adjusting Start_Column to account for them.
2825
2826         Blanks_Loop : loop
2827            if Source (Scan_Ptr) = ' ' then
2828               if Source (Scan_Ptr + 1) = ' ' then
2829                  if Source (Scan_Ptr + 2) = ' ' then
2830                     if Source (Scan_Ptr + 3) = ' ' then
2831                        if Source (Scan_Ptr + 4) = ' ' then
2832                           if Source (Scan_Ptr + 5) = ' ' then
2833                              if Source (Scan_Ptr + 6) = ' ' then
2834                                 Scan_Ptr := Scan_Ptr + 7;
2835                                 Start_Column := Start_Column + 7;
2836                              else
2837                                 Scan_Ptr := Scan_Ptr + 6;
2838                                 Start_Column := Start_Column + 6;
2839                                 exit Blanks_Loop;
2840                              end if;
2841                           else
2842                              Scan_Ptr := Scan_Ptr + 5;
2843                              Start_Column := Start_Column + 5;
2844                              exit Blanks_Loop;
2845                           end if;
2846                        else
2847                           Scan_Ptr := Scan_Ptr + 4;
2848                           Start_Column := Start_Column + 4;
2849                           exit Blanks_Loop;
2850                        end if;
2851                     else
2852                        Scan_Ptr := Scan_Ptr + 3;
2853                        Start_Column := Start_Column + 3;
2854                        exit Blanks_Loop;
2855                     end if;
2856                  else
2857                     Scan_Ptr := Scan_Ptr + 2;
2858                     Start_Column := Start_Column + 2;
2859                     exit Blanks_Loop;
2860                  end if;
2861               else
2862                  Scan_Ptr := Scan_Ptr + 1;
2863                  Start_Column := Start_Column + 1;
2864                  exit Blanks_Loop;
2865               end if;
2866            else
2867               exit Blanks_Loop;
2868            end if;
2869         end loop Blanks_Loop;
2870
2871         --  Outer loop keeps going only if a horizontal tab follows
2872
2873         if Source (Scan_Ptr) = HT then
2874            if Style_Check then
2875               Style.Check_HT;
2876            end if;
2877
2878            Scan_Ptr := Scan_Ptr + 1;
2879            Start_Column := (Start_Column / 8) * 8 + 8;
2880         else
2881            exit Tabs_Loop;
2882         end if;
2883      end loop Tabs_Loop;
2884
2885      return Start_Column;
2886
2887   --  A constraint error can happen only if we have a compiler with checks on
2888   --  and a line with a ludicrous number of tabs or spaces at the start. In
2889   --  such a case, we really don't care if Start_Column is right or not.
2890
2891   exception
2892      when Constraint_Error =>
2893         return Start_Column;
2894   end Set_Start_Column;
2895
2896end Scng;
2897