| blob | a53dd61c81e1b6fdbc8b092888ec3c19cc953c36 |
1 ;;; cc-engine.el --- core syntax guessing engine for CC mode
3 ;; Copyright (C) 1985, 1987, 1992, 1993, 1994, 1995, 1996, 1997, 1998,
4 ;; 1999, 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011
5 ;; Free Software Foundation, Inc.
7 ;; Authors: 2001- Alan Mackenzie
8 ;; 1998- Martin Stjernholm
9 ;; 1992-1999 Barry A. Warsaw
10 ;; 1987 Dave Detlefs
11 ;; 1987 Stewart Clamen
12 ;; 1985 Richard M. Stallman
13 ;; Maintainer: bug-cc-mode@gnu.org
14 ;; Created: 22-Apr-1997 (split from cc-mode.el)
15 ;; Version: See cc-mode.el
16 ;; Keywords: c languages oop
18 ;; This file is part of GNU Emacs.
20 ;; GNU Emacs is free software: you can redistribute it and/or modify
21 ;; it under the terms of the GNU General Public License as published by
22 ;; the Free Software Foundation, either version 3 of the License, or
23 ;; (at your option) any later version.
25 ;; GNU Emacs is distributed in the hope that it will be useful,
26 ;; but WITHOUT ANY WARRANTY; without even the implied warranty of
27 ;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
28 ;; GNU General Public License for more details.
30 ;; You should have received a copy of the GNU General Public License
31 ;; along with GNU Emacs. If not, see <http://www.gnu.org/licenses/>.
33 ;;; Commentary:
35 ;; The functions which have docstring documentation can be considered
36 ;; part of an API which other packages can use in CC Mode buffers.
37 ;; Otoh, undocumented functions and functions with the documentation
38 ;; in comments are considered purely internal and can change semantics
39 ;; or even disappear in the future.
40 ;;
41 ;; (This policy applies to CC Mode as a whole, not just this file. It
42 ;; probably also applies to many other Emacs packages, but here it's
43 ;; clearly spelled out.)
45 ;; Hidden buffer changes
46 ;;
47 ;; Various functions in CC Mode use text properties for caching and
48 ;; syntactic markup purposes, and those of them that might modify such
49 ;; properties but still don't modify the buffer in a visible way are
50 ;; said to do "hidden buffer changes". They should be used within
51 ;; `c-save-buffer-state' or a similar function that saves and restores
52 ;; buffer modifiedness, disables buffer change hooks, etc.
53 ;;
54 ;; Interactive functions are assumed to not do hidden buffer changes,
55 ;; except in the specific parts of them that do real changes.
56 ;;
57 ;; Lineup functions are assumed to do hidden buffer changes. They
58 ;; must not do real changes, though.
59 ;;
60 ;; All other functions that do hidden buffer changes have that noted
61 ;; in their doc string or comment.
62 ;;
63 ;; The intention with this system is to avoid wrapping every leaf
64 ;; function that do hidden buffer changes inside
65 ;; `c-save-buffer-state'. It should be used as near the top of the
66 ;; interactive functions as possible.
67 ;;
68 ;; Functions called during font locking are allowed to do hidden
69 ;; buffer changes since the font-lock package run them in a context
70 ;; similar to `c-save-buffer-state' (in fact, that function is heavily
71 ;; inspired by `save-buffer-state' in the font-lock package).
73 ;; Use of text properties
74 ;;
75 ;; CC Mode uses several text properties internally to mark up various
76 ;; positions, e.g. to improve speed and to eliminate glitches in
77 ;; interactive refontification.
78 ;;
79 ;; Note: This doc is for internal use only. Other packages should not
80 ;; assume that these text properties are used as described here.
81 ;;
82 ;; 'syntax-table
83 ;; Used to modify the syntax of some characters. It is used to
84 ;; mark the "<" and ">" of angle bracket parens with paren syntax, and
85 ;; to "hide" obtrusive characters in preprocessor lines.
86 ;;
87 ;; This property is used on single characters and is therefore
88 ;; always treated as front and rear nonsticky (or start and end open
89 ;; in XEmacs vocabulary). It's therefore installed on
90 ;; `text-property-default-nonsticky' if that variable exists (Emacs
91 ;; >= 21).
92 ;;
93 ;; 'c-is-sws and 'c-in-sws
94 ;; Used by `c-forward-syntactic-ws' and `c-backward-syntactic-ws' to
95 ;; speed them up. See the comment blurb before `c-put-is-sws'
96 ;; below for further details.
97 ;;
98 ;; 'c-type
99 ;; This property is used on single characters to mark positions with
100 ;; special syntactic relevance of various sorts. Its primary use is
101 ;; to avoid glitches when multiline constructs are refontified
102 ;; interactively (on font lock decoration level 3). It's cleared in
103 ;; a region before it's fontified and is then put on relevant chars
104 ;; in that region as they are encountered during the fontification.
105 ;; The value specifies the kind of position:
106 ;;
107 ;; 'c-decl-arg-start
108 ;; Put on the last char of the token preceding each declaration
109 ;; inside a declaration style arglist (typically in a function
110 ;; prototype).
111 ;;
112 ;; 'c-decl-end
113 ;; Put on the last char of the token preceding a declaration.
114 ;; This is used in cases where declaration boundaries can't be
115 ;; recognized simply by looking for a token like ";" or "}".
116 ;; `c-type-decl-end-used' must be set if this is used (see also
117 ;; `c-find-decl-spots').
118 ;;
119 ;; 'c-<>-arg-sep
120 ;; Put on the commas that separate arguments in angle bracket
121 ;; arglists like C++ template arglists.
122 ;;
123 ;; 'c-decl-id-start and 'c-decl-type-start
124 ;; Put on the last char of the token preceding each declarator
125 ;; in the declarator list of a declaration. They are also used
126 ;; between the identifiers cases like enum declarations.
127 ;; 'c-decl-type-start is used when the declarators are types,
128 ;; 'c-decl-id-start otherwise.
129 ;;
130 ;; 'c-awk-NL-prop
131 ;; Used in AWK mode to mark the various kinds of newlines. See
132 ;; cc-awk.el.
134 ;;; Code:
141 load-path)))
148 ;; Silence the compiler.
151 \f
152 ;; Make declarations for all the `c-lang-defvar' variables in cc-langs.
165 \f
166 ;;; Internal state variables.
168 ;; Internal state of hungry delete key feature
172 ;; The electric flag (toggled by `c-toggle-electric-state').
173 ;; If t, electric actions (like automatic reindentation, and (if
174 ;; c-auto-newline is also set) auto newlining) will happen when an electric
175 ;; key like `{' is pressed (or an electric keyword like `else').
179 ;; Internal state of auto newline feature.
183 ;; Included in the mode line to indicate the active submodes.
184 ;; (defvar c-submode-indicators nil)
185 ;; (make-variable-buffer-local 'c-submode-indicators)
188 ;; Calculate the new state of PREVSTATE, t or nil, based on arg. If
189 ;; arg is nil or zero, toggle the state. If arg is negative, turn
190 ;; the state off, and if arg is positive, turn the state on
196 ;; Dynamically bound cache for `c-in-literal'.
199 \f
200 ;; Basic handling of preprocessor directives.
202 ;; This is a dynamically bound cache used together with
203 ;; `c-query-macro-start' and `c-query-and-set-macro-start'. It only
204 ;; works as long as point doesn't cross a macro boundary.
213 c-macro-start))
221 c-macro-start))
224 "Go to the beginning of a preprocessor directive.
225 Leave point at the beginning of the directive and return t if in one,
226 otherwise return nil and leave point unchanged.
228 Note that this function might do hidden buffer changes. See the
229 comment at the start of cc-engine.el for more info."
240 t
242 nil)))))
245 "Go to the end of a preprocessor directive.
246 More accurately, move the point to the end of the closest following
247 line that doesn't end with a line continuation backslash - no check is
248 done that the point is inside a cpp directive to begin with.
250 Note that this function might do hidden buffer changes. See the
251 comment at the start of cc-engine.el for more info."
257 t))))
260 ;; Assuming point is at the "#" that introduces a preprocessor
261 ;; directive, it's moved forward to the end of the identifier which is
262 ;; "#define"d (or whatever c-opt-cpp-macro-define specifies). Non-nil
263 ;; is returned in this case, in all other cases nil is returned and
264 ;; point isn't moved.
265 ;;
266 ;; This function might do hidden buffer changes.
272 ;; Assuming point is at the "#" that introduces a preprocessor
273 ;; directive, it's moved forward to the start of the definition body
274 ;; if it's a "#define" (or whatever c-opt-cpp-macro-define
275 ;; specifies). Non-nil is returned in this case, in all other cases
276 ;; nil is returned and point isn't moved.
277 ;;
278 ;; This function might do hidden buffer changes.
284 \f
285 ;;; Basic utility functions.
288 ;; Return the given region as a string where all syntactic
289 ;; whitespace is removed or, where necessary, replaced with a single
290 ;; space. If PAREN-LEVEL is given then all parens in the region are
291 ;; collapsed to "()", "[]" etc.
292 ;;
293 ;; This function might do hidden buffer changes.
337 ;; Shift the indentation of the current line with the specified
338 ;; amount (positive inwards). The buffer is modified only if
339 ;; SHIFT-AMT isn't equal to zero.
342 tmp-char-inserted)
344 nil
345 ;; If we're on an empty line inside a macro, we take the point
346 ;; to be at the current indentation and shift it to the
347 ;; appropriate column. This way we don't treat the extra
348 ;; whitespace out to the line continuation as indentation.
364 ;; If initial point was within line's indentation and we're not on
365 ;; a line with a line continuation in a macro, position after the
366 ;; indentation. Else stay at same point in text.
374 ;; Return non-nil if the string KEYWORD is a known keyword. More
375 ;; precisely, the value is the symbol for the keyword in
376 ;; `c-keywords-obarray'.
380 ;; Return non-nil if the symbol KEYWORD-SYM, as returned by
381 ;; `c-keyword-sym', is a member of LANG-CONSTANT, which is the name
382 ;; of a language constant that ends with "-kwds". If KEYWORD-SYM is
383 ;; nil then the result is nil.
386 ;; String syntax chars, suitable for skip-syntax-(forward|backward).
391 ;; Regexp matching string limit syntax.
396 ;; Regexp matching WS followed by string limit syntax.
400 ;; Holds formatted error strings for the few cases where parse errors
401 ;; are reported.
408 c-parsing-error)
410 ;; Faces given to comments and string literals. This is used in some
411 ;; situations to speed up recognition; it isn't mandatory that font
412 ;; locking is in use. This variable is extended with the face in
413 ;; `c-doc-face-name' when fontification is activated in cc-fonts.el.
417 ;; New in Emacs 22.
421 ;; Put a c-type property with the given value at POS.
425 ;; Remove all occurrences of the c-type property that has the given
426 ;; value in the region between FROM and TO. VALUE is assumed to not
427 ;; be nil.
428 ;;
429 ;; Note: This assumes that c-type is put on single chars only; it's
430 ;; very inefficient if matching properties cover large regions.
439 \f
440 ;; Some debug tools to visualize various special positions. This
441 ;; debug code isn't as portable as the rest of CC Mode.
477 \f
478 ;; `c-beginning-of-statement-1' and accompanying stuff.
480 ;; KLUDGE ALERT: c-maybe-labelp is used to pass information between
481 ;; c-crosses-statement-barrier-p and c-beginning-of-statement-1. A
482 ;; better way should be implemented, but this will at least shut up
483 ;; the byte compiler.
486 ;; New awk-compatible version of c-beginning-of-statement-1, ACM 2002/6/22
488 ;; Macros used internally in c-beginning-of-statement-1 for the
489 ;; automaton actions.
492 stack)))
497 t
498 ,do-if-done
503 ;; Throw nil if stack is empty, else throw non-nil.
527 noerror comma-delim)
528 "Move to the start of the current statement or declaration, or to
529 the previous one if already at the beginning of one. Only
530 statements/declarations on the same level are considered, i.e. don't
531 move into or out of sexps (not even normal expression parentheses).
533 If point is already at the earliest statement within braces or parens,
534 this function doesn't move back into any whitespace preceding it; it
535 returns 'same in this case.
539 is within the continuation. If starting at such a token, move to the
540 corresponding statement start. If at the beginning of a statement,
541 move to the closest containing statement if there is any. This might
542 also stop at a continuation clause.
544 Labels are treated as part of the following statements if
545 IGNORE-LABELS is non-nil. (FIXME: Doesn't work if we stop at a known
546 statement start keyword.) Otherwise, each label is treated as a
547 separate statement.
549 Macros are ignored \(i.e. skipped over) unless point is within one, in
550 which case the content of the macro is treated as normal code. Aside
551 from any normal statement starts found in it, stop at the first token
554 macros before leaving them.
556 Return:
558 'same if stopped at the beginning of the current statement;
559 'up if stepped to a containing statement;
560 'previous if stepped to a preceding statement;
561 'beginning if stepped from a statement continuation clause to
562 its start clause; or
563 'macro if stepped to a macro start.
564 Note that 'same and not 'label is returned if stopped at the same
565 label without crossing the colon character.
567 LIM may be given to limit the search. If the search hits the limit,
568 point will be left at the closest following token, or at the start
569 position if that is less ('same is returned in this case).
571 NOERROR turns off error logging to `c-parsing-error'.
573 Normally only ';' and virtual semicolons are considered to delimit
574 statements, but if COMMA-DELIM is non-nil then ',' is treated
575 as a delimiter too.
577 Note that this function might do hidden buffer changes. See the
578 comment at the start of cc-engine.el for more info."
580 ;; The bulk of this function is a pushdown automaton that looks at statement
581 ;; boundaries and the tokens (such as "while") in c-opt-block-stmt-key. Its
582 ;; purpose is to keep track of nested statements, ensuring that such
583 ;; statements are skipped over in their entirety (somewhat akin to what C-M-p
584 ;; does with nested braces/brackets/parentheses).
585 ;;
586 ;; Note: The position of a boundary is the following token.
587 ;;
588 ;; Beginning with the current token (the one following point), move back one
589 ;; sexp at a time (where a sexp is, more or less, either a token or the
590 ;; entire contents of a brace/bracket/paren pair). Each time a statement
591 ;; boundary is crossed or a "while"-like token is found, update the state of
592 ;; the PDA. Stop at the beginning of a statement when the stack (holding
593 ;; nested statement info) is empty and the position has been moved.
594 ;;
595 ;; The following variables constitute the PDA:
596 ;;
597 ;; sym: This is either the "while"-like token (e.g. 'for) we've just
598 ;; scanned back over, 'boundary if we've just gone back over a
599 ;; statement boundary, or nil otherwise.
600 ;; state: takes one of the values (nil else else-boundary while
601 ;; while-boundary catch catch-boundary).
602 ;; nil means "no "while"-like token yet scanned".
603 ;; 'else, for example, means "just gone back over an else".
604 ;; 'else-boundary means "just gone back over a statement boundary
605 ;; immediately after having gone back over an else".
606 ;; saved-pos: A vector of either saved positions (tok ptok pptok, etc.) or
607 ;; of error reporting information.
608 ;; stack: The stack onto which the PDA pushes its state. Each entry
609 ;; consists of a saved value of state and saved-pos. An entry is
610 ;; pushed when we move back over a "continuation" token (e.g. else)
611 ;; and popped when we encounter the corresponding opening token
612 ;; (e.g. if).
613 ;;
614 ;;
615 ;; The following diagram briefly outlines the PDA.
616 ;;
617 ;; Common state:
618 ;; "else": Push state, goto state `else'.
619 ;; "while": Push state, goto state `while'.
620 ;; "catch" or "finally": Push state, goto state `catch'.
621 ;; boundary: Pop state.
622 ;; other: Do nothing special.
623 ;;
624 ;; State `else':
625 ;; boundary: Goto state `else-boundary'.
626 ;; other: Error, pop state, retry token.
627 ;;
628 ;; State `else-boundary':
629 ;; "if": Pop state.
630 ;; boundary: Error, pop state.
631 ;; other: See common state.
632 ;;
633 ;; State `while':
634 ;; boundary: Save position, goto state `while-boundary'.
635 ;; other: Pop state, retry token.
636 ;;
637 ;; State `while-boundary':
638 ;; "do": Pop state.
639 ;; boundary: Restore position if it's not at start, pop state. [*see below]
640 ;; other: See common state.
641 ;;
642 ;; State `catch':
643 ;; boundary: Goto state `catch-boundary'.
644 ;; other: Error, pop state, retry token.
645 ;;
646 ;; State `catch-boundary':
647 ;; "try": Pop state.
648 ;; "catch": Goto state `catch'.
649 ;; boundary: Error, pop state.
650 ;; other: See common state.
651 ;;
652 ;; [*] In the `while-boundary' state, we had pushed a 'while state, and were
653 ;; searching for a "do" which would have opened a do-while. If we didn't
654 ;; find it, we discard the analysis done since the "while", go back to this
655 ;; token in the buffer and restart the scanning there, this time WITHOUT
656 ;; pushing the 'while state onto the stack.
657 ;;
658 ;; In addition to the above there is some special handling of labels
659 ;; and macros.
663 macro-start
666 c-stmt-delim-chars-with-comma
667 c-stmt-delim-chars))
668 c-in-literal-cache c-maybe-labelp after-case:-pos saved
669 ;; Current position.
670 pos
671 ;; Position of last stmt boundary character (e.g. ;).
672 boundary-pos
673 ;; The position of the last sexp or bound that follows the
674 ;; first found colon, i.e. the start of the nonlabel part of
675 ;; the statement. It's `start' if a colon is found just after
676 ;; the start.
677 after-labels-pos
678 ;; Like `after-labels-pos', but the first such position inside
679 ;; a label, i.e. the start of the last label before the start
680 ;; of the nonlabel part of the statement.
681 last-label-pos
682 ;; The last position where a label is possible provided the
683 ;; statement started there. It's nil as long as no invalid
684 ;; label content has been found (according to
685 ;; `c-nonlabel-token-key'. It's `start' if no valid label
686 ;; content was found in the label. Note that we might still
687 ;; regard it a label if it starts with `c-label-kwds'.
688 label-good-pos
689 ;; Symbol just scanned back over (e.g. 'while or 'boundary).
690 ;; See above.
691 sym
692 ;; Current state in the automaton. See above.
693 state
694 ;; Current saved positions. See above.
695 saved-pos
696 ;; Stack of conses (state . saved-pos).
697 stack
698 ;; Regexp which matches "for", "if", etc.
701 ;; Return value.
703 ;; Positions of the last three sexps or bounds we've stopped at.
704 tok ptok pptok)
714 ;; Try to skip back over unary operator characters, to register
715 ;; that we've moved.
719 ;; Protect post-++/-- operators just before a virtual semicolon.
723 ;; Skip back over any semicolon here. If it was a bare semicolon, we're
724 ;; done. Later on we ignore the boundaries for statements that don't
725 ;; contain any sexp. The only thing that is affected is that the error
726 ;; checking is a little less strict, and we really don't bother.
736 pos saved)
738 ;; Begin at start and not pos to detect macros if we stand
739 ;; directly after the #.
742 ;; Record this as the first token if not starting inside it.
745 ;; The following while loop goes back one sexp (balanced parens,
746 ;; etc. with contents, or symbol or suchlike) each iteration. This
747 ;; movement is accomplished with a call to scan-sexps approx 130 lines
748 ;; below.
766 ;; Stop at the first token in the content of the macro.
770 ret 'macro
771 ignore-labels t))
774 ;; Do a round through the automaton if we've just passed a
775 ;; statement boundary or passed a "while"-like token.
783 ;; The PDA state handling.
784 ;;
785 ;; Refer to the description of the PDA in the opening
786 ;; comments. In the following OR form, the first leaf
787 ;; attempts to handles one of the specific actions detailed
788 ;; (e.g., finding token "if" whilst in state `else-boundary').
789 ;; We drop through to the second leaf (which handles common
790 ;; state) if no specific handler is found in the first cond.
791 ;; If a parsing error is detected (e.g. an "else" with no
792 ;; preceding "if"), we throw to the enclosing catch.
793 ;;
794 ;; Note that the (eq state 'else) means
795 ;; "we've just passed an else", NOT "we're looking for an
796 ;; else".
813 ;; Since this can cause backtracking we do a
814 ;; little more careful analysis to avoid it:
815 ;; If there's a label in front of the while
816 ;; it can't be part of a do-while.
845 ;; This is state common. We get here when the previous
846 ;; cond statement found no particular state handler.
848 ;; If we have a boundary at the start
849 ;; position we push a frame to go to the
850 ;; previous statement.
859 ;; Is this a real while, or a do-while?
860 ;; The next `when' triggers unless we are SURE that
861 ;; the `while' is not the tailend of a `do-while'.
864 ;; The following kludge is to prevent
865 ;; infinite recursion when called from
866 ;; c-awk-after-if-for-while-condition-p,
867 ;; or the like.
871 ;; Since this can cause backtracking we do a
872 ;; little more careful analysis to avoid it: If
873 ;; the while isn't followed by a (possibly
874 ;; virtual) semicolon it can't be a do-while.
883 ;; We're either past a statement boundary or at the
884 ;; start of a statement, so throw away any label data
885 ;; for the previous one.
887 last-label-pos nil
888 c-maybe-labelp nil))))
890 ;; Step to the previous sexp, but not if we crossed a
891 ;; boundary, since that doesn't consume an sexp.
895 ;; HERE IS THE SINGLE PLACE INSIDE THE PDA LOOP WHERE WE MOVE
896 ;; BACKWARDS THROUGH THE SOURCE.
898 ;; This is typically fast with the caching done by
899 ;; c-(backward|forward)-sws.
903 ;; Set as long as we have to continue jumping by sexps.
904 ;; It's the position to use as end in the next round.
905 sexp-loop-continue-pos
906 ;; The end position of the area to search for statement
907 ;; barriers in this round.
910 ;; The following while goes back one sexp per iteration.
914 ;; Give up if we hit an unbalanced block. Since the
915 ;; stack won't be empty the code below will report a
916 ;; suitable error.
919 ;; Check if the sexp movement crossed a statement or
920 ;; declaration boundary. But first modify the point
921 ;; so that `c-crosses-statement-barrier-p' only looks
922 ;; at the non-sexp chars following the sexp.
925 boundary-pos
928 nil
931 ;; Set continuation position in case
932 ;; `c-crosses-statement-barrier-p'
933 ;; doesn't detect anything below.
935 ;; If the sexp movement took us into a
936 ;; macro then there were only some non-sexp
937 ;; chars after it. Skip out of the macro
938 ;; to analyze them but not the non-sexp
939 ;; chars that might be inside the macro.
947 ;; Passed a block sexp. That's a boundary
948 ;; alright.
952 ;; Passed some other paren. Only analyze
953 ;; the non-sexp chars after it.
955 before-sws-pos)))
956 ;; We're at a valid token start position
957 ;; (outside the `save-excursion') if
958 ;; `c-crosses-statement-barrier-p' failed.
963 ;; Passed a symbol sexp or line
964 ;; continuation. It doesn't matter that
965 ;; it's included in the analyzed region.
968 t
969 ;; If it was a line continuation then we
970 ;; have to continue looping.
973 nil))))
976 ptok tok
977 tok boundary-pos
979 ;; Like a C "continue". Analyze the next sexp.
985 sexp-loop-continue-pos nil))))
987 ;; ObjC method def?
994 ;; Handle labels.
997 ;; `c-crosses-statement-barrier-p' has found a colon, so we
998 ;; might be in a label now. Have we got a real label
999 ;; (including a case label) or something like C++'s "public:"?
1000 ;; A case label might use an expression rather than a token.
1012 ; been found.
1014 ;; We're in a potential label and it's the first
1015 ;; time we've found something that isn't allowed in
1016 ;; one.
1019 ;; We've moved back by a sexp, so update the token positions.
1021 pptok ptok
1022 ptok tok
1026 ;; If the stack isn't empty there might be errors to report.
1035 ;; Need to investigate closer whether we've crossed
1036 ;; between a substatement and its containing statement.
1038 ptok
1039 pptok))
1046 after-labels-pos
1052 last-label-pos
1053 pos))
1055 ;; We're in a label. Maybe we should step to the statement
1056 ;; after it.
1061 ;; Might have jumped over several labels. Go to the last one.
1064 ;; Have we got "case <expression>:"?
1074 ;; Skip over the unary operators that can start the statement.
1077 ;; protect AWK post-inc/decrement operators, etc.
1082 ret)))
1085 "Return non-nil if buffer positions FROM to TO cross one or more
1086 statement or declaration boundaries. The returned value is actually
1087 the position of the earliest boundary char. FROM must not be within
1088 a string or comment.
1090 The variable `c-maybe-labelp' is set to the position of the first `:' that
1091 might start a label (i.e. not part of `::' and not preceded by `?'). If a
1092 single `?' is found, then `c-maybe-labelp' is cleared.
1094 For AWK, a statement which is terminated by an EOL (not a \; or a }) is
1096 the line. If this virtual semicolon is _at_ from, the function recognizes it.
1098 Note that this function might do hidden buffer changes. See the
1099 comment at the start of cc-engine.el for more info."
1101 lit-range)
1114 ;; Ignore scope operators.
1118 ;; A question mark. Can't be a label, so stop
1119 ;; looking for more : and ?.
1130 ;; In trailing space after an as yet undetected virtual semicolon?
1135 nil)))))
1138 "Return non-nil if the point is at the first token in a statement
1139 or somewhere in the syntactic whitespace before it.
1142 Any kind of declaration-like construct that occur outside function
1145 Note that this function might do hidden buffer changes. See the
1146 comment at the start of cc-engine.el for more info."
1150 c-maybe-labelp)
1161 "Return non-nil if the point is at the first token in an expression or
1162 statement, or somewhere in the syntactic whitespace before it.
1165 grammar sense: It's any sequence of expression tokens except commas,
1166 unless they are enclosed inside parentheses of some kind. Also, an
1167 expression never continues past an enclosing parenthesis, but it might
1168 contain parenthesis pairs of any sort except braces.
1170 Since expressions never cross statement boundaries, this function also
1171 recognizes statement beginnings, just like `c-at-statement-start-p'.
1173 Note that this function might do hidden buffer changes. See the
1174 comment at the start of cc-engine.el for more info."
1179 c-maybe-labelp)
1187 \f
1188 ;; A set of functions that covers various idiosyncrasies in
1189 ;; implementations of `forward-comment'.
1191 ;; Note: Some emacsen considers incorrectly that any line comment
1192 ;; ending with a backslash continues to the next line. I can't think
1193 ;; of any way to work around that in a reliable way without changing
1194 ;; the buffer, though. Suggestions welcome. ;) (No, temporarily
1195 ;; changing the syntax for backslash doesn't work since we must treat
1196 ;; escapes in string literals correctly.)
1199 "Move forward past whitespace and the closest following comment, if any.
1200 Return t if a comment was found, nil otherwise. In either case, the
1201 point is moved past the following whitespace. Line continuations,
1202 i.e. a backslashes followed by line breaks, are treated as whitespace.
1203 The line breaks that end line comments are considered to be the
1204 comment enders, so the point will be put on the beginning of the next
1205 line if it moved past a line comment.
1207 This function does not do any hidden buffer changes."
1215 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1216 ;; forwards at eob.
1217 nil
1219 ;; Emacs includes the ending newline in a b-style (c++)
1220 ;; comment, but XEmacs doesn't. We depend on the Emacs
1221 ;; behavior (which also is symmetric).
1225 t))))
1228 "Move forward past all following whitespace and comments.
1229 Line continuations, i.e. a backslashes followed by line breaks, are
1230 treated as whitespace.
1232 Note that this function might do hidden buffer changes. See the
1233 comment at the start of cc-engine.el for more info."
1236 ;; If forward-comment in at least XEmacs 21 is given a large
1237 ;; positive value, it'll loop all the way through if it hits
1238 ;; eob.
1240 ;; Some emacsen (e.g. XEmacs 21) return t when moving
1241 ;; forwards at eob.
1246 t))))
1249 "Move backward past whitespace and the closest preceding comment, if any.
1250 Return t if a comment was found, nil otherwise. In either case, the
1251 point is moved past the preceding whitespace. Line continuations,
1252 i.e. a backslashes followed by line breaks, are treated as whitespace.
1253 The line breaks that end line comments are considered to be the
1254 comment enders, so the point cannot be at the end of the same line to
1255 move over a line comment.
1257 This function does not do any hidden buffer changes."
1260 ;; When we got newline terminated comments, forward-comment in all
1261 ;; supported emacsen so far will stop at eol of each line not
1262 ;; ending with a comment when moving backwards. This corrects for
1263 ;; that, and at the same time handles line continuations.
1271 ;; Some emacsen (e.g. Emacs 19.34) return t when moving
1272 ;; backwards at bob.
1273 nil
1275 ;; Leave point after the closest following newline if we've
1276 ;; backed up over any above, since forward-comment won't move
1277 ;; backward over a line comment if point is at the end of the
1278 ;; same line.
1283 ;; If forward-comment above succeeded and we're at eol
1284 ;; then the newline we moved over above didn't end a
1285 ;; line comment, so we give it another go.
1287 t))
1289 ;; Emacs <= 20 and XEmacs move back over the closer of a
1290 ;; block comment that lacks an opener.
1293 t)))))
1296 "Move backward past all preceding whitespace and comments.
1297 Line continuations, i.e. a backslashes followed by line breaks, are
1298 treated as whitespace. The line breaks that end line comments are
1299 considered to be the comment enders, so the point cannot be at the end
1300 of the same line to move over a line comment. Unlike
1301 c-backward-syntactic-ws, this function doesn't move back over
1302 preprocessor directives.
1304 Note that this function might do hidden buffer changes. See the
1305 comment at the start of cc-engine.el for more info."
1309 ;; `forward-comment' in some emacsen (e.g. XEmacs 21.4)
1310 ;; return t when moving backwards at bob.
1315 ;; Emacs <= 20 and XEmacs move back over the
1316 ;; closer of a block comment that lacks an opener.
1318 t)
1320 ;; XEmacs treats line continuations as whitespace but
1321 ;; only in the backward direction, which seems a bit
1322 ;; odd. Anyway, this is necessary for Emacs.
1327 t))))))
1329 \f
1330 ;; Tools for skipping over syntactic whitespace.
1332 ;; The following functions use text properties to cache searches over
1333 ;; large regions of syntactic whitespace. It works as follows:
1334 ;;
1335 ;; o If a syntactic whitespace region contains anything but simple
1336 ;; whitespace (i.e. space, tab and line breaks), the text property
1337 ;; `c-in-sws' is put over it. At places where we have stopped
1338 ;; within that region there's also a `c-is-sws' text property.
1339 ;; That since there typically are nested whitespace inside that
1340 ;; must be handled separately, e.g. whitespace inside a comment or
1341 ;; cpp directive. Thus, from one point with `c-is-sws' it's safe
1342 ;; to jump to another point with that property within the same
1343 ;; `c-in-sws' region. It can be likened to a ladder where
1344 ;; `c-in-sws' marks the bars and `c-is-sws' the rungs.
1345 ;;
1346 ;; o The `c-is-sws' property is put on the simple whitespace chars at
1347 ;; a "rung position" and also maybe on the first following char.
1348 ;; As many characters as can be conveniently found in this range
1349 ;; are marked, but no assumption can be made that the whole range
1350 ;; is marked (it could be clobbered by later changes, for
1351 ;; instance).
1352 ;;
1353 ;; Note that some part of the beginning of a sequence of simple
1354 ;; whitespace might be part of the end of a preceding line comment
1355 ;; or cpp directive and must not be considered part of the "rung".
1356 ;; Such whitespace is some amount of horizontal whitespace followed
1357 ;; by a newline. In the case of cpp directives it could also be
1358 ;; two newlines with horizontal whitespace between them.
1359 ;;
1360 ;; The reason to include the first following char is to cope with
1361 ;; "rung positions" that doesn't have any ordinary whitespace. If
1362 ;; `c-is-sws' is put on a token character it does not have
1363 ;; `c-in-sws' set simultaneously. That's the only case when that
1364 ;; can occur, and the reason for not extending the `c-in-sws'
1365 ;; region to cover it is that the `c-in-sws' region could then be
1366 ;; accidentally merged with a following one if the token is only
1367 ;; one character long.
1368 ;;
1369 ;; o On buffer changes the `c-in-sws' and `c-is-sws' properties are
1370 ;; removed in the changed region. If the change was inside
1371 ;; syntactic whitespace that means that the "ladder" is broken, but
1372 ;; a later call to `c-forward-sws' or `c-backward-sws' will use the
1373 ;; parts on either side and use an ordinary search only to "repair"
1374 ;; the gap.
1375 ;;
1376 ;; Special care needs to be taken if a region is removed: If there
1377 ;; are `c-in-sws' on both sides of it which do not connect inside
1378 ;; the region then they can't be joined. If e.g. a marked macro is
1379 ;; broken, syntactic whitespace inside the new text might be
1380 ;; marked. If those marks would become connected with the old
1381 ;; `c-in-sws' range around the macro then we could get a ladder
1382 ;; with one end outside the macro and the other at some whitespace
1383 ;; within it.
1384 ;;
1385 ;; The main motivation for this system is to increase the speed in
1386 ;; skipping over the large whitespace regions that can occur at the
1387 ;; top level in e.g. header files that contain a lot of comments and
1388 ;; cpp directives. For small comments inside code it's probably
1389 ;; slower than using `forward-comment' straightforwardly, but speed is
1390 ;; not a significant factor there anyway.
1392 ; (defface c-debug-is-sws-face
1393 ; '((t (:background "GreenYellow")))
1394 ; "Debug face to mark the `c-is-sws' property.")
1395 ; (defface c-debug-in-sws-face
1396 ; '((t (:underline t)))
1397 ; "Debug face to mark the `c-in-sws' property.")
1399 ; (defun c-debug-put-sws-faces ()
1400 ; ;; Put the sws debug faces on all the `c-is-sws' and `c-in-sws'
1401 ; ;; properties in the buffer.
1402 ; (interactive)
1403 ; (save-excursion
1404 ; (c-save-buffer-state (in-face)
1405 ; (goto-char (point-min))
1406 ; (setq in-face (if (get-text-property (point) 'c-is-sws)
1407 ; (point)))
1408 ; (while (progn
1409 ; (goto-char (next-single-property-change
1410 ; (point) 'c-is-sws nil (point-max)))
1411 ; (if in-face
1412 ; (progn
1413 ; (c-debug-add-face in-face (point) 'c-debug-is-sws-face)
1414 ; (setq in-face nil))
1415 ; (setq in-face (point)))
1416 ; (not (eobp))))
1417 ; (goto-char (point-min))
1418 ; (setq in-face (if (get-text-property (point) 'c-in-sws)
1419 ; (point)))
1420 ; (while (progn
1421 ; (goto-char (next-single-property-change
1422 ; (point) 'c-in-sws nil (point-max)))
1423 ; (if in-face
1424 ; (progn
1425 ; (c-debug-add-face in-face (point) 'c-debug-in-sws-face)
1426 ; (setq in-face nil))
1427 ; (setq in-face (point)))
1428 ; (not (eobp)))))))
1431 ;;`(message ,@args)
1432 )
1435 ;; This macro does a hidden buffer change.
1442 ;; This macro does a hidden buffer change.
1449 ;; This macro does a hidden buffer change.
1456 ;; This macro does a hidden buffer change.
1463 ;; This macro does a hidden buffer change.
1471 ;; Called from `after-change-functions'. Note that if
1472 ;; `c-forward-sws' or `c-backward-sws' are used outside
1473 ;; `c-save-buffer-state' or similar then this will remove the cache
1474 ;; properties right after they're added.
1475 ;;
1476 ;; This function does hidden buffer changes.
1479 ;; Adjust the end to remove the properties in any following simple
1480 ;; ws up to and including the next line break, if there is any
1481 ;; after the changed region. This is necessary e.g. when a rung
1482 ;; marked empty line is converted to a line comment by inserting
1483 ;; "//" before the line break. In that case the line break would
1484 ;; keep the rung mark which could make a later `c-backward-sws'
1485 ;; move into the line comment instead of over it.
1495 ;; Ensure that an `c-in-sws' range gets broken. Note that it isn't
1496 ;; safe to keep a range that was continuous before the change. E.g:
1497 ;;
1498 ;; #define foo
1499 ;; \
1500 ;; bar
1501 ;;
1502 ;; There can be a "ladder" between "#" and "b". Now, if the newline
1503 ;; after "foo" is removed then "bar" will become part of the cpp
1504 ;; directive instead of a syntactically relevant token. In that
1505 ;; case there's no longer syntactic ws from "#" to "b".
1512 ;; Used by `c-forward-syntactic-ws' to implement the unbounded search.
1513 ;;
1514 ;; This function might do hidden buffer changes.
1517 ;; unmarked part of the simple ws region.
1519 rung-is-marked next-rung-is-marked simple-ws-end
1520 ;; `safe-start' is set when it's safe to cache the start position.
1521 ;; It's not set if we've initially skipped over comments and line
1522 ;; continuations since we might have gone out through the end of a
1523 ;; macro then. This provision makes `c-forward-sws' not populate the
1524 ;; cache in the majority of cases, but otoh is `c-backward-sws' by far
1525 ;; more common.
1526 safe-start)
1528 ;; Skip simple ws and do a quick check on the following character to see
1529 ;; if it's anything that can't start syntactic ws, so we can bail out
1530 ;; early in the majority of cases when there just are a few ws chars.
1537 ;; Find the last rung position to avoid setting properties in all
1538 ;; the cases when the marked rung is complete.
1539 ;; (`next-single-property-change' is certain to move at least one
1540 ;; step forward.)
1543 ;; Got no marked rung here. Since the simple ws might have started
1544 ;; inside a line comment or cpp directive we must set `rung-pos' as
1545 ;; high as possible.
1554 ;; The following search is the main reason that `c-in-sws'
1555 ;; and `c-is-sws' aren't combined to one property.
1559 ;; If the `c-in-sws' region extended past the last
1560 ;; `c-is-sws' char we have to go back a bit.
1567 "c-forward-sws cached move %s -> %s (max %s)"
1574 ;; We'll loop here if there is simple ws after the last rung.
1575 ;; That means that there's been some change in it and it's
1576 ;; possible that we've stepped into another ladder, so extend
1577 ;; the previous one to join with it if there is one, and try to
1578 ;; use the cache again.
1580 "c-forward-sws extending rung with [%s..%s] (max %s)"
1583 ;; Remove any `c-in-sws' property from the last char of
1584 ;; the rung before we mark it with `c-is-sws', so that we
1585 ;; won't connect with the remains of a broken "ladder".
1591 last-put-in-sws-pos rung-pos)))
1598 ;; Skipped over comments. Don't cache at eob in case the buffer
1599 ;; is narrowed.
1610 ;; Skip a preprocessor directive.
1617 ;; Don't cache at eob in case the buffer is narrowed.
1620 ;; We've searched over a piece of non-white syntactic ws. See if this
1621 ;; can be cached.
1627 ;; Cache if we haven't skipped comments only, and if we started
1628 ;; either from a marked rung or from a completely uncached
1629 ;; position.
1634 ;; See if there's a marked rung in the encountered simple ws. If
1635 ;; so then we can cache, unless `safe-start' is nil. Even then
1636 ;; we need to do this to check if the cache can be used for the
1637 ;; next step.
1641 safe-start))
1645 "c-forward-sws caching [%s..%s] - [%s..%s] (max %s)"
1649 ;; Remove the properties for any nested ws that might be cached.
1650 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1651 ;; anyway.
1659 last-put-in-sws-pos rung-pos))
1661 ;; Remove any `c-in-sws' property from the last char of
1662 ;; the rung before we mark it with `c-is-sws', so that we
1663 ;; won't connect with the remains of a broken "ladder".
1666 rung-end-pos))
1669 "c-forward-sws not caching [%s..%s] - [%s..%s] (max %s)"
1673 ;; Set `rung-pos' for the next rung. It's the same thing here as
1674 ;; initially, except that the rung position is set as early as
1675 ;; possible since we can't be in the ending ws of a line comment or
1676 ;; cpp directive now.
1683 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1684 ;; another one after the point (which might occur when editing inside a
1685 ;; comment or macro).
1689 "c-forward-sws clearing at %s for cache separation"
1690 last-put-in-sws-pos)
1694 ;; If at eob we have to clear the last character before the end
1695 ;; instead since the buffer might be narrowed and there might
1696 ;; be a `c-in-sws' after (point-max). In this case it's
1697 ;; necessary to clear both properties.
1699 "c-forward-sws clearing thoroughly at %s for cache separation"
1702 last-put-in-sws-pos))))
1703 )))
1706 ;; Used by `c-backward-syntactic-ws' to implement the unbounded search.
1707 ;;
1708 ;; This function might do hidden buffer changes.
1711 ;; part of the simple ws region.
1713 rung-is-marked simple-ws-beg cmt-skip-pos)
1715 ;; Skip simple horizontal ws and do a quick check on the preceding
1716 ;; character to see if it's anying that can't end syntactic ws, so we can
1717 ;; bail out early in the majority of cases when there just are a few ws
1718 ;; chars. Newlines are complicated in the backward direction, so we can't
1719 ;; skip over them.
1726 ;; Try to find a rung position in the simple ws preceding point, so that
1727 ;; we can get a cache hit even if the last bit of the simple ws has
1728 ;; changed recently.
1734 ;; `rung-pos' will be the earliest marked position, which means that
1735 ;; there might be later unmarked parts in the simple ws region.
1736 ;; It's not worth the effort to fix that; the last part of the
1737 ;; simple ws is also typically edited often, so it could be wasted.
1748 ;; The following search is the main reason that `c-in-sws'
1749 ;; and `c-is-sws' aren't combined to one property.
1753 ;; If the `c-in-sws' region extended past the first
1754 ;; `c-is-sws' char we have to go forward a bit.
1759 "c-backward-sws cached move %s <- %s (min %s)"
1771 t
1773 nil))
1775 ;; We'll loop here if there is simple ws before the first rung.
1776 ;; That means that there's been some change in it and it's
1777 ;; possible that we've stepped into another ladder, so extend
1778 ;; the previous one to join with it if there is one, and try to
1779 ;; use the cache again.
1781 "c-backward-sws extending rung with [%s..%s] (min %s)"
1784 ;; Remove any `c-in-sws' property from the last char of
1785 ;; the rung before we mark it with `c-is-sws', so that we
1786 ;; won't connect with the remains of a broken "ladder".
1789 rung-pos)
1793 last-put-in-sws-pos rung-pos))
1802 ;; Inside a cpp directive. See if it should be skipped over.
1805 ;; Move back over all line continuations in the region skipped
1806 ;; over by `c-backward-comments'. If we go past it then we
1807 ;; started inside the cpp directive.
1816 ;; Don't move past the cpp directive if we began inside
1817 ;; it. Note that the position at the end of the last line
1818 ;; of the macro is also considered to be within it.
1820 nil)
1822 ;; It's worthwhile to spend a little bit of effort on finding
1823 ;; the end of the macro, to get a good `simple-ws-beg'
1824 ;; position for the cache. Note that `c-backward-comments'
1825 ;; could have stepped over some comments before going into
1826 ;; the macro, and then `simple-ws-beg' must be kept on the
1827 ;; same side of those comments.
1834 ;; Might happen if comments after the macro were skipped
1835 ;; over.
1839 t)))
1844 simple-ws-beg)
1845 ;; Skipped over comments. Must put point at the end of
1846 ;; the simple ws at point since we might be after a line
1847 ;; comment or cpp directive that's been partially
1848 ;; narrowed out, and we can't risk marking the simple ws
1849 ;; at the end of it.
1851 t)))
1853 ;; We've searched over a piece of non-white syntactic ws. See if this
1854 ;; can be cached.
1859 ;; Cache if we started either from a marked rung or from a
1860 ;; completely uncached position.
1861 rung-is-marked
1864 ;; Cache if there's a marked rung in the encountered simple ws.
1872 "c-backward-sws caching [%s..%s] - [%s..%s] (min %s)"
1877 ;; Remove the properties for any nested ws that might be cached.
1878 ;; Only necessary for `c-is-sws' since `c-in-sws' will be set
1879 ;; anyway.
1884 ;; Remove any `c-in-sws' property from the last char of
1885 ;; the rung before we mark it with `c-is-sws', so that we
1886 ;; won't connect with the remains of a broken "ladder".
1889 rung-end-pos)
1892 last-put-in-sws-pos simple-ws-beg)
1893 rung-pos)
1898 "c-backward-sws not caching [%s..%s] - [%s..%s] (min %s)"
1904 ))
1906 ;; Make sure that the newly marked `c-in-sws' region doesn't connect to
1907 ;; another one before the point (which might occur when editing inside a
1908 ;; comment or macro).
1912 "c-backward-sws clearing at %s for cache separation"
1915 last-put-in-sws-pos))
1917 ;; If at bob and the buffer is narrowed, we have to clear the
1918 ;; character we're standing on instead since there might be a
1919 ;; `c-in-sws' before (point-min). In this case it's necessary
1920 ;; to clear both properties.
1922 "c-backward-sws clearing thoroughly at %s for cache separation"
1923 last-put-in-sws-pos)
1926 )))
1928 \f
1929 ;; Other whitespace tools
1931 ;; Is the region (beg end) WS, and is there WS (or BOB/EOB) next to the
1932 ;; region? This is a "heuristic" function. .....
1933 ;;
1934 ;; The motivation for the second bit is to check whether removing this
1935 ;; region would coalesce two symbols.
1936 ;;
1937 ;; FIXME!!! This function doesn't check virtual semicolons in any way. Be
1938 ;; careful about using this function for, e.g. AWK. (2007/3/7)
1947 \f
1948 ;; A system for finding noteworthy parens before the point.
1952 ;; The state cache used by `c-parse-state' to cut down the amount of
1953 ;; searching. It's the result from some earlier `c-parse-state' call.
1954 ;;
1955 ;; The use of the cached info is more effective if the next
1956 ;; `c-parse-state' call is on a line close by the one the cached state
1957 ;; was made at; the cache can actually slow down a little if the
1958 ;; cached state was made very far back in the buffer. The cache is
1959 ;; most effective if `c-parse-state' is used on each line while moving
1960 ;; forward.
1964 ;; This is (point-min) when `c-state-cache' was calculated, since a
1965 ;; change of narrowing is likely to affect the parens that are visible
1966 ;; before the point.
1970 ;; This is a position where `c-state-cache' is known to be correct.
1971 ;; It's a position inside one of the recorded unclosed parens or the
1972 ;; top level, but not further nested inside any literal or subparen
1973 ;; that is closed before the last recorded position.
1974 ;;
1975 ;; The exact position is chosen to try to be close to yet earlier than
1976 ;; the position where `c-state-cache' will be called next. Right now
1977 ;; the heuristic is to set it to the position after the last found
1978 ;; closing paren (of any type) before the line on which
1979 ;; `c-parse-state' was called. That is chosen primarily to work well
1980 ;; with refontification of the current line.
1983 ;; Invalidate all info on `c-state-cache' that applies to the buffer
1984 ;; at POS or higher. This is much like `c-whack-state-after', but
1985 ;; it never changes a paren pair element into an open paren element.
1986 ;; Doing that would mean that the new open paren wouldn't have the
1987 ;; required preceding paren pair element.
1991 nil))
1997 nil))
2001 nil)))))
2005 ;; Return the start position for building `c-state-cache' from
2006 ;; scratch.
2008 ;; Go back 2 bods, but ignore any bogus positions returned by
2009 ;; beginning-of-defun (i.e. open paren in column zero).
2019 ;; Find and record all noteworthy parens between some good point
2020 ;; earlier in the file and point. That good point is at least the
2021 ;; beginning of the top-level construct we are in, or the beginning
2022 ;; of the preceding top-level construct if we aren't in one.
2023 ;;
2024 ;; The returned value is a list of the noteworthy parens with the
2025 ;; last one first. If an element in the list is an integer, it's
2026 ;; the position of an open paren which has not been closed before
2027 ;; the point. If an element is a cons, it gives the position of a
2028 ;; closed brace paren pair; the car is the start paren position and
2029 ;; the cdr is the position following the closing paren. Only the
2030 ;; last closed brace paren pair before each open paren and before
2031 ;; the point is recorded, and thus the state never contains two cons
2032 ;; elements in succession.
2033 ;;
2034 ;; Currently no characters which are given paren syntax with the
2035 ;; syntax-table property are recorded, i.e. angle bracket arglist
2036 ;; parens are never present here. Note that this might change.
2037 ;;
2038 ;; BUG: This function doesn't cope entirely well with unbalanced
2039 ;; parens in macros. E.g. in the following case the brace before
2040 ;; the macro isn't balanced with the one after it:
2041 ;;
2042 ;; {
2043 ;; #define X {
2044 ;; }
2045 ;;
2046 ;; This function might do hidden buffer changes.
2053 old-state last-pos brace-pair-open brace-pair-close
2054 pos save-pos)
2057 ;; If the minimum position has changed due to narrowing then we
2058 ;; have to fix the tail of `c-state-cache' accordingly.
2061 ;; If point-min has moved forward then we just need to cut
2062 ;; off a bit of the tail.
2073 ;; If point-min has moved backward then we drop the state
2074 ;; completely. It's possible to do a better job here and
2075 ;; recalculate the top only.
2080 ;; Get the latest position we know are directly inside the
2081 ;; closest containing paren of the cached state.
2089 ;; Take the opportunity to move the cached good position
2090 ;; further down.
2094 ;; Check if `last-pos' is in a macro. If it is, and we're not
2095 ;; in the same macro, we must discard everything on
2096 ;; `c-state-cache' that is inside the macro before using it.
2102 ;; Set `last-pos' again just like above except that there's
2103 ;; no use looking at `c-state-cache-good-pos' here.
2110 ;; If we've moved very far from the last cached position then
2111 ;; it's probably better to redo it from scratch, otherwise we
2112 ;; might spend a lot of time searching from `last-pos' down to
2113 ;; here.
2115 ;; First get the fallback start position. If it turns out
2116 ;; that it's so far back that the cached state is closer then
2117 ;; we'll keep it afterall.
2122 c-state-cache nil
2125 ;; Find the start position for the forward search. (Can't
2126 ;; search in the backward direction since the point might be in
2127 ;; some kind of literal.)
2132 ;; There's a cached state with a containing paren. Pop off
2133 ;; the stale containing sexps from it by going forward out of
2134 ;; parens as far as possible.
2148 ;; The last paren pair we moved out from was a brace
2149 ;; pair. Modify the state to record this as a closed
2150 ;; pair now.
2154 c-state-cache))))
2156 ;; Check if the preceding balanced paren is within a
2157 ;; macro; it should be ignored if we're outside the
2158 ;; macro. There's no need to check any further upwards;
2159 ;; if the macro contains an unbalanced opening paren then
2160 ;; we're smoked anyway.
2170 ;; Have to adjust the cached good position if state has been
2171 ;; popped off.
2178 old-state c-state-cache))
2183 ;; Get the fallback start position.
2186 c-state-cache nil
2193 brace-pair-open nil)
2195 ;; Find the balanced brace pairs. This loop is hot, so it
2196 ;; does ugly tricks to go faster.
2206 ;; Update the cached good position and record the brace
2207 ;; pair, whichever is applicable for the paren we've
2208 ;; just jumped over. But first check that it isn't
2209 ;; inside a macro and the point isn't inside the same
2210 ;; one.
2220 brace-pair-close pos))))))
2222 ;; Record the last brace pair.
2230 brace-pair-close)
2231 c-state-cache)))))
2234 ;; Prepare to loop, but record the open paren only if it's
2235 ;; outside a macro or within the same macro as point, and
2236 ;; if it is a legitimate open paren and not some character
2237 ;; that got an open paren syntax-table property.
2244 ;; Check for known types of parens that we
2245 ;; want to record. The syntax table is not to
2246 ;; be trusted here since the caller might be
2247 ;; using e.g. `c++-template-syntax-table'.
2254 ;; Found a close paren without a corresponding opening
2255 ;; one. Maybe we didn't go back far enough, so try to
2256 ;; scan backward for the start paren and then start over.
2259 c-state-cache nil
2260 c-state-cache-good-pos c-state-cache-start)
2262 ;; Emacs (up to at least 21.2) can get confused by
2263 ;; open parens in column zero inside comments: The
2264 ;; sexp functions can then misbehave and bring us
2265 ;; back to the same point again. Check this so that
2266 ;; we don't get an infinite loop.
2269 c-parsing-error
2275 ;;(message "c-parse-state: %S end: %S" c-state-cache c-state-cache-good-pos)
2276 c-state-cache)))
2278 ;; Debug tool to catch cache inconsistencies.
2290 ;; The cache can actually go further back due to the ad-hoc way
2291 ;; the first paren is found, so try to whack off a bit of its
2292 ;; start before complaining.
2301 res1 res2))))
2302 res1))
2307 'c-debug-parse-state
2314 ;; Whack off any state information from PAREN-STATE which lies
2315 ;; before BUFPOS. Not destructive on PAREN-STATE.
2318 car)
2329 ;; Whack off any state information from PAREN-STATE which lies at or
2330 ;; after BUFPOS. Not destructive on PAREN-STATE.
2335 ;; just check the car, because in a balanced brace
2336 ;; expression, it must be impossible for the corresponding
2337 ;; close brace to be before point, but the open brace to
2338 ;; be after.
2340 nil ; whack it off
2342 ;; its possible that the open brace is before
2343 ;; bufpos, but the close brace is after. In that
2344 ;; case, convert this to a non-cons element. The
2345 ;; rest of the state is before bufpos, so we're
2346 ;; done.
2348 ;; we know that both the open and close braces are
2349 ;; before bufpos, so we also know that everything else
2350 ;; on state is before bufpos.
2353 nil ; whack it off
2354 ;; it's before bufpos, so everything else should too.
2357 nil)))
2360 ;; Return the bufpos of the innermost enclosing open paren before
2361 ;; bufpos, or nil if none was found.
2371 enclosingp))
2374 ;; Return the bufpos of the outermost enclosing open paren, or nil
2375 ;; if none was found.
2382 pos))
2385 ;; Return the closest "safe" position recorded on PAREN-STATE that
2386 ;; is higher up than BUFPOS. Return nil if PAREN-STATE doesn't
2387 ;; contain any. Return nil if BUFPOS is nil, which is useful to
2388 ;; find the closest limit before a given limit that might be nil.
2389 ;;
2390 ;; A "safe" position is a position at or after a recorded open
2391 ;; paren, or after a recorded close paren. The returned position is
2392 ;; thus either the first position after a close brace, or the first
2393 ;; position after an enclosing paren, or at the enclosing paren in
2394 ;; case BUFPOS is immediately after it.
2404 ;; See below.
2407 ;; elem is the position at and not after the opening paren, so
2408 ;; we can go forward one more step unless it's equal to
2409 ;; bufpos. This is useful in some cases avoid an extra paren
2410 ;; level between the safe position and bufpos.
2415 ;; This is used for `font-lock-beginning-of-syntax-function'. It
2416 ;; goes to the closest previous point that is known to be outside
2417 ;; any string literal or comment. `c-state-cache' is used if it has
2418 ;; a position in the vicinity.
2420 elem
2423 ;; Note: Similar code in `c-safe-position'. The
2424 ;; difference is that we accept a safe position at
2425 ;; the point and don't bother to go forward past open
2426 ;; parens.
2441 ;; The position is far back. Try `c-beginning-of-defun-1'
2442 ;; (although we can't be entirely sure it will go to a position
2443 ;; outside a comment or string in current emacsen). FIXME:
2444 ;; Consult `syntax-ppss' here.
2449 \f
2450 ;; Tools for scanning identifiers and other tokens.
2453 "Return non-nil if the point is on or directly after an identifier.
2454 Keywords are recognized and not considered identifiers. If an
2455 identifier is detected, the returned value is its starting position.
2456 If an identifier ends at the point and another begins at it \(can only
2457 happen in Pike) then the point for the preceding one is returned.
2459 Note that this function might do hidden buffer changes. See the
2460 comment at the start of cc-engine.el for more info."
2462 ;; FIXME: Shouldn't this function handle "operator" in C++?
2469 ;; Check for a normal (non-keyword) identifier.
2475 ;; Handle the `<operator> syntax in Pike.
2479 t
2486 ;; Handle the "operator +" syntax in C++.
2503 )))
2506 ;; If the point is at the end of a symbol then skip backward to the
2507 ;; beginning of it. Don't move otherwise. Return non-nil if point
2508 ;; moved.
2509 ;;
2510 ;; This function might do hidden buffer changes.
2513 ;; Handle the `<operator> syntax in Pike.
2519 t
2521 nil)))))
2524 ;; Move to the beginning of the current token. Do not move if not
2525 ;; in the middle of one. BACK-LIMIT may be used to bound the
2526 ;; backward search; if given it's assumed to be at the boundary
2527 ;; between two tokens. Return non-nil if the point is moved, nil
2528 ;; otherwise.
2529 ;;
2530 ;; This function might do hidden buffer changes.
2537 ;; `c-nonsymbol-token-regexp' should always match
2538 ;; since we've skipped backward over punctuator
2539 ;; or paren syntax, but consume one char in case
2540 ;; it doesn't so that we don't leave point before
2541 ;; some earlier incorrect token.
2548 ;; Move to the end of the current token. Do not move if not in the
2549 ;; middle of one. BACK-LIMIT may be used to bound the backward
2550 ;; search; if given it's assumed to be at the boundary between two
2551 ;; tokens. Return non-nil if the point is moved, nil otherwise.
2552 ;;
2553 ;; This function might do hidden buffer changes.
2561 ;; `c-nonsymbol-token-regexp' should always match since
2562 ;; we've skipped backward over punctuator or paren
2563 ;; syntax, but move forward in case it doesn't so that
2564 ;; we don't leave point earlier than we started with.
2580 "Move forward by tokens.
2581 A token is defined as all symbols and identifiers which aren't
2583 treated properly). Point is always either left at the beginning of a
2584 token or not moved at all. COUNT specifies the number of tokens to
2585 move; a negative COUNT moves in the opposite direction. A COUNT of 0
2586 moves to the next token beginning only if not already at one. If
2587 BALANCED is true, move over balanced parens, otherwise move into them.
2588 Also, if BALANCED is true, never move out of an enclosing paren.
2590 LIMIT sets the limit for the movement and defaults to the point limit.
2591 The case when LIMIT is set in the middle of a token, comment or macro
2592 is handled correctly, i.e. the point won't be left there.
2594 Return the number of tokens left to move \(positive or negative). If
2595 BALANCED is true, a move over a balanced paren counts as one. Note
2596 that if COUNT is 0 and no appropriate token beginning is found, 1 will
2597 be returned. Thus, a return value of 0 guarantees that point is at
2598 the requested position and a return value less \(without signs) than
2599 COUNT guarantees that point is at the beginning of some token.
2601 Note that this function might do hidden buffer changes. See the
2602 comment at the start of cc-engine.el for more info."
2609 c-jump-syntax-balanced
2610 c-jump-syntax-unbalanced))
2615 ;; If count is zero we should jump if in the middle of a token.
2622 ;; Skip whitespace. Count this as a move if we did in
2623 ;; fact move.
2627 ;; Moved out of bounds. Make sure the returned count isn't zero.
2632 ;; Use `condition-case' to avoid having the limit tests
2633 ;; inside the loop.
2641 t)
2644 t)
2645 ;; `c-nonsymbol-token-regexp' above should always
2646 ;; match if there are correct tokens. Try to
2647 ;; widen to see if the limit was set in the
2648 ;; middle of one, else fall back to treating
2649 ;; the offending thing as a one character token.
2654 nil)
2657 t))))
2667 count)))
2670 "Move backward by tokens.
2671 See `c-forward-token-2' for details."
2679 c-jump-syntax-balanced
2680 c-jump-syntax-unbalanced))
2684 ;; The count is zero so try to skip to the beginning of the
2685 ;; current token.
2689 ;; The limit is inside the same token, so return 1.
2692 ;; We're not in the middle of a token. If there's
2693 ;; whitespace after the point then we must move backward,
2694 ;; so set count to 1 in that case.
2696 ;; If we're looking at a '#' that might start a cpp
2697 ;; directive then we have to do a more elaborate check.
2710 ;; Use `condition-case' to avoid having to check for buffer
2711 ;; limits in `backward-char', `scan-sexps' and `goto-char' below.
2720 ;; This can be very inefficient if there's a long
2721 ;; sequence of operator tokens without any separation.
2722 ;; That doesn't happen in practice, anyway.
2732 count)))
2735 "Like `c-forward-token-2' but doesn't treat multicharacter operator
2737 characters are jumped over character by character. This function is
2738 for compatibility only; it's only a wrapper over `c-forward-token-2'."
2743 "Like `c-backward-token-2' but doesn't treat multicharacter operator
2745 characters are jumped over character by character. This function is
2746 for compatibility only; it's only a wrapper over `c-backward-token-2'."
2750 \f
2751 ;; Tools for doing searches restricted to syntactically relevant text.
2754 paren-level not-inside-token
2755 lookbehind-submatch)
2756 "Like `re-search-forward', but only report matches that are found
2757 in syntactically significant text. I.e. matches in comments, macros
2758 or string literals are ignored. The start point is assumed to be
2759 outside any comment, macro or string literal, or else the content of
2760 that region is taken as syntactically significant text.
2762 If PAREN-LEVEL is non-nil, an additional restriction is added to
2763 ignore matches in nested paren sexps. The search will also not go
2764 outside the current list sexp, which has the effect that if the point
2765 should be moved to BOUND when no match is found \(i.e. NOERROR is
2766 neither nil nor t), then it will be at the closing paren if the end of
2767 the current list sexp is encountered first.
2769 If NOT-INSIDE-TOKEN is non-nil, matches in the middle of tokens are
2770 ignored. Things like multicharacter operators and special symbols
2772 constants.
2774 If LOOKBEHIND-SUBMATCH is non-nil, it's taken as a number of a
2775 subexpression in REGEXP. The end of that submatch is used as the
2776 position to check for syntactic significance. If LOOKBEHIND-SUBMATCH
2777 isn't used or if that subexpression didn't match then the start
2779 subexpression is never tested before the starting position, so it
2783 subexpression can match the end of nonwhite syntactic whitespace,
2784 i.e. the end of comments or cpp directives. This since the function
2785 skips over such things before resuming the search. It's on the other
2787 matches syntactic whitespace.
2789 Bug: Unbalanced parens inside cpp directives are currently not handled
2790 correctly \(i.e. they don't get ignored as they should) when
2791 PAREN-LEVEL is set.
2793 Note that this function might do hidden buffer changes. See the
2794 comment at the start of cc-engine.el for more info."
2799 ;;(message "c-syntactic-re-search-forward %s %s %S" (point) bound regexp)
2802 tmp
2803 ;; Start position for the last search.
2804 search-pos
2805 ;; The `parse-partial-sexp' state between the start position
2806 ;; and the point.
2807 state
2808 ;; The current position after the last state update. The next
2809 ;; `parse-partial-sexp' continues from here.
2811 ;; The position at which to check the state and the state
2812 ;; there. This is separate from `state-pos' since we might
2813 ;; need to back up before doing the next search round.
2814 check-pos check-state
2815 ;; Last position known to end a token.
2817 ;; Set when a valid match is found.
2818 found)
2836 state-pos check-pos paren-level nil state))
2838 check-state state))
2840 ;; NOTE: If we got a look behind subexpression and get
2841 ;; an insignificant match in something that isn't
2842 ;; syntactic whitespace (i.e. strings or in nested
2843 ;; parentheses), then we can never skip more than a
2844 ;; single character from the match start position
2845 ;; (i.e. `state-pos' here) before continuing the
2846 ;; search. That since the look behind subexpression
2847 ;; might match the end of the insignificant region in
2848 ;; the next search.
2852 ;; Match inside a line comment. Skip to eol. Use
2853 ;; `re-search-forward' instead of `skip-chars-forward' to get
2854 ;; the right bound behavior.
2858 ;; Match inside a block comment. Skip to the '*/'.
2865 ;; Match in the middle of the opener of a block or line
2866 ;; comment.
2871 ;; The last `parse-partial-sexp' above might have
2872 ;; stopped short of the real check position if the end
2873 ;; of the current sexp was encountered in paren-level
2874 ;; mode. The checks above are always false in that
2875 ;; case, and since they can do better skipping in
2876 ;; lookbehind-submatch mode, we do them before
2877 ;; checking the paren level.
2881 ;; Check the paren level first since we're short of the
2882 ;; syntactic checking position if the end of the
2883 ;; current sexp was encountered by `parse-partial-sexp'.
2886 ;; Inside a nested paren sexp.
2888 ;; See the NOTE above.
2890 ;; Skip out of the paren quickly.
2894 ;; Have exited the current paren sexp.
2897 ;; The last `parse-partial-sexp' call above
2898 ;; has left us just after the closing paren
2899 ;; in this case, so we can modify the bound
2900 ;; to leave the point at the right position
2901 ;; upon return.
2903 nil)
2907 ;; Match inside a string.
2910 ;; See the NOTE above.
2912 ;; Skip to the end of the string before continuing.
2921 continue)))
2926 ;; Match inside a macro. Skip to the end of it.
2934 (< check-pos
2940 ;; Inside a token.
2942 ;; See the NOTE above.
2947 ;; A real match.
2949 nil)))
2951 ;; Should loop to search again, but take care to avoid
2952 ;; looping on the same spot.
2956 nil
2959 t))))
2965 ;;(message "c-syntactic-re-search-forward done %s" (or (match-end 0) (point)))
2972 ;; Search failed. Set point as appropriate.
2976 nil)))
2981 ;; Return the start of the literal point is in, or nil.
2982 ;; We read and write the variables `safe-pos', `safe-pos-list', `state'
2983 ;; bound in the caller.
2985 ;; Use `parse-partial-sexp' from a safe position down to the point to check
2986 ;; if it's outside comments and strings.
2989 ;; Pick a safe position as close to the point as possible.
2990 ;;
2991 ;; FIXME: Consult `syntax-ppss' here if our cache doesn't give a good
2992 ;; position.
3005 ;; Cache positions along the way to use if we have to back up more. We
3006 ;; cache every closing paren on the same level. If the paren cache is
3007 ;; relevant in this region then we're typically already on the same
3008 ;; level as the target position. Note that we might cache positions
3009 ;; after opening parens in case safe-pos is in a nested list. That's
3010 ;; both uncommon and harmless.
3018 ;; If the state contains the start of the containing sexp we cache that
3019 ;; position too, so that parse-partial-sexp in the next run has a bigger
3020 ;; chance of starting at the same level as the target position and thus
3021 ;; will get more good safe positions into the list.
3027 ;; Inside string or comment. Continue search at the
3028 ;; beginning of it.
3032 "Like `skip-chars-backward' but only look at syntactically relevant chars,
3033 i.e. don't stop at positions inside syntactic whitespace or string
3034 literals. Preprocessor directives are also ignored, with the exception
3035 of the one that the point starts within, if any. If LIMIT is given,
3036 it's assumed to be at a syntactically relevant position.
3038 If PAREN-LEVEL is non-nil, the function won't stop in nested paren
3039 sexps, and the search will also not go outside the current paren sexp.
3040 However, if LIMIT or the buffer limit is reached inside a nested paren
3041 then the point will be left at the limit.
3043 Non-nil is returned if the point moved, nil otherwise.
3045 Note that this function might do hidden buffer changes. See the
3046 comment at the start of cc-engine.el for more info."
3049 state-2
3050 ;; A list of syntactically relevant positions in descending
3051 ;; order. It's used to avoid scanning repeatedly over
3052 ;; potentially large regions with `parse-partial-sexp' to verify
3053 ;; each position. Used in `c-ssb-lit-begin'
3054 safe-pos-list
3055 ;; The result from `c-beginning-of-macro' at the start position or the
3056 ;; start position itself if it isn't within a macro. Evaluated on
3057 ;; demand.
3058 start-macro-beg
3059 ;; The earliest position after the current one with the same paren
3060 ;; level. Used only when `paren-level' is set.
3061 lit-beg
3066 ;; The next loop "tries" to find the end point each time round,
3067 ;; loops when it hasn't succeeded.
3075 ;; Don't stop inside a literal
3078 t)
3086 ;; Not at the right level.
3089 ;; We stop above if we go out of a paren.
3090 ;; Now check whether it precedes or is
3091 ;; nested in the starting sexp.
3095 pps-end-pos paren-level-pos
3096 nil nil state-2))
3099 ;; We've stopped short of the starting position
3100 ;; so the hit was inside a nested list. Go up
3101 ;; until we are at the right level.
3110 nil)
3111 t))
3114 nil))
3116 ;; The hit was outside the list at the start
3117 ;; position. Go to the start of the list and exit.
3119 nil))
3122 ;; Inside a macro.
3130 t
3132 ;; It's inside the same macro we started in so it's
3133 ;; a relevant match.
3135 nil))))))
3139 ;; Skip syntactic ws afterwards so that we don't stop at the
3140 ;; end of a comment if `skip-chars' is something like "^/".
3144 ;; We might want to extend this with more useful return values in
3145 ;; the future.
3148 ;; The following is an alternative implementation of
3149 ;; `c-syntactic-skip-backward' that uses backward movement to keep
3150 ;; track of the syntactic context. It turned out to be generally
3151 ;; slower than the one above which uses forward checks from earlier
3152 ;; safe positions.
3153 ;;
3154 ;;(defconst c-ssb-stop-re
3155 ;; ;; The regexp matching chars `c-syntactic-skip-backward' needs to
3156 ;; ;; stop at to avoid going into comments and literals.
3157 ;; (concat
3158 ;; ;; Match comment end syntax and string literal syntax. Also match
3159 ;; ;; '/' for block comment endings (not covered by comment end
3160 ;; ;; syntax).
3161 ;; "\\s>\\|/\\|\\s\""
3162 ;; (if (memq 'gen-string-delim c-emacs-features)
3163 ;; "\\|\\s|"
3164 ;; "")
3165 ;; (if (memq 'gen-comment-delim c-emacs-features)
3166 ;; "\\|\\s!"
3167 ;; "")))
3168 ;;
3169 ;;(defconst c-ssb-stop-paren-re
3170 ;; ;; Like `c-ssb-stop-re' but also stops at paren chars.
3171 ;; (concat c-ssb-stop-re "\\|\\s(\\|\\s)"))
3172 ;;
3173 ;;(defconst c-ssb-sexp-end-re
3174 ;; ;; Regexp matching the ending syntax of a complex sexp.
3175 ;; (concat c-string-limit-regexp "\\|\\s)"))
3176 ;;
3177 ;;(defun c-syntactic-skip-backward (skip-chars &optional limit paren-level)
3178 ;; "Like `skip-chars-backward' but only look at syntactically relevant chars,
3179 ;;i.e. don't stop at positions inside syntactic whitespace or string
3180 ;;literals. Preprocessor directives are also ignored. However, if the
3181 ;;point is within a comment, string literal or preprocessor directory to
3182 ;;begin with, its contents is treated as syntactically relevant chars.
3183 ;;If LIMIT is given, it limits the backward search and the point will be
3184 ;;left there if no earlier position is found.
3185 ;;
3186 ;;If PAREN-LEVEL is non-nil, the function won't stop in nested paren
3187 ;;sexps, and the search will also not go outside the current paren sexp.
3188 ;;However, if LIMIT or the buffer limit is reached inside a nested paren
3189 ;;then the point will be left at the limit.
3190 ;;
3191 ;;Non-nil is returned if the point moved, nil otherwise.
3192 ;;
3193 ;;Note that this function might do hidden buffer changes. See the
3194 ;;comment at the start of cc-engine.el for more info."
3195 ;;
3196 ;; (save-restriction
3197 ;; (when limit
3198 ;; (narrow-to-region limit (point-max)))
3199 ;;
3200 ;; (let ((start (point)))
3201 ;; (catch 'done
3202 ;; (while (let ((last-pos (point))
3203 ;; (stop-pos (progn
3204 ;; (skip-chars-backward skip-chars)
3205 ;; (point))))
3206 ;;
3207 ;; ;; Skip back over the same region as
3208 ;; ;; `skip-chars-backward' above, but keep to
3209 ;; ;; syntactically relevant positions.
3210 ;; (goto-char last-pos)
3211 ;; (while (and
3212 ;; ;; `re-search-backward' with a single char regexp
3213 ;; ;; should be fast.
3214 ;; (re-search-backward
3215 ;; (if paren-level c-ssb-stop-paren-re c-ssb-stop-re)
3216 ;; stop-pos 'move)
3217 ;;
3218 ;; (progn
3219 ;; (cond
3220 ;; ((looking-at "\\s(")
3221 ;; ;; `paren-level' is set and we've found the
3222 ;; ;; start of the containing paren.
3223 ;; (forward-char)
3224 ;; (throw 'done t))
3225 ;;
3226 ;; ((looking-at c-ssb-sexp-end-re)
3227 ;; ;; We're at the end of a string literal or paren
3228 ;; ;; sexp (if `paren-level' is set).
3229 ;; (forward-char)
3230 ;; (condition-case nil
3231 ;; (c-backward-sexp)
3232 ;; (error
3233 ;; (goto-char limit)
3234 ;; (throw 'done t))))
3235 ;;
3236 ;; (t
3237 ;; (forward-char)
3238 ;; ;; At the end of some syntactic ws or possibly
3239 ;; ;; after a plain '/' operator.
3240 ;; (let ((pos (point)))
3241 ;; (c-backward-syntactic-ws)
3242 ;; (if (= pos (point))
3243 ;; ;; Was a plain '/' operator. Go past it.
3244 ;; (backward-char)))))
3245 ;;
3246 ;; (> (point) stop-pos))))
3247 ;;
3248 ;; ;; Now the point is either at `stop-pos' or at some
3249 ;; ;; position further back if `stop-pos' was at a
3250 ;; ;; syntactically irrelevant place.
3251 ;;
3252 ;; ;; Skip additional syntactic ws so that we don't stop
3253 ;; ;; at the end of a comment if `skip-chars' is
3254 ;; ;; something like "^/".
3255 ;; (c-backward-syntactic-ws)
3256 ;;
3257 ;; (< (point) stop-pos))))
3258 ;;
3259 ;; ;; We might want to extend this with more useful return values
3260 ;; ;; in the future.
3261 ;; (/= (point) start))))
3263 \f
3264 ;; Tools for handling comments and string literals.
3267 "Return the type of literal point is in, if any.
3268 The return value is `c' if in a C-style comment, `c++' if in a C++
3269 style comment, `string' if in a string literal, `pound' if DETECT-CPP
3270 is non-nil and in a preprocessor line, or nil if somewhere else.
3271 Optional LIM is used as the backward limit of the search. If omitted,
3272 or nil, `c-beginning-of-defun' is used.
3274 The last point calculated is cached if the cache is enabled, i.e. if
3275 `c-in-literal-cache' is bound to a two element vector.
3277 Note that this function might do hidden buffer changes. See the
3278 comment at the start of cc-engine.el for more info."
3294 ;; cache this result if the cache is enabled
3297 rtn)))
3299 ;; XEmacs has a built-in function that should make this much quicker.
3300 ;; I don't think we even need the cache, which makes our lives more
3301 ;; complicated anyway. In this case, lim is only used to detect
3302 ;; cpp directives.
3303 ;;
3304 ;; Note that there is a bug in Xemacs's buffer-syntactic-context when used in
3305 ;; conjunction with syntax-table-properties. The bug is present in, e.g.,
3306 ;; Xemacs 21.4.4. It manifested itself thus:
3307 ;;
3308 ;; Starting with an empty AWK Mode buffer, type
3309 ;; /regexp/ {<C-j>
3310 ;; Point gets wrongly left at column 0, rather than being indented to tab-width.
3311 ;;
3312 ;; AWK Mode is designed such that when the first / is typed, it gets the
3313 ;; syntax-table property "string fence". When the second / is typed, BOTH /s
3314 ;; are given the s-t property "string". However, buffer-syntactic-context
3315 ;; fails to take account of the change of the s-t property on the opening / to
3316 ;; "string", and reports that the { is within a string started by the second /.
3317 ;;
3318 ;; The workaround for this is for the AWK Mode initialisation to switch the
3319 ;; defalias for c-in-literal to c-slow-in-literal. This will slow down other
3320 ;; cc-modes in Xemacs whenever an awk-buffer has been initialised.
3321 ;;
3322 ;; (Alan Mackenzie, 2003/4/30).
3325 ;; This function might do hidden buffer changes.
3338 ;; The defalias above isn't enough to shut up the byte compiler.
3342 "Return a cons of the beginning and end positions of the comment or
3343 string surrounding point (including both delimiters), or nil if point
3345 to start parsing from. If NEAR is non-nil, then the limits of any
3347 spaces and tabs between point and the literal. The search for such a
3348 literal is done first in forward direction. If NOT-IN-DELIMITER is
3349 non-nil, the case when point is inside a starting delimiter won't be
3350 recognized. This only has effect for comments which have starting
3351 delimiters with more than one character.
3353 Note that this function might do hidden buffer changes. See the
3354 comment at the start of cc-engine.el for more info."
3376 ;; We're standing in a comment starter.
3383 ;; Search forward for a literal.
3395 ;; Search backward.
3406 ;; Block comment. Due to the nature of line
3407 ;; comments, they will always be covered by the
3408 ;; normal case above.
3411 ;; If LIM is bogus, beg will be bogus.
3415 ))))
3417 ;; In case external callers use this; it did have a docstring.
3421 "If the argument is a cons of two buffer positions (such as returned by
3422 `c-literal-limits'), and that range contains a C++ style line comment,
3423 then an extended range is returned that contains all adjacent line
3424 comments (i.e. all comments that starts in the same column with no
3425 empty lines or non-whitespace characters between them). Otherwise the
3426 argument is returned.
3428 Note that this function might do hidden buffer changes. See the
3429 comment at the start of cc-engine.el for more info."
3440 ;; Got to take care in the backward direction to handle
3441 ;; comments which are preceded by code.
3455 range)
3459 "Convenience function that given the result of `c-literal-limits',
3460 returns nil or the type of literal that the range surrounds, one
3461 of the symbols 'c, 'c++ or 'string. It's much faster than using
3462 `c-in-literal' and is intended to be used when you need both the
3463 type of a literal and its limits.
3465 Note that this function might do hidden buffer changes. See the
3466 comment at the start of cc-engine.el for more info."
3477 range))
3479 \f
3480 ;; `c-find-decl-spots' and accompanying stuff.
3482 ;; Variables used in `c-find-decl-spots' to cache the search done for
3483 ;; the first declaration in the last call. When that function starts,
3484 ;; it needs to back up over syntactic whitespace to look at the last
3485 ;; token before the region being searched. That can sometimes cause
3486 ;; moves back and forth over a quite large region of comments and
3487 ;; macros, which would be repeated for each changed character when
3488 ;; we're called during fontification, since font-lock refontifies the
3489 ;; current line for each change. Thus it's worthwhile to cache the
3490 ;; first match.
3491 ;;
3492 ;; `c-find-decl-syntactic-pos' is a syntactically relevant position in
3493 ;; the syntactic whitespace less or equal to some start position.
3494 ;; There's no cached value if it's nil.
3495 ;;
3496 ;; `c-find-decl-match-pos' is the match position if
3497 ;; `c-find-decl-prefix-search' matched before the syntactic whitespace
3498 ;; at `c-find-decl-syntactic-pos', or nil if there's no such match.
3509 ; (defface c-debug-decl-spot-face
3510 ; '((t (:background "Turquoise")))
3511 ; "Debug face to mark the spots where `c-find-decl-spots' stopped.")
3512 ; (defface c-debug-decl-sws-face
3513 ; '((t (:background "Khaki")))
3514 ; "Debug face to mark the syntactic whitespace between the declaration
3515 ; spots and the preceding token end.")
3531 ;; Macro used inside `c-find-decl-spots'. It ought to be a defun,
3532 ;; but it contains lots of free variables that refer to things
3533 ;; inside `c-find-decl-spots'. The point is left at `cfd-match-pos'
3534 ;; if there is a match, otherwise at `cfd-limit'.
3535 ;;
3536 ;; This macro might do hidden buffer changes.
3539 ;; Find the next property match position if we haven't got one already.
3550 ;; Find the next `c-decl-prefix-or-start-re' match if we haven't
3551 ;; got one already.
3561 ;; Match. Check if it's inside a comment or string literal.
3564 ;; Matched the end of a token preceding a decl spot.
3568 ;; Matched a token that start a decl spot.
3571 c-literal-faces)
3573 ;; No match. Finish up and exit the loop.
3575 nil)
3577 ;; Skip out of comments and string literals.
3584 ;; If we matched at the decl start, we have to back up over the
3585 ;; preceding syntactic ws to set `cfd-match-pos' and to catch
3586 ;; any decl spots in the syntactic ws.
3591 ;; Choose whichever match is closer to the start.
3594 cfd-re-match nil)
3596 cfd-prop-match nil))
3601 ;; Skip forward past comments only so we don't skip macros.
3603 ;; Set the position to continue at. We can avoid going over
3604 ;; the comments skipped above a second time, but it's possible
3605 ;; that the comment skipping has taken us past `cfd-prop-match'
3606 ;; since the property might be used inside comments.
3612 ;; Call CFD-FUN for each possible spot for a declaration, cast or
3613 ;; label from the point to CFD-LIMIT.
3614 ;;
3615 ;; CFD-FUN is called with point at the start of the spot. It's
3616 ;; passed two arguments: The first is the end position of the token
3617 ;; preceding the spot, or 0 for the implicit match at bob. The
3618 ;; second is a flag that is t when the match is inside a macro. If
3619 ;; CFD-FUN adds `c-decl-end' properties somewhere below the current
3620 ;; spot, it should return non-nil to ensure that the next search
3621 ;; will find them.
3622 ;;
3623 ;; Such a spot is:
3624 ;; o The first token after bob.
3625 ;; o The first token after the end of submatch 1 in
3626 ;; `c-decl-prefix-or-start-re' when that submatch matches.
3627 ;; o The start of each `c-decl-prefix-or-start-re' match when
3628 ;; submatch 1 doesn't match.
3629 ;; o The first token after the end of each occurrence of the
3630 ;; `c-type' text property with the value `c-decl-end', provided
3631 ;; `c-type-decl-end-used' is set.
3632 ;;
3633 ;; Only a spot that match CFD-DECL-RE and whose face is in the
3634 ;; CFD-FACE-CHECKLIST list causes CFD-FUN to be called. The face
3635 ;; check is disabled if CFD-FACE-CHECKLIST is nil.
3636 ;;
3637 ;; If the match is inside a macro then the buffer is narrowed to the
3638 ;; end of it, so that CFD-FUN can investigate the following tokens
3639 ;; without matching something that begins inside a macro and ends
3640 ;; outside it. It's to avoid this work that the CFD-DECL-RE and
3641 ;; CFD-FACE-CHECKLIST checks exist.
3642 ;;
3643 ;; The spots are visited approximately in order from top to bottom.
3644 ;; It's however the positions where `c-decl-prefix-or-start-re'
3645 ;; matches and where `c-decl-end' properties are found that are in
3646 ;; order. Since the spots often are at the following token, they
3647 ;; might be visited out of order insofar as more spots are reported
3648 ;; later on within the syntactic whitespace between the match
3649 ;; positions and their spots.
3650 ;;
3651 ;; It's assumed that comments and strings are fontified in the
3652 ;; searched range.
3653 ;;
3654 ;; This is mainly used in fontification, and so has an elaborate
3655 ;; cache to handle repeated calls from the same start position; see
3656 ;; the variables above.
3657 ;;
3658 ;; All variables in this function begin with `cfd-' to avoid name
3659 ;; collision with the (dynamically bound) variables used in CFD-FUN.
3660 ;;
3661 ;; This function might do hidden buffer changes.
3665 ;; The end of the token preceding the decl spot last found
3666 ;; with `c-decl-prefix-or-start-re'. `cfd-limit' if there's
3667 ;; no match.
3668 cfd-re-match
3669 ;; The end position of the last `c-decl-prefix-or-start-re'
3670 ;; match. If this is greater than `cfd-continue-pos', the
3671 ;; next regexp search is started here instead.
3673 ;; The end of the last `c-decl-end' found by
3674 ;; `c-find-decl-prefix-search'. `cfd-limit' if there's no
3675 ;; match. If searching for the property isn't needed then we
3676 ;; disable it by setting it to `cfd-limit' directly.
3678 ;; The end of the token preceding the decl spot last found by
3679 ;; `c-find-decl-prefix-search'. 0 for the implicit match at
3680 ;; bob. `cfd-limit' if there's no match. In other words,
3681 ;; this is the minimum of `cfd-re-match' and `cfd-prop-match'.
3683 ;; The position to continue searching at.
3684 cfd-continue-pos
3685 ;; The position of the last "real" token we've stopped at.
3686 ;; This can be greater than `cfd-continue-pos' when we get
3687 ;; hits inside macros or at `c-decl-end' positions inside
3688 ;; comments.
3690 ;; The end position of the last entered macro.
3693 ;; Initialize by finding a syntactically relevant start position
3694 ;; before the point, and do the first `c-decl-prefix-or-start-re'
3695 ;; search unless we're at bob.
3698 ;; Must back up a bit since we look for the end of the previous
3699 ;; statement or declaration, which is earlier than the first
3700 ;; returned match.
3703 ;; First we need to move to a syntactically relevant position.
3704 ;; Begin by backing out of comment or string literals.
3707 ;; Try to use the faces to back up to the start of the
3708 ;; literal. FIXME: What if the point is on a declaration
3709 ;; inside a comment?
3715 ;; XEmacs doesn't fontify the quotes surrounding string
3716 ;; literals.
3725 ;; Don't trust the literal to contain only literal faces
3726 ;; (the font lock package might not have fontified the
3727 ;; start of it at all, for instance) so check that we have
3728 ;; arrived at something that looks like a start or else
3729 ;; resort to `c-literal-limits'.
3736 ;; The start is in a literal. If the limit is in the same
3737 ;; one we don't have to find a syntactic position etc. We
3738 ;; only check that if the limit is at or before bonl to save
3739 ;; time; it covers the by far most common case when font-lock
3740 ;; refontifies the current line only.
3751 ;; Completely inside a literal. Set up variables to trig the
3752 ;; (< cfd-continue-pos cfd-start-pos) case below and it'll
3753 ;; find a suitable start position.
3756 ;; Check if the region might be completely inside a macro, to
3757 ;; optimize that like the completely-inside-literal above.
3764 ;; (Maybe) completely inside a macro. Only need to trig the
3765 ;; (< cfd-continue-pos cfd-start-pos) case below to make it
3766 ;; set things up.
3768 start-in-macro t))
3771 ;; Back out of any macro so we don't miss any declaration
3772 ;; that could follow after it.
3776 ;; Now we're at a proper syntactically relevant position so we
3777 ;; can use the cache. But first clear it if it applied
3778 ;; further down.
3783 ;; Don't have to do this if the cache is relevant here,
3784 ;; typically if the same line is refontified again. If
3785 ;; we're just some syntactic whitespace further down we can
3786 ;; still use the cache to limit the skipping.
3789 ;; If we hit `c-find-decl-syntactic-pos' and
3790 ;; `c-find-decl-match-pos' is set then we install the cached
3791 ;; values. If we hit `c-find-decl-syntactic-pos' and
3792 ;; `c-find-decl-match-pos' is nil then we know there's no decl
3793 ;; prefix in the whitespace before `c-find-decl-syntactic-pos'
3794 ;; and so we can continue the search from this point. If we
3795 ;; didn't hit `c-find-decl-syntactic-pos' then we're now in
3796 ;; the right spot to begin searching anyway.
3798 c-find-decl-match-pos)
3800 cfd-continue-pos syntactic-pos)
3805 ;; Always consider bob a match to get the first
3806 ;; declaration in the file. Do this separately instead of
3807 ;; letting `c-decl-prefix-or-start-re' match bob, so that
3808 ;; regexp always can consume at least one character to
3809 ;; ensure that we won't get stuck in an infinite loop.
3814 ;; Do an initial search now. In the bob case above it's
3815 ;; only done to search for a `c-decl-end' spot.
3819 cfd-match-pos)))))
3821 ;; Advance `cfd-continue-pos' if it's before the start position.
3822 ;; The closest continue position that might have effect at or
3823 ;; after the start depends on what we started in. This also
3824 ;; finds a suitable start position in the special cases when the
3825 ;; region is completely within a literal or macro.
3830 ;; If we're in a macro then it's the closest preceding token
3831 ;; in the macro. Check this before `start-in-literal',
3832 ;; since if we're inside a literal in a macro, the preceding
3833 ;; token is earlier than any `c-decl-end' spot inside the
3834 ;; literal (comment).
3836 ;; The only syntactic ws in macros are comments.
3842 ;; If we're in a comment it can only be the closest
3843 ;; preceding `c-decl-end' position within that comment, if
3844 ;; any. Go back to the beginning of such a property so that
3845 ;; `c-find-decl-prefix-search' will find the end of it.
3846 ;; (Can't stop at the end and install it directly on
3847 ;; `cfd-prop-match' since that variable might be cleared
3848 ;; after `cfd-fun' below.)
3849 ;;
3850 ;; Note that if the literal is a string then the property
3851 ;; search will simply skip to the beginning of it right
3852 ;; away.
3864 ;; Didn't find any property inside the comment, so we can
3865 ;; skip it entirely. (This won't skip past a string, but
3866 ;; that'll be handled quickly by the next
3867 ;; `c-find-decl-prefix-search' anyway.)
3873 ;; If we started in normal code, the only match that might
3874 ;; apply before the start is what we already got in
3875 ;; `cfd-match-pos' so we can continue at the start position.
3876 ;; (Note that we don't get here if the first match is below
3877 ;; it.)
3880 ;; Delete found matches if they are before our new continue
3881 ;; position, so that `c-find-decl-prefix-search' won't back up
3882 ;; to them later on.
3890 ;; This is the normal case and we got a proper syntactic
3891 ;; position. If there's a match then it's always outside
3892 ;; macros and comments, so advance to the next token and set
3893 ;; `cfd-token-pos'. The loop below will later go back using
3894 ;; `cfd-continue-pos' to fix declarations inside the
3895 ;; syntactic ws.
3903 ;; Have one of the special cases when the region is completely
3904 ;; within a literal or macro. `cfd-continue-pos' is set to a
3905 ;; good start position for the search, so do it.
3908 ;; Now loop. Round what? (ACM, 2006/7/5). We already got the first match.
3915 ;; Kludge to filter out matches on the "<" that
3916 ;; aren't open parens, for the sake of languages
3917 ;; that got `c-recognize-<>-arglists' set.
3922 ;; If `cfd-continue-pos' is less or equal to
3923 ;; `cfd-token-pos', we've got a hit inside a macro
3924 ;; that's in the syntactic whitespace before the last
3925 ;; "real" declaration we've checked. If they're equal
3926 ;; we've arrived at the declaration a second time, so
3927 ;; there's nothing to do.
3931 ;; If `cfd-continue-pos' is less than `cfd-token-pos'
3932 ;; we're still searching for declarations embedded in
3933 ;; the syntactic whitespace. In that case we need
3934 ;; only to skip comments and not macros, since they
3935 ;; can't be nested, and that's already been done in
3936 ;; `c-find-decl-prefix-search'.
3941 ;; Continue if the following token fails the
3942 ;; CFD-DECL-RE and CFD-FACE-CHECKLIST checks.
3949 t)))
3960 ;; Narrow to the end of the macro if we got a hit inside
3961 ;; one, to avoid recognizing things that start inside the
3962 ;; macro and end outside it.
3964 ;; Not in the same macro as in the previous round.
3975 t
3978 t)
3979 ;; The matched token was the last thing in the macro,
3980 ;; so the whole match is bogus.
3982 nil))))
3989 ;; Restore limits if we did macro narrowment above.
3997 \f
3998 ;; A cache for found types.
4000 ;; Buffer local variable that contains an obarray with the types we've
4001 ;; found. If a declaration is recognized somewhere we record the
4002 ;; fully qualified identifier in it to recognize it as a type
4003 ;; elsewhere in the file too. This is not accurate since we do not
4004 ;; bother with the scoping rules of the languages, but in practice the
4005 ;; same name is seldom used as both a type and something else in a
4006 ;; file, and we only use this as a last resort in ambiguous cases (see
4007 ;; `c-forward-decl-or-cast-1').
4008 ;;
4009 ;; Not every type need be in this cache. However, things which have
4010 ;; ceased to be types must be removed from it.
4011 ;;
4012 ;; Template types in C++ are added here too but with the template
4013 ;; arglist replaced with "<>" in references or "<" for the one in the
4014 ;; primary type. E.g. the type "Foo<A,B>::Bar<C>" is stored as
4015 ;; "Foo<>::Bar<". This avoids storing very long strings (since C++
4016 ;; template specs can be fairly sized programs in themselves) and
4017 ;; improves the hit ratio (it's a type regardless of the template
4018 ;; args; it's just not the same type, but we're only interested in
4019 ;; recognizing types, not telling distinct types apart). Note that
4020 ;; template types in references are added here too; from the example
4021 ;; above there will also be an entry "Foo<".
4026 ;; Clears `c-found-types'.
4030 ;; Add the given region as a type in `c-found-types'. If the region
4031 ;; doesn't match an existing type but there is a type which is equal
4032 ;; to the given one except that the last character is missing, then
4033 ;; the shorter type is removed. That's done to avoid adding all
4034 ;; prefixes of a type as it's being entered and font locked. This
4035 ;; doesn't cover cases like when characters are removed from a type
4036 ;; or added in the middle. We'd need the position of point when the
4037 ;; font locking is invoked to solve this well.
4038 ;;
4039 ;; This function might do hidden buffer changes.
4046 ;; Remove the "NAME" from c-found-types, if present.
4050 ;; Return non-nil if the given region contains a type in
4051 ;; `c-found-types'.
4052 ;;
4053 ;; This function might do hidden buffer changes.
4055 c-found-types))
4058 ;; Return all the types in `c-found-types' as a sorted list of
4059 ;; strings.
4063 type-list)))
4064 c-found-types)
4067 ;; Shut up the byte compiler.
4071 ;; An after change function which, in conjunction with the info in
4072 ;; c-maybe-stale-found-type (set in c-before-change), removes a type
4073 ;; from `c-found-types', should this type have become stale. For
4074 ;; example, this happens to "foo" when "foo \n bar();" becomes
4075 ;; "foo(); \n bar();". Such stale types, if not removed, foul up
4076 ;; the fontification.
4077 ;;
4078 ;; Have we, perhaps, added non-ws characters to the front/back of a found
4079 ;; type?
4099 ;; Changing the amount of (already existing) whitespace - don't do anything.
4104 ;; The syntactic relationship which defined a "found type" has been
4105 ;; destroyed.
4108 ;; ((eq (car c-maybe-stale-found-type) 'c-decl-type-start) FIXME!!!
4109 )))
4111 \f
4112 ;; Handling of small scale constructs like types and names.
4115 ;; This is called from `after-change-functions' when
4116 ;; c-recognize-<>-arglists' is set. It ensures that no "<" or ">"
4117 ;; chars with paren syntax become part of another operator like "<<"
4118 ;; or ">=".
4119 ;;
4120 ;; This function might do hidden buffer changes.
4152 ;; Dynamically bound variable that instructs `c-forward-type' to also
4153 ;; treat possible types (i.e. those that it normally returns 'maybe or
4154 ;; 'found for) as actual types (and always return 'found for them).
4155 ;; This means that it records them in `c-record-type-identifiers' if
4156 ;; that is set, and that it adds them to `c-found-types'.
4159 ;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
4160 ;; mark up successfully parsed arglists with paren syntax properties on
4161 ;; the surrounding angle brackets and with `c-<>-arg-sep' in the
4162 ;; `c-type' property of each argument separating comma.
4163 ;;
4164 ;; Setting this variable also makes `c-forward-<>-arglist' recurse into
4165 ;; all arglists for side effects (i.e. recording types), otherwise it
4166 ;; exploits any existing paren syntax properties to quickly jump to the
4167 ;; end of already parsed arglists.
4168 ;;
4169 ;; Marking up the arglists is not the default since doing that correctly
4170 ;; depends on a proper value for `c-restricted-<>-arglists'.
4173 ;; Dynamically bound variable that instructs `c-forward-<>-arglist' to
4174 ;; not accept arglists that contain binary operators.
4175 ;;
4176 ;; This is primarily used to handle C++ template arglists. C++
4177 ;; disambiguates them by checking whether the preceding name is a
4178 ;; template or not. We can't do that, so we assume it is a template
4179 ;; if it can be parsed as one. That usually works well since
4180 ;; comparison expressions on the forms "a < b > c" or "a < b, c > d"
4181 ;; in almost all cases would be pointless.
4182 ;;
4183 ;; However, in function arglists, e.g. in "foo (a < b, c > d)", we
4184 ;; should let the comma separate the function arguments instead. And
4185 ;; in a context where the value of the expression is taken, e.g. in
4186 ;; "if (a < b || c > d)", it's probably not a template.
4189 ;; Dynamically bound variables that instructs
4190 ;; `c-forward-keyword-clause', `c-forward-<>-arglist',
4191 ;; `c-forward-name', `c-forward-type', `c-forward-decl-or-cast-1', and
4192 ;; `c-forward-label' to record the ranges of all the type and
4193 ;; reference identifiers they encounter. They will build lists on
4194 ;; these variables where each element is a cons of the buffer
4195 ;; positions surrounding each identifier. This recording is only
4196 ;; activated when `c-record-type-identifiers' is non-nil.
4197 ;;
4198 ;; All known types that can't be identifiers are recorded, and also
4199 ;; other possible types if `c-promote-possible-types' is set.
4200 ;; Recording is however disabled inside angle bracket arglists that
4201 ;; are encountered inside names and other angle bracket arglists.
4202 ;; Such occurrences are taken care of by `c-font-lock-<>-arglists'
4203 ;; instead.
4204 ;;
4205 ;; Only the names in C++ template style references (e.g. "tmpl" in
4206 ;; "tmpl<a,b>::foo") are recorded as references, other references
4207 ;; aren't handled here.
4208 ;;
4209 ;; `c-forward-label' records the label identifier(s) on
4210 ;; `c-record-ref-identifiers'.
4214 ;; This variable will receive a cons cell of the range of the last
4215 ;; single identifier symbol stepped over by `c-forward-name' if it's
4216 ;; successful. This is the range that should be put on one of the
4217 ;; record lists above by the caller. It's assigned nil if there's no
4218 ;; such symbol in the name.
4223 ;; Always true.
4233 ;; Always true.
4241 ;; Dynamically bound variable that instructs `c-forward-type' to
4242 ;; record the ranges of types that only are found. Behaves otherwise
4243 ;; like `c-record-type-identifiers'.
4247 ;; Used internally in `c-forward-keyword-clause' to move forward
4248 ;; over a type (if TYPE is 'type) or a name (otherwise) which
4249 ;; possibly is prefixed by keywords and their associated clauses.
4250 ;; Try with a type/name first to not trip up on those that begin
4251 ;; with a keyword. Return t if a known or found type is moved
4252 ;; over. The point is clobbered if nil is returned. If range
4253 ;; recording is enabled, the identifier is recorded on as a type
4254 ;; if TYPE is 'type or as a reference if TYPE is 'ref.
4255 ;;
4256 ;; This macro might do hidden buffer changes.
4261 nil
4268 t)))
4271 ;; Used internally in `c-forward-keyword-clause' to move forward
4272 ;; over a comma separated list of types or names using
4273 ;; `c-forward-keyword-prefixed-id'.
4274 ;;
4275 ;; This macro might do hidden buffer changes.
4286 ;; Submatch MATCH in the current match data is assumed to surround a
4287 ;; token. If it's a keyword, move over it and any immediately
4288 ;; following clauses associated with it, stopping at the start of
4289 ;; the next token. t is returned in that case, otherwise the point
4290 ;; stays and nil is returned. The kind of clauses that are
4291 ;; recognized are those specified by `c-type-list-kwds',
4292 ;; `c-ref-list-kwds', `c-colon-type-list-kwds',
4293 ;; `c-paren-nontype-kwds', `c-paren-type-kwds', `c-<>-type-kwds',
4294 ;; and `c-<>-arglist-kwds'.
4295 ;;
4296 ;; This function records identifier ranges on
4297 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
4298 ;; `c-record-type-identifiers' is non-nil.
4299 ;;
4300 ;; Note that for `c-colon-type-list-kwds', which doesn't necessary
4301 ;; apply directly after the keyword, the type list is moved over
4302 ;; only when there is no unaccounted token before it (i.e. a token
4303 ;; that isn't moved over due to some other keyword list). The
4304 ;; identifier ranges in the list are still recorded if that should
4305 ;; be done, though.
4306 ;;
4307 ;; This function might do hidden buffer changes.
4310 ;; The call to `c-forward-<>-arglist' below is made after
4311 ;; `c-<>-sexp-kwds' keywords, so we're certain they actually
4312 ;; are angle bracket arglists and `c-restricted-<>-arglists'
4313 ;; should therefore be nil.
4325 ;; There's a type directly after a keyword in `c-type-list-kwds'.
4330 ;; There's a name directly after a keyword in `c-ref-list-kwds'.
4335 ;; There's an open paren after a keyword in `c-paren-any-kwds'.
4342 ;; Use `c-forward-type' on every identifier we can find
4343 ;; inside the paren, to record the types.
4368 ;; If we are at the colon already, we move over the type
4369 ;; list after it.
4375 ;; Not at the colon, so stop here. But the identifier
4376 ;; ranges in the type list later on should still be
4377 ;; recorded.
4380 ;; If a keyword matched both one of the types above and
4381 ;; this one, we match `c-colon-type-list-re' after the
4382 ;; clause matched above.
4389 ;; There's a type after the `c-colon-type-list-re' match
4390 ;; after a keyword in `c-colon-type-list-kwds'.
4394 t)))
4399 ;; The point is assumed to be at a "<". Try to treat it as the open
4400 ;; paren of an angle bracket arglist and move forward to the
4401 ;; corresponding ">". If successful, the point is left after the
4402 ;; ">" and t is returned, otherwise the point isn't moved and nil is
4403 ;; returned. If ALL-TYPES is t then all encountered arguments in
4404 ;; the arglist that might be types are treated as found types.
4405 ;;
4406 ;; The variable `c-parse-and-markup-<>-arglists' controls how this
4407 ;; function handles text properties on the angle brackets and argument
4408 ;; separating commas.
4409 ;;
4410 ;; `c-restricted-<>-arglists' controls how lenient the template
4411 ;; arglist recognition should be.
4412 ;;
4413 ;; This function records identifier ranges on
4414 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
4415 ;; `c-record-type-identifiers' is non-nil.
4416 ;;
4417 ;; This function might do hidden buffer changes.
4420 ;; If `c-record-type-identifiers' is set then activate
4421 ;; recording of any found types that constitute an argument in
4422 ;; the arglist.
4431 ;; `nconc' doesn't mind that the tail of
4432 ;; `c-record-found-types' is t.
4434 t)
4437 nil)))
4441 ;; Temporary workaround for Bug#7722.
4448 ;; Recursive part of `c-forward-<>-arglist'.
4449 ;;
4450 ;; This function might do hidden buffer changes.
4453 ;; Cover this so that any recorded found type ranges are
4454 ;; automatically lost if it turns out to not be an angle
4455 ;; bracket arglist. It's propagated through the return value
4456 ;; on successful completion.
4458 ;; List that collects the positions after the argument
4459 ;; separating ',' in the arglist.
4460 arg-start-pos)
4462 ;; If the '<' has paren open syntax then we've marked it as an angle
4463 ;; bracket arglist before, so skip to the end.
4471 t
4473 ;; Got unmatched paren angle brackets. We don't clear the paren
4474 ;; syntax properties and retry, on the basis that it's very
4475 ;; unlikely that paren angle brackets become operators by code
4476 ;; manipulation. It's far more likely that it doesn't match due
4477 ;; to narrowing or some temporary change.
4479 nil))
4489 ;; All encountered identifiers are types, so set the
4490 ;; promote flag and parse the type.
4497 ;; Check if this arglist argument is a sole type. If
4498 ;; it's known then it's recorded in
4499 ;; `c-record-type-identifiers'. If it only is found
4500 ;; then it's recorded in `c-record-found-types' which we
4501 ;; might roll back if it turns out that this isn't an
4502 ;; angle bracket arglist afterall.
4508 ;; A found type was recorded but it's not the
4509 ;; only thing in the arglist argument, so reset
4510 ;; `c-record-found-types'.
4512 orig-record-found-types))))))
4516 ;; Must check for '>' at the very start separately,
4517 ;; since the regexp below has to avoid ">>" without
4518 ;; using \\=.
4520 t)
4522 ;; Note: These regexps exploit the match order in \| so
4523 ;; that "<>" is matched by "<" rather than "[^>:-]>".
4526 ;; Stop on ',', '|', '&', '+' and '-' to catch
4527 ;; common binary operators that could be between
4528 ;; two comparison expressions "a<b" and "c>d".
4530 ;; Otherwise we still stop on ',' to find the
4531 ;; argument start positions.
4535 ;; If the arglist starter has lost its open paren
4536 ;; syntax but not the closer, we won't find the
4537 ;; closer above since we only search in the
4538 ;; balanced sexp. In that case we stop just short
4539 ;; of it so check if the following char is the closer.
4542 t)))
4546 ;; Either an operator starting with '>' or the end of
4547 ;; the angle bracket arglist.
4554 ;; The angle bracket arglist is finished.
4566 ;; Either an operator starting with '<' or a nested arglist.
4578 ;; There's always an identifier before an angle
4579 ;; bracket arglist, or a keyword in
4580 ;; `c-<>-type-kwds' or `c-<>-arglist-kwds'.
4597 )))
4599 ;; It was not an angle bracket arglist.
4602 ;; It was an angle bracket arglist.
4605 ;; Record the identifier before the template as a type
4606 ;; or reference depending on whether the arglist is last
4607 ;; in a qualified identifier.
4616 t)
4620 ;; Just another argument. Record the position. The
4621 ;; type check stuff that made us stop at it is at
4622 ;; the top of the loop.
4626 ;; Got a character that can't be in an angle bracket
4627 ;; arglist argument. Abort using `throw', since
4628 ;; it's useless to try to find a surrounding arglist
4629 ;; if we're nested.
4636 ;; The point is assumed to be directly after a ">". Try to treat it
4637 ;; as the close paren of an angle bracket arglist and move back to
4638 ;; the corresponding "<". If successful, the point is left at
4639 ;; the "<" and t is returned, otherwise the point isn't moved and
4640 ;; nil is returned. ALL-TYPES is passed on to
4641 ;; `c-forward-<>-arglist'.
4642 ;;
4643 ;; If the optional LIMIT is given, it bounds the backward search.
4644 ;; It's then assumed to be at a syntactically relevant position.
4645 ;;
4646 ;; This is a wrapper around `c-forward-<>-arglist'. See that
4647 ;; function for more details.
4656 t
4657 ;; See corresponding note in `c-forward-<>-arglist'.
4659 nil)
4666 t
4667 ;; Stopped at bob or a char that isn't allowed in an
4668 ;; arglist, so we've failed.
4670 nil)
4675 ;; If we moved then the "<" was part of some
4676 ;; multicharacter token.
4677 t
4683 ;; Matched the arglist. Break the while.
4685 nil)
4687 ;; We started from a non-paren ">" inside an
4688 ;; arglist.
4690 nil)
4692 ;; Matched a shorter arglist. Can be a nested
4693 ;; one so continue looking.
4695 t))
4696 t))))))
4701 ;; Move forward over a complete name if at the beginning of one,
4702 ;; stopping at the next following token. If the point is not at
4703 ;; something that are recognized as name then it stays put. A name
4704 ;; could be something as simple as "foo" in C or something as
4705 ;; complex as "X<Y<class A<int>::B, BIT_MAX >> b>, ::operator<> ::
4706 ;; Z<(a>b)> :: operator const X<&foo>::T Q::G<unsigned short
4707 ;; int>::*volatile const" in C++ (this function is actually little
4708 ;; more than a `looking-at' call in all modes except those that,
4709 ;; like C++, have `c-recognize-<>-arglists' set). Return nil if no
4710 ;; name is found, 'template if it's an identifier ending with an
4711 ;; angle bracket arglist, 'operator of it's an operator identifier,
4712 ;; or t if it's some other kind of name.
4713 ;;
4714 ;; This function records identifier ranges on
4715 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
4716 ;; `c-record-type-identifiers' is non-nil.
4717 ;;
4718 ;; This function might do hidden buffer changes.
4721 ;; Turn off `c-promote-possible-types' here since we might
4722 ;; call `c-forward-<>-arglist' and we don't want it to promote
4723 ;; every suspect thing in the arglist to a type. We're
4724 ;; typically called from `c-forward-type' in this case, and
4725 ;; the caller only wants the top level type that it finds to
4726 ;; be promoted.
4727 c-promote-possible-types)
4733 ;; Check for keyword. We go to the last symbol in
4734 ;; `c-identifier-key' first.
4747 ;; "template" is only valid inside an
4748 ;; identifier if preceded by "::".
4753 t))
4755 ;; Handle a C++ operator or template identifier.
4759 ;; Got "... ::template".
4763 res subres))))
4766 ;; Got a cast operator.
4770 ;; Now we should match a sequence of either
4771 ;; '*', '&' or a name followed by ":: *",
4772 ;; where each can be followed by a sequence
4773 ;; of `c-opt-type-modifier-key'.
4776 t)
4786 t))))
4794 ;; Got some other operator.
4802 nil)
4804 ;; `id-start' is equal to `id-end' if we've jumped over
4805 ;; an identifier that doesn't end with a symbol token.
4806 ;; That can occur e.g. for Java import directives on the
4807 ;; form "foo.bar.*".
4814 res t)))
4824 ;; Got a concatenated identifier. This handles the
4825 ;; cases with tricky syntactic whitespace that aren't
4826 ;; covered in `c-identifier-key'.
4829 t)
4833 ;; Maybe an angle bracket arglist.
4836 c-record-found-types)
4842 c-last-identifier-range nil)
4847 ;; Continue if there's an identifier concatenation
4848 ;; operator after the template argument.
4854 t)
4859 nil)))
4860 )))))
4863 res))
4866 ;; Move forward over a type spec if at the beginning of one,
4867 ;; stopping at the next following token. Return t if it's a known
4868 ;; type that can't be a name or other expression, 'known if it's an
4869 ;; otherwise known type (according to `*-font-lock-extra-types'),
4870 ;; 'prefix if it's a known prefix of a type, 'found if it's a type
4871 ;; that matches one in `c-found-types', 'maybe if it's an identfier
4872 ;; that might be a type, or nil if it can't be a type (the point
4873 ;; isn't moved then). The point is assumed to be at the beginning
4874 ;; of a token.
4875 ;;
4876 ;; Note that this function doesn't skip past the brace definition
4877 ;; that might be considered part of the type, e.g.
4878 ;; "enum {a, b, c} foo".
4879 ;;
4880 ;; This function records identifier ranges on
4881 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
4882 ;; `c-record-type-identifiers' is non-nil.
4883 ;;
4884 ;; This function might do hidden buffer changes.
4888 ;; Skip leading type modifiers. If any are found we know it's a
4889 ;; prefix of a type.
4898 ;; Looking at a keyword that prefixes a type identifier,
4899 ;; e.g. "class".
4906 ;; In many languages the name can be used without the
4907 ;; prefix, so we add it to `c-found-types'.
4910 c-last-identifier-range)
4913 ;; Invalid syntax.
4925 id-range c-last-identifier-range))))
4937 id-end)
4939 ;; Looking at a primitive or known type identifier. We've
4940 ;; checked for a name first so that we don't go here if the
4941 ;; known type match only is a prefix of another name.
4952 ;; There might be more keywords for the type.
4980 ;; A normal identifier.
4990 ;; It's an identifier that has been used as
4991 ;; a type somewhere else.
4992 'found
4993 ;; It's an identifier that might be a type.
4996 ;; A template is a type.
5000 ;; Otherwise it's an operator identifier, which is not a type.
5005 ;; Skip trailing type modifiers. If any are found we know it's
5006 ;; a type.
5013 ;; Step over any type suffix operator. Do not let the existence
5014 ;; of these alter the classification of the found type, since
5015 ;; these operators typically are allowed in normal expressions
5016 ;; too.
5023 ;; Look for a trailing operator that concatenates the type
5024 ;; with a following one, and if so step past that one through
5025 ;; a recursive call. Note that we don't record concatenated
5026 ;; types in `c-found-types' - it's the component types that
5027 ;; are recorded when appropriate.
5030 c-promote-possible-types))
5031 ;; If we can't promote then set `c-record-found-types' so that
5032 ;; we can merge in the types from the second part afterwards if
5033 ;; it turns out to be a known type there.
5036 subres)
5045 ;; If either operand certainly is a type then both are, but we
5046 ;; don't let the existence of the operator itself promote two
5047 ;; uncertain types to a certain one.
5066 ;; Merge in the ranges of any types found by the second
5067 ;; `c-forward-type'.
5069 ;; `nconc' doesn't mind that the tail of
5070 ;; `c-record-found-types' is t.
5072 c-record-type-identifiers))))
5080 ;;(message "c-forward-type %s -> %s: %s" start (point) res)
5082 res))
5084 \f
5085 ;; Handling of large scale constructs like statements and declarations.
5087 ;; Macro used inside `c-forward-decl-or-cast-1'. It ought to be a
5088 ;; defsubst or perhaps even a defun, but it contains lots of free
5089 ;; variables that refer to things inside `c-forward-decl-or-cast-1'.
5091 ;; `c-forward-decl-or-cast-1' can consume an arbitrary length list
5092 ;; of types when parsing a declaration, which means that it
5093 ;; sometimes consumes the identifier in the declaration as a type.
5094 ;; This is used to "backtrack" and make the last type be treated as
5095 ;; an identifier instead.
5098 ;; These identifiers are bound only in the inner let.
5100 identifier-start type-start
5101 got-parens nil
5102 got-identifier t
5103 got-suffix t
5104 got-suffix-after-parens id-start
5108 t
5109 backup-at-type))
5111 id-start backup-id-start)
5113 id-start start-pos))
5115 ;; When these flags already are set we've found specifiers that
5116 ;; unconditionally signal these attributes - backtracking doesn't
5117 ;; change that. So keep them set in that case.
5124 ;; This identifier is bound only in the inner let.
5128 ;; Move forward over a declaration or a cast if at the start of one.
5129 ;; The point is assumed to be at the start of some token. Nil is
5130 ;; returned if no declaration or cast is recognized, and the point
5131 ;; is clobbered in that case.
5132 ;;
5133 ;; If a declaration is parsed:
5134 ;;
5135 ;; The point is left at the first token after the first complete
5136 ;; declarator, if there is one. The return value is a cons where
5137 ;; the car is the position of the first token in the declarator. (See
5138 ;; below for the cdr.)
5139 ;; Some examples:
5140 ;;
5141 ;; void foo (int a, char *b) stuff ...
5142 ;; car ^ ^ point
5143 ;; float (*a)[], b;
5144 ;; car ^ ^ point
5145 ;; unsigned int a = c_style_initializer, b;
5146 ;; car ^ ^ point
5147 ;; unsigned int a (cplusplus_style_initializer), b;
5148 ;; car ^ ^ point (might change)
5149 ;; class Foo : public Bar {}
5150 ;; car ^ ^ point
5151 ;; class PikeClass (int a, string b) stuff ...
5152 ;; car ^ ^ point
5153 ;; enum bool;
5154 ;; car ^ ^ point
5155 ;; enum bool flag;
5156 ;; car ^ ^ point
5157 ;; void cplusplus_function (int x) throw (Bad);
5158 ;; car ^ ^ point
5159 ;; Foo::Foo (int b) : Base (b) {}
5160 ;; car ^ ^ point
5161 ;;
5162 ;; The cdr of the return value is non-nil iff a `c-typedef-decl-kwds'
5163 ;; specifier (e.g. class, struct, enum, typedef) is found in the
5164 ;; declaration, i.e. the declared identifier(s) are types.
5165 ;;
5166 ;; If a cast is parsed:
5167 ;;
5168 ;; The point is left at the first token after the closing paren of
5169 ;; the cast. The return value is `cast'. Note that the start
5170 ;; position must be at the first token inside the cast parenthesis
5171 ;; to recognize it.
5172 ;;
5173 ;; PRECEDING-TOKEN-END is the first position after the preceding
5174 ;; token, i.e. on the other side of the syntactic ws from the point.
5175 ;; Use a value less than or equal to (point-min) if the point is at
5176 ;; the first token in (the visible part of) the buffer.
5177 ;;
5178 ;; CONTEXT is a symbol that describes the context at the point:
5179 ;; 'decl In a comma-separated declaration context (typically
5180 ;; inside a function declaration arglist).
5181 ;; '<> In an angle bracket arglist.
5182 ;; 'arglist Some other type of arglist.
5183 ;; nil Some other context or unknown context. Includes
5184 ;; within the parens of an if, for, ... construct.
5185 ;;
5186 ;; LAST-CAST-END is the first token after the closing paren of a
5187 ;; preceding cast, or nil if none is known. If
5188 ;; `c-forward-decl-or-cast-1' is used in succession, it should be
5189 ;; the position after the closest preceding call where a cast was
5190 ;; matched. In that case it's used to discover chains of casts like
5191 ;; "(a) (b) c".
5192 ;;
5193 ;; This function records identifier ranges on
5194 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
5195 ;; `c-record-type-identifiers' is non-nil.
5196 ;;
5197 ;; This function might do hidden buffer changes.
5200 ;; first type, i.e. after any leading specifiers. It might
5201 ;; also point at the beginning of the preceding syntactic
5202 ;; whitespace.
5204 ;; Set to the result of `c-forward-type'.
5205 at-type
5206 ;; The position of the first token in what we currently
5207 ;; believe is the type in the declaration or cast, after any
5208 ;; specifiers and their associated clauses.
5209 type-start
5210 ;; The position of the first token in what we currently
5211 ;; believe is the declarator for the first identifier. Set
5212 ;; when the type is found, and moved forward over any
5213 ;; `c-decl-hangon-kwds' and their associated clauses that
5214 ;; occurs after the type.
5215 id-start
5216 ;; These store `at-type', `type-start' and `id-start' of the
5217 ;; identifier before the one in those variables. The previous
5218 ;; identifier might turn out to be the real type in a
5219 ;; declaration if the last one has to be the declarator in it.
5220 ;; If `backup-at-type' is nil then the other variables have
5221 ;; undefined values.
5222 backup-at-type backup-type-start backup-id-start
5223 ;; Set if we've found a specifier that makes the defined
5224 ;; identifier(s) types.
5225 at-type-decl
5226 ;; Set if we've found a specifier that can start a declaration
5227 ;; where there's no type.
5228 maybe-typeless
5229 ;; If a specifier is found that also can be a type prefix,
5230 ;; these flags are set instead of those above. If we need to
5231 ;; back up an identifier, they are copied to the real flag
5232 ;; variables. Thus they only take effect if we fail to
5233 ;; interpret it as a type.
5234 backup-at-type-decl backup-maybe-typeless
5235 ;; Whether we've found a declaration or a cast. We might know
5236 ;; this before we've found the type in it. It's 'ids if we've
5237 ;; found two consecutive identifiers (usually a sure sign, but
5238 ;; we should allow that in labels too), and t if we've found a
5239 ;; specifier keyword (a 100% sure sign).
5240 at-decl-or-cast
5241 ;; Set when we need to back up to parse this as a declaration
5242 ;; but not as a cast.
5243 backup-if-not-cast
5244 ;; For casts, the return position.
5245 cast-end
5246 ;; Save `c-record-type-identifiers' and
5247 ;; `c-record-ref-identifiers' since ranges are recorded
5248 ;; speculatively and should be thrown away if it turns out
5249 ;; that it isn't a declaration or cast.
5253 ;; Check for a type. Unknown symbols are treated as possible
5254 ;; types, but they could also be specifiers disguised through
5255 ;; macros like __INLINE__, so we recognize both types and known
5256 ;; specifiers after them too.
5260 ;; Look for a specifier keyword clause.
5268 ;; Found a known or possible type or a prefix of a known type.
5271 ;; Got two identifiers with nothing but whitespace
5272 ;; between them. That can only happen in declarations.
5276 ;; If the previous identifier is a found type we
5277 ;; record it as a real one; it might be some sort of
5278 ;; alias for a prefix like "unsigned".
5285 backup-type-start type-start
5286 backup-id-start id-start
5287 at-type found-type
5288 type-start start
5290 ;; The previous ambiguous specifier/type turned out
5291 ;; to be a type since we've parsed another one after
5292 ;; it, so clear these backup flags.
5293 backup-at-type-decl nil
5294 backup-maybe-typeless nil))
5298 ;; Handle known specifier keywords and
5299 ;; `c-decl-hangon-kwds' which can occur after known
5300 ;; types.
5303 ;; It's a hang-on keyword that can occur anywhere.
5307 ;; Move the identifier start position if
5308 ;; we've passed a type.
5310 ;; Otherwise treat this as a specifier and
5311 ;; move the fallback position.
5315 ;; It's an ordinary specifier so we know that
5316 ;; anything before this can't be the type.
5318 start-pos kwd-clause-end)
5321 ;; It's ambiguous whether this keyword is a
5322 ;; specifier or a type prefix, so set the backup
5323 ;; flags. (It's assumed that `c-forward-type'
5324 ;; moved further than `c-forward-keyword-clause'.)
5336 ;; Haven't matched a type so it's an umambiguous
5337 ;; specifier keyword and we know we're in a
5338 ;; declaration.
5343 ;; If the type isn't known we continue so that we'll jump
5344 ;; over all specifiers and type identifiers. The reason
5345 ;; to do this for a known type prefix is to make things
5346 ;; like "unsigned INT16" work.
5351 ;; If a known type was found, we still need to skip over any
5352 ;; hangon keyword clauses after it. Otherwise it has already
5353 ;; been done in the loop above.
5359 ;; A prefix type is itself a primitive type when it's not
5360 ;; followed by another type.
5364 ;; Got no type but set things up to continue anyway to handle
5365 ;; the various cases when a declaration doesn't start with a
5366 ;; type.
5371 ;; If it's C++ then check if the last "type" ends on the form
5372 ;; "foo::foo" or "foo::~foo", i.e. if it's the name of a
5373 ;; (con|de)structor.
5384 ;; Cheating in the handling of syntactic ws below.
5391 name))
5392 ;; It is a (con|de)structor name. In that case the
5393 ;; declaration is typeless so zap out any preceding
5394 ;; identifier(s) that we might have taken as types.
5397 backup-at-type nil
5398 id-start type-start))))))
5400 ;; Check for and step over a type decl expression after the thing
5401 ;; that is or might be a type. This can't be skipped since we
5402 ;; need the correct end position of the declarator for
5403 ;; `max-type-decl-end-*'.
5405 ;; True if there's a non-open-paren match of
5406 ;; `c-type-decl-prefix-key'.
5407 got-prefix
5408 ;; True if the declarator is surrounded by a parenthesis pair.
5409 got-parens
5410 ;; True if there is an identifier in the declarator.
5411 got-identifier
5412 ;; True if there's a non-close-paren match of
5413 ;; `c-type-decl-suffix-key'.
5414 got-suffix
5415 ;; True if there's a prefix match outside the outermost
5416 ;; paren pair that surrounds the declarator.
5417 got-prefix-before-parens
5418 ;; True if there's a suffix match outside the outermost
5419 ;; paren pair that surrounds the declarator. The value is
5420 ;; the position of the first suffix match.
5421 got-suffix-after-parens
5422 ;; True if we've parsed the type decl to a token that is
5423 ;; known to end declarations in this context.
5424 at-decl-end
5425 ;; The earlier values of `at-type' and `type-start' if we've
5426 ;; shifted the type backwards.
5427 identifier-type identifier-start
5428 ;; If `c-parse-and-markup-<>-arglists' is set we need to
5429 ;; turn it off during the name skipping below to avoid
5430 ;; getting `c-type' properties that might be bogus. That
5431 ;; can happen since we don't know if
5432 ;; `c-restricted-<>-arglists' will be correct inside the
5433 ;; arglist paren that gets entered.
5438 ;; Skip over type decl prefix operators. (Note similar code in
5439 ;; `c-font-lock-declarators'.)
5443 ;; If the second submatch matches in C++ then
5444 ;; we're looking at an identifier that's a
5445 ;; prefix only if it specifies a member pointer.
5448 ;; We only check for a trailing "::" and
5449 ;; let the "*" that should follow be
5450 ;; matched in the next round.
5452 ;; It turned out to be the real identifier,
5453 ;; so stop.
5454 nil))
5455 t))
5469 ;; Skip over an identifier.
5474 ;; Skip over type decl suffix operators.
5481 t)
5485 t)
5491 ;; No suffix matched. We might have matched the
5492 ;; identifier as a type and the open paren of a
5493 ;; function arglist as a type decl prefix. In that
5494 ;; case we should "backtrack": Reinterpret the last
5495 ;; type as the identifier, move out of the arglist and
5496 ;; continue searching for suffix operators.
5497 ;;
5498 ;; Do this even if there's no preceding type, to cope
5499 ;; with old style function declarations in K&R C,
5500 ;; (con|de)structors in C++ and `c-typeless-decl-kwds'
5501 ;; style declarations. That isn't applicable in an
5502 ;; arglist context, though.
5507 maybe-typeless
5508 backup-maybe-typeless
5515 t))
5522 at-type)
5523 ;; Have found no identifier but `c-typeless-decl-kwds' has
5524 ;; matched so we know we're inside a declaration. The
5525 ;; preceding type must be the identifier instead.
5529 at-decl-or-cast
5532 ;; CASE 1
5534 ;; Encountered something inside parens that isn't matched by
5535 ;; the `c-type-decl-*' regexps, so it's not a type decl
5536 ;; expression. Try to skip out to the same paren depth to
5537 ;; not confuse the cast check below.
5539 ;; If we've found a specifier keyword then it's a
5540 ;; declaration regardless.
5548 ;; Now we've collected info about various characteristics of
5549 ;; the construct we're looking at. Below follows a decision
5550 ;; tree based on that. It's ordered to check more certain
5551 ;; signs before less certain ones.
5556 ;; CASE 2
5559 ;; Got another identifier directly after the type, so it's a
5560 ;; declaration.
5567 maybe-typeless
5568 backup-maybe-typeless))
5569 ;; Got a declaration of the form "foo bar (gnu);" where we've
5570 ;; recognized "bar" as the type and "gnu" as the declarator.
5571 ;; In this case it's however more likely that "bar" is the
5572 ;; declarator and "gnu" a function argument or initializer (if
5573 ;; `c-recognize-paren-inits' is set), since the parens around
5574 ;; "gnu" would be superfluous if it's a declarator. Shift the
5575 ;; type one step backward.
5578 ;; Found no identifier.
5583 ;; CASE 3
5585 ;; Got a plain list of identifiers. If a colon follows it's
5586 ;; a valid label. Otherwise the last one probably is the
5587 ;; declared identifier and we should back up to the previous
5588 ;; type, providing it isn't a cast.
5590 ;; If we've found a specifier keyword then it's a
5591 ;; declaration regardless.
5596 ;; CASE 4
5600 ;; Got a plain list of identifiers followed by some suffix.
5601 ;; If this isn't a cast then the last identifier probably is
5602 ;; the declared one and we should back up to the previous
5603 ;; type.
5607 ;; CASE 5
5609 ;; If the type is known we know that there can't be any
5610 ;; identifier somewhere else, and it's only in declarations in
5611 ;; e.g. function prototypes and in casts that the identifier may
5612 ;; be left out.
5616 ;; Only got a single identifier (parsed as a type so far).
5617 ;; CASE 6
5619 ;; Check that the identifier isn't at the start of an
5620 ;; expression.
5621 at-decl-end
5624 ;; Inside an arglist that contains declarations. If K&R
5625 ;; style declarations and parenthesis style initializers
5626 ;; aren't allowed then the single identifier must be a
5627 ;; type, else we require that it's known or found
5628 ;; (primitive types are handled above).
5633 ;; Inside a template arglist. Accept known and found
5634 ;; types; other identifiers could just as well be
5635 ;; constants in C++.
5638 ;; CASE 7
5639 ;; Can't be a valid declaration or cast, but if we've found a
5640 ;; specifier it can't be anything else either, so treat it as
5641 ;; an invalid/unfinished declaration or cast.
5649 maybe-typeless
5650 backup-maybe-typeless
5654 c-after-suffixed-type-maybe-decl-key))))))
5655 ;; Got an empty paren pair and a preceding type that probably
5656 ;; really is the identifier. Shift the type backwards to make
5657 ;; the last one the identifier. This is analogous to the
5658 ;; "backtracking" done inside the `c-type-decl-suffix-key' loop
5659 ;; above.
5660 ;;
5661 ;; Exception: In addition to the conditions in that
5662 ;; "backtracking" code, do not shift backward if we're not
5663 ;; looking at either `c-after-suffixed-type-decl-key' or "[;,]".
5664 ;; Since there's no preceding type, the shift would mean that
5665 ;; the declaration is typeless. But if the regexp doesn't match
5666 ;; then we will simply fall through in the tests below and not
5667 ;; recognize it at all, so it's better to try it as an abstract
5668 ;; declarator instead.
5671 ;; Still no identifier.
5672 ;; CASE 8
5674 ;; Require `got-prefix' together with either `got-parens' or
5675 ;; `got-suffix' to recognize it as an abstract declarator:
5676 ;; `got-parens' only is probably an empty function call.
5677 ;; `got-suffix' only can build an ordinary expression together
5678 ;; with the preceding identifier which we've taken as a type.
5679 ;; We could actually accept on `got-prefix' only, but that can
5680 ;; easily occur temporarily while writing an expression so we
5681 ;; avoid that case anyway. We could do a better job if we knew
5682 ;; the point when the fontification was invoked.
5685 ;; CASE 9
5689 got-suffix-after-parens
5691 ;; Got a type, no declarator but a paren suffix. I.e. it's a
5692 ;; normal function call afterall (or perhaps a C++ style object
5693 ;; instantiation expression).
5696 ;; CASE 10
5698 ;; By now we've located the type in the declaration that we know
5699 ;; we're in.
5702 ;; CASE 11
5710 ;; Shift the type backward in the case that there's a
5711 ;; single identifier inside parens. That can only
5712 ;; occur in K&R style function declarations so it's
5713 ;; more likely that it really is a function call.
5714 ;; Therefore we only do this after
5715 ;; `c-after-suffixed-type-decl-key' has matched.
5717 got-suffix-after-parens))
5718 ;; A declaration according to `c-after-suffixed-type-decl-key'.
5721 ;; CASE 12
5724 ;; It's a declaration if a known type precedes it and it can't be a
5725 ;; function call.
5728 ;; If we get here we can't tell if this is a type decl or a normal
5729 ;; expression by looking at it alone. (That's under the assumption
5730 ;; that normal expressions always can look like type decl expressions,
5731 ;; which isn't really true but the cases where it doesn't hold are so
5732 ;; uncommon (e.g. some placements of "const" in C++) it's not worth
5733 ;; the effort to look for them.)
5736 ;; If this is a declaration it should end here or its initializer(*)
5737 ;; should start here, so check for allowed separation tokens. Note
5738 ;; that this rule doesn't work e.g. with a K&R arglist after a
5739 ;; function header.
5740 ;;
5741 ;; *) Don't check for C++ style initializers using parens
5742 ;; since those already have been matched as suffixes.
5743 ;;
5744 ;; If `at-decl-or-cast' is then we've found some other sign that
5745 ;; it's a declaration or cast, so then it's probably an
5746 ;; invalid/unfinished one.
5749 ;; Below are tests that only should be applied when we're certain to
5750 ;; not have parsed halfway through an expression.
5752 ;; CASE 14
5754 ;; The expression starts with a known type so treat it as a
5755 ;; declaration.
5758 ;; CASE 15
5760 ;; In C++ we check if the identifier is a known type, since
5761 ;; (con|de)structors use the class name as identifier.
5762 ;; We've always shifted over the identifier as a type and
5763 ;; then backed up again in this case.
5764 identifier-type
5767 ;; `at-type' probably won't be 'found for
5768 ;; destructors since the "~" is then part of the
5769 ;; type name being checked against the list of
5770 ;; known types, so do a check without that
5771 ;; operator.
5776 c-identifier-syntax-table
5780 ;; We have already parsed the type earlier,
5781 ;; so it'd be possible to cache the end
5782 ;; position instead of redoing it here, but
5783 ;; then we'd need to keep track of another
5784 ;; position everywhere.
5792 ;; CASE 16
5794 at-type
5798 ;; Got something like "foo * bar;". Since we're not inside an
5799 ;; arglist it would be a meaningless expression because the
5800 ;; result isn't used. We therefore choose to recognize it as
5801 ;; a declaration. Do not allow a suffix since it could then
5802 ;; be a function call.
5805 ;; CASE 17
5810 ;; Got something like "a (*b) (c);" or "a (b) = c;". It could
5811 ;; be an odd expression or it could be a declaration. Treat
5812 ;; it as a declaration if "a" has been used as a type
5813 ;; somewhere else (if it's a known type we won't get here).
5816 ;; CASE 18
5822 ;; Got a type followed by an abstract declarator. If `got-prefix'
5823 ;; is set it's something like "a *" without anything after it. If
5824 ;; `got-parens' or `got-suffix' is set it's "a()", "a[]", "a()[]",
5825 ;; or similar, which we accept only if the context rules out
5826 ;; expressions.
5829 ;; If we had a complete symbol table here (which rules out
5830 ;; `c-found-types') we should return t due to the disambiguation rule
5831 ;; (in at least C++) that anything that can be parsed as a declaration
5832 ;; is a declaration. Now we're being more defensive and prefer to
5833 ;; highlight things like "foo (bar);" as a declaration only if we're
5834 ;; inside an arglist that contains declarations.
5837 ;; The point is now after the type decl expression.
5840 ;; Check for a cast.
5843 c-cast-parens
5845 ;; Should be the first type/identifier in a cast paren.
5849 ;; The closing paren should follow.
5854 ;; There should be a primary expression after it.
5863 ;; Check if the expression begins with a prefix keyword.
5866 ;; Expression begins with an ambiguous operator. Treat
5867 ;; it as a cast if it's a type decl or if we've
5868 ;; recognized the type somewhere else.
5871 ;; Unless it's a keyword, it's the beginning of a primary
5872 ;; expression.
5874 ;; If `c-primary-expr-regexp' matched a nonsymbol token, check
5875 ;; that it matched a whole one so that we don't e.g. confuse
5876 ;; the operator '-' with '->'. It's ok if it matches further,
5877 ;; though, since it e.g. can match the float '.5' while the
5878 ;; operator regexp only matches '.'.
5882 ;; There should either be a cast before it or something that isn't an
5883 ;; identifier or close paren.
5891 ;; It's a symbol. Accept it only if it's one of the
5892 ;; keywords that can precede an expression (without
5893 ;; surrounding parens).
5896 ;; Check that it isn't a close paren (block close is ok,
5897 ;; though).
5899 ;; Check that it isn't a nonsymbol identifier.
5902 ;; Handle the cast.
5912 ;; We're at a declaration. Highlight the type and the following
5913 ;; declarators.
5919 ;; Make sure to propagate the `c-decl-arg-start' property to
5920 ;; the next argument if it's set in this one, to cope with
5921 ;; interactive refontification.
5933 ;; False alarm. Restore the recorded ranges.
5935 c-record-ref-identifiers save-rec-ref-ids)
5936 nil))))
5939 ;; Assuming that point is at the beginning of a token, check if it starts a
5940 ;; label and if so move over it and return non-nil (t in default situations,
5941 ;; specific symbols (see below) for interesting situations), otherwise don't
5942 ;; move and return nil. "Label" here means "most things with a colon".
5943 ;;
5944 ;; More precisely, a "label" is regarded as one of:
5945 ;; (i) a goto target like "foo:" - returns the symbol `goto-target';
5946 ;; (ii) A case label - either the entire construct "case FOO:", or just the
5947 ;; bare "case", should the colon be missing. We return t;
5948 ;; (iii) a keyword which needs a colon, like "default:" or "private:"; We
5949 ;; return t;
5950 ;; (iv) One of QT's "extended" C++ variants of
5951 ;; "private:"/"protected:"/"public:"/"more:" looking like "public slots:".
5952 ;; Returns the symbol `qt-2kwds-colon'.
5953 ;; (v) QT's construct "signals:". Returns the symbol `qt-1kwd-colon'.
5954 ;; (vi) One of the keywords matched by `c-opt-extra-label-key' (without any
5955 ;; colon). Currently (2006-03), this applies only to Objective C's
5956 ;; keywords "@private", "@protected", and "@public". Returns t.
5957 ;;
5958 ;; One of the things which will NOT be recognised as a label is a bit-field
5959 ;; element of a struct, something like "int foo:5".
5960 ;;
5961 ;; The end of the label is taken to be just after the colon, or the end of
5962 ;; the first submatch in `c-opt-extra-label-key'. The point is directly
5963 ;; after the end on return. The terminating char gets marked with
5964 ;; `c-decl-end' to improve recognition of the following declaration or
5965 ;; statement.
5966 ;;
5967 ;; If ASSUME-MARKUP is non-nil, it's assumed that the preceding
5968 ;; label, if any, has already been marked up like that.
5969 ;;
5970 ;; If PRECEDING-TOKEN-END is given, it should be the first position
5971 ;; after the preceding token, i.e. on the other side of the
5972 ;; syntactic ws from the point. Use a value less than or equal to
5973 ;; (point-min) if the point is at the first token in (the visible
5974 ;; part of) the buffer.
5975 ;;
5976 ;; The optional LIMIT limits the forward scan for the colon.
5977 ;;
5978 ;; This function records the ranges of the label symbols on
5979 ;; `c-record-ref-identifiers' if `c-record-type-identifiers' (!) is
5980 ;; non-nil.
5981 ;;
5982 ;; This function might do hidden buffer changes.
5985 label-end
5986 qt-symbol-idx
5987 macro-start ; if we're in one.
5988 label-type
5989 kwd)
5991 ;; "case" or "default" (Doesn't apply to AWK).
5994 ;; Record only the keyword itself for fontification, since in
5995 ;; case labels the following is a constant expression and not
5996 ;; a label.
6000 ;; Find the label end.
6004 ;; Stop on chars that aren't allowed in expressions,
6005 ;; and on operator chars that would be meaningless
6006 ;; there. FIXME: This doesn't cope with ?: operators.
6014 t)
6016 ;; It's an unfinished label. We consider the keyword enough
6017 ;; to recognize it as a label, so that it gets fontified.
6018 ;; Leave the point at the end of it, but don't put any
6019 ;; `c-decl-end' marker.
6021 t))))
6023 ;; @private, @protected, @public, in Objective C, or similar.
6026 ;; For a `c-opt-extra-label-key' match, we record the whole
6027 ;; thing for fontification. That's to get the leading '@' in
6028 ;; Objective-C protection labels fontified.
6035 ;; All other cases of labels.
6038 ;; A colon label must have something before the colon.
6041 ;; Check that we're not after a token that can't precede a label.
6043 ;; Trivially succeeds when there's no preceding token.
6051 ;; Check if we're after a label, if we're after a closing
6052 ;; paren that belong to statement, and with
6053 ;; `c-label-prefix-re'. It's done in different order
6054 ;; depending on `assume-markup' since the checks have
6055 ;; different expensiveness.
6082 ;; Might be after another label, so check it recursively.
6086 ;; Essentially the same as the
6087 ;; `c-syntactic-re-search-forward' regexp below.
6093 ;; Note: the following should work instead of the
6094 ;; narrow-to-region above. Investigate why not,
6095 ;; sometime. ACM, 2006-03-31.
6096 ;; (c-syntactic-skip-backward "^-]:?;}=*/%&|,<>!@+"
6097 ;; macro-start t)
6099 ;; If the caller turned on recording for us,
6100 ;; it shouldn't apply when we check the
6101 ;; preceding label.
6102 c-record-type-identifiers)
6103 ;; A label can't start at a cpp directive. Check for
6104 ;; this, since c-forward-syntactic-ws would foul up on it.
6109 ;; Point is still at the beginning of the possible label construct.
6110 ;;
6111 ;; Check that the next nonsymbol token is ":", or that we're in one
6112 ;; of QT's "slots" declarations. Allow '(' for the sake of macro
6113 ;; arguments. FIXME: Should build this regexp from the language
6114 ;; constants.
6116 ;; public: protected: private:
6126 ;; QT double keyword like "protected slots:" or goto target.
6145 'qt-1kwd-colon
6157 ;; Check that `c-nonlabel-token-key' doesn't match anywhere.
6167 t)
6170 ;; Record the identifiers in the label for fontification, unless
6171 ;; it begins with `c-label-kwds' in which case the following
6172 ;; identifiers are part of a (constant) expression that
6173 ;; shouldn't be fontified.
6185 ;; Not a label.
6187 label-type))
6190 ;; Assuming the point is at the beginning of a token, try to move
6191 ;; forward to the end of the Objective-C directive that starts
6192 ;; there. Return t if a directive was fully recognized, otherwise
6193 ;; the point is moved as far as one could be successfully parsed and
6194 ;; nil is returned.
6195 ;;
6196 ;; This function records identifier ranges on
6197 ;; `c-record-type-identifiers' and `c-record-ref-identifiers' if
6198 ;; `c-record-type-identifiers' is non-nil.
6199 ;;
6200 ;; This function might do hidden buffer changes.
6203 start-char
6205 ;; Turn off recognition of angle bracket arglists while parsing
6206 ;; types here since the protocol reference list might then be
6207 ;; considered part of the preceding name or superclass-name.
6218 t)
6227 ;; Handle the name of the class itself.
6229 ; (c-forward-token-2) ; 2006/1/13 This doesn't move if the token's
6230 ; at EOB.
6236 ;; Look for ": superclass-name" or "( category-name )".
6247 ;; Look for a protocol reference list.
6253 t))))
6259 t)
6262 nil)))
6265 ;; Go to the first non-whitespace after the colon that starts a
6266 ;; multiple inheritance introduction. Optional LIM is the farthest
6267 ;; back we should search.
6268 ;;
6269 ;; This function might do hidden buffer changes.
6277 ;; Return nil if we aren't in a method definition, otherwise the
6278 ;; position of the initial [+-].
6279 ;;
6280 ;; This function might do hidden buffer changes.
6286 ))
6288 ;; Contributed by Kevin Ryde <user42@zip.com.au>.
6290 ;; Return non-nil if point is within a gcc \"asm\" block.
6291 ;;
6292 ;; This should be called with point inside an argument list.
6293 ;;
6294 ;; Only one level of enclosing parentheses is considered, so for
6295 ;; instance `nil' is returned when in a function call within an asm
6296 ;; operand.
6297 ;;
6298 ;; This function might do hidden buffer changes.
6310 a function.
6312 More precisely, being at the top-level means that point is either
6313 outside any enclosing block (such as a function definition), or
6314 directly inside a class, namespace or other block that contains
6315 another declaration level.
6317 If point is not at the top-level (e.g. it is inside a method
6318 definition), then nil is returned. Otherwise, if point is at a
6319 top-level not enclosed within a class definition, t is returned.
6320 Otherwise, a 2-vector is returned where the zeroth element is the
6321 buffer position of the start of the class declaration, and the first
6322 element is the buffer position of the enclosing class's opening
6323 brace.
6325 Note that this function might do hidden buffer changes. See the
6326 comment at the start of cc-engine.el for more info."
6332 ;; Return non-nil if the point is in the region after the argument
6333 ;; list of a function and its opening brace (or semicolon in case it
6334 ;; got no body). If there are K&R style argument declarations in
6335 ;; that region, the point has to be inside the first one for this
6336 ;; function to recognize it.
6337 ;;
6338 ;; If successful, the point is moved to the first token after the
6339 ;; function header (see `c-forward-decl-or-cast-1' for details) and
6340 ;; the position of the opening paren of the function arglist is
6341 ;; returned.
6342 ;;
6343 ;; The point is clobbered if not successful.
6344 ;;
6345 ;; LIM is used as bound for backward buffer searches.
6346 ;;
6347 ;; This function might do hidden buffer changes.
6360 ;; There should not be a '=' or ',' between beg and the
6361 ;; start of the declaration since that means we were in the
6362 ;; "expression part" of the declaration.
6367 ;; Check that there's an arglist paren in the
6368 ;; declaration.
6371 ;; The declarator is a paren expression, so skip past it
6372 ;; so that we don't get stuck on that instead of the
6373 ;; function arglist.
6377 ;; Don't trip up on "operator ()".
6384 ;; Return the position of the first argument declaration if point is
6385 ;; inside a K&R style argument declaration list, nil otherwise.
6386 ;; `c-recognize-knr-p' is not checked. If LIM is non-nil, it's a
6387 ;; position that bounds the backward search for the argument list.
6388 ;;
6389 ;; Point must be within a possible K&R region, e.g. just before a top-level
6390 ;; "{". It must be outside of parens and brackets. The test can return
6391 ;; false positives otherwise.
6392 ;;
6393 ;; This function might do hidden buffer changes.
6397 ;; If we're in a macro, our search range is restricted to it. Narrow to
6398 ;; the searchable range.
6401 before-lparen after-rparen
6405 ;; Search backwards for the defun's argument list. We give up if we
6406 ;; encounter a "}" (end of a previous defun) or BOB.
6407 ;;
6408 ;; The criterion for a paren structure being the arg list is:
6409 ;; o - there is non-WS stuff after it but before any "{"; AND
6410 ;; o - the token after it isn't a ";" AND
6411 ;; o - it is preceded by either an identifier (the function name) or
6412 ;; a macro expansion like "DEFUN (...)"; AND
6413 ;; o - its content is a non-empty comma-separated list of identifiers
6414 ;; (an empty arg list won't have a knr region).
6415 ;;
6416 ;; The following snippet illustrates these rules:
6417 ;; int foo (bar, baz, yuk)
6418 ;; int bar [] ;
6419 ;; int (*baz) (my_type) ;
6420 ;; int (*) (void) (*yuk) (void) ;
6421 ;; {
6435 ;; We're inside a paren. Could it be our argument list....?
6441 ;; FIXME!!! What about macros between the parens? 2007/01/20
6444 ;; It can't be the arg list if next token is ; or {
6449 ;; Is the thing preceding the list an identifier (the
6450 ;; function name), or a macro expansion?
6460 ;; Have we got a non-empty list of comma-separated
6461 ;; identifiers?
6475 ;; ...Yes. We've identified the function's argument list.
6481 ;; ...No. The current parens aren't the function's arg list.
6488 ;; skip forward over conditional at point, including any predicate
6489 ;; statements in parentheses. No error checking is performed.
6490 ;;
6491 ;; This function might do hidden buffer changes.
6493 ;; else if()
6498 ;; do, else, try, finally
6503 ;; for, if, while, switch, catch, synchronized, foreach
6507 ;; If looking at the token after a conditional then return the
6508 ;; position of its start, otherwise return nil.
6509 ;;
6510 ;; This function might do hidden buffer changes.
6520 ;; If the point is after an operator identifier that isn't handled
6521 ;; like an ordinary symbol (i.e. like "operator =" in C++) then the
6522 ;; position of the start of that identifier is returned. nil is
6523 ;; returned otherwise. The point may be anywhere in the syntactic
6524 ;; whitespace after the last token of the operator identifier.
6525 ;;
6526 ;; This function might do hidden buffer changes.
6538 ;; Assuming point is at a brace that opens a statement block of some
6539 ;; kind, move to the proper anchor point for that block. It might
6540 ;; need to be adjusted further by c-add-stmt-syntax, but the
6541 ;; position at return is suitable as start position for that
6542 ;; function.
6543 ;;
6544 ;; This function might do hidden buffer changes.
6551 ;; Assuming point is at a brace that opens the block of a top level
6552 ;; declaration of some kind, move to the proper anchor point for
6553 ;; that block.
6554 ;;
6555 ;; This function might do hidden buffer changes.
6560 ;; Search forward for the end of the "header" of the current
6561 ;; declaration. That's the position where the definition body
6562 ;; starts, or the first variable initializer, or the ending
6563 ;; semicolon. I.e. search forward for the closest following
6564 ;; (syntactically relevant) '{', '=' or ';' token. Point is left
6565 ;; _after_ the first found token, or at point-max if none is found.
6566 ;;
6567 ;; This function might do hidden buffer changes.
6572 ;; In C++ we need to take special care to handle operator
6573 ;; tokens and those pesky template brackets.
6578 ;; Handle operator identifiers, i.e. ignore any
6579 ;; operator token preceded by "operator".
6586 t
6588 nil)))))
6597 ;; Go to the beginning of the current declaration, or the beginning
6598 ;; of the previous one if already at the start of it. Point won't
6599 ;; be moved out of any surrounding paren. Return a cons cell of the
6600 ;; form (MOVE . KNR-POS). MOVE is like the return value from
6601 ;; `c-beginning-of-statement-1'. If point skipped over some K&R
6602 ;; style argument declarations (and they are to be recognized) then
6603 ;; KNR-POS is set to the start of the first such argument
6604 ;; declaration, otherwise KNR-POS is nil. If LIM is non-nil, it's a
6605 ;; position that bounds the backward search.
6606 ;;
6607 ;; NB: Cases where the declaration continues after the block, as in
6608 ;; "struct foo { ... } bar;", are currently recognized as two
6609 ;; declarations, e.g. "struct foo { ... }" and "bar;" in this case.
6610 ;;
6611 ;; This function might do hidden buffer changes.
6617 ;; `c-beginning-of-statement-1' stops at a block start, but we
6618 ;; want to continue if the block doesn't begin a top level
6619 ;; construct, i.e. if it isn't preceded by ';', '}', ':', bob,
6620 ;; or an open paren.
6622 ;; Go back one "statement" each time round the loop until we're just
6623 ;; after a ;, }, or :, or at BOB or the start of a macro or start of
6624 ;; an ObjC method. This will move over a multiple declaration whose
6625 ;; components are comma separated.
6627 ;; Must check with c-opt-method-key in ObjC mode.
6637 ;; Check that we don't move from the first thing in a
6638 ;; macro to its header.
6644 move tentative-move))
6649 ;; Handle K&R argdecls. Back up after the "statement" jumped
6650 ;; over by `c-beginning-of-statement-1', unless it was the
6651 ;; function body, in which case we're sitting on the opening
6652 ;; brace now. Then test if we're in a K&R argdecl region and
6653 ;; that we started at the other side of the first argdecl in
6654 ;; it.
6664 'previous
6666 knr-argdecl-start))
6669 ;; `c-beginning-of-statement-1' counts each brace block as a separate
6670 ;; statement, so the result will be 'previous if we've moved over any.
6671 ;; So change our result back to 'same if necessary.
6672 ;;
6673 ;; If they were brace list initializers we might not have moved over a
6674 ;; declaration boundary though, so change it to 'same if we've moved
6675 ;; past a '=' before '{', but not ';'. (This ought to be integrated
6676 ;; into `c-beginning-of-statement-1', so we avoid this extra pass which
6677 ;; potentially can search over a large amount of text.). Take special
6678 ;; pains not to get mislead by C++'s "operator=", and the like.
6681 c++-template-syntax-table
6687 ; no more, or get to one which isn't an "operator ="
6690 c-overloadable-operators-regexp
6691 c-opt-op-identifier-prefix
6706 ;; Assuming point is at the start of a declaration (as detected by
6707 ;; e.g. `c-beginning-of-decl-1'), go to the end of it. Unlike
6708 ;; `c-beginning-of-decl-1', this function handles the case when a
6709 ;; block is followed by identifiers in e.g. struct declarations in C
6710 ;; or C++. If a proper end was found then t is returned, otherwise
6711 ;; point is moved as far as possible within the current sexp and nil
6712 ;; is returned. This function doesn't handle macros; use
6713 ;; `c-end-of-macro' instead in those cases.
6714 ;;
6715 ;; This function might do hidden buffer changes.
6718 c++-template-syntax-table
6726 ;; Stopped at the ';' in a K&R argdecl section which is
6727 ;; detected using the same criteria as in
6728 ;; `c-beginning-of-decl-1'. Move to the following block
6729 ;; start.
6733 ;; Encountered a block in the declaration. Jump over it.
6744 ;; Check for `c-opt-block-decls-with-vars-key'
6745 ;; before the first paren.
6748 c-opt-block-decls-with-vars-key
6750 lim t t t)
6753 ;; Check that the first following paren is
6754 ;; the block.
6756 lim t t t)
6758 ;; The declaration doesn't have any of the
6759 ;; `c-opt-block-decls-with-vars' keywords in the
6760 ;; beginning, so it ends here at the end of the block.
6768 nil)))
6771 ;; Assuming the point is at an open brace, check if it starts a
6772 ;; block that contains another declaration level, i.e. that isn't a
6773 ;; statement block or a brace list, and if so return non-nil.
6774 ;;
6775 ;; If the check is successful, the return value is the start of the
6776 ;; keyword that tells what kind of construct it is, i.e. typically
6777 ;; what `c-decl-block-key' matched. Also, if GOTO-START is set then
6778 ;; the point will be at the start of the construct, before any
6779 ;; leading specifiers, otherwise it's at the returned position.
6780 ;;
6781 ;; The point is clobbered if the check is unsuccessful.
6782 ;;
6783 ;; CONTAINING-SEXP is the position of the open of the surrounding
6784 ;; paren, or nil if none.
6785 ;;
6786 ;; The optional LIMIT limits the backward search for the start of
6787 ;; the construct. It's assumed to be at a syntactically relevant
6788 ;; position.
6789 ;;
6790 ;; If any template arglists are found in the searched region before
6791 ;; the open brace, they get marked with paren syntax.
6792 ;;
6793 ;; This function might do hidden buffer changes.
6800 ;; Could be at the end of a template arglist.
6811 ;; Note: Can't get bogus hits inside template arglists below since they
6812 ;; have gotten paren syntax above.
6814 ;; If `goto-start' is set we begin by searching for the
6815 ;; first possible position of a leading specifier list.
6816 ;; The `c-decl-block-key' search continues from there since
6817 ;; we know it can't match earlier.
6820 open-brace t t)
6822 t)
6823 t)
6830 ;; Found a keyword that can't be a type?
6833 ;; Can be a type too, in which case it's the return type of a
6834 ;; function (under the assumption that no declaration level
6835 ;; block construct starts with a type).
6838 ;; Jumped over a type, but it could be a declaration keyword
6839 ;; followed by the declared identifier that we've jumped over
6840 ;; instead (e.g. in "class Foo {"). If it indeed is a type
6841 ;; then we should be at the declarator now, so check for a
6842 ;; valid declarator start.
6843 ;;
6844 ;; Note: This doesn't cope with the case when a declared
6845 ;; identifier is followed by e.g. '(' in a language where '('
6846 ;; also might be part of a declarator expression. Currently
6847 ;; there's no such language.
6851 ;; In Pike a list of modifiers may be followed by a brace
6852 ;; to make them apply to many identifiers. Note that the
6853 ;; match data will be empty on return in this case.
6859 ;; Use this variant to avoid yet another special regexp.
6863 t)))
6865 ;; Got a match.
6868 ;; Back up over any preceding specifiers and their clauses
6869 ;; by going forward from `first-specifier-pos', which is the
6870 ;; earliest possible position where the specifier list can
6871 ;; start.
6877 ;; Accept any plain symbol token on the ground that
6878 ;; it's a specifier masked through a macro (just
6879 ;; like `c-forward-decl-or-cast-1' skip forward over
6880 ;; such tokens).
6881 ;;
6882 ;; Could be more restrictive wrt invalid keywords,
6883 ;; but that'd only occur in invalid code so there's
6884 ;; no use spending effort on it.
6890 ;; Can't parse a declaration preamble and is still
6891 ;; before `kwd-start'. That means `first-specifier-pos'
6892 ;; was in some earlier construct. Search again.
6896 ;; Got no preamble before the block declaration keyword.
6902 kwd-start)))
6905 ;; Check if the closest containing paren sexp is a declaration
6906 ;; block, returning a 2 element vector in that case. Aref 0
6907 ;; contains the bufpos at boi of the class key line, and aref 1
6908 ;; contains the bufpos of the open brace. This function is an
6909 ;; obsolete wrapper for `c-looking-at-decl-block'.
6910 ;;
6911 ;; This function might do hidden buffer changes.
6919 nil))
6924 ;; return the buffer position of the beginning of the brace list
6925 ;; statement if we're inside a brace list, otherwise return nil.
6926 ;; CONTAINING-SEXP is the buffer pos of the innermost containing
6927 ;; paren. PAREN-STATE is the remainder of the state of enclosing
6928 ;; braces
6929 ;;
6930 ;; N.B.: This algorithm can potentially get confused by cpp macros
6931 ;; placed in inconvenient locations. It's a trade-off we make for
6932 ;; speed.
6933 ;;
6934 ;; This function might do hidden buffer changes.
6936 ;; This will pick up brace list declarations.
6949 ;; this will pick up array/aggregate init lists, even if they are nested.
6952 ;; Pike can have class definitions anywhere, so we must
6953 ;; check for the class key here.
6955 c-decl-block-key))
6956 bufpos braceassignp lim next-containing)
6958 containing-sexp)
6969 ;; We're in an in-expression block of some kind. Do not
6970 ;; check nesting. We deliberately set the limit to the
6971 ;; containing sexp, so that c-looking-at-inexpr-block
6972 ;; doesn't check for an identifier before it.
6974 ;; see if the open brace is preceded by = or [...] in
6975 ;; this statement, but watch out for operator=
6978 ;; Checks to do only on the first sexp before the brace.
6981 ;; In Java, an initialization brace list may follow
6982 ;; directly after "new Foo[]", so check for a "new"
6983 ;; earlier.
6989 ;; Carry on looking if this is an
6990 ;; identifier (may contain "." in Java)
6991 ;; or another "[]" sexp.
6994 ;; Checks to do on all sexps before the brace, up to the
6995 ;; beginning of the statement.
7003 ;; We've seen a =, but must check earlier tokens so
7004 ;; that it isn't something that should be ignored.
7010 ;; Check for operator =
7013 nil)
7014 ;; Check for `<opchar>= in Pike.
7017 ;; Special case for Pikes
7018 ;; `[]=, since '[' is not in
7019 ;; the punctuation class.
7022 nil)
7024 ;; make sure we're not in a C++ template
7025 ;; argument assignment
7037 ))))
7038 nil)
7045 ;; Brace lists can't contain a semicolon, so we're done.
7047 ;; Go up one level.
7049 lim nil
7050 next-containing nil))
7051 ;; we've hit the beginning of the aggregate list
7055 )
7056 bufpos))
7057 ))
7060 ;; If we're looking at the start of a pike-style list, ie `({Â })',
7061 ;; `([Â ])', `(<Â >)' etc, a cons of a cons of its starting and ending
7062 ;; positions and its entry in c-special-brace-lists is returned, nil
7063 ;; otherwise. The ending position is nil if the list is still open.
7064 ;; LIM is the limit for forward search. The point may either be at
7065 ;; the `(' or at the following paren character. Tries to check the
7066 ;; matching closer, but assumes it's correct if no balanced paren is
7067 ;; found (i.e. the case `({ ... } ... )' is detected as _not_ being
7068 ;; a special brace list).
7069 ;;
7070 ;; This function might do hidden buffer changes.
7075 inner-beg end type)
7090 nil)))))
7100 ;; Check balancing of the inner paren
7101 ;; below.
7104 t)
7105 ;; If the inner char isn't a paren then
7106 ;; we can't check balancing, so just
7107 ;; check the char before the outer
7108 ;; closing paren.
7123 ;; Return non-nil if between two statements or declarations, assuming
7124 ;; point is not inside a literal or comment.
7125 ;;
7126 ;; Obsolete - `c-at-statement-start-p' or `c-at-expression-start-p'
7127 ;; are recommended instead.
7128 ;;
7129 ;; This function might do hidden buffer changes.
7134 ;; Return non-nil if we're looking at the beginning of a block
7135 ;; inside an expression. The value returned is actually a cons of
7136 ;; either 'inlambda, 'inexpr-statement or 'inexpr-class and the
7137 ;; position of the beginning of the construct.
7138 ;;
7139 ;; LIM limits the backward search. CONTAINING-SEXP is the start
7140 ;; position of the closest containing list. If it's nil, the
7141 ;; containing paren isn't used to decide whether we're inside an
7142 ;; expression or not. If both LIM and CONTAINING-SEXP are used, LIM
7143 ;; needs to be farther back.
7144 ;;
7145 ;; If CHECK-AT-END is non-nil then extra checks at the end of the
7146 ;; brace block might be done. It should only be used when the
7147 ;; construct can be assumed to be complete, i.e. when the original
7148 ;; starting position was further down than that.
7149 ;;
7150 ;; This function might do hidden buffer changes.
7155 ;; Look at the character after point only as a last resort
7156 ;; when we can't disambiguate.
7176 ;; If the class definition is at the start of
7177 ;; a statement, we don't consider it an
7178 ;; in-expression class.
7186 ;; Also, in Pike we treat it as an
7187 ;; in-expression class if it's used in an
7188 ;; object clone expression.
7204 nil)
7212 ;; Accept several square bracket sexps for
7213 ;; Java array initializations.
7221 block-follows
7222 containing-sexp
7231 nil
7234 res))))
7237 ;; Returns non-nil if we're looking at the end of an in-expression
7238 ;; block, otherwise the same as `c-looking-at-inexpr-block'.
7239 ;; PAREN-STATE is the paren state relevant at the current position.
7240 ;;
7241 ;; This function might do hidden buffer changes.
7243 ;; We currently only recognize a block.
7246 containing-sexp)
7255 paren-state)
7256 containing-sexp)))))
7258 \f
7259 ;; `c-guess-basic-syntax' and the functions that precedes it below
7260 ;; implements the main decision tree for determining the syntactic
7261 ;; analysis of the current line of code.
7263 ;; Dynamically bound to t when `c-guess-basic-syntax' is called during
7264 ;; auto newline analysis.
7268 ;; BRACEPOS is the position of a brace in a construct like "namespace
7269 ;; Bar {". Return the anchor point in this construct; this is the
7270 ;; earliest symbol on the brace's line which isn't earlier than
7271 ;; "namespace".
7272 ;;
7273 ;; Currently (2007-08-17), "like namespace" means "matches
7274 ;; c-other-block-decl-kwds". It doesn't work with "class" or "struct"
7275 ;; or anything like that.
7285 ;; A simple function to prepend a new syntax element to
7286 ;; `c-syntactic-context'. Using `setq' on it is unsafe since it
7287 ;; should always be dynamically bound but since we read it first
7288 ;; we'll fail properly anyway if this function is misused.
7290 c-syntactic-context)))
7293 ;; Like `c-add-syntax' but appends to the end of the syntax list.
7294 ;; (Normally not necessary.)
7299 syntax-extra-args
7300 stop-at-boi-only
7301 containing-sexp
7302 paren-state)
7303 ;; Add the indicated SYNTAX-SYMBOL to `c-syntactic-context', extending it as
7304 ;; needed with further syntax elements of the types `substatement',
7305 ;; `inexpr-statement', `arglist-cont-nonempty', `statement-block-intro', and
7306 ;; `defun-block-intro'.
7307 ;;
7308 ;; Do the generic processing to anchor the given syntax symbol on
7309 ;; the preceding statement: Skip over any labels and containing
7310 ;; statements on the same line, and then search backward until we
7311 ;; find a statement or block start that begins at boi without a
7312 ;; label or comment.
7313 ;;
7314 ;; Point is assumed to be at the prospective anchor point for the
7315 ;; given SYNTAX-SYMBOL. More syntax entries are added if we need to
7316 ;; skip past open parens and containing statements. Most of the added
7317 ;; syntax elements will get the same anchor point - the exception is
7318 ;; for an anchor in a construct like "namespace"[*] - this is as early
7319 ;; as possible in the construct but on the same line as the {.
7320 ;;
7321 ;; [*] i.e. with a keyword matching c-other-block-decl-kwds.
7322 ;;
7323 ;; SYNTAX-EXTRA-ARGS are a list of the extra arguments for the
7324 ;; syntax symbol. They are appended after the anchor point.
7325 ;;
7326 ;; If STOP-AT-BOI-ONLY is nil, we can stop in the middle of the line
7327 ;; if the current statement starts there.
7328 ;;
7329 ;; Note: It's not a problem if PAREN-STATE "overshoots"
7330 ;; CONTAINING-SEXP, i.e. contains info about parens further down.
7331 ;;
7332 ;; This function might do hidden buffer changes.
7335 ;; This is by far the most common case, so let's give it special
7336 ;; treatment.
7341 ;; Set when we're on a label, so that we don't stop there.
7342 ;; FIXME: To be complete we should check if we're on a label
7343 ;; now at the start.
7344 on-label)
7346 ;; Use point as the anchor point for "namespace", "extern", etc.
7350 syntax-extra-args)
7352 ;; Loop while we have to back out of containing blocks.
7357 ;; Loop while we have to back up statements.
7359 on-label
7362 ;; Skip past any comments that stands between the
7363 ;; statement start and boi.
7371 ;; Skip to the beginning of this statement or backward
7372 ;; another one.
7380 ;; If we didn't move we're at the start of a block and
7381 ;; have to continue outside it.
7390 ;; Special case to avoid deeper and deeper indentation
7391 ;; of "else if" clauses.
7392 )
7397 ;; If stop-at-boi-only is nil, we shouldn't back up
7398 ;; over previous or containing statements to try to
7399 ;; reach boi, so go back to the last position and
7400 ;; exit.
7407 ;; If we've moved into another statement then we
7408 ;; should no longer try to stop in the middle of a
7409 ;; line.
7412 ;; Record this as a substatement if we skipped up one
7413 ;; level.
7416 )))
7418 containing-sexp)
7420 ;; Now we have to go out of this block.
7423 ;; Don't stop in the middle of a special brace list opener
7424 ;; like "({".
7436 ;; Analyze the construct in front of the block we've stepped out
7437 ;; from and add the right syntactic element for it.
7440 step-type)
7443 ;; Stepped out of a parenthesis block, so we're in an
7444 ;; expression now.
7456 ;; Stepped out of an in-expression statement. This
7457 ;; syntactic element won't get an anchor pos.
7460 ;; A parenthesis normally belongs to an arglist.
7468 on-label nil))
7470 ;; Stepped out of a brace block.
7482 containing-sexp)))
7484 'defun-block-intro
7486 nil))
7490 c-other-decl-block-key-in-symbols-alist))
7497 ;; Always done if the open brace was at boi. The
7498 ;; c-beginning-of-statement-1 call above is necessary
7499 ;; anyway, to decide the type of block-intro to add.
7502 ))
7504 ;; Fill in the current point as the anchor for all the symbols
7505 ;; added above.
7514 )))
7517 containing-decl-open
7518 containing-decl-start
7519 containing-decl-kwd
7520 paren-state)
7521 ;; The inclass and class-close syntactic symbols are added in
7522 ;; several places and some work is needed to fix everything.
7523 ;; Therefore it's collected here.
7524 ;;
7525 ;; This function might do hidden buffer changes.
7530 containing-decl-open)
7532 ;; Ought to use `c-add-stmt-syntax' instead of backing up to boi
7533 ;; here, but we have to do like this for compatibility.
7543 char-after-ip
7544 beg-of-same-or-containing-stmt
7545 containing-sexp
7546 paren-state)
7547 ;; This function contains the decision tree reached through both
7548 ;; cases 18 and 10. It's a continued statement or top level
7549 ;; construct of some kind.
7550 ;;
7551 ;; This function might do hidden buffer changes.
7558 ;; (CASE A removed.)
7559 ;; CASE B: open braces for class or brace-lists
7566 ;; CASE B.1: class-open
7573 ;; CASE B.2: brace-list-open
7578 indent-point t t t)))
7579 ;; The most semantically accurate symbol here is
7580 ;; brace-list-open, but we normally report it simply as a
7581 ;; statement-cont. The reason is that one normally adjusts
7582 ;; brace-list-open for brace lists as top-level constructs,
7583 ;; and brace lists inside statements is a completely different
7584 ;; context. C.f. case 5A.3.
7587 ;; Turn off the dwim above when we're
7588 ;; analyzing the nature of the brace
7589 ;; for the auto newline feature.
7590 'brace-list-open
7592 nil nil
7593 containing-sexp paren-state))
7595 ;; CASE B.3: The body of a function declared inside a normal
7596 ;; block. Can occur e.g. in Pike and when using gcc
7597 ;; extensions, but watch out for macros followed by blocks.
7598 ;; C.f. cases E, 16F and 17G.
7604 ;; Turn off recognition of constructs that lacks a
7605 ;; type in this case, since that's more likely to be
7606 ;; a macro followed by a block.
7609 containing-sexp paren-state))
7611 ;; CASE B.4: Continued statement with block open. The most
7612 ;; accurate analysis is perhaps `statement-cont' together with
7613 ;; `block-open' but we play DWIM and use `substatement-open'
7614 ;; instead. The rationaly is that this typically is a macro
7615 ;; followed by a block which makes it very similar to a
7616 ;; statement with a substatement block.
7619 containing-sexp paren-state))
7620 ))
7622 ;; CASE C: iostream insertion or extraction operator
7626 ;; If there is no preceding streamop in the statement
7627 ;; then indent this line as a normal statement-cont.
7631 t))))
7633 ;; CASE E: In the "K&R region" of a function declared inside a
7634 ;; normal block. C.f. case B.3.
7636 ;; Check that the next token is a '{'. This works as
7637 ;; long as no language that allows nested function
7638 ;; definitions allows stuff like member init lists, K&R
7639 ;; declarations or throws clauses there.
7640 ;;
7641 ;; Note that we do a forward search for something ahead
7642 ;; of the indentation line here. That's not good since
7643 ;; the user might not have typed it yet. Unfortunately
7644 ;; it's exceedingly tricky to recognize a function
7645 ;; prototype in a code block without resorting to this.
7653 ;; Turn off recognition of constructs that lacks a
7654 ;; type in this case, since that's more likely to be
7655 ;; a macro followed by a block.
7658 containing-sexp paren-state))
7660 ;; CASE D: continued statement.
7664 containing-sexp paren-state))
7665 )))
7667 ;; The next autoload was added by RMS on 2005/8/9 - don't know why (ACM,
7668 ;; 2005/11/29).
7669 ;;;###autoload
7671 "Return the syntactic context of the current line."
7677 ;; A whole ugly bunch of various temporary variables. Have
7678 ;; to declare them here since it's not possible to declare
7679 ;; a variable with only the scope of a cond test and the
7680 ;; following result clauses, and most of this function is a
7681 ;; single gigantic cond. :P
7682 literal char-before-ip before-ws-ip char-after-ip macro-start
7683 in-macro-expr c-syntactic-context placeholder c-in-literal-cache
7684 step-type tmpsymbol keyword injava-inher special-brace-list tmp-pos
7685 containing-<
7686 ;; The following record some positions for the containing
7687 ;; declaration block if we're directly within one:
7688 ;; `containing-decl-open' is the position of the open
7689 ;; brace. `containing-decl-start' is the start of the
7690 ;; declaration. `containing-decl-kwd' is the keyword
7691 ;; symbol of the keyword that tells what kind of block it
7692 ;; is.
7693 containing-decl-open
7694 containing-decl-start
7695 containing-decl-kwd
7696 ;; The open paren of the closest surrounding sexp or nil if
7697 ;; there is none.
7698 containing-sexp
7699 ;; The position after the closest preceding brace sexp
7700 ;; (nested sexps are ignored), or the position after
7701 ;; `containing-sexp' if there is none, or (point-min) if
7702 ;; `containing-sexp' is nil.
7703 lim
7704 ;; The paren state outside `containing-sexp', or at
7705 ;; `indent-point' if `containing-sexp' is nil.
7707 ;; There's always at most one syntactic element which got
7708 ;; an anchor pos. It's stored in syntactic-relpos.
7709 syntactic-relpos
7712 ;; Check if we're directly inside an enclosing declaration
7713 ;; level block.
7722 containing-sexp)
7723 t)))
7726 containing-sexp nil)
7731 ;; Init some position variables.
7740 ;; Ignore balanced paren. The next entry
7741 ;; can't be another one.
7744 ;; If there is no surrounding open paren then
7745 ;; put the last balanced pair back on paren-state.
7747 containing-sexp nil)))
7751 ;; If we're in a parenthesis list then ',' delimits the
7752 ;; "statements" rather than being an operator (with the
7753 ;; exception of the "for" clause). This difference is
7754 ;; typically only noticeable when statements are used in macro
7755 ;; arglists.
7760 ;; cache char before and after indent point, and move point to
7761 ;; the most likely position to perform the majority of tests
7770 ;; are we in a literal?
7773 ;; now figure out syntactic qualities of the current line
7776 ;; CASE 1: in a string.
7780 ;; CASE 2: in a C or C++ style comment.
7782 ;; This is a kludge for XEmacs where we use
7783 ;; `buffer-syntactic-context', which doesn't correctly
7784 ;; recognize "\*/" to end a block comment.
7785 ;; `parse-partial-sexp' which is used by
7786 ;; `c-literal-limits' will however do that in most
7787 ;; versions, which results in that we get nil from
7788 ;; `c-literal-limits' even when `c-in-literal' claims
7789 ;; we're inside a comment.
7793 ;; CASE 3: in a cpp preprocessor macro continuation.
7803 ;; If at the beginning of the body of a #define
7804 ;; directive then analyze as cpp-define-intro
7805 ;; only. Go on with the syntactic analysis
7806 ;; otherwise. in-macro-expr is set if we're in a
7807 ;; cpp expression, i.e. before the #define body
7808 ;; or anywhere in a non-#define directive.
7815 nil)))))
7819 ;; CASE 11: an else clause?
7823 containing-sexp paren-state))
7825 ;; CASE 12: while closure of a do/while construct?
7833 containing-sexp paren-state))
7835 ;; CASE 13: A catch or finally clause? This case is simpler
7836 ;; than if-else and do-while, because a block is required
7837 ;; after every try, catch and finally.
7845 t)
7849 t)
7852 t))
7857 containing-sexp paren-state))
7859 ;; CASE 18: A substatement we can recognize by keyword.
7869 ;; Ought to cache the result from the
7870 ;; c-beginning-of-statement-1 calls here.
7880 ;; Step up to the containing statement if we
7881 ;; stayed in the same one.
7889 ;; There was no containing statement afterall.
7891 placeholder))
7893 ;; Require a parenthesis after these keywords.
7894 ;; Necessary to catch e.g. synchronized in Java,
7895 ;; which can be used both as statement and
7896 ;; modifier.
7902 ;; CASE 18A: Simple substatement.
7908 containing-sexp paren-state))
7914 containing-sexp paren-state))
7917 containing-sexp paren-state))))
7919 ;; CASE 18B: Some other substatement. This is shared
7920 ;; with case 10.
7922 char-after-ip
7923 placeholder
7924 lim
7925 paren-state)))
7927 ;; CASE 14: A case or default label
7933 containing-sexp))
7936 ;; Got a bogus label at the top level. In lack of better
7937 ;; alternatives, anchor it on (point-min).
7940 ;; CASE 15: any other label
7947 containing-decl-open
7948 containing-decl-start
7949 containing-decl-kwd
7950 paren-state))
7951 ;; Append access-label with the same anchor point as
7952 ;; inclass gets.
7958 containing-sexp))
7963 ;; If the surrounding statement is a switch then
7964 ;; let's analyze all labels as switch labels, so
7965 ;; that they get lined up consistently.
7966 'case-label
7972 ;; A label on the top level. Treat it as a class
7973 ;; context. (point-min) is the closest we get to the
7974 ;; class open brace.
7977 ;; CASE 4: In-expression statement. C.f. cases 7B, 16A and
7978 ;; 17E.
7981 containing-sexp
7982 ;; Have to turn on the heuristics after
7983 ;; the point even though it doesn't work
7984 ;; very well. C.f. test case class-16.pike.
7985 t))
7990 ;; It's a statement block or an anonymous class.
7992 ;; It's a Pike lambda. Check whether we are between the
7993 ;; lambda keyword and the argument list or at the defun
7994 ;; opener.
7996 'inline-open
8002 paren-state)
8006 ;; CASE 5: Line is inside a declaration level block or at top level.
8010 ;; CASE 5A: we are looking at a defun, brace list, class,
8011 ;; or inline-inclass method opening brace
8018 ;; CASE 5A.1: Non-class declaration block open.
8033 ;; Special case for extern-lang-open.
8034 'extern-lang-open
8036 nil t containing-sexp paren-state))
8038 ;; CASE 5A.2: we are looking at a class opening brace
8047 ;; CASE 5A.3: brace list open
8072 ))))
8076 ;; We're in Java and have found that the open brace
8077 ;; belongs to a "new Foo[]" initialization list,
8078 ;; which means the brace list is part of an
8079 ;; expression and not a top level definition. We
8080 ;; therefore treat it as any topmost continuation
8081 ;; even though the semantically correct symbol still
8082 ;; is brace-list-open, on the same grounds as in
8083 ;; case B.2.
8089 ;; CASE 5A.4: inline defun open
8095 containing-decl-open
8096 containing-decl-start
8097 containing-decl-kwd
8098 paren-state))
8100 ;; CASE 5A.5: ordinary defun open
8108 ;; Bogus to use bol here, but it's the legacy. (Resolved,
8109 ;; 2007-11-09)
8110 ))))
8112 ;; CASE 5B: After a function header but before the body (or
8113 ;; the ending semicolon if there's no body).
8119 ;; CASE 5B.1: Member init list.
8124 ;; There is no preceding member init clause.
8125 ;; Indent relative to the beginning of indentation
8126 ;; for the topmost-intro line that contains the
8127 ;; prototype's open paren.
8130 ;; Indent relative to the first member init clause.
8135 ;; CASE 5B.2: K&R arg decl intro
8142 containing-decl-open
8143 containing-decl-start
8144 containing-decl-kwd
8145 paren-state)))
8147 ;; CASE 5B.4: Nether region after a C++ or Java func
8148 ;; decl, which could include a `throws' declaration.
8152 )))
8154 ;; CASE 5C: inheritance line. could be first inheritance
8155 ;; line, or continuation of a multiple inheritance
8163 ;; watch out for scope operator
8168 )))
8177 ;; for Java
8182 cont done)
8187 done t))
8191 )
8193 injava-inher)
8196 ))
8199 ;; CASE 5C.1: non-hanging colon on an inher intro
8203 ;; don't add inclass symbol since relative point already
8204 ;; contains any class offset
8205 )
8207 ;; CASE 5C.2: hanging colon on an inher intro
8213 containing-decl-open
8214 containing-decl-start
8215 containing-decl-kwd
8216 paren-state)))
8218 ;; CASE 5C.3: in a Java implements/extends
8232 )))
8234 ;; CASE 5C.4: a continued inheritance line
8238 ;; don't add inclass symbol since relative point already
8239 ;; contains any class offset
8240 )))
8242 ;; CASE 5D: this could be a top-level initialization, a
8243 ;; member init list continuation, or a template argument
8244 ;; list continuation.
8246 ;; Note: We use the fact that lim is always after any
8247 ;; preceding brace sexp.
8257 t))
8262 t)
8268 nil)
8269 t))
8271 ;; NB: No c-after-special-operator-id stuff in this
8272 ;; clause - we assume only C++ needs it.
8277 ;; CASE 5D.3: perhaps a template list continuation?
8292 ;; In a nested template arglist.
8298 ;; FIXME: Should use c-add-stmt-syntax, but it's not yet
8299 ;; template aware.
8302 ;; CASE 5D.4: perhaps a multiple inheritance line?
8314 t)
8319 ;; CASE 5D.5: Continuation of the "expression part" of a
8320 ;; top level construct. Or, perhaps, an unrecognised construct.
8332 ;; A preceding comma at the top level means that a
8333 ;; new variable declaration starts here. Use
8334 ;; topmost-intro-cont for it, for consistency with
8335 ;; the first variable declaration. C.f. case 5N.
8338 nil nil containing-sexp paren-state))
8339 ))
8341 ;; CASE 5F: Close of a non-class declaration level block.
8345 ;; This is inconsistent: Should use `containing-decl-open'
8346 ;; here if it's at boi, like in case 5J.
8350 ;; Special case for compatibility with the
8351 ;; extern-lang syntactic symbols.
8352 'extern-lang-close
8355 nil t
8357 paren-state))
8359 ;; CASE 5G: we are looking at the brace which closes the
8360 ;; enclosing nested class decl
8365 containing-decl-open
8366 containing-decl-start
8367 containing-decl-kwd
8368 paren-state))
8370 ;; CASE 5H: we could be looking at subsequent knr-argdecls
8377 ;; Do an extra check to avoid tripping up on
8378 ;; statements that occur in invalid contexts
8379 ;; (e.g. in macro bodies where we don't really
8380 ;; know the context of what we're looking at).
8387 ;; CASE 5I: ObjC method definition.
8392 ;; Handle the case when it's the first (non-comment)
8393 ;; thing in the buffer. Can't look for a 'same return
8394 ;; value from cbos1 since ObjC directives currently
8395 ;; aren't recognized fully, so that we get 'same
8396 ;; instead of 'previous if it moved over a preceding
8397 ;; directive.
8401 ;; CASE 5P: AWK pattern or function or continuation
8402 ;; thereof.
8408 'topmost-intro-cont
8410 nil nil
8411 containing-sexp paren-state))
8413 ;; CASE 5N: At a variable declaration that follows a class
8414 ;; definition or some other block declaration that doesn't
8415 ;; end at the closing '}'. C.f. case 5D.5.
8424 ;; Speed up the backward search a bit.
8429 ;; The '}' is unbalanced.
8430 nil
8435 containing-sexp paren-state))
8437 ;; NOTE: The point is at the end of the previous token here.
8439 ;; CASE 5J: we are at the topmost level, make
8440 ;; sure we skip back past any access specifiers
8442 ;; A macro continuation line is never at top level.
8464 nil)))))
8465 ;; For historic reasons we anchor at bol of the last
8466 ;; line of the previous declaration. That's clearly
8467 ;; highly bogus and useless, and it makes our lives hard
8468 ;; to remain compatible. :P
8479 ;; Special case for compatibility with the
8480 ;; extern-lang syntactic symbols.
8481 'inextern-lang
8484 nil t
8486 paren-state))
8488 containing-decl-open
8489 containing-decl-start
8490 containing-decl-kwd
8491 paren-state)))
8493 macro-start
8498 ;; CASE 5K: we are at an ObjC method definition
8499 ;; continuation line.
8508 ;; CASE 5L: we are at the first argument of a template
8509 ;; arglist that begins on the previous line.
8518 ;; CASE 5Q: we are at a statement within a macro.
8523 ;; CASE 5M: we are at a topmost continuation line
8534 ))
8536 ;; (CASE 6 has been removed.)
8538 ;; CASE 19: line is an expression, not a statement, and is directly
8539 ;; contained by a template delimiter. Most likely, we are in a
8540 ;; template arglist within a statement. This case is based on CASE
8541 ;; 7. At some point in the future, we may wish to create more
8542 ;; syntactic symbols such as `template-intro',
8543 ;; `template-cont-nonempty', etc., and distinguish between them as we
8544 ;; do for `arglist-intro' etc. (2009-12-07).
8550 ; '<') before indent-point.
8558 paren-state))
8561 ;; CASE 7: line is an expression, not a statement. Most
8562 ;; likely we are either in a function prototype or a function
8563 ;; call argument list, or a template argument list.
8572 ;; CASE 7A: we are looking at the arglist closing paren.
8573 ;; C.f. case 7F.
8585 paren-state))
8587 ;; CASE 7B: Looking at the opening brace of an
8588 ;; in-expression block or brace list. C.f. cases 4, 16A
8589 ;; and 17E.
8593 paren-state))
8597 placeholder
8600 paren-state)
8601 containing-sexp)))
8602 ;; placeholder is nil if it's a block directly in
8603 ;; a function arglist. That makes us skip out of
8604 ;; this case.
8605 )))
8610 paren-state)
8614 ;; CASE 7C: we are looking at the first argument in an empty
8615 ;; argument list. Use arglist-close if we're actually
8616 ;; looking at a close paren or bracket.
8628 paren-state))
8630 ;; CASE 7D: we are inside a conditional test clause. treat
8631 ;; these things as statements
8641 ))
8643 ;; CASE 7E: maybe a continued ObjC method call. This is the
8644 ;; case when we are inside a [] bracketed exp, and what
8645 ;; precede the opening bracket is not an identifier.
8653 )))
8655 ;; CASE 7F: we are looking at an arglist continuation line,
8656 ;; but the preceding argument is on the same line as the
8657 ;; opening paren. This case includes multi-line
8658 ;; mathematical paren groupings, but we could be on a
8659 ;; for-list continuation line. C.f. case 7A.
8676 paren-state))
8678 ;; CASE 7G: we are looking at just a normal arglist
8679 ;; continuation line
8682 ))
8684 ;; CASE 8: func-local multi-inheritance line
8694 ;; CASE 8A: non-hanging colon on an inher intro
8699 ;; CASE 8B: hanging colon on an inher intro
8703 ;; CASE 8C: a continued inheritance line
8707 )))
8709 ;; CASE 9: we are inside a brace-list
8719 ;; CASE 9A: In the middle of a special brace list opener.
8739 ;; CASE 9B: brace-list-close brace
8741 ;; Check special brace list closer.
8748 ;; We were between the special close char and the `)'.
8751 ;; We were before the special close char.
8755 ;; Normal brace list check.
8766 ;; Prepare for the rest of the cases below by going to the
8767 ;; token following the opening brace
8780 ;; CASE 9C: we're looking at the first line in a brace-list
8791 ;; CASE 9D: this is just a later brace-list-entry or
8792 ;; brace-entry-open
8801 ))
8802 ))))
8804 ;; CASE 10: A continued statement or top level construct.
8814 ;; This is shared with case 18.
8816 char-after-ip
8817 placeholder
8818 containing-sexp
8819 paren-state))
8821 ;; CASE 16: block close brace, possibly closing the defun or
8822 ;; the class
8824 ;; From here on we have the next containing sexp in lim.
8829 ;; CASE 16E: Closing a statement block? This catches
8830 ;; cases where it's preceded by a statement keyword,
8831 ;; which works even when used in an "invalid" context,
8832 ;; e.g. a macro argument.
8837 ;; CASE 16A: closing a lambda defun or an in-expression
8838 ;; block? C.f. cases 4, 7B and 17E.
8841 nil))
8843 'inline-close
8853 paren-state)
8857 ;; CASE 16B: does this close an inline or a function in
8858 ;; a non-class declaration level block?
8865 nil))
8875 ;; CASE 16F: Can be a defun-close of a function declared
8876 ;; in a statement block, e.g. in Pike or when using gcc
8877 ;; extensions, but watch out for macros followed by
8878 ;; blocks. Let it through to be handled below.
8879 ;; C.f. cases B.3 and 17G.
8885 ;; Turn off recognition of constructs that
8886 ;; lacks a type in this case, since that's more
8887 ;; likely to be a macro followed by a block.
8894 ;; CASE 16C: If there is an enclosing brace then this is
8895 ;; a block close since defun closes inside declaration
8896 ;; level blocks have been handled above.
8898 ;; If the block is preceded by a case/switch label on
8899 ;; the same line, we anchor at the first preceding label
8900 ;; at boi. The default handling in c-add-stmt-syntax
8901 ;; really fixes it better, but we do like this to keep
8902 ;; the indentation compatible with version 5.28 and
8903 ;; earlier. C.f. case 17H.
8910 ;; c-backward-to-block-anchor not necessary here; those
8911 ;; situations are handled in case 16E above.
8914 ;; CASE 16D: Only top level defun close left.
8920 paren-state))
8921 ))
8923 ;; CASE 17: Statement or defun catchall.
8926 ;; Back up statements until we find one that starts at boi.
8929 containing-sexp)))
8933 nil)
8938 ;; CASE 17B: continued statement
8942 containing-sexp paren-state))
8944 ;; CASE 17A: After a case/default label?
8952 'statement-case-open
8954 nil t containing-sexp paren-state))
8956 ;; CASE 17D: any old statement
8963 containing-sexp paren-state)
8967 ;; CASE 17I: Inside a substatement block.
8969 ;; The following tests are all based on containing-sexp.
8971 ;; From here on we have the next containing sexp in lim.
8976 lim paren-state)
8980 ;; CASE 17E: first statement in an in-expression block.
8981 ;; C.f. cases 4, 7B and 16A.
8984 nil))
8986 'defun-block-intro
8995 paren-state)
9001 ;; CASE 17F: first statement in an inline, or first
9002 ;; statement in a top-level defun. we can tell this is it
9003 ;; if there are no enclosing braces that haven't been
9004 ;; narrowed out by a class (i.e. don't use bod here).
9007 paren-state)))
9013 nil))))
9018 ;; CASE 17G: First statement in a function declared inside
9019 ;; a normal block. This can occur in Pike and with
9020 ;; e.g. the gcc extensions, but watch out for macros
9021 ;; followed by blocks. C.f. cases B.3 and 16F.
9027 ;; Turn off recognition of constructs that lacks
9028 ;; a type in this case, since that's more likely
9029 ;; to be a macro followed by a block.
9035 lim paren-state))
9037 ;; CASE 17H: First statement in a block.
9039 ;; If the block is preceded by a case/switch label on the
9040 ;; same line, we anchor at the first preceding label at
9041 ;; boi. The default handling in c-add-stmt-syntax is
9042 ;; really fixes it better, but we do like this to keep the
9043 ;; indentation compatible with version 5.28 and earlier.
9044 ;; C.f. case 16C.
9051 ;; c-backward-to-block-anchor not necessary here; those
9052 ;; situations are handled in case 17I above.
9054 lim paren-state))
9057 ))
9058 )
9060 ;; now we need to look at any modifiers
9064 ;; are we looking at a comment only line?
9069 ;; we might want to give additional offset to friends (in C++).
9074 ;; Set syntactic-relpos.
9080 nil)
9081 t))
9084 ;; Start of or a continuation of a preprocessor directive?
9099 ;; If inside a cpp expression, i.e. anywhere in a
9100 ;; cpp directive except a #define body, we only let
9101 ;; through the syntactic analysis that is internal
9102 ;; in the expression. That means the arglist
9103 ;; elements, if they are anchored inside the cpp
9104 ;; expression.
9113 ;; Inside a #define body and the syntactic analysis is
9114 ;; anchored on the start of the #define. In this case
9115 ;; we add cpp-define-intro to get the extra
9116 ;; indentation of the #define body.
9119 ;; return the syntax
9120 c-syntactic-context)))
9122 \f
9123 ;; Indentation calculation.
9126 ;; offset can be a number, a function, a variable, a list, or one of
9127 ;; the symbols + or -
9128 ;;
9129 ;; This function might do hidden buffer changes.
9148 langelem symbol))
9154 offset symbol)
9155 nil)
9169 Error evaluating offset %S for %s: \
9170 Cannot combine absolute offset %S with relative %S in `%s' method"
9176 Error evaluating offset %S for %s: \
9177 Cannot combine relative offset %S with absolute %S in `%s' method"
9182 res))
9200 Error evaluating offset %S for %s: \
9201 Cannot combine absolute offsets %S and %S in `add' method"
9205 res))
9214 res))))
9221 nil))))
9225 res
9227 offset symbol res)
9228 nil)))
9231 ;; Get offset from LANGELEM which is a list beginning with the
9232 ;; syntactic symbol and followed by any analysis data it provides.
9233 ;; That data may be zero or more elements, but if at least one is
9234 ;; given then the first is the anchor position (or nil). The symbol
9235 ;; is matched against `c-offsets-alist' and the offset calculated
9236 ;; from that is returned.
9237 ;;
9238 ;; This function might do hidden buffer changes.
9248 offset
9252 ))
9255 ;; This is a compatibility wrapper for `c-calc-offset' in case
9256 ;; someone is calling it directly. It takes an old style syntactic
9257 ;; element on the form (SYMBOL . ANCHOR-POS) and converts it to the
9258 ;; new list form.
9259 ;;
9260 ;; This function might do hidden buffer changes.
9267 ;; Calculate the syntactic indentation from a syntactic description
9268 ;; as returned by `c-guess-syntax'.
9269 ;;
9270 ;; Note that topmost-intro always has an anchor position at bol, for
9271 ;; historical reasons. It's often used together with other symbols
9272 ;; that has more sane positions. Since we always use the first
9273 ;; found anchor position, we rely on that these other symbols always
9274 ;; precede topmost-intro in the LANGELEMS list.
9275 ;;
9276 ;; This function might do hidden buffer changes.
9284 ;; Got an absolute column that overrides any indentation
9285 ;; we've collected so far, but not the relative
9286 ;; indentation we might get for the nested structures
9287 ;; further down the langelems list.
9291 ;; Got a relative change of the current calculated
9292 ;; indentation.
9295 ;; Use the anchor position from the first syntactic
9296 ;; element with one.
9306 indent)))
9308 \f
9311 ;; arch-tag: 149add18-4673-4da5-ac47-6805e4eae089
9312 ;;; cc-engine.el ends here