1 ------------------------------------------------------------------------------
3 -- GNAT COMPILER COMPONENTS --
9 -- Copyright (C) 1992-2003 Free Software Foundation, Inc. --
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 2, 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 COPYING. If not, write --
19 -- to the Free Software Foundation, 59 Temple Place - Suite 330, Boston, --
20 -- MA 02111-1307, USA. --
22 -- GNAT was originally developed by the GNAT team at New York University. --
23 -- Extensive contributions were provided by Ada Core Technologies Inc. --
25 ------------------------------------------------------------------------------
27 with Atree; use Atree;
28 with Casing; use Casing;
29 with Debug; use Debug;
30 with Elists; use Elists;
31 with Errout; use Errout;
32 with Fname; use Fname;
34 with Namet; use Namet;
35 with Nlists; use Nlists;
36 with Nmake; use Nmake;
38 with Output; use Output;
39 with Scans; use Scans;
41 with Sinput; use Sinput;
42 with Sinput.L; use Sinput.L;
43 with Sinfo; use Sinfo;
44 with Snames; use Snames;
47 with Tbuild; use Tbuild;
49 function Par (Configuration_Pragmas : Boolean) return List_Id is
51 Num_Library_Units : Natural := 0;
52 -- Count number of units parsed (relevant only in syntax check only mode,
53 -- since in semantics check mode only a single unit is permitted anyway)
55 Save_Config_Switches : Config_Switches_Type;
56 -- Variable used to save values of config switches while we parse the
57 -- new unit, to be restored on exit for proper recursive behavior.
59 Loop_Block_Count : Nat := 0;
60 -- Counter used for constructing loop/block names (see the routine
61 -- Par.Ch5.Get_Loop_Block_Name)
67 -- When an error is encountered, a call is made to one of the Error_Msg
68 -- routines to record the error. If the syntax scan is not derailed by the
69 -- error (e.g. a complaint that logical operators are inconsistent in an
70 -- EXPRESSION), then control returns from the Error_Msg call, and the
71 -- parse continues unimpeded.
73 -- If on the other hand, the Error_Msg represents a situation from which
74 -- the parser cannot recover locally, the exception Error_Resync is raised
75 -- immediately after the call to Error_Msg. Handlers for Error_Resync
76 -- are located at strategic points to resynchronize the parse. For example,
77 -- when an error occurs in a statement, the handler skips to the next
78 -- semicolon and continues the scan from there.
80 -- Each parsing procedure contains a note with the heading "Error recovery"
81 -- which shows if it can propagate the Error_Resync exception. In order
82 -- not to propagate the exception, a procedure must either contain its own
83 -- handler for this exception, or it must not call any other routines which
84 -- propagate the exception.
86 -- Note: the arrangement of Error_Resync handlers is such that it should
87 -- never be possible to transfer control through a procedure which made
88 -- an entry in the scope stack, invalidating the contents of the stack.
90 Error_Resync : exception;
91 -- Exception raised on error that is not handled locally, see above.
93 Last_Resync_Point : Source_Ptr;
94 -- The resynchronization routines in Par.Sync run a risk of getting
95 -- stuck in an infinite loop if they do not skip a token, and the caller
96 -- keeps repeating the same resync call. On the other hand, if they skip
97 -- a token unconditionally, some recovery opportunities are missed. The
98 -- variable Last_Resync_Point records the token location previously set
99 -- by a Resync call, and if a subsequent Resync call occurs at the same
100 -- location, then the Resync routine does guarantee to skip a token.
102 --------------------------------------------
103 -- Handling Semicolon Used in Place of IS --
104 --------------------------------------------
106 -- The following global variables are used in handling the error situation
107 -- of using a semicolon in place of IS in a subprogram declaration as in:
109 -- procedure X (Y : Integer);
115 -- The two contexts in which this can appear are at the outer level, and
116 -- within a declarative region. At the outer level, we know something is
117 -- wrong as soon as we see the Q (or begin, if there are no declarations),
118 -- and we can immediately decide that the semicolon should have been IS.
120 -- The situation in a declarative region is more complex. The declaration
121 -- of Q could belong to the outer region, and we do not know that we have
122 -- an error until we hit the begin. It is still not clear at this point
123 -- from a syntactic point of view that something is wrong, because the
124 -- begin could belong to the enclosing subprogram or package. However, we
125 -- can incorporate a bit of semantic knowledge and note that the body of
126 -- X is missing, so we definitely DO have an error. We diagnose this error
127 -- as semicolon in place of IS on the subprogram line.
129 -- There are two styles for this diagnostic. If the begin immediately
130 -- follows the semicolon, then we can place a flag (IS expected) right
131 -- on the semicolon. Otherwise we do not detect the error until we hit
132 -- the begin which refers back to the line with the semicolon.
134 -- To control the process in the second case, the following global
135 -- variables are set to indicate that we have a subprogram declaration
136 -- whose body is required and has not yet been found. The prefix SIS
137 -- stands for "Subprogram IS" handling.
139 SIS_Entry_Active : Boolean;
140 -- Set True to indicate that an entry is active (i.e. that a subprogram
141 -- declaration has been encountered, and no body for this subprogram has
142 -- been encountered). The remaining fields are valid only if this is True.
145 -- Subprogram designator
147 SIS_Sloc : Source_Ptr;
148 -- Source location of FUNCTION/PROCEDURE keyword
150 SIS_Ecol : Column_Number;
151 -- Column number of FUNCTION/PROCEDURE keyword
153 SIS_Semicolon_Sloc : Source_Ptr;
154 -- Source location of semicolon at end of subprogram declaration
156 SIS_Declaration_Node : Node_Id;
157 -- Pointer to tree node for subprogram declaration
159 SIS_Missing_Semicolon_Message : Error_Msg_Id;
160 -- Used to save message ID of missing semicolon message (which will be
161 -- modified to missing IS if necessary). Set to No_Error_Msg in the
162 -- normal (non-error) case.
164 -- Five things can happen to an active SIS entry
166 -- 1. If a BEGIN is encountered with an SIS entry active, then we have
167 -- exactly the situation in which we know the body of the subprogram is
168 -- missing. After posting an error message, we change the spec to a body,
169 -- rechaining the declarations that intervened between the spec and BEGIN.
171 -- 2. Another subprogram declaration or body is encountered. In this
172 -- case the entry gets overwritten with the information for the new
173 -- subprogram declaration. We don't catch some nested cases this way,
174 -- but it doesn't seem worth the effort.
176 -- 3. A nested declarative region (e.g. package declaration or package
177 -- body) is encountered. The SIS active indication is reset at the start
178 -- of such a nested region. Again, like case 2, this causes us to miss
179 -- some nested cases, but it doesn't seen worth the effort to stack and
180 -- unstack the SIS information. Maybe we will reconsider this if we ever
181 -- get a complaint about a missed case :-)
183 -- 4. We encounter a valid pragma INTERFACE or IMPORT that effectively
184 -- supplies the missing body. In this case we reset the entry.
186 -- 5. We encounter the end of the declarative region without encoutering
187 -- a BEGIN first. In this situation we simply reset the entry. We know
188 -- that there is a missing body, but it seems more reasonable to let the
189 -- later semantic checking discover this.
191 ----------------------------------------------------
192 -- Handling of Reserved Words Used as Identifiers --
193 ----------------------------------------------------
195 -- Note: throughout the parser, the terms reserved word and keyword
196 -- are used interchangably to refer to the same set of reserved
197 -- keywords (including until, protected, etc).
199 -- If a reserved word is used in place of an identifier, the parser
200 -- where possible tries to recover gracefully. In particular, if the
201 -- keyword is clearly spelled using identifier casing, e.g. Until in
202 -- a source program using mixed case identifiers and lower case keywords,
203 -- then the keyword is treated as an identifier if it appears in a place
204 -- where an identifier is required.
206 -- The situation is more complex if the keyword is spelled with normal
207 -- keyword casing. In this case, the parser is more reluctant to
208 -- consider it to be intended as an identifier, unless it has some
209 -- further confirmation.
211 -- In the case of an identifier appearing in the identifier list of a
212 -- declaration, the appearence of a comma or colon right after the
213 -- keyword on the same line is taken as confirmation. For an enumeration
214 -- literal, a comma or right paren right after the identifier is also
215 -- treated as adequate confirmation.
217 -- The following type is used in calls to Is_Reserved_Identifier and
218 -- also to P_Defining_Identifier and P_Identifier. The default for all
219 -- these functins is that reserved words in reserved word case are not
220 -- considered to be reserved identifiers. The Id_Check value indicates
221 -- tokens, which if they appear immediately after the identifier, are
222 -- taken as confirming that the use of an identifier was expected
226 -- Default, no special token test
229 -- Consider as identifier if followed by comma or right paren
232 -- Consider as identifier if followed by comma or colon
235 -- Consider as identifier if followed by DO
238 -- Consider as identifier if followed by period
241 -- Consider as identifier if followed by >>
244 -- Consider as identifier if followed by IN
247 -- Consider as identifier if followed by IS
249 C_Left_Paren_Semicolon,
250 -- Consider as identifier if followed by left paren or semicolon
253 -- Consider as identifier if followed by USE
255 C_Vertical_Bar_Arrow);
256 -- Consider as identifier if followed by | or =>
258 --------------------------------------------
259 -- Handling IS Used in Place of Semicolon --
260 --------------------------------------------
262 -- This is a somewhat trickier situation, and we can't catch it in all
263 -- cases, but we do our best to detect common situations resulting from
264 -- a "cut and paste" operation which forgets to change the IS to semicolon.
265 -- Consider the following example:
280 -- The trouble is that the section of text from PROCEDURE B through END;
281 -- consitutes a valid procedure body, and the danger is that we find out
282 -- far too late that something is wrong (indeed most compilers will behave
283 -- uncomfortably on the above example).
285 -- We have two approaches to helping to control this situation. First we
286 -- make every attempt to avoid swallowing the last END; if we can be
287 -- sure that some error will result from doing so. In particular, we won't
288 -- accept the END; unless it is exactly correct (in particular it must not
289 -- have incorrect name tokens), and we won't accept it if it is immediately
290 -- followed by end of file, WITH or SEPARATE (all tokens that unmistakeably
291 -- signal the start of a compilation unit, and which therefore allow us to
292 -- reserve the END; for the outer level.) For more details on this aspect
293 -- of the handling, see package Par.Endh.
295 -- If we can avoid eating up the END; then the result in the absense of
296 -- any additional steps would be to post a missing END referring back to
297 -- the subprogram with the bogus IS. Similarly, if the enclosing package
298 -- has no BEGIN, then the result is a missing BEGIN message, which again
299 -- refers back to the subprogram header.
301 -- Such an error message is not too bad (it's already a big improvement
302 -- over what many parsers do), but it's not ideal, because the declarations
303 -- following the IS have been absorbed into the wrong scope. In the above
304 -- case, this could result for example in a bogus complaint that the body
305 -- of D was missing from the package.
307 -- To catch at least some of these cases, we take the following additional
308 -- steps. First, a subprogram body is marked as having a suspicious IS if
309 -- the declaration line is followed by a line which starts with a symbol
310 -- that can start a declaration in the same column, or to the left of the
311 -- column in which the FUNCTION or PROCEDURE starts (normal style is to
312 -- indent any declarations which really belong a subprogram). If such a
313 -- subprogram encounters a missing BEGIN or missing END, then we decide
314 -- that the IS should have been a semicolon, and the subprogram body node
315 -- is marked (by setting the Bad_Is_Detected flag true. Note that we do
316 -- not do this for library level procedures, only for nested procedures,
317 -- since for library level procedures, we must have a body.
319 -- The processing for a declarative part checks to see if the last
320 -- declaration scanned is marked in this way, and if it is, the tree
321 -- is modified to reflect the IS being interpreted as a semicolon.
323 ---------------------------------------------------
324 -- Parser Type Definitions and Control Variables --
325 ---------------------------------------------------
327 -- The following variable and associated type declaration are used by the
328 -- expression parsing routines to return more detailed information about
329 -- the categorization of a parsed expression.
331 type Expr_Form_Type is (
332 EF_Simple_Name, -- Simple name, i.e. possibly qualified identifier
333 EF_Name, -- Simple expression which could also be a name
334 EF_Simple, -- Simple expression which is not call or name
335 EF_Range_Attr, -- Range attribute reference
336 EF_Non_Simple); -- Expression that is not a simple expression
338 Expr_Form : Expr_Form_Type;
340 -- The following type is used for calls to P_Subprogram, P_Package, P_Task,
341 -- P_Protected to indicate which of several possibilities is acceptable.
343 type Pf_Rec is record
344 Spcn : Boolean; -- True if specification OK
345 Decl : Boolean; -- True if declaration OK
346 Gins : Boolean; -- True if generic instantiation OK
347 Pbod : Boolean; -- True if proper body OK
348 Rnam : Boolean; -- True if renaming declaration OK
349 Stub : Boolean; -- True if body stub OK
350 Fil1 : Boolean; -- Filler to fill to 8 bits
351 Fil2 : Boolean; -- Filler to fill to 8 bits
353 pragma Pack (Pf_Rec);
355 function T return Boolean renames True;
356 function F return Boolean renames False;
358 Pf_Decl_Gins_Pbod_Rnam_Stub : constant Pf_Rec :=
359 Pf_Rec'(F, T, T, T, T, T, F, F);
360 Pf_Decl : constant Pf_Rec :=
361 Pf_Rec'(F, T, F, F, F, F, F, F);
362 Pf_Decl_Gins_Pbod_Rnam : constant Pf_Rec :=
363 Pf_Rec'(F, T, T, T, T, F, F, F);
364 Pf_Decl_Pbod : constant Pf_Rec :=
365 Pf_Rec'(F, T, F, T, F, F, F, F);
366 Pf_Pbod : constant Pf_Rec :=
367 Pf_Rec'(F, F, F, T, F, F, F, F);
368 Pf_Spcn : constant Pf_Rec :=
369 Pf_Rec'(T, F, F, F, F, F, F, F);
370 -- The above are the only allowed values of Pf_Rec arguments
372 type SS_Rec is record
373 Eftm : Boolean; -- ELSIF can terminate sequence
374 Eltm : Boolean; -- ELSE can terminate sequence
375 Extm : Boolean; -- EXCEPTION can terminate sequence
376 Ortm : Boolean; -- OR can terminate sequence
377 Sreq : Boolean; -- at least one statement required
378 Tatm : Boolean; -- THEN ABORT can terminate sequence
379 Whtm : Boolean; -- WHEN can terminate sequence
380 Unco : Boolean; -- Unconditional terminate after one statement
382 pragma Pack (SS_Rec);
384 SS_Eftm_Eltm_Sreq : constant SS_Rec := SS_Rec'(T, T, F, F, T, F, F, F);
385 SS_Eltm_Ortm_Tatm : constant SS_Rec := SS_Rec'(F, T, F, T, F, T, F, F);
386 SS_Extm_Sreq : constant SS_Rec := SS_Rec'(F, F, T, F, T, F, F, F);
387 SS_None : constant SS_Rec := SS_Rec'(F, F, F, F, F, F, F, F);
388 SS_Ortm_Sreq : constant SS_Rec := SS_Rec'(F, F, F, T, T, F, F, F);
389 SS_Sreq : constant SS_Rec := SS_Rec'(F, F, F, F, T, F, F, F);
390 SS_Sreq_Whtm : constant SS_Rec := SS_Rec'(F, F, F, F, T, F, T, F);
391 SS_Whtm : constant SS_Rec := SS_Rec'(F, F, F, F, F, F, T, F);
392 SS_Unco : constant SS_Rec := SS_Rec'(F, F, F, F, F, F, F, T);
394 Label_List : Elist_Id;
395 -- List of label nodes for labels appearing in the current compilation.
396 -- Used by Par.Labl to construct the corresponding implicit declarations.
402 -- The scope table, also referred to as the scope stack, is used to
403 -- record the current scope context. It is organized as a stack, with
404 -- inner nested entries corresponding to higher entries on the stack.
405 -- An entry is made when the parser encounters the opening of a nested
406 -- construct (such as a record, task, package etc.), and then package
407 -- Par.Endh uses this stack to deal with END lines (including properly
408 -- dealing with END nesting errors).
411 -- Type of end entry required for this scope. The last two entries are
412 -- used only in the subprogram body case to mark the case of a suspicious
413 -- IS, or a bad IS (i.e. suspicions confirmed by missing BEGIN or END).
414 -- See separate section on dealing with IS used in place of semicolon.
415 -- Note that for many purposes E_Name, E_Suspicious_Is and E_Bad_Is are
416 -- treated the same (E_Suspicious_Is and E_Bad_Is are simply special cases
417 -- of E_Name). They are placed at the end of the enumeration so that a
418 -- test for >= E_Name catches all three cases efficiently.
420 (E_Dummy, -- dummy entry at outer level
424 E_Record, -- END RECORD;
425 E_Select, -- END SELECT;
426 E_Name, -- END [name];
427 E_Suspicious_Is, -- END [name]; (case of suspicious IS)
428 E_Bad_Is); -- END [name]; (case of bad IS)
430 -- The following describes a single entry in the scope table
432 type Scope_Table_Entry is record
434 -- Type of end entry, as per above description
437 -- A flag indicating whether the label, if present, is required to
438 -- appear on the end line. It is referenced only in the case of
439 -- Etyp = E_Name or E_Suspicious_Is where the name may or may not be
440 -- required (yes for labeled block, no in other cases). Note that for
441 -- all cases except begin, the question of whether a label is required
442 -- can be determined from the other fields (for loop, it is required if
443 -- it is present, and for the other constructs it is never required or
446 Ecol : Column_Number;
447 -- Contains the absolute column number (with tabs expanded) of the
448 -- the expected column of the end assuming normal Ada indentation
449 -- usage. If the RM_Column_Check mode is set, this value is used for
450 -- generating error messages about indentation. Otherwise it is used
451 -- only to control heuristic error recovery actions.
454 -- This field is used only for the LOOP and BEGIN cases, and is the
455 -- Node_Id value of the label name. For all cases except child units,
456 -- this value is an entity whose Chars field contains the name pointer
457 -- that identifies the label uniquely. For the child unit case the Labl
458 -- field references an N_Defining_Program_Unit_Name node for the name.
459 -- For cases other than LOOP or BEGIN, the Label field is set to Error,
460 -- indicating that it is an error to have a label on the end line.
461 -- (this is really a misuse of Error since there is no Error ???)
464 -- Points to the list of declarations (i.e. the declarative part)
465 -- associated with this construct. It is set only in the END [name]
466 -- cases, and is set to No_List for all other cases which do not have a
467 -- declarative unit associated with them. This is used for determining
468 -- the proper location for implicit label declarations.
471 -- Empty except in the case of entries for IF and CASE statements,
472 -- in which case it contains the N_If_Statement or N_Case_Statement
473 -- node. This is used for setting the End_Span field.
476 -- Source location of the opening token of the construct. This is
477 -- used to refer back to this line in error messages (such as missing
478 -- or incorrect end lines). The Sloc field is not used, and is not set,
479 -- if a label is present (the Labl field provides the text name of the
480 -- label in this case, which is fine for error messages).
483 -- S_Is is relevant only if Etyp is set to E_Suspicious_Is or
484 -- E_Bad_Is. It records the location of the IS that is considered
488 -- A boolean flag that is set true if the opening entry is the dubious
489 -- result of some prior error, e.g. a record entry where the record
490 -- keyword was missing. It is used to suppress the issuing of a
491 -- corresponding junk complaint about the end line (we do not want
492 -- to complain about a missing end record when there was no record).
495 -- The following declares the scope table itself. The Last field is the
496 -- stack pointer, so that Scope.Table (Scope.Last) is the top entry. The
497 -- oldest entry, at Scope_Stack (0), is a dummy entry with Etyp set to
498 -- E_Dummy, and the other fields undefined. This dummy entry ensures that
499 -- Scope_Stack (Scope_Stack_Ptr).Etyp can always be tested, and that the
500 -- scope stack pointer is always in range.
502 package Scope is new Table.Table (
503 Table_Component_Type => Scope_Table_Entry,
504 Table_Index_Type => Int,
505 Table_Low_Bound => 0,
507 Table_Increment => 100,
508 Table_Name => "Scope");
510 ---------------------------------
511 -- Parsing Routines by Chapter --
512 ---------------------------------
514 -- Uncommented declarations in this section simply parse the construct
515 -- corresponding to their name, and return an ID value for the Node or
516 -- List that is created.
519 function P_Pragma return Node_Id;
521 function P_Identifier (C : Id_Check := None) return Node_Id;
522 -- Scans out an identifier. The parameter C determines the treatment
523 -- of reserved identifiers. See declaration of Id_Check for details.
525 function P_Pragmas_Opt return List_Id;
526 -- This function scans for a sequence of pragmas in other than a
527 -- declaration sequence or statement sequence context. All pragmas
528 -- can appear except pragmas Assert and Debug, which are only allowed
529 -- in a declaration or statement sequence context.
531 procedure P_Pragmas_Misplaced;
532 -- Skips misplaced pragmas with a complaint
534 procedure P_Pragmas_Opt (List : List_Id);
535 -- Parses optional pragmas and appends them to the List
539 Missing_Begin_Msg : Error_Msg_Id;
540 -- This variable is set by a call to P_Declarative_Part. Normally it
541 -- is set to No_Error_Msg, indicating that no special processing is
542 -- required by the caller. The special case arises when a statement
543 -- is found in the sequence of declarations. In this case the Id of
544 -- the message issued ("declaration expected") is preserved in this
545 -- variable, then the caller can change it to an appropriate missing
546 -- begin message if indeed the BEGIN is missing.
548 function P_Access_Definition return Node_Id;
549 function P_Access_Type_Definition return Node_Id;
550 function P_Array_Type_Definition return Node_Id;
551 function P_Basic_Declarative_Items return List_Id;
552 function P_Constraint_Opt return Node_Id;
553 function P_Declarative_Part return List_Id;
554 function P_Discrete_Choice_List return List_Id;
555 function P_Discrete_Range return Node_Id;
556 function P_Discrete_Subtype_Definition return Node_Id;
557 function P_Known_Discriminant_Part_Opt return List_Id;
558 function P_Signed_Integer_Type_Definition return Node_Id;
559 function P_Range return Node_Id;
560 function P_Range_Or_Subtype_Mark return Node_Id;
561 function P_Range_Constraint return Node_Id;
562 function P_Record_Definition return Node_Id;
563 function P_Subtype_Indication return Node_Id;
564 function P_Subtype_Mark return Node_Id;
565 function P_Subtype_Mark_Resync return Node_Id;
566 function P_Unknown_Discriminant_Part_Opt return Boolean;
568 procedure P_Component_Items (Decls : List_Id);
569 -- Scan out one or more component items and append them to the
570 -- given list. Only scans out more than one declaration in the
571 -- case where the source has a single declaration with multiple
572 -- defining identifiers.
574 function P_Defining_Identifier (C : Id_Check := None) return Node_Id;
575 -- Scan out a defining identifier. The parameter C controls the
576 -- treatment of errors in case a reserved word is scanned. See the
577 -- declaration of this type for details.
579 function Init_Expr_Opt (P : Boolean := False) return Node_Id;
580 -- If an initialization expression is present (:= expression), then
581 -- it is scanned out and returned, otherwise Empty is returned if no
582 -- initialization expression is present. This procedure also handles
583 -- certain common error cases cleanly. The parameter P indicates if
584 -- a right paren can follow the expression (default = no right paren
587 procedure Skip_Declaration (S : List_Id);
588 -- Used when scanning statements to skip past a mispaced declaration
589 -- The declaration is scanned out and appended to the given list.
590 -- Token is known to be a declaration token (in Token_Class_Declk)
591 -- on entry, so there definition is a declaration to be scanned.
593 function P_Subtype_Indication (Subtype_Mark : Node_Id) return Node_Id;
594 -- This version of P_Subtype_Indication is called when the caller has
595 -- already scanned out the subtype mark which is passed as a parameter.
597 function P_Subtype_Mark_Attribute (Type_Node : Node_Id) return Node_Id;
598 -- Parse a subtype mark attribute. The caller has already parsed the
599 -- subtype mark, which is passed in as the argument, and has checked
600 -- that the current token is apostrophe.
605 function P_Aggregate return Node_Id;
606 function P_Expression return Node_Id;
607 function P_Expression_No_Right_Paren return Node_Id;
608 function P_Expression_Or_Range_Attribute return Node_Id;
609 function P_Function_Name return Node_Id;
610 function P_Name return Node_Id;
611 function P_Qualified_Simple_Name return Node_Id;
612 function P_Qualified_Simple_Name_Resync return Node_Id;
613 function P_Simple_Expression return Node_Id;
614 function P_Simple_Expression_Or_Range_Attribute return Node_Id;
616 function P_Qualified_Expression
617 (Subtype_Mark : Node_Id)
619 -- This routine scans out a qualified expression when the caller has
620 -- already scanned out the name and apostrophe of the construct.
626 function P_Statement_Name (Name_Node : Node_Id) return Node_Id;
627 -- Given a node representing a name (which is a call), converts it
628 -- to the syntactically corresponding procedure call statement.
630 function P_Sequence_Of_Statements (SS_Flags : SS_Rec) return List_Id;
631 -- The argument indicates the acceptable termination tokens.
632 -- See body in Par.Ch5 for details of the use of this parameter.
634 procedure Parse_Decls_Begin_End (Parent : Node_Id);
635 -- Parses declarations and handled statement sequence, setting
636 -- fields of Parent node appropriately.
641 function P_Designator return Node_Id;
642 function P_Defining_Program_Unit_Name return Node_Id;
643 function P_Formal_Part return List_Id;
644 function P_Parameter_Profile return List_Id;
645 function P_Return_Statement return Node_Id;
646 function P_Subprogram_Specification return Node_Id;
648 procedure P_Mode (Node : Node_Id);
649 -- Sets In_Present and/or Out_Present flags in Node scanning past
650 -- IN, OUT or IN OUT tokens in the source.
652 function P_Subprogram (Pf_Flags : Pf_Rec) return Node_Id;
653 -- Scans out any construct starting with either of the keywords
654 -- PROCEDURE or FUNCTION. The parameter indicates which possible
655 -- possible kinds of construct (body, spec, instantiation etc.)
656 -- are permissible in the current context.
661 function P_Package (Pf_Flags : Pf_Rec) return Node_Id;
662 -- Scans out any construct starting with the keyword PACKAGE. The
663 -- parameter indicates which possible kinds of construct (body, spec,
664 -- instantiation etc.) are permissible in the current context.
668 function P_Use_Clause return Node_Id;
672 function P_Abort_Statement return Node_Id;
673 function P_Abortable_Part return Node_Id;
674 function P_Accept_Statement return Node_Id;
675 function P_Delay_Statement return Node_Id;
676 function P_Entry_Body return Node_Id;
677 function P_Protected return Node_Id;
678 function P_Requeue_Statement return Node_Id;
679 function P_Select_Statement return Node_Id;
680 function P_Task return Node_Id;
681 function P_Terminate_Alternative return Node_Id;
685 function P_Compilation_Unit return Node_Id;
686 -- Note: this function scans a single compilation unit, and
687 -- checks that an end of file follows this unit, diagnosing
688 -- any unexpected input as an error, and then skipping it, so
689 -- that Token is set to Tok_EOF on return. An exception is in
690 -- syntax-only mode, where multiple compilation units are
691 -- permitted. In this case, P_Compilation_Unit does not check
692 -- for end of file and there may be more compilation units to
693 -- scan. The caller can uniquely detect this situation by the
694 -- fact that Token is not set to Tok_EOF on return.
698 function P_Handled_Sequence_Of_Statements return Node_Id;
699 function P_Raise_Statement return Node_Id;
701 function Parse_Exception_Handlers return List_Id;
702 -- Parses the partial construct EXCEPTION followed by a list of
703 -- exception handlers which appears in a number of productions,
704 -- and returns the list of exception handlers.
709 function P_Generic return Node_Id;
710 function P_Generic_Actual_Part_Opt return List_Id;
714 function P_Representation_Clause return Node_Id;
716 function P_Code_Statement (Subtype_Mark : Node_Id) return Node_Id;
717 -- Function to parse a code statement. The caller has scanned out
718 -- the name to be used as the subtype mark (but has not checked that
719 -- it is suitable for use as a subtype mark, i.e. is either an
720 -- identifier or a selected component). The current token is an
721 -- apostrophe and the following token is either a left paren or
722 -- RANGE (the latter being an error to be caught by P_Code_Statement.
725 -- Note: the parsing for annexe J features (i.e. obsolescent features)
726 -- is found in the logical section where these features would be if
727 -- they were not obsolescent. In particular:
729 -- Delta constraint is parsed by P_Delta_Constraint (3.5.9)
730 -- At clause is parsed by P_At_Clause (13.1)
731 -- Mod clause is parsed by P_Mod_Clause (13.5.1)
737 -- Routines for handling end lines, including scope recovery
741 function Check_End return Boolean;
742 -- Called when an end sequence is required. In the absence of an error
743 -- situation, Token contains Tok_End on entry, but in a missing end
744 -- case, this may not be the case. Pop_End_Context is used to determine
745 -- the appropriate action to be taken. The returned result is True if
746 -- an End sequence was encountered and False if no End sequence was
747 -- present. This occurs if the END keyword encountered was determined
748 -- to be improper and deleted (i.e. Pop_End_Context set End_Action to
749 -- Skip_And_Reject). Note that the END sequence includes a semicolon,
750 -- except in the case of END RECORD, where a semicolon follows the END
751 -- RECORD, but is not part of the record type definition itself.
754 -- Skip past an end sequence. On entry Token contains Tok_End, and we
755 -- we know that the end sequence is syntactically incorrect, and that
756 -- an appropriate error message has already been posted. The mission
757 -- is simply to position the scan pointer to be the best guess of the
758 -- position after the end sequence. We do not issue any additional
759 -- error messages while carrying this out.
761 procedure End_Statements (Parent : Node_Id := Empty);
762 -- Called when an end is required or expected to terminate a sequence
763 -- of statements. The caller has already made an appropriate entry in
764 -- the Scope.Table to describe the expected form of the end. This can
765 -- only be used in cases where the only appropriate terminator is end.
766 -- If Parent is non-empty, then if a correct END line is encountered,
767 -- the End_Label field of Parent is set appropriately.
771 ------------------------------------
772 -- Resynchronization After Errors --
773 ------------------------------------
775 -- These procedures are used to resynchronize after errors. Following an
776 -- error which is not immediately locally recoverable, the exception
777 -- Error_Resync is raised. The handler for Error_Resync typically calls
778 -- one of these recovery procedures to resynchronize the source position
779 -- to a point from which parsing can be restarted.
781 -- Note: these procedures output an information message that tokens are
782 -- being skipped, but this message is output only if the option for
783 -- Multiple_Errors_Per_Line is set in Options.
787 procedure Resync_Choice;
788 -- Used if an error occurs scanning a choice. The scan pointer is
789 -- advanced to the next vertical bar, arrow, or semicolon, whichever
790 -- comes first. We also quit if we encounter an end of file.
792 procedure Resync_Expression;
793 -- Used if an error is detected during the parsing of an expression.
794 -- It skips past tokens until either a token which cannot be part of
795 -- an expression is encountered (an expression terminator), or if a
796 -- comma or right parenthesis or vertical bar is encountered at the
797 -- current parenthesis level (a parenthesis level counter is maintained
798 -- to carry out this test).
800 procedure Resync_Past_Semicolon;
801 -- Used if an error occurs while scanning a sequence of declarations.
802 -- The scan pointer is positioned past the next semicolon and the scan
803 -- resumes. The scan is also resumed on encountering a token which
804 -- starts a declaration (but we make sure to skip at least one token
805 -- in this case, to avoid getting stuck in a loop).
807 procedure Resync_To_Semicolon;
808 -- Similar to Resync_Past_Semicolon, except that the scan pointer is
809 -- left pointing to the semicolon rather than past it.
811 procedure Resync_Past_Semicolon_Or_To_Loop_Or_Then;
812 -- Used if an error occurs while scanning a sequence of statements.
813 -- The scan pointer is positioned past the next semicolon, or to the
814 -- next occurrence of either then or loop, and the scan resumes.
816 procedure Resync_To_When;
817 -- Used when an error occurs scanning an entry index specification.
818 -- The scan pointer is positioned to the next WHEN (or to IS or
819 -- semicolon if either of these appear before WHEN, indicating
820 -- another error has occurred).
822 procedure Resync_Semicolon_List;
823 -- Used if an error occurs while scanning a parenthesized list of items
824 -- separated by semicolons. The scan pointer is advanced to the next
825 -- semicolon or right parenthesis at the outer parenthesis level, or
826 -- to the next is or RETURN keyword occurence, whichever comes first.
828 procedure Resync_Cunit;
829 -- Synchronize to next token which could be the start of a compilation
830 -- unit, or to the end of file token.
834 -------------------------
835 -- Token Scan Routines --
836 -------------------------
838 -- Routines to check for expected tokens
842 -- Procedures with names of the form T_xxx, where Tok_xxx is a token
843 -- name, check that the current token matches the required token, and
844 -- if so, scan past it. If not, an error is issued indicating that
845 -- the required token is not present (xxx expected). In most cases, the
846 -- scan pointer is not moved in the not-found case, but there are some
847 -- exceptions to this, see for example T_Id, where the scan pointer is
848 -- moved across a literal appearing where an identifier is expected.
856 procedure T_Colon_Equal;
860 procedure T_Greater_Greater;
861 procedure T_Identifier;
864 procedure T_Left_Paren;
873 procedure T_Right_Paren;
874 procedure T_Semicolon;
881 -- Procedures have names of the form TF_xxx, where Tok_xxx is a token
882 -- name check that the current token matches the required token, and
883 -- if so, scan past it. If not, an error message is issued indicating
884 -- that the required token is not present (xxx expected).
886 -- If the missing token is at the end of the line, then control returns
887 -- immediately after posting the message. If there are remaining tokens
888 -- on the current line, a search is conducted to see if the token
889 -- appears later on the current line, as follows:
891 -- A call to Scan_Save is issued and a forward search for the token
892 -- is carried out. If the token is found on the current line before a
893 -- semicolon, then it is scanned out and the scan continues from that
894 -- point. If not the scan is restored to the point where it was missing.
900 procedure TF_Semicolon;
906 ----------------------
907 -- Utility Routines --
908 ----------------------
912 function Bad_Spelling_Of (T : Token_Type) return Boolean;
913 -- This function is called in an error situation. It checks if the
914 -- current token is an identifier whose name is a plausible bad
915 -- spelling of the given keyword token, and if so, issues an error
916 -- message, sets Token from T, and returns True. Otherwise Token is
917 -- unchanged, and False is returned.
919 procedure Check_Bad_Layout;
920 -- Check for bad indentation in RM checking mode. Used for statements
921 -- and declarations. Checks if current token is at start of line and
922 -- is exdented from the current expected end column, and if so an
923 -- error message is generated.
925 procedure Check_Misspelling_Of (T : Token_Type);
926 pragma Inline (Check_Misspelling_Of);
927 -- This is similar to the function above, except that it does not
928 -- return a result. It is typically used in a situation where any
929 -- identifier is an error, and it makes sense to simply convert it
930 -- to the given token if it is a plausible misspelling of it.
932 procedure Check_95_Keyword (Token_95, Next : Token_Type);
933 -- This routine checks if the token after the current one matches the
934 -- Next argument. If so, the scan is backed up to the current token
935 -- and Token_Type is changed to Token_95 after issuing an appropriate
936 -- error message ("(Ada 83) keyword xx cannot be used"). If not,
937 -- the scan is backed up with Token_Type unchanged. This routine
938 -- is used to deal with an attempt to use a 95 keyword in Ada 83
939 -- mode. The caller has typically checked that the current token,
940 -- an identifier, matches one of the 95 keywords.
942 procedure Check_Simple_Expression (E : Node_Id);
943 -- Given an expression E, that has just been scanned, so that Expr_Form
944 -- is still set, outputs an error if E is a non-simple expression. E is
945 -- not modified by this call.
947 procedure Check_Simple_Expression_In_Ada_83 (E : Node_Id);
948 -- Like Check_Simple_Expression, except that the error message is only
949 -- given when operating in Ada 83 mode, and includes "in Ada 83".
951 function Check_Subtype_Mark (Mark : Node_Id) return Node_Id;
952 -- Called to check that a node representing a name (or call) is
953 -- suitable for a subtype mark, i.e, that it is an identifier or
954 -- a selected component. If so, or if it is already Error, then
955 -- it is returned unchanged. Otherwise an error message is issued
956 -- and Error is returned.
958 function Comma_Present return Boolean;
959 -- Used in comma delimited lists to determine if a comma is present, or
960 -- can reasonably be assumed to have been present (an error message is
961 -- generated in the latter case). If True is returned, the scan has been
962 -- positioned past the comma. If False is returned, the scan position
963 -- is unchanged. Note that all comma-delimited lists are terminated by
964 -- a right paren, so the only legitimate tokens when Comma_Present is
965 -- called are right paren and comma. If some other token is found, then
966 -- Comma_Present has the job of deciding whether it is better to pretend
967 -- a comma was present, post a message for a missing comma and return
968 -- True, or return False and let the caller diagnose the missing right
971 procedure Discard_Junk_Node (N : Node_Id);
972 procedure Discard_Junk_List (L : List_Id);
973 pragma Inline (Discard_Junk_Node);
974 pragma Inline (Discard_Junk_List);
975 -- These procedures do nothing at all, their effect is simply to discard
976 -- the argument. A typical use is to skip by some junk that is not
977 -- expected in the current context.
979 procedure Ignore (T : Token_Type);
980 -- If current token matches T, then give an error message and skip
981 -- past it, otherwise the call has no effect at all. T may be any
982 -- reserved word token, or comma, left or right paren, or semicolon.
984 function Is_Reserved_Identifier (C : Id_Check := None) return Boolean;
985 -- Test if current token is a reserved identifier. This test is based
986 -- on the token being a keyword and being spelled in typical identifier
987 -- style (i.e. starting with an upper case letter). The parameter C
988 -- determines the special treatment if a reserved word is encountered
989 -- that has the normal casing of a reserved word.
991 procedure Merge_Identifier (Prev : Node_Id; Nxt : Token_Type);
992 -- Called when the previous token is an identifier (whose Token_Node
993 -- value is given by Prev) to check if current token is an identifier
994 -- that can be merged with the previous one adding an underscore. The
995 -- merge is only attempted if the following token matches Nxt. If all
996 -- conditions are met, an error message is issued, and the merge is
997 -- carried out, modifying the Chars field of Prev.
999 procedure No_Constraint;
1000 -- Called in a place where no constraint is allowed, but one might
1001 -- appear due to a common error (e.g. after the type mark in a procedure
1002 -- parameter. If a constraint is present, an error message is posted,
1003 -- and the constraint is scanned and discarded.
1005 function No_Right_Paren (Expr : Node_Id) return Node_Id;
1006 -- Function to check for no right paren at end of expression, returns
1007 -- its argument if no right paren, else flags paren and returns Error.
1009 procedure Push_Scope_Stack;
1010 pragma Inline (Push_Scope_Stack);
1011 -- Push a new entry onto the scope stack. Scope.Last (the stack pointer)
1012 -- is incremented. The Junk field is preinitialized to False. The caller
1013 -- is expected to fill in all remaining entries of the new new top stack
1014 -- entry at Scope.Table (Scope.Last).
1016 procedure Pop_Scope_Stack;
1017 -- Pop an entry off the top of the scope stack. Scope_Last (the scope
1018 -- table stack pointer) is decremented by one. It is a fatal error to
1019 -- try to pop off the dummy entry at the bottom of the stack (i.e.
1020 -- Scope.Last must be non-zero at the time of call).
1022 function Separate_Present return Boolean;
1023 -- Determines if the current token is either Tok_Separate, or an
1024 -- identifier that is a possible misspelling of "separate" followed
1025 -- by a semicolon. True is returned if so, otherwise False.
1027 procedure Signal_Bad_Attribute;
1028 -- The current token is an identifier that is supposed to be an
1029 -- attribute identifier but is not. This routine posts appropriate
1030 -- error messages, including a check for a near misspelling.
1032 function Token_Is_At_Start_Of_Line return Boolean;
1033 pragma Inline (Token_Is_At_Start_Of_Line);
1034 -- Determines if the current token is the first token on the line
1036 function Token_Is_At_End_Of_Line return Boolean;
1037 -- Determines if the current token is the last token on the line
1041 ---------------------------------------
1042 -- Specialized Syntax Check Routines --
1043 ---------------------------------------
1045 function Prag (Pragma_Node : Node_Id; Semi : Source_Ptr) return Node_Id;
1046 -- This function is passed a tree for a pragma that has been scanned out.
1047 -- The pragma is syntactically well formed according to the general syntax
1048 -- for pragmas and the pragma identifier is for one of the recognized
1049 -- pragmas. It performs specific syntactic checks for specific pragmas.
1050 -- The result is the input node if it is OK, or Error otherwise. The
1051 -- reason that this is separated out is to facilitate the addition
1052 -- of implementation defined pragmas. The second parameter records the
1053 -- location of the semicolon following the pragma (this is needed for
1054 -- correct processing of the List and Page pragmas). The returned value
1055 -- is a copy of Pragma_Node, or Error if an error is found. Note that
1056 -- at the point where Prag is called, the right paren ending the pragma
1057 -- has been scanned out, and except in the case of pragma Style_Checks,
1058 -- so has the following semicolon. For Style_Checks, the caller delays
1059 -- the scanning of the semicolon so that it will be scanned using the
1060 -- settings from the Style_Checks pragma preceding it.
1062 -------------------------
1063 -- Subsidiary Routines --
1064 -------------------------
1067 -- This procedure creates implicit label declarations for all label that
1068 -- are declared in the current unit. Note that this could conceptually
1069 -- be done at the point where the labels are declared, but it is tricky
1070 -- to do it then, since the tree is not hooked up at the point where the
1071 -- label is declared (e.g. a sequence of statements is not yet attached
1072 -- to its containing scope at the point a label in the sequence is found)
1075 -- This procedure loads all subsidiary units that are required by this
1076 -- unit, including with'ed units, specs for bodies, and parents for child
1077 -- units. It does not load bodies for inlined procedures and generics,
1078 -- since we don't know till semantic analysis is complete what is needed.
1084 -- The package bodies can see all routines defined in all other subpackages
1104 package body Ch2 is separate;
1105 package body Ch3 is separate;
1106 package body Ch4 is separate;
1107 package body Ch5 is separate;
1108 package body Ch6 is separate;
1109 package body Ch7 is separate;
1110 package body Ch8 is separate;
1111 package body Ch9 is separate;
1112 package body Ch10 is separate;
1113 package body Ch11 is separate;
1114 package body Ch12 is separate;
1115 package body Ch13 is separate;
1117 package body Endh is separate;
1118 package body Tchk is separate;
1119 package body Sync is separate;
1120 package body Util is separate;
1122 function Prag (Pragma_Node : Node_Id; Semi : Source_Ptr) return Node_Id
1125 procedure Labl is separate;
1126 procedure Load is separate;
1132 -- This function is the parse routine called at the outer level. It parses
1133 -- the current compilation unit and adds implicit label declarations.
1136 -- Deal with configuration pragmas case first
1138 if Configuration_Pragmas then
1140 Ecount : constant Int := Serious_Errors_Detected;
1141 Pragmas : constant List_Id := Empty_List;
1146 if Token = Tok_EOF then
1149 elsif Token /= Tok_Pragma then
1150 Error_Msg_SC ("only pragmas allowed in configuration file");
1156 if Serious_Errors_Detected > Ecount then
1160 if Chars (P_Node) > Last_Configuration_Pragma_Name
1161 and then Chars (P_Node) /= Name_Source_Reference
1164 ("only configuration pragmas allowed " &
1165 "in configuration file");
1169 Append (P_Node, Pragmas);
1174 -- Normal case of compilation unit
1177 Save_Opt_Config_Switches (Save_Config_Switches);
1179 -- Special processing for language defined units. For this purpose
1180 -- we do NOT consider the renamings in annex J as predefined. That
1181 -- allows users to compile their own versions of these files, and
1182 -- in particular, in the VMS implementation, the DEC versions can
1183 -- be substituted for the standard Ada 95 versions.
1185 if Is_Predefined_File_Name
1186 (Fname => File_Name (Current_Source_File),
1187 Renamings_Included => False)
1189 Set_Opt_Config_Switches
1190 (Is_Internal_File_Name (File_Name (Current_Source_File)));
1192 -- If this is the main unit, disallow compilation unless the -gnatg
1193 -- (GNAT mode) switch is set (from a user point of view, the rule is
1194 -- that language defined units cannot be recompiled).
1196 -- However, an exception is s-rpc, and its children. We test this
1197 -- by looking at the character after the minus, the rule is that
1198 -- System.RPC and its children are the only children in System
1199 -- whose second level name can start with the letter r.
1201 Get_Name_String (File_Name (Current_Source_File));
1203 if (Name_Len < 3 or else Name_Buffer (1 .. 3) /= "s-r")
1204 and then Current_Source_Unit = Main_Unit
1205 and then not GNAT_Mode
1206 and then Operating_Mode = Generate_Code
1208 Error_Msg_SC ("language defined units may not be recompiled");
1212 -- The following loop runs more than once only in syntax check mode
1213 -- where we allow multiple compilation units in the same file.
1216 Set_Opt_Config_Switches
1217 (Is_Internal_File_Name (File_Name (Current_Source_File)));
1219 -- Initialize scope table and other parser control variables
1221 Compiler_State := Parsing;
1223 Scope.Increment_Last;
1224 Scope.Table (0).Etyp := E_Dummy;
1225 SIS_Entry_Active := False;
1226 Last_Resync_Point := No_Location;
1228 Label_List := New_Elmt_List;
1229 Discard_Node (P_Compilation_Unit);
1231 -- If we are not at an end of file, then this means that we are
1232 -- in syntax scan mode, and we can have another compilation unit,
1233 -- otherwise we will exit from the loop.
1235 exit when Token = Tok_EOF;
1236 Restore_Opt_Config_Switches (Save_Config_Switches);
1239 -- Now that we have completely parsed the source file, we can
1240 -- complete the source file table entry.
1242 Complete_Source_File_Entry;
1244 -- An internal error check, the scope stack should now be empty
1246 pragma Assert (Scope.Last = 0);
1248 -- Remaining steps are to create implicit label declarations and to
1249 -- load required subsidiary sources. These steps are required only
1250 -- if we are doing semantic checking.
1252 if Operating_Mode /= Check_Syntax or else Debug_Flag_F then
1257 -- Restore settings of switches saved on entry
1259 Restore_Opt_Config_Switches (Save_Config_Switches);
1260 Set_Comes_From_Source_Default (False);