1 ------------------------------------------------------------------------------
3 -- GNAT COMPILER COMPONENTS --
9 -- Copyright (C) 1992-2004 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;
53 function Par (Configuration_Pragmas : Boolean) return List_Id is
55 Num_Library_Units : Natural := 0;
56 -- Count number of units parsed (relevant only in syntax check only mode,
57 -- since in semantics check mode only a single unit is permitted anyway)
59 Save_Config_Switches : Config_Switches_Type;
60 -- Variable used to save values of config switches while we parse the
61 -- new unit, to be restored on exit for proper recursive behavior.
63 Loop_Block_Count : Nat := 0;
64 -- Counter used for constructing loop/block names (see the routine
65 -- Par.Ch5.Get_Loop_Block_Name)
71 -- When an error is encountered, a call is made to one of the Error_Msg
72 -- routines to record the error. If the syntax scan is not derailed by the
73 -- error (e.g. a complaint that logical operators are inconsistent in an
74 -- EXPRESSION), then control returns from the Error_Msg call, and the
75 -- parse continues unimpeded.
77 -- If on the other hand, the Error_Msg represents a situation from which
78 -- the parser cannot recover locally, the exception Error_Resync is raised
79 -- immediately after the call to Error_Msg. Handlers for Error_Resync
80 -- are located at strategic points to resynchronize the parse. For example,
81 -- when an error occurs in a statement, the handler skips to the next
82 -- semicolon and continues the scan from there.
84 -- Each parsing procedure contains a note with the heading "Error recovery"
85 -- which shows if it can propagate the Error_Resync exception. In order
86 -- not to propagate the exception, a procedure must either contain its own
87 -- handler for this exception, or it must not call any other routines which
88 -- propagate the exception.
90 -- Note: the arrangement of Error_Resync handlers is such that it should
91 -- never be possible to transfer control through a procedure which made
92 -- an entry in the scope stack, invalidating the contents of the stack.
94 Error_Resync : exception;
95 -- Exception raised on error that is not handled locally, see above.
97 Last_Resync_Point : Source_Ptr;
98 -- The resynchronization routines in Par.Sync run a risk of getting
99 -- stuck in an infinite loop if they do not skip a token, and the caller
100 -- keeps repeating the same resync call. On the other hand, if they skip
101 -- a token unconditionally, some recovery opportunities are missed. The
102 -- variable Last_Resync_Point records the token location previously set
103 -- by a Resync call, and if a subsequent Resync call occurs at the same
104 -- location, then the Resync routine does guarantee to skip a token.
106 --------------------------------------------
107 -- Handling Semicolon Used in Place of IS --
108 --------------------------------------------
110 -- The following global variables are used in handling the error situation
111 -- of using a semicolon in place of IS in a subprogram declaration as in:
113 -- procedure X (Y : Integer);
119 -- The two contexts in which this can appear are at the outer level, and
120 -- within a declarative region. At the outer level, we know something is
121 -- wrong as soon as we see the Q (or begin, if there are no declarations),
122 -- and we can immediately decide that the semicolon should have been IS.
124 -- The situation in a declarative region is more complex. The declaration
125 -- of Q could belong to the outer region, and we do not know that we have
126 -- an error until we hit the begin. It is still not clear at this point
127 -- from a syntactic point of view that something is wrong, because the
128 -- begin could belong to the enclosing subprogram or package. However, we
129 -- can incorporate a bit of semantic knowledge and note that the body of
130 -- X is missing, so we definitely DO have an error. We diagnose this error
131 -- as semicolon in place of IS on the subprogram line.
133 -- There are two styles for this diagnostic. If the begin immediately
134 -- follows the semicolon, then we can place a flag (IS expected) right
135 -- on the semicolon. Otherwise we do not detect the error until we hit
136 -- the begin which refers back to the line with the semicolon.
138 -- To control the process in the second case, the following global
139 -- variables are set to indicate that we have a subprogram declaration
140 -- whose body is required and has not yet been found. The prefix SIS
141 -- stands for "Subprogram IS" handling.
143 SIS_Entry_Active : Boolean;
144 -- Set True to indicate that an entry is active (i.e. that a subprogram
145 -- declaration has been encountered, and no body for this subprogram has
146 -- been encountered). The remaining fields are valid only if this is True.
149 -- Subprogram designator
151 SIS_Sloc : Source_Ptr;
152 -- Source location of FUNCTION/PROCEDURE keyword
154 SIS_Ecol : Column_Number;
155 -- Column number of FUNCTION/PROCEDURE keyword
157 SIS_Semicolon_Sloc : Source_Ptr;
158 -- Source location of semicolon at end of subprogram declaration
160 SIS_Declaration_Node : Node_Id;
161 -- Pointer to tree node for subprogram declaration
163 SIS_Missing_Semicolon_Message : Error_Msg_Id;
164 -- Used to save message ID of missing semicolon message (which will be
165 -- modified to missing IS if necessary). Set to No_Error_Msg in the
166 -- normal (non-error) case.
168 -- Five things can happen to an active SIS entry
170 -- 1. If a BEGIN is encountered with an SIS entry active, then we have
171 -- exactly the situation in which we know the body of the subprogram is
172 -- missing. After posting an error message, we change the spec to a body,
173 -- rechaining the declarations that intervened between the spec and BEGIN.
175 -- 2. Another subprogram declaration or body is encountered. In this
176 -- case the entry gets overwritten with the information for the new
177 -- subprogram declaration. We don't catch some nested cases this way,
178 -- but it doesn't seem worth the effort.
180 -- 3. A nested declarative region (e.g. package declaration or package
181 -- body) is encountered. The SIS active indication is reset at the start
182 -- of such a nested region. Again, like case 2, this causes us to miss
183 -- some nested cases, but it doesn't seen worth the effort to stack and
184 -- unstack the SIS information. Maybe we will reconsider this if we ever
185 -- get a complaint about a missed case :-)
187 -- 4. We encounter a valid pragma INTERFACE or IMPORT that effectively
188 -- supplies the missing body. In this case we reset the entry.
190 -- 5. We encounter the end of the declarative region without encoutering
191 -- a BEGIN first. In this situation we simply reset the entry. We know
192 -- that there is a missing body, but it seems more reasonable to let the
193 -- later semantic checking discover this.
195 ----------------------------------------------------
196 -- Handling of Reserved Words Used as Identifiers --
197 ----------------------------------------------------
199 -- Note: throughout the parser, the terms reserved word and keyword
200 -- are used interchangably to refer to the same set of reserved
201 -- keywords (including until, protected, etc).
203 -- If a reserved word is used in place of an identifier, the parser
204 -- where possible tries to recover gracefully. In particular, if the
205 -- keyword is clearly spelled using identifier casing, e.g. Until in
206 -- a source program using mixed case identifiers and lower case keywords,
207 -- then the keyword is treated as an identifier if it appears in a place
208 -- where an identifier is required.
210 -- The situation is more complex if the keyword is spelled with normal
211 -- keyword casing. In this case, the parser is more reluctant to
212 -- consider it to be intended as an identifier, unless it has some
213 -- further confirmation.
215 -- In the case of an identifier appearing in the identifier list of a
216 -- declaration, the appearence of a comma or colon right after the
217 -- keyword on the same line is taken as confirmation. For an enumeration
218 -- literal, a comma or right paren right after the identifier is also
219 -- treated as adequate confirmation.
221 -- The following type is used in calls to Is_Reserved_Identifier and
222 -- also to P_Defining_Identifier and P_Identifier. The default for all
223 -- these functins is that reserved words in reserved word case are not
224 -- considered to be reserved identifiers. The Id_Check value indicates
225 -- tokens, which if they appear immediately after the identifier, are
226 -- taken as confirming that the use of an identifier was expected
230 -- Default, no special token test
233 -- Consider as identifier if followed by comma or right paren
236 -- Consider as identifier if followed by comma or colon
239 -- Consider as identifier if followed by DO
242 -- Consider as identifier if followed by period
245 -- Consider as identifier if followed by >>
248 -- Consider as identifier if followed by IN
251 -- Consider as identifier if followed by IS
253 C_Left_Paren_Semicolon,
254 -- Consider as identifier if followed by left paren or semicolon
257 -- Consider as identifier if followed by USE
259 C_Vertical_Bar_Arrow);
260 -- Consider as identifier if followed by | or =>
262 --------------------------------------------
263 -- Handling IS Used in Place of Semicolon --
264 --------------------------------------------
266 -- This is a somewhat trickier situation, and we can't catch it in all
267 -- cases, but we do our best to detect common situations resulting from
268 -- a "cut and paste" operation which forgets to change the IS to semicolon.
269 -- Consider the following example:
284 -- The trouble is that the section of text from PROCEDURE B through END;
285 -- consitutes a valid procedure body, and the danger is that we find out
286 -- far too late that something is wrong (indeed most compilers will behave
287 -- uncomfortably on the above example).
289 -- We have two approaches to helping to control this situation. First we
290 -- make every attempt to avoid swallowing the last END; if we can be
291 -- sure that some error will result from doing so. In particular, we won't
292 -- accept the END; unless it is exactly correct (in particular it must not
293 -- have incorrect name tokens), and we won't accept it if it is immediately
294 -- followed by end of file, WITH or SEPARATE (all tokens that unmistakeably
295 -- signal the start of a compilation unit, and which therefore allow us to
296 -- reserve the END; for the outer level.) For more details on this aspect
297 -- of the handling, see package Par.Endh.
299 -- If we can avoid eating up the END; then the result in the absense of
300 -- any additional steps would be to post a missing END referring back to
301 -- the subprogram with the bogus IS. Similarly, if the enclosing package
302 -- has no BEGIN, then the result is a missing BEGIN message, which again
303 -- refers back to the subprogram header.
305 -- Such an error message is not too bad (it's already a big improvement
306 -- over what many parsers do), but it's not ideal, because the declarations
307 -- following the IS have been absorbed into the wrong scope. In the above
308 -- case, this could result for example in a bogus complaint that the body
309 -- of D was missing from the package.
311 -- To catch at least some of these cases, we take the following additional
312 -- steps. First, a subprogram body is marked as having a suspicious IS if
313 -- the declaration line is followed by a line which starts with a symbol
314 -- that can start a declaration in the same column, or to the left of the
315 -- column in which the FUNCTION or PROCEDURE starts (normal style is to
316 -- indent any declarations which really belong a subprogram). If such a
317 -- subprogram encounters a missing BEGIN or missing END, then we decide
318 -- that the IS should have been a semicolon, and the subprogram body node
319 -- is marked (by setting the Bad_Is_Detected flag true. Note that we do
320 -- not do this for library level procedures, only for nested procedures,
321 -- since for library level procedures, we must have a body.
323 -- The processing for a declarative part checks to see if the last
324 -- declaration scanned is marked in this way, and if it is, the tree
325 -- is modified to reflect the IS being interpreted as a semicolon.
327 ---------------------------------------------------
328 -- Parser Type Definitions and Control Variables --
329 ---------------------------------------------------
331 -- The following variable and associated type declaration are used by the
332 -- expression parsing routines to return more detailed information about
333 -- the categorization of a parsed expression.
335 type Expr_Form_Type is (
336 EF_Simple_Name, -- Simple name, i.e. possibly qualified identifier
337 EF_Name, -- Simple expression which could also be a name
338 EF_Simple, -- Simple expression which is not call or name
339 EF_Range_Attr, -- Range attribute reference
340 EF_Non_Simple); -- Expression that is not a simple expression
342 Expr_Form : Expr_Form_Type;
344 -- The following type is used for calls to P_Subprogram, P_Package, P_Task,
345 -- P_Protected to indicate which of several possibilities is acceptable.
347 type Pf_Rec is record
348 Spcn : Boolean; -- True if specification OK
349 Decl : Boolean; -- True if declaration OK
350 Gins : Boolean; -- True if generic instantiation OK
351 Pbod : Boolean; -- True if proper body OK
352 Rnam : Boolean; -- True if renaming declaration OK
353 Stub : Boolean; -- True if body stub OK
354 Fil1 : Boolean; -- Filler to fill to 8 bits
355 Fil2 : Boolean; -- Filler to fill to 8 bits
357 pragma Pack (Pf_Rec);
359 function T return Boolean renames True;
360 function F return Boolean renames False;
362 Pf_Decl_Gins_Pbod_Rnam_Stub : constant Pf_Rec :=
363 Pf_Rec'(F, T, T, T, T, T, F, F);
364 Pf_Decl : constant Pf_Rec :=
365 Pf_Rec'(F, T, F, F, F, F, F, F);
366 Pf_Decl_Gins_Pbod_Rnam : constant Pf_Rec :=
367 Pf_Rec'(F, T, T, T, T, F, F, F);
368 Pf_Decl_Pbod : constant Pf_Rec :=
369 Pf_Rec'(F, T, F, T, F, F, F, F);
370 Pf_Pbod : constant Pf_Rec :=
371 Pf_Rec'(F, F, F, T, F, F, F, F);
372 Pf_Spcn : constant Pf_Rec :=
373 Pf_Rec'(T, F, F, F, F, F, F, F);
374 -- The above are the only allowed values of Pf_Rec arguments
376 type SS_Rec is record
377 Eftm : Boolean; -- ELSIF can terminate sequence
378 Eltm : Boolean; -- ELSE can terminate sequence
379 Extm : Boolean; -- EXCEPTION can terminate sequence
380 Ortm : Boolean; -- OR can terminate sequence
381 Sreq : Boolean; -- at least one statement required
382 Tatm : Boolean; -- THEN ABORT can terminate sequence
383 Whtm : Boolean; -- WHEN can terminate sequence
384 Unco : Boolean; -- Unconditional terminate after one statement
386 pragma Pack (SS_Rec);
388 SS_Eftm_Eltm_Sreq : constant SS_Rec := SS_Rec'(T, T, F, F, T, F, F, F);
389 SS_Eltm_Ortm_Tatm : constant SS_Rec := SS_Rec'(F, T, F, T, F, T, F, F);
390 SS_Extm_Sreq : constant SS_Rec := SS_Rec'(F, F, T, F, T, F, F, F);
391 SS_None : constant SS_Rec := SS_Rec'(F, F, F, F, F, F, F, F);
392 SS_Ortm_Sreq : constant SS_Rec := SS_Rec'(F, F, F, T, T, F, F, F);
393 SS_Sreq : constant SS_Rec := SS_Rec'(F, F, F, F, T, F, F, F);
394 SS_Sreq_Whtm : constant SS_Rec := SS_Rec'(F, F, F, F, T, F, T, F);
395 SS_Whtm : constant SS_Rec := SS_Rec'(F, F, F, F, F, F, T, F);
396 SS_Unco : constant SS_Rec := SS_Rec'(F, F, F, F, F, F, F, T);
398 Label_List : Elist_Id;
399 -- List of label nodes for labels appearing in the current compilation.
400 -- Used by Par.Labl to construct the corresponding implicit declarations.
406 -- The scope table, also referred to as the scope stack, is used to
407 -- record the current scope context. It is organized as a stack, with
408 -- inner nested entries corresponding to higher entries on the stack.
409 -- An entry is made when the parser encounters the opening of a nested
410 -- construct (such as a record, task, package etc.), and then package
411 -- Par.Endh uses this stack to deal with END lines (including properly
412 -- dealing with END nesting errors).
415 -- Type of end entry required for this scope. The last two entries are
416 -- used only in the subprogram body case to mark the case of a suspicious
417 -- IS, or a bad IS (i.e. suspicions confirmed by missing BEGIN or END).
418 -- See separate section on dealing with IS used in place of semicolon.
419 -- Note that for many purposes E_Name, E_Suspicious_Is and E_Bad_Is are
420 -- treated the same (E_Suspicious_Is and E_Bad_Is are simply special cases
421 -- of E_Name). They are placed at the end of the enumeration so that a
422 -- test for >= E_Name catches all three cases efficiently.
424 (E_Dummy, -- dummy entry at outer level
428 E_Record, -- END RECORD;
429 E_Select, -- END SELECT;
430 E_Name, -- END [name];
431 E_Suspicious_Is, -- END [name]; (case of suspicious IS)
432 E_Bad_Is); -- END [name]; (case of bad IS)
434 -- The following describes a single entry in the scope table
436 type Scope_Table_Entry is record
438 -- Type of end entry, as per above description
441 -- A flag indicating whether the label, if present, is required to
442 -- appear on the end line. It is referenced only in the case of
443 -- Etyp = E_Name or E_Suspicious_Is where the name may or may not be
444 -- required (yes for labeled block, no in other cases). Note that for
445 -- all cases except begin, the question of whether a label is required
446 -- can be determined from the other fields (for loop, it is required if
447 -- it is present, and for the other constructs it is never required or
450 Ecol : Column_Number;
451 -- Contains the absolute column number (with tabs expanded) of the
452 -- the expected column of the end assuming normal Ada indentation
453 -- usage. If the RM_Column_Check mode is set, this value is used for
454 -- generating error messages about indentation. Otherwise it is used
455 -- only to control heuristic error recovery actions.
458 -- This field is used only for the LOOP and BEGIN cases, and is the
459 -- Node_Id value of the label name. For all cases except child units,
460 -- this value is an entity whose Chars field contains the name pointer
461 -- that identifies the label uniquely. For the child unit case the Labl
462 -- field references an N_Defining_Program_Unit_Name node for the name.
463 -- For cases other than LOOP or BEGIN, the Label field is set to Error,
464 -- indicating that it is an error to have a label on the end line.
465 -- (this is really a misuse of Error since there is no Error ???)
468 -- Points to the list of declarations (i.e. the declarative part)
469 -- associated with this construct. It is set only in the END [name]
470 -- cases, and is set to No_List for all other cases which do not have a
471 -- declarative unit associated with them. This is used for determining
472 -- the proper location for implicit label declarations.
475 -- Empty except in the case of entries for IF and CASE statements,
476 -- in which case it contains the N_If_Statement or N_Case_Statement
477 -- node. This is used for setting the End_Span field.
480 -- Source location of the opening token of the construct. This is
481 -- used to refer back to this line in error messages (such as missing
482 -- or incorrect end lines). The Sloc field is not used, and is not set,
483 -- if a label is present (the Labl field provides the text name of the
484 -- label in this case, which is fine for error messages).
487 -- S_Is is relevant only if Etyp is set to E_Suspicious_Is or
488 -- E_Bad_Is. It records the location of the IS that is considered
492 -- A boolean flag that is set true if the opening entry is the dubious
493 -- result of some prior error, e.g. a record entry where the record
494 -- keyword was missing. It is used to suppress the issuing of a
495 -- corresponding junk complaint about the end line (we do not want
496 -- to complain about a missing end record when there was no record).
499 -- The following declares the scope table itself. The Last field is the
500 -- stack pointer, so that Scope.Table (Scope.Last) is the top entry. The
501 -- oldest entry, at Scope_Stack (0), is a dummy entry with Etyp set to
502 -- E_Dummy, and the other fields undefined. This dummy entry ensures that
503 -- Scope_Stack (Scope_Stack_Ptr).Etyp can always be tested, and that the
504 -- scope stack pointer is always in range.
506 package Scope is new Table.Table (
507 Table_Component_Type => Scope_Table_Entry,
508 Table_Index_Type => Int,
509 Table_Low_Bound => 0,
511 Table_Increment => 100,
512 Table_Name => "Scope");
514 ---------------------------------
515 -- Parsing Routines by Chapter --
516 ---------------------------------
518 -- Uncommented declarations in this section simply parse the construct
519 -- corresponding to their name, and return an ID value for the Node or
520 -- List that is created.
527 function P_Pragma return Node_Id;
529 function P_Identifier (C : Id_Check := None) return Node_Id;
530 -- Scans out an identifier. The parameter C determines the treatment
531 -- of reserved identifiers. See declaration of Id_Check for details.
533 function P_Pragmas_Opt return List_Id;
534 -- This function scans for a sequence of pragmas in other than a
535 -- declaration sequence or statement sequence context. All pragmas
536 -- can appear except pragmas Assert and Debug, which are only allowed
537 -- in a declaration or statement sequence context.
539 procedure P_Pragmas_Misplaced;
540 -- Skips misplaced pragmas with a complaint
542 procedure P_Pragmas_Opt (List : List_Id);
543 -- Parses optional pragmas and appends them to the List
551 Missing_Begin_Msg : Error_Msg_Id;
552 -- This variable is set by a call to P_Declarative_Part. Normally it
553 -- is set to No_Error_Msg, indicating that no special processing is
554 -- required by the caller. The special case arises when a statement
555 -- is found in the sequence of declarations. In this case the Id of
556 -- the message issued ("declaration expected") is preserved in this
557 -- variable, then the caller can change it to an appropriate missing
558 -- begin message if indeed the BEGIN is missing.
560 function P_Access_Definition return Node_Id;
561 function P_Access_Type_Definition return Node_Id;
562 function P_Array_Type_Definition return Node_Id;
563 function P_Basic_Declarative_Items return List_Id;
564 function P_Constraint_Opt return Node_Id;
565 function P_Declarative_Part return List_Id;
566 function P_Discrete_Choice_List return List_Id;
567 function P_Discrete_Range return Node_Id;
568 function P_Discrete_Subtype_Definition return Node_Id;
569 function P_Known_Discriminant_Part_Opt return List_Id;
570 function P_Signed_Integer_Type_Definition return Node_Id;
571 function P_Range return Node_Id;
572 function P_Range_Or_Subtype_Mark return Node_Id;
573 function P_Range_Constraint return Node_Id;
574 function P_Record_Definition return Node_Id;
575 function P_Subtype_Mark return Node_Id;
576 function P_Subtype_Mark_Resync return Node_Id;
577 function P_Unknown_Discriminant_Part_Opt return Boolean;
579 procedure P_Component_Items (Decls : List_Id);
580 -- Scan out one or more component items and append them to the
581 -- given list. Only scans out more than one declaration in the
582 -- case where the source has a single declaration with multiple
583 -- defining identifiers.
585 function P_Defining_Identifier (C : Id_Check := None) return Node_Id;
586 -- Scan out a defining identifier. The parameter C controls the
587 -- treatment of errors in case a reserved word is scanned. See the
588 -- declaration of this type for details.
590 function P_Null_Exclusion return Boolean;
591 -- Ada 0Y (AI-231): Parse the null-excluding part. True indicates
592 -- that the null-excluding part was present.
594 function P_Subtype_Indication
595 (Not_Null_Present : Boolean := False) return Node_Id;
596 -- Ada 0Y (AI-231): The flag Not_Null_Present indicates that the
597 -- null-excluding part has been scanned out and it was present.
599 function Init_Expr_Opt (P : Boolean := False) return Node_Id;
600 -- If an initialization expression is present (:= expression), then
601 -- it is scanned out and returned, otherwise Empty is returned if no
602 -- initialization expression is present. This procedure also handles
603 -- certain common error cases cleanly. The parameter P indicates if
604 -- a right paren can follow the expression (default = no right paren
607 procedure Skip_Declaration (S : List_Id);
608 -- Used when scanning statements to skip past a mispaced declaration
609 -- The declaration is scanned out and appended to the given list.
610 -- Token is known to be a declaration token (in Token_Class_Declk)
611 -- on entry, so there definition is a declaration to be scanned.
613 function P_Subtype_Indication
614 (Subtype_Mark : Node_Id;
615 Not_Null_Present : Boolean := False) return Node_Id;
616 -- This version of P_Subtype_Indication is called when the caller has
617 -- already scanned out the subtype mark which is passed as a parameter.
618 -- Ada 0Y (AI-231): The flag Not_Null_Present indicates that the
619 -- null-excluding part has been scanned out and it was present.
621 function P_Subtype_Mark_Attribute (Type_Node : Node_Id) return Node_Id;
622 -- Parse a subtype mark attribute. The caller has already parsed the
623 -- subtype mark, which is passed in as the argument, and has checked
624 -- that the current token is apostrophe.
632 function P_Aggregate return Node_Id;
633 function P_Expression return Node_Id;
634 function P_Expression_No_Right_Paren return Node_Id;
635 function P_Expression_Or_Range_Attribute return Node_Id;
636 function P_Function_Name return Node_Id;
637 function P_Name return Node_Id;
638 function P_Qualified_Simple_Name return Node_Id;
639 function P_Qualified_Simple_Name_Resync return Node_Id;
640 function P_Simple_Expression return Node_Id;
641 function P_Simple_Expression_Or_Range_Attribute return Node_Id;
643 function P_Qualified_Expression
644 (Subtype_Mark : Node_Id)
646 -- This routine scans out a qualified expression when the caller has
647 -- already scanned out the name and apostrophe of the construct.
655 function P_Statement_Name (Name_Node : Node_Id) return Node_Id;
656 -- Given a node representing a name (which is a call), converts it
657 -- to the syntactically corresponding procedure call statement.
659 function P_Sequence_Of_Statements (SS_Flags : SS_Rec) return List_Id;
660 -- The argument indicates the acceptable termination tokens.
661 -- See body in Par.Ch5 for details of the use of this parameter.
663 procedure Parse_Decls_Begin_End (Parent : Node_Id);
664 -- Parses declarations and handled statement sequence, setting
665 -- fields of Parent node appropriately.
673 function P_Designator return Node_Id;
674 function P_Defining_Program_Unit_Name return Node_Id;
675 function P_Formal_Part return List_Id;
676 function P_Parameter_Profile return List_Id;
677 function P_Return_Statement return Node_Id;
678 function P_Subprogram_Specification return Node_Id;
680 procedure P_Mode (Node : Node_Id);
681 -- Sets In_Present and/or Out_Present flags in Node scanning past
682 -- IN, OUT or IN OUT tokens in the source.
684 function P_Subprogram (Pf_Flags : Pf_Rec) return Node_Id;
685 -- Scans out any construct starting with either of the keywords
686 -- PROCEDURE or FUNCTION. The parameter indicates which possible
687 -- possible kinds of construct (body, spec, instantiation etc.)
688 -- are permissible in the current context.
696 function P_Package (Pf_Flags : Pf_Rec) return Node_Id;
697 -- Scans out any construct starting with the keyword PACKAGE. The
698 -- parameter indicates which possible kinds of construct (body, spec,
699 -- instantiation etc.) are permissible in the current context.
707 function P_Use_Clause return Node_Id;
715 function P_Abort_Statement return Node_Id;
716 function P_Abortable_Part return Node_Id;
717 function P_Accept_Statement return Node_Id;
718 function P_Delay_Statement return Node_Id;
719 function P_Entry_Body return Node_Id;
720 function P_Protected return Node_Id;
721 function P_Requeue_Statement return Node_Id;
722 function P_Select_Statement return Node_Id;
723 function P_Task return Node_Id;
724 function P_Terminate_Alternative return Node_Id;
732 function P_Compilation_Unit return Node_Id;
733 -- Note: this function scans a single compilation unit, and
734 -- checks that an end of file follows this unit, diagnosing
735 -- any unexpected input as an error, and then skipping it, so
736 -- that Token is set to Tok_EOF on return. An exception is in
737 -- syntax-only mode, where multiple compilation units are
738 -- permitted. In this case, P_Compilation_Unit does not check
739 -- for end of file and there may be more compilation units to
740 -- scan. The caller can uniquely detect this situation by the
741 -- fact that Token is not set to Tok_EOF on return.
743 -- The Ignore parameter is normally set False. It is set True
744 -- in multiple unit per file mode if we are skipping past a unit
745 -- that we are not interested in.
753 function P_Handled_Sequence_Of_Statements return Node_Id;
754 function P_Raise_Statement return Node_Id;
756 function Parse_Exception_Handlers return List_Id;
757 -- Parses the partial construct EXCEPTION followed by a list of
758 -- exception handlers which appears in a number of productions,
759 -- and returns the list of exception handlers.
767 function P_Generic return Node_Id;
768 function P_Generic_Actual_Part_Opt return List_Id;
776 function P_Representation_Clause return Node_Id;
778 function P_Code_Statement (Subtype_Mark : Node_Id) return Node_Id;
779 -- Function to parse a code statement. The caller has scanned out
780 -- the name to be used as the subtype mark (but has not checked that
781 -- it is suitable for use as a subtype mark, i.e. is either an
782 -- identifier or a selected component). The current token is an
783 -- apostrophe and the following token is either a left paren or
784 -- RANGE (the latter being an error to be caught by P_Code_Statement.
787 -- Note: the parsing for annexe J features (i.e. obsolescent features)
788 -- is found in the logical section where these features would be if
789 -- they were not obsolescent. In particular:
791 -- Delta constraint is parsed by P_Delta_Constraint (3.5.9)
792 -- At clause is parsed by P_At_Clause (13.1)
793 -- Mod clause is parsed by P_Mod_Clause (13.5.1)
799 -- Routines for handling end lines, including scope recovery
802 function Check_End return Boolean;
803 -- Called when an end sequence is required. In the absence of an error
804 -- situation, Token contains Tok_End on entry, but in a missing end
805 -- case, this may not be the case. Pop_End_Context is used to determine
806 -- the appropriate action to be taken. The returned result is True if
807 -- an End sequence was encountered and False if no End sequence was
808 -- present. This occurs if the END keyword encountered was determined
809 -- to be improper and deleted (i.e. Pop_End_Context set End_Action to
810 -- Skip_And_Reject). Note that the END sequence includes a semicolon,
811 -- except in the case of END RECORD, where a semicolon follows the END
812 -- RECORD, but is not part of the record type definition itself.
815 -- Skip past an end sequence. On entry Token contains Tok_End, and we
816 -- we know that the end sequence is syntactically incorrect, and that
817 -- an appropriate error message has already been posted. The mission
818 -- is simply to position the scan pointer to be the best guess of the
819 -- position after the end sequence. We do not issue any additional
820 -- error messages while carrying this out.
822 procedure End_Statements (Parent : Node_Id := Empty);
823 -- Called when an end is required or expected to terminate a sequence
824 -- of statements. The caller has already made an appropriate entry in
825 -- the Scope.Table to describe the expected form of the end. This can
826 -- only be used in cases where the only appropriate terminator is end.
827 -- If Parent is non-empty, then if a correct END line is encountered,
828 -- the End_Label field of Parent is set appropriately.
835 -- These procedures are used to resynchronize after errors. Following an
836 -- error which is not immediately locally recoverable, the exception
837 -- Error_Resync is raised. The handler for Error_Resync typically calls
838 -- one of these recovery procedures to resynchronize the source position
839 -- to a point from which parsing can be restarted.
841 -- Note: these procedures output an information message that tokens are
842 -- being skipped, but this message is output only if the option for
843 -- Multiple_Errors_Per_Line is set in Options.
846 procedure Resync_Choice;
847 -- Used if an error occurs scanning a choice. The scan pointer is
848 -- advanced to the next vertical bar, arrow, or semicolon, whichever
849 -- comes first. We also quit if we encounter an end of file.
851 procedure Resync_Expression;
852 -- Used if an error is detected during the parsing of an expression.
853 -- It skips past tokens until either a token which cannot be part of
854 -- an expression is encountered (an expression terminator), or if a
855 -- comma or right parenthesis or vertical bar is encountered at the
856 -- current parenthesis level (a parenthesis level counter is maintained
857 -- to carry out this test).
859 procedure Resync_Past_Semicolon;
860 -- Used if an error occurs while scanning a sequence of declarations.
861 -- The scan pointer is positioned past the next semicolon and the scan
862 -- resumes. The scan is also resumed on encountering a token which
863 -- starts a declaration (but we make sure to skip at least one token
864 -- in this case, to avoid getting stuck in a loop).
866 procedure Resync_To_Semicolon;
867 -- Similar to Resync_Past_Semicolon, except that the scan pointer is
868 -- left pointing to the semicolon rather than past it.
870 procedure Resync_Past_Semicolon_Or_To_Loop_Or_Then;
871 -- Used if an error occurs while scanning a sequence of statements.
872 -- The scan pointer is positioned past the next semicolon, or to the
873 -- next occurrence of either then or loop, and the scan resumes.
875 procedure Resync_To_When;
876 -- Used when an error occurs scanning an entry index specification.
877 -- The scan pointer is positioned to the next WHEN (or to IS or
878 -- semicolon if either of these appear before WHEN, indicating
879 -- another error has occurred).
881 procedure Resync_Semicolon_List;
882 -- Used if an error occurs while scanning a parenthesized list of items
883 -- separated by semicolons. The scan pointer is advanced to the next
884 -- semicolon or right parenthesis at the outer parenthesis level, or
885 -- to the next is or RETURN keyword occurence, whichever comes first.
887 procedure Resync_Cunit;
888 -- Synchronize to next token which could be the start of a compilation
889 -- unit, or to the end of file token.
896 -- Routines to check for expected tokens
900 -- Procedures with names of the form T_xxx, where Tok_xxx is a token
901 -- name, check that the current token matches the required token, and
902 -- if so, scan past it. If not, an error is issued indicating that
903 -- the required token is not present (xxx expected). In most cases, the
904 -- scan pointer is not moved in the not-found case, but there are some
905 -- exceptions to this, see for example T_Id, where the scan pointer is
906 -- moved across a literal appearing where an identifier is expected.
914 procedure T_Colon_Equal;
918 procedure T_Greater_Greater;
919 procedure T_Identifier;
922 procedure T_Left_Paren;
931 procedure T_Right_Paren;
932 procedure T_Semicolon;
939 -- Procedures have names of the form TF_xxx, where Tok_xxx is a token
940 -- name check that the current token matches the required token, and
941 -- if so, scan past it. If not, an error message is issued indicating
942 -- that the required token is not present (xxx expected).
944 -- If the missing token is at the end of the line, then control returns
945 -- immediately after posting the message. If there are remaining tokens
946 -- on the current line, a search is conducted to see if the token
947 -- appears later on the current line, as follows:
949 -- A call to Scan_Save is issued and a forward search for the token
950 -- is carried out. If the token is found on the current line before a
951 -- semicolon, then it is scanned out and the scan continues from that
952 -- point. If not the scan is restored to the point where it was missing.
958 procedure TF_Semicolon;
968 function Bad_Spelling_Of (T : Token_Type) return Boolean;
969 -- This function is called in an error situation. It checks if the
970 -- current token is an identifier whose name is a plausible bad
971 -- spelling of the given keyword token, and if so, issues an error
972 -- message, sets Token from T, and returns True. Otherwise Token is
973 -- unchanged, and False is returned.
975 procedure Check_Bad_Layout;
976 -- Check for bad indentation in RM checking mode. Used for statements
977 -- and declarations. Checks if current token is at start of line and
978 -- is exdented from the current expected end column, and if so an
979 -- error message is generated.
981 procedure Check_Misspelling_Of (T : Token_Type);
982 pragma Inline (Check_Misspelling_Of);
983 -- This is similar to the function above, except that it does not
984 -- return a result. It is typically used in a situation where any
985 -- identifier is an error, and it makes sense to simply convert it
986 -- to the given token if it is a plausible misspelling of it.
988 procedure Check_95_Keyword (Token_95, Next : Token_Type);
989 -- This routine checks if the token after the current one matches the
990 -- Next argument. If so, the scan is backed up to the current token
991 -- and Token_Type is changed to Token_95 after issuing an appropriate
992 -- error message ("(Ada 83) keyword xx cannot be used"). If not,
993 -- the scan is backed up with Token_Type unchanged. This routine
994 -- is used to deal with an attempt to use a 95 keyword in Ada 83
995 -- mode. The caller has typically checked that the current token,
996 -- an identifier, matches one of the 95 keywords.
998 procedure Check_Simple_Expression (E : Node_Id);
999 -- Given an expression E, that has just been scanned, so that Expr_Form
1000 -- is still set, outputs an error if E is a non-simple expression. E is
1001 -- not modified by this call.
1003 procedure Check_Simple_Expression_In_Ada_83 (E : Node_Id);
1004 -- Like Check_Simple_Expression, except that the error message is only
1005 -- given when operating in Ada 83 mode, and includes "in Ada 83".
1007 function Check_Subtype_Mark (Mark : Node_Id) return Node_Id;
1008 -- Called to check that a node representing a name (or call) is
1009 -- suitable for a subtype mark, i.e, that it is an identifier or
1010 -- a selected component. If so, or if it is already Error, then
1011 -- it is returned unchanged. Otherwise an error message is issued
1012 -- and Error is returned.
1014 function Comma_Present return Boolean;
1015 -- Used in comma delimited lists to determine if a comma is present, or
1016 -- can reasonably be assumed to have been present (an error message is
1017 -- generated in the latter case). If True is returned, the scan has been
1018 -- positioned past the comma. If False is returned, the scan position
1019 -- is unchanged. Note that all comma-delimited lists are terminated by
1020 -- a right paren, so the only legitimate tokens when Comma_Present is
1021 -- called are right paren and comma. If some other token is found, then
1022 -- Comma_Present has the job of deciding whether it is better to pretend
1023 -- a comma was present, post a message for a missing comma and return
1024 -- True, or return False and let the caller diagnose the missing right
1027 procedure Discard_Junk_Node (N : Node_Id);
1028 procedure Discard_Junk_List (L : List_Id);
1029 pragma Inline (Discard_Junk_Node);
1030 pragma Inline (Discard_Junk_List);
1031 -- These procedures do nothing at all, their effect is simply to discard
1032 -- the argument. A typical use is to skip by some junk that is not
1033 -- expected in the current context.
1035 procedure Ignore (T : Token_Type);
1036 -- If current token matches T, then give an error message and skip
1037 -- past it, otherwise the call has no effect at all. T may be any
1038 -- reserved word token, or comma, left or right paren, or semicolon.
1040 function Is_Reserved_Identifier (C : Id_Check := None) return Boolean;
1041 -- Test if current token is a reserved identifier. This test is based
1042 -- on the token being a keyword and being spelled in typical identifier
1043 -- style (i.e. starting with an upper case letter). The parameter C
1044 -- determines the special treatment if a reserved word is encountered
1045 -- that has the normal casing of a reserved word.
1047 procedure Merge_Identifier (Prev : Node_Id; Nxt : Token_Type);
1048 -- Called when the previous token is an identifier (whose Token_Node
1049 -- value is given by Prev) to check if current token is an identifier
1050 -- that can be merged with the previous one adding an underscore. The
1051 -- merge is only attempted if the following token matches Nxt. If all
1052 -- conditions are met, an error message is issued, and the merge is
1053 -- carried out, modifying the Chars field of Prev.
1055 procedure No_Constraint;
1056 -- Called in a place where no constraint is allowed, but one might
1057 -- appear due to a common error (e.g. after the type mark in a procedure
1058 -- parameter. If a constraint is present, an error message is posted,
1059 -- and the constraint is scanned and discarded.
1061 function No_Right_Paren (Expr : Node_Id) return Node_Id;
1062 -- Function to check for no right paren at end of expression, returns
1063 -- its argument if no right paren, else flags paren and returns Error.
1065 procedure Push_Scope_Stack;
1066 pragma Inline (Push_Scope_Stack);
1067 -- Push a new entry onto the scope stack. Scope.Last (the stack pointer)
1068 -- is incremented. The Junk field is preinitialized to False. The caller
1069 -- is expected to fill in all remaining entries of the new new top stack
1070 -- entry at Scope.Table (Scope.Last).
1072 procedure Pop_Scope_Stack;
1073 -- Pop an entry off the top of the scope stack. Scope_Last (the scope
1074 -- table stack pointer) is decremented by one. It is a fatal error to
1075 -- try to pop off the dummy entry at the bottom of the stack (i.e.
1076 -- Scope.Last must be non-zero at the time of call).
1078 function Separate_Present return Boolean;
1079 -- Determines if the current token is either Tok_Separate, or an
1080 -- identifier that is a possible misspelling of "separate" followed
1081 -- by a semicolon. True is returned if so, otherwise False.
1083 procedure Signal_Bad_Attribute;
1084 -- The current token is an identifier that is supposed to be an
1085 -- attribute identifier but is not. This routine posts appropriate
1086 -- error messages, including a check for a near misspelling.
1088 function Token_Is_At_Start_Of_Line return Boolean;
1089 pragma Inline (Token_Is_At_Start_Of_Line);
1090 -- Determines if the current token is the first token on the line
1092 function Token_Is_At_End_Of_Line return Boolean;
1093 -- Determines if the current token is the last token on the line
1100 -- The processing for pragmas is split off from chapter 2
1102 function Prag (Pragma_Node : Node_Id; Semi : Source_Ptr) return Node_Id;
1103 -- This function is passed a tree for a pragma that has been scanned out.
1104 -- The pragma is syntactically well formed according to the general syntax
1105 -- for pragmas and the pragma identifier is for one of the recognized
1106 -- pragmas. It performs specific syntactic checks for specific pragmas.
1107 -- The result is the input node if it is OK, or Error otherwise. The
1108 -- reason that this is separated out is to facilitate the addition
1109 -- of implementation defined pragmas. The second parameter records the
1110 -- location of the semicolon following the pragma (this is needed for
1111 -- correct processing of the List and Page pragmas). The returned value
1112 -- is a copy of Pragma_Node, or Error if an error is found. Note that
1113 -- at the point where Prag is called, the right paren ending the pragma
1114 -- has been scanned out, and except in the case of pragma Style_Checks,
1115 -- so has the following semicolon. For Style_Checks, the caller delays
1116 -- the scanning of the semicolon so that it will be scanned using the
1117 -- settings from the Style_Checks pragma preceding it.
1124 -- This procedure creates implicit label declarations for all label that
1125 -- are declared in the current unit. Note that this could conceptually
1126 -- be done at the point where the labels are declared, but it is tricky
1127 -- to do it then, since the tree is not hooked up at the point where the
1128 -- label is declared (e.g. a sequence of statements is not yet attached
1129 -- to its containing scope at the point a label in the sequence is found)
1136 -- This procedure loads all subsidiary units that are required by this
1137 -- unit, including with'ed units, specs for bodies, and parents for child
1138 -- units. It does not load bodies for inlined procedures and generics,
1139 -- since we don't know till semantic analysis is complete what is needed.
1145 -- The package bodies can see all routines defined in all other subpackages
1165 package body Ch2 is separate;
1166 package body Ch3 is separate;
1167 package body Ch4 is separate;
1168 package body Ch5 is separate;
1169 package body Ch6 is separate;
1170 package body Ch7 is separate;
1171 package body Ch8 is separate;
1172 package body Ch9 is separate;
1173 package body Ch10 is separate;
1174 package body Ch11 is separate;
1175 package body Ch12 is separate;
1176 package body Ch13 is separate;
1178 package body Endh is separate;
1179 package body Tchk is separate;
1180 package body Sync is separate;
1181 package body Util is separate;
1183 function Prag (Pragma_Node : Node_Id; Semi : Source_Ptr) return Node_Id
1186 procedure Labl is separate;
1187 procedure Load is separate;
1189 -- Start of processing for Par
1193 -- Deal with configuration pragmas case first
1195 if Configuration_Pragmas then
1197 Ecount : constant Int := Serious_Errors_Detected;
1198 Pragmas : constant List_Id := Empty_List;
1203 if Token = Tok_EOF then
1206 elsif Token /= Tok_Pragma then
1207 Error_Msg_SC ("only pragmas allowed in configuration file");
1213 if Serious_Errors_Detected > Ecount then
1217 if Chars (P_Node) > Last_Configuration_Pragma_Name
1218 and then Chars (P_Node) /= Name_Source_Reference
1221 ("only configuration pragmas allowed " &
1222 "in configuration file");
1226 Append (P_Node, Pragmas);
1231 -- Normal case of compilation unit
1234 Save_Opt_Config_Switches (Save_Config_Switches);
1236 -- The following loop runs more than once in syntax check mode
1237 -- where we allow multiple compilation units in the same file
1238 -- and in Multiple_Unit_Per_file mode where we skip units till
1239 -- we get to the unit we want.
1241 for Ucount in Pos loop
1242 Set_Opt_Config_Switches
1243 (Is_Internal_File_Name (File_Name (Current_Source_File)));
1245 -- Initialize scope table and other parser control variables
1247 Compiler_State := Parsing;
1249 Scope.Increment_Last;
1250 Scope.Table (0).Etyp := E_Dummy;
1251 SIS_Entry_Active := False;
1252 Last_Resync_Point := No_Location;
1254 Label_List := New_Elmt_List;
1256 -- If in multiple unit per file mode, skip past ignored unit
1258 if Ucount < Multiple_Unit_Index then
1260 -- We skip in syntax check only mode, since we don't want
1261 -- to do anything more than skip past the unit and ignore it.
1262 -- This causes processing like setting up a unit table entry
1266 Save_Operating_Mode : constant Operating_Mode_Type :=
1269 Save_Style_Check : constant Boolean := Style_Check;
1273 Operating_Mode := Check_Syntax;
1274 Style_Check := False;
1275 Discard_Node (P_Compilation_Unit);
1276 Operating_Mode := Save_Operating_Mode;
1277 Style_Check := Save_Style_Check;
1279 -- If we are at an end of file, and not yet at the right
1280 -- unit, then we have a fatal error. The unit is missing.
1282 if Token = Tok_EOF then
1283 Error_Msg_SC ("file has too few compilation units");
1284 raise Unrecoverable_Error;
1288 -- Here if we are not skipping a file in multiple unit per file
1289 -- mode. Parse the unit that we are interested in. Note that in
1290 -- check syntax mode we are interested in all units in the file.
1294 Comp_Unit_Node : constant Node_Id := P_Compilation_Unit;
1297 -- If parsing was successful and we are not in check syntax
1298 -- mode, check that language defined units are compiled in
1299 -- GNAT mode. For this purpose we do NOT consider renamings
1300 -- in annex J as predefined. That allows users to compile
1301 -- their own versions of these files, and in particular,
1302 -- in the VMS implementation, the DEC versions can be
1303 -- substituted for the standard Ada 95 versions. Another
1304 -- exception is System.RPC and its children. This allows
1305 -- a user to supply their own communication layer.
1307 if Comp_Unit_Node /= Error
1308 and then Operating_Mode = Generate_Code
1309 and then Current_Source_Unit = Main_Unit
1310 and then not GNAT_Mode
1313 Uname : constant String :=
1315 (Unit_Name (Current_Source_Unit));
1316 Name : String (1 .. Uname'Length - 2);
1319 -- Because Unit_Name includes "%s" or "%b", we need to
1320 -- strip the last two characters to get the real unit
1323 Name := Uname (Uname'First .. Uname'Last - 2);
1325 if (Name = "ada" or else
1326 Name = "calendar" or else
1327 Name = "interfaces" or else
1328 Name = "system" or else
1329 Name = "machine_code" or else
1330 Name = "unchecked_conversion" or else
1331 Name = "unchecked_deallocation"
1332 or else (Name'Length > 4
1334 Name (Name'First .. Name'First + 3) =
1336 or else (Name'Length > 11
1338 Name (Name'First .. Name'First + 10) =
1340 or else (Name'Length > 7
1342 Name (Name'First .. Name'First + 6) =
1344 and then Name /= "system.rpc"
1347 or else Name (Name'First .. Name'First + 10) /=
1351 ("language defined units may not be recompiled",
1352 Sloc (Unit (Comp_Unit_Node)));
1358 -- All done if at end of file
1360 exit when Token = Tok_EOF;
1362 -- If we are not at an end of file, it means we are in syntax
1363 -- check only mode, and we keep the loop going to parse all
1364 -- remaining units in the file.
1368 Restore_Opt_Config_Switches (Save_Config_Switches);
1371 -- Now that we have completely parsed the source file, we can
1372 -- complete the source file table entry.
1374 Complete_Source_File_Entry;
1376 -- An internal error check, the scope stack should now be empty
1378 pragma Assert (Scope.Last = 0);
1380 -- Remaining steps are to create implicit label declarations and to
1381 -- load required subsidiary sources. These steps are required only
1382 -- if we are doing semantic checking.
1384 if Operating_Mode /= Check_Syntax or else Debug_Flag_F then
1389 -- Restore settings of switches saved on entry
1391 Restore_Opt_Config_Switches (Save_Config_Switches);
1392 Set_Comes_From_Source_Default (False);
projects / gcc.git / blob