master 9b44824620: CC Mode: Add accurate handling for backslash in C line and block comments

[Alan Mackenzie]

branch: master
commit 9b44824620fd500b9e7094bd1a8ca23608cb2e5b
Author: Alan Mackenzie <acm@muc.de>
Commit: Alan Mackenzie <acm@muc.de>

    CC Mode: Add accurate handling for backslash in C line and block comments
    
    This is needed to handle the idiosyncratic meaning of backslash in comments 
in
    the C and C++ standards.
    
    * lisp/progmodes/cc-engine.el: Correct a spelling error.
    
    * lisp/progmodes/cc-mode.el (c-before-change-fix-comment-escapes)
    (c-after-change-fix-comment-escapes): New functions.
    
    * lisp/progmodes/cc-langs.el (c-get-state-before-change-functions): Add
    c-before-change-fix-comment-escapes to the C/Objc and C++ values.
    (c-before-font-lock-functions): Add c-after-change-fix-comment-escapes to 
the
    C/Objc and C++ values.
---
 lisp/progmodes/cc-engine.el |  2 +-
 lisp/progmodes/cc-langs.el  |  8 +++--
 lisp/progmodes/cc-mode.el   | 82 ++++++++++++++++++++++++++++++++++++++++++++-
 3 files changed, 88 insertions(+), 4 deletions(-)

diff --git a/lisp/progmodes/cc-engine.el b/lisp/progmodes/cc-engine.el
index f9c33f5149..cfbb668bae 100644
--- a/lisp/progmodes/cc-engine.el
+++ b/lisp/progmodes/cc-engine.el
@@ -6837,7 +6837,7 @@ comment at the start of cc-engine.el for more info."
   (let ((type (c-syntactic-content from to c-recognize-<>-arglists)))
     (unless (gethash type c-found-types)
       (puthash type t c-found-types)
-      (when (and (not c-record-found-types) ; Only call 
`c-fontify-new-fount-type'
+      (when (and (not c-record-found-types) ; Only call 
`c-fontify-new-found-type'
                                        ; when we haven't "bound" c-found-types
                                        ; to itself in c-forward-<>-arglist.
                 (eq (string-match c-symbol-key type) 0)
diff --git a/lisp/progmodes/cc-langs.el b/lisp/progmodes/cc-langs.el
index 68070cd058..c5964165c8 100644
--- a/lisp/progmodes/cc-langs.el
+++ b/lisp/progmodes/cc-langs.el
@@ -458,12 +458,14 @@ so that all identifiers are recognized as words.")
        c-before-change-check-<>-operators
        c-truncate-bs-cache
        c-before-change-check-unbalanced-strings
-       c-parse-quotes-before-change)
+       c-parse-quotes-before-change
+       c-before-change-fix-comment-escapes)
   (c objc) '(c-extend-region-for-CPP
             c-depropertize-CPP
             c-truncate-bs-cache
             c-before-change-check-unbalanced-strings
-            c-parse-quotes-before-change)
+            c-parse-quotes-before-change
+            c-before-change-fix-comment-escapes)
   java '(c-parse-quotes-before-change
         c-before-change-check-unbalanced-strings
         c-before-change-check-<>-operators)
@@ -500,6 +502,7 @@ parameters \(point-min) and \(point-max).")
       c-after-change-mark-abnormal-strings
       c-change-expand-fl-region)
   (c objc) '(c-depropertize-new-text
+            c-after-change-fix-comment-escapes
             c-after-change-escape-NL-in-string
             c-parse-quotes-after-change
             c-after-change-mark-abnormal-strings
@@ -507,6 +510,7 @@ parameters \(point-min) and \(point-max).")
             c-neutralize-syntax-in-CPP
             c-change-expand-fl-region)
   c++ '(c-depropertize-new-text
+       c-after-change-fix-comment-escapes
        c-after-change-escape-NL-in-string
        c-after-change-unmark-ml-strings
        c-parse-quotes-after-change
diff --git a/lisp/progmodes/cc-mode.el b/lisp/progmodes/cc-mode.el
index ae96cdbd2f..70fc1cb73a 100644
--- a/lisp/progmodes/cc-mode.el
+++ b/lisp/progmodes/cc-mode.el
@@ -1979,6 +1979,87 @@ Note that this is a strict tail, so won't match, e.g. 
\"0x....\".")
 (defvar c-new-id-is-type nil)
 (make-variable-buffer-local 'c-new-id-is-type)
 
+(defun c-before-change-fix-comment-escapes (beg end)
+  "Remove punctuation syntax-table text properties from C/C++ comment markers.
+This is to handle the rare case of two or more backslashes at an
+end of line in a // comment or the equally rare case of a
+backslash preceding the terminator of a /* comment, as \\*/.
+
+This function is used solely as a member of
+`c-get-state-before-change-functions', where it should appear
+late in that variable, and it must be used only together with
+`c-after-change-fix-comment-escapes'.
+
+Note that the function currently only handles comments beginning
+with // and /*, not more generic line and block comments."
+  (c-save-buffer-state (end-state)
+    (setq end-state (c-full-pp-to-literal end))
+    (when (memq (cadr end-state) '(c c++))
+      (goto-char (max (- beg 2) (point-min)))
+      (if (eq (cadr end-state) 'c)
+         (when (search-forward "\\*/"
+                               (or (cdr (caddr end-state)) (point-max)) t)
+           (c-clear-char-property (match-beginning 0) 'syntax-table)
+           (c-truncate-lit-pos-cache (match-beginning 0)))
+       (while (search-forward "\\\\\n"
+                              (or (cdr (caddr end-state)) (point-max)) t)
+         (c-clear-char-property (match-beginning 0) 'syntax-table)
+         (c-truncate-lit-pos-cache (match-beginning 0)))))))
+
+(defun c-after-change-fix-comment-escapes (beg end _old-len)
+  "Apply punctuation syntax-table text properties to C/C++ comment markers.
+This is to handle the rare case of two or more backslashes at an
+end of line in a // comment or the equally rare case of a
+backslash preceding the terminator of a /* comment, as \\*/.
+
+This function is used solely as a member of
+`c-before-font-lock-functions', where it should appear early in
+that variable, and it must be used only together with
+`c-before-change-fix-comment-escapes'.
+
+Note that the function currently only handles comments beginning
+with // and /*, not more generic line and block comments."
+  (c-save-buffer-state (state)
+    ;; We cannot use `c-full-pp-to-literal' in this function, since the
+    ;; `syntax-table' text properties after point are not yet in a consistent
+    ;; state.
+    (setq state (c-semi-pp-to-literal beg))
+    (goto-char (if (memq (cadr state) '(c c++))
+                  (caddr state)
+                (max (- beg 2) (point-min))))
+    (while
+       (re-search-forward "\\\\\\(\\(\\\\\n\\)\\|\\(\\*/\\)\\)"
+                          (min (+ end 2) (point-max)) t)
+      (setq state (c-semi-pp-to-literal (match-beginning 0)))
+      (when (cond
+            ((eq (cadr state) 'c)
+             (match-beginning 3))
+            ((eq (cadr state) 'c++)
+             (match-beginning 2)))
+       (c-put-char-property (match-beginning 0) 'syntax-table '(1))
+       (c-truncate-lit-pos-cache (match-beginning 0))))
+
+    (goto-char end)
+    (setq state (c-semi-pp-to-literal (point)))
+    (cond
+     ((eq (cadr state) 'c)
+      (when (search-forward "*/" nil t)
+       (when (eq (char-before (match-beginning 0)) ?\\)
+         (c-put-char-property (1- (match-beginning 0)) 'syntax-table '(1))
+         (c-truncate-lit-pos-cache (1- (match-beginning 0))))))
+     ((eq (cadr state) 'c++)
+      (while
+         (progn
+           (end-of-line)
+           (and (eq (char-before) ?\\)
+                (progn
+                  (when (eq (char-before (1- (point))) ?\\)
+                    (c-put-char-property (- (point) 2) 'syntax-table '(1))
+                    (c-truncate-lit-pos-cache (1- (point))))
+                  t)
+                (not (eobp))))
+       (forward-char))))))
+
 (defun c-update-new-id (end)
   ;; Note the bounds of any identifier that END is in or just after, in
   ;; `c-new-id-start' and `c-new-id-end'.  Otherwise set these variables to
@@ -1990,7 +2071,6 @@ Note that this is a strict tail, so won't match, e.g. 
\"0x....\".")
            c-new-id-end (and id-beg
                              (progn (c-end-of-current-token) (point)))))))
 
-
 (defun c-post-command ()
   ;; If point was inside of a new identifier and no longer is, record that
   ;; fact.