Treat interpreter-mode-alist as alist of regexps, not literals

Cf http://lists.gnu.org/archive/html/emacs-devel/2005-08/msg00472.html

* lisp/files.el (interpreter-mode-alist): Convert to regexps.
(set-auto-mode): Adapt for this. 

* lisp/progmodes/cperl-mode.el (cperl-clobber-mode-lists):
Comment out unused variable.

* lisp/progmodes/cc-mode.el (interpreter-mode-alist):
* lisp/progmodes/python.el (interpreter-mode-alist):
* lisp/progmodes/ruby-mode.el (interpreter-mode-alist): Convert to regexps.

* lisp/progmodes/sh-script.el (sh-set-shell):
No longer use interpreter-mode-alist to get list of shells.

* etc/NEWS: Mention this.

Fixes: debbugs:15306

8 files changed, 99 insertions, 75 deletions

diff --git a/etc/NEWS b/etc/NEWS
index 9914da31859..78f99dbc621 100644
--- a/etc/NEWS
+++ b/etc/NEWS
@@ -614,6 +614,14 @@ file using `set-file-extended-attributes'.
 ** `visited-file-modtime' now returns -1 for nonexistent files.
 Formerly it returned a list (-1 LOW USEC PSEC), but this was ambiguous
 in the presence of files with negative time stamps.
+
+** The cars of the elements in `interpreter-mode-alist' are now treated
+as regexps rather than literal strings.  For the time being, any
+element whose car does not start with "\\" is still treated as a
+literal string, so this change should not cause any incompatibilities
+if you have code that just adds elements to the list, only if you are
+actually using interpreter-mode-alist for something.
+
 
 * Lisp Changes in Emacs 24.4
 
diff --git a/lisp/ChangeLog b/lisp/ChangeLog
index 9282aa16a2d..32ea5e72efd 100644
--- a/lisp/ChangeLog
+++ b/lisp/ChangeLog
@@ -1,5 +1,15 @@
 2013-09-10  Glenn Morris  <rgm@gnu.org>
 
+	* files.el (interpreter-mode-alist): Convert to regexps.
+	(set-auto-mode): Adapt for this.  (Bug#15306)
+	* progmodes/cperl-mode.el (cperl-clobber-mode-lists):
+	Comment out unused variable.
+	* progmodes/cc-mode.el (interpreter-mode-alist):
+	* progmodes/python.el (interpreter-mode-alist):
+	* progmodes/ruby-mode.el (interpreter-mode-alist): Convert to regexps.
+	* progmodes/sh-script.el (sh-set-shell):
+	No longer use interpreter-mode-alist to get list of shells.
+
 	* progmodes/cc-mode.el (awk-mode): Remove duplicate autoload.
 
 2013-09-10  Stefan Monnier  <monnier@iro.umontreal.ca>
diff --git a/lisp/files.el b/lisp/files.el
index f9ff3c936bd..2315448c549 100644
--- a/lisp/files.el
+++ b/lisp/files.el
@@ -2447,48 +2447,43 @@ and `magic-mode-alist', which determines modes based on file contents.")
   (mapcar
    (lambda (l)
      (cons (purecopy (car l)) (cdr l)))
-   '(("perl" . perl-mode)
-     ("perl5" . perl-mode)
-     ("miniperl" . perl-mode)
-     ("wish" . tcl-mode)
-     ("wishx" . tcl-mode)
-     ("tcl" . tcl-mode)
-     ("tclsh" . tcl-mode)
-     ("expect" . tcl-mode)
-     ("scm" . scheme-mode)
-     ("ash" . sh-mode)
-     ("bash" . sh-mode)
-     ("bash2" . sh-mode)
-     ("csh" . sh-mode)
-     ("dtksh" . sh-mode)
-     ("es" . sh-mode)
-     ("itcsh" . sh-mode)
-     ("jsh" . sh-mode)
-     ("ksh" . sh-mode)
-     ("oash" . sh-mode)
-     ("pdksh" . sh-mode)
-     ("rbash" . sh-mode)
-     ("rc" . sh-mode)
-     ("rpm" . sh-mode)
-     ("sh" . sh-mode)
-     ("sh5" . sh-mode)
-     ("tcsh" . sh-mode)
-     ("wksh" . sh-mode)
-     ("wsh" . sh-mode)
-     ("zsh" . sh-mode)
-     ("tail" . text-mode)
-     ("more" . text-mode)
-     ("less" . text-mode)
-     ("pg" . text-mode)
-     ("make" . makefile-gmake-mode)		; Debian uses this
-     ("guile" . scheme-mode)
-     ("clisp" . lisp-mode)
-     ("emacs" . emacs-lisp-mode)))
+   '(("\\`\\(mini\\)?perl5?\\'" . perl-mode)
+     ("\\`wishx?\\'" . tcl-mode)
+     ("\\`tcl\\(sh\\)?\\'" . tcl-mode)
+     ("\\`expect\\'" . tcl-mode)
+     ("\\`scm\\'" . scheme-mode)
+     ("\\`[acjkwz]sh\\'" . sh-mode)
+     ("\\`bash2?\\'" . sh-mode)
+     ("\\`dtksh\\'" . sh-mode)
+     ("\\`es\\'" . sh-mode)
+     ("\\`itcsh\\'" . sh-mode)
+     ("\\`oash\\'" . sh-mode)
+     ("\\`pdksh\\'" . sh-mode)
+     ("\\`rbash\\'" . sh-mode)
+     ("\\`rc\\'" . sh-mode)
+     ("\\`rpm\\'" . sh-mode)
+     ("\\`sh5?\\'" . sh-mode)
+     ("\\`tcsh\\'" . sh-mode)
+     ("\\`wksh\\'" . sh-mode)
+     ("\\`tail\\'" . text-mode)
+     ("\\`more\\'" . text-mode)
+     ("\\`less\\'" . text-mode)
+     ("\\`pg\\'" . text-mode)
+     ("\\`make\\'" . makefile-gmake-mode)		; Debian uses this
+     ("\\`guile\\'" . scheme-mode)
+     ("\\`clisp\\'" . lisp-mode)
+     ("\\`emacs\\'" . emacs-lisp-mode)))
   "Alist mapping interpreter names to major modes.
 This is used for files whose first lines match `auto-mode-interpreter-regexp'.
-Each element looks like (INTERPRETER . MODE).
-If INTERPRETER matches the name of the interpreter specified in the first line
-of a script, mode MODE is enabled.
+Each element looks like (REGEXP . MODE).
+If REGEXP matches the name (minus any directory part) of the interpreter
+specified in the first line of a script, enable major mode MODE.
+
+Emacs versions earlier than 24.4 treat the car of each element as a
+literal string that must match the entire name, rather than a regexp.
+For backwards compatibility, any REGEXP that does not begin with \"\\\\\"
+continues to be treated in this way.  This behavior may be removed in
+future and should not be relied upon.
 
 See also `auto-mode-alist'.")
 
@@ -2683,19 +2678,27 @@ we don't actually set it to the same mode the buffer already has."
     ;; If we didn't, look for an interpreter specified in the first line.
     ;; As a special case, allow for things like "#!/bin/env perl", which
     ;; finds the interpreter anywhere in $PATH.
-    (unless done
-      (setq mode (save-excursion
-		   (goto-char (point-min))
-		   (if (looking-at auto-mode-interpreter-regexp)
-		       (match-string 2)
-		     ""))
-	    ;; Map interpreter name to a mode, signaling we're done at the
-	    ;; same time.
-	    done (assoc (file-name-nondirectory mode)
-			interpreter-mode-alist))
-      ;; If we found an interpreter mode to use, invoke it now.
-      (if done
-	  (set-auto-mode-0 (cdr done) keep-mode-if-same)))
+    (and (not done)
+	 (setq mode (save-excursion
+		      (goto-char (point-min))
+		      (if (looking-at auto-mode-interpreter-regexp)
+			  (match-string 2))))
+	 ;; Map interpreter name to a mode, signaling we're done at the
+	 ;; same time.
+	 (setq done (assoc-default
+		     (file-name-nondirectory mode)
+		     ;; Backwards compat: if car of i-m-alist does not start
+		     ;; with "\\", treat as literal string.
+		     (mapcar (lambda (e)
+			       (if (string-match-p "\\`\\\\" (car e))
+				   e
+				 (cons
+				  (format "\\`%s\\'" (regexp-quote (car e)))
+				  (cdr e))))
+			     interpreter-mode-alist)
+		     #'string-match-p))
+	 ;; If we found an interpreter mode to use, invoke it now.
+	 (set-auto-mode-0 done keep-mode-if-same))
     ;; Next try matching the buffer beginning against magic-mode-alist.
     (unless done
       (if (setq done (save-excursion
diff --git a/lisp/progmodes/cc-mode.el b/lisp/progmodes/cc-mode.el
index e977a415d62..bffc5b95d25 100644
--- a/lisp/progmodes/cc-mode.el
+++ b/lisp/progmodes/cc-mode.el
@@ -1555,7 +1555,7 @@ Key bindings:
 		  (cons "Pike" (c-lang-const c-mode-menu pike)))
 
 ;;;###autoload (add-to-list 'auto-mode-alist '("\\.\\(u?lpc\\|pike\\|pmod\\(\\.in\\)?\\)\\'" . pike-mode))
-;;;###autoload (add-to-list 'interpreter-mode-alist '("pike" . pike-mode))
+;;;###autoload (add-to-list 'interpreter-mode-alist '("\\`pike\\'" . pike-mode))
 
 ;;;###autoload
 (define-derived-mode pike-mode prog-mode "Pike"
@@ -1589,10 +1589,7 @@ Key bindings:
 ;; Support for AWK
 
 ;;;###autoload (add-to-list 'auto-mode-alist '("\\.awk\\'" . awk-mode))
-;;;###autoload (add-to-list 'interpreter-mode-alist '("awk" . awk-mode))
-;;;###autoload (add-to-list 'interpreter-mode-alist '("mawk" . awk-mode))
-;;;###autoload (add-to-list 'interpreter-mode-alist '("nawk" . awk-mode))
-;;;###autoload (add-to-list 'interpreter-mode-alist '("gawk" . awk-mode))
+;;;###autoload (add-to-list 'interpreter-mode-alist '("\\`[gmn]?awk\\'" . awk-mode))
 
 (c-define-abbrev-table 'awk-mode-abbrev-table
   '(("else" "else" c-electric-continued-statement 0)
diff --git a/lisp/progmodes/cperl-mode.el b/lisp/progmodes/cperl-mode.el
index 770e78bb3b1..1eea6972707 100644
--- a/lisp/progmodes/cperl-mode.el
+++ b/lisp/progmodes/cperl-mode.el
@@ -412,15 +412,15 @@ Affects: `cperl-font-lock', `cperl-electric-lbrace-space',
  "use cperl-vc-rcs-header or cperl-vc-sccs-header instead."
  "22.1")
 
-(defcustom cperl-clobber-mode-lists
-  (not
-   (and
-    (boundp 'interpreter-mode-alist)
-    (assoc "miniperl" interpreter-mode-alist)
-    (assoc "\\.\\([pP][Llm]\\|al\\)$" auto-mode-alist)))
-  "*Whether to install us into `interpreter-' and `extension' mode lists."
-  :type 'boolean
-  :group 'cperl)
+;;; (defcustom cperl-clobber-mode-lists
+;;;   (not
+;;;    (and
+;;;     (boundp 'interpreter-mode-alist)
+;;;     (assoc "miniperl" interpreter-mode-alist)
+;;;     (assoc "\\.\\([pP][Llm]\\|al\\)$" auto-mode-alist)))
+;;;   "*Whether to install us into `interpreter-' and `extension' mode lists."
+;;;   :type 'boolean
+;;;   :group 'cperl)
 
 (defcustom cperl-info-on-command-no-prompt nil
   "*Not-nil (and non-null) means not to prompt on C-h f.
diff --git a/lisp/progmodes/python.el b/lisp/progmodes/python.el
index 05d707acf43..bba4453cce8 100644
--- a/lisp/progmodes/python.el
+++ b/lisp/progmodes/python.el
@@ -225,7 +225,7 @@
 ;;;###autoload
 (add-to-list 'auto-mode-alist (cons (purecopy "\\.py\\'")  'python-mode))
 ;;;###autoload
-(add-to-list 'interpreter-mode-alist (cons (purecopy "python") 'python-mode))
+(add-to-list 'interpreter-mode-alist (cons (purecopy "\\`python[23.]*\\'") 'python-mode))
 
 (defgroup python nil
   "Python Language's flying circus support for Emacs."
diff --git a/lisp/progmodes/ruby-mode.el b/lisp/progmodes/ruby-mode.el
index 0f868255589..9dc2c4fb6c9 100644
--- a/lisp/progmodes/ruby-mode.el
+++ b/lisp/progmodes/ruby-mode.el
@@ -1968,8 +1968,7 @@ The variable `ruby-indent-level' controls the amount of indentation.
                                      "\\)\\'")) 'ruby-mode))
 
 ;;;###autoload
-(dolist (name (list "ruby" "rbx" "jruby" "ruby1.9" "ruby1.8"))
-  (add-to-list 'interpreter-mode-alist (cons (purecopy name) 'ruby-mode)))
+(add-to-list 'interpreter-mode-alist (cons (purecopy "\\`\\(rbx\\|j?ruby\\(1\\.[89]\\)?\\)\\'") 'ruby-mode))
 
 (provide 'ruby-mode)
 
diff --git a/lisp/progmodes/sh-script.el b/lisp/progmodes/sh-script.el
index 292bc2369a6..3ea2afb6fc3 100644
--- a/lisp/progmodes/sh-script.el
+++ b/lisp/progmodes/sh-script.el
@@ -2170,11 +2170,18 @@ the visited file executable, and NO-QUERY-FLAG (the second argument)
 controls whether to query about making the visited file executable.
 
 Calls the value of `sh-set-shell-hook' if set."
-  (interactive (list (completing-read (format "Shell \(default %s\): "
- 					      sh-shell-file)
-  				      interpreter-mode-alist
- 				      (lambda (x) (eq (cdr x) 'sh-mode))
- 				      nil nil nil sh-shell-file)
+  (interactive (list (completing-read
+                      (format "Shell \(default %s\): "
+                              sh-shell-file)
+                      ;; This used to use interpreter-mode-alist, but that is
+                      ;; no longer appropriate now that uses regexps.
+                      ;; Maybe there could be a separate variable that lists
+                      ;; the shells, used here and to construct i-mode-alist.
+                      ;; But the following is probably good enough:
+                      (append (mapcar (lambda (e) (symbol-name (car e)))
+                                      sh-ancestor-alist)
+                              '("csh" "rc" "sh"))
+                      nil nil nil nil sh-shell-file)
 		     (eq executable-query 'function)
 		     t))
   (if (string-match "\\.exe\\'" shell)