bug#18378: wrong fontification of Java code

bug#18378: wrong fontification of Java code

[Paul Pogonyshev]

[-- Attachment #1: Type: text/plain, Size: 299 bytes --]

class X
{
    public void x ()
    {
        for (int this_is_not_a_function_name : getX ()) {
        }
    }
}

In current trunk (compiled five minutes ago) 'this_is_not_a_function_name'
above is fontified as a function name. This seems to depend on whether I
have '()' after 'getX' or not.

Paul

[-- Attachment #2: Type: text/html, Size: 429 bytes --]

bug#18378: [Patch] wrong fontification of Java code

[Jordon Biondo]

I have a potential solution for this bug, as it has impacted me as well.

Not sure what the best solution would be but I have included a patch that does solve the problem. From what I can tell there are no tests for java fortification but I have not seen any issues arising in my own java projects. The issue was that in java, a “:” can mark the end of a declaration in a foreach, not just a “;” or “,”.


=== modified file 'lisp/ChangeLog'
*** lisp/ChangeLog	2014-09-01 15:03:45 +0000
--- lisp/ChangeLog	2014-09-01 19:11:46 +0000
***************
*** 1,3 ****
--- 1,8 ----
+ 2014-09-01  Jordon Biondo  <jordonbiondo@gmail.com>
+ 
+ 	* progmodes/cc-fonts.el (c-font-lock-declarators): Special
+ 	case for declarations inside Java foreach loops (Bug#18378)
+ 
  2014-09-01  Eli Zaretskii  <eliz@gnu.org>
  
  	* ls-lisp.el (ls-lisp-use-string-collate)

=== modified file 'lisp/progmodes/cc-fonts.el'
*** lisp/progmodes/cc-fonts.el	2014-08-24 20:50:11 +0000
--- lisp/progmodes/cc-fonts.el	2014-09-01 19:11:59 +0000
*************** casts and declarations are fontified.  U
*** 1116,1123 ****
  	    ;; initializing brace lists.
  	    (let (found)
  	      (while
! 		  (and (setq found (c-syntactic-re-search-forward
! 			     "[;,]\\|\\s)\\|\\'\\|\\(=\\|\\s(\\)" limit t t))
  		       (eq (char-before) ?\[)
  		       (c-go-up-list-forward))
  		     (setq brackets-after-id t))
--- 1116,1128 ----
  	    ;; initializing brace lists.
  	    (let (found)
  	      (while
!                   (and (setq found (c-syntactic-re-search-forward
!                                     (concat "[;,]\\|\\s)\\|\\'\\|\\(=\\|"
!                                             (if (c-major-mode-is 'java-mode)
!                                                 ":\\|"
!                                               "")
!                                             "\\s(\\)")
!                                     limit t t))
  		       (eq (char-before) ?\[)
  		       (c-go-up-list-forward))
  		     (setq brackets-after-id t))

bug#18378: [Patch] wrong fontification of Java code

[Stefan Kangas]

Hi Alan,

Could you please help review the below patch for cc-fonts.el?

Thanks in advance.

Best regards,
Stefan Kangas

Paul Pogonyshev <pogonyshev@gmail.com> writes:

> class X
> {
>     public void x ()
>     {
>         for (int this_is_not_a_function_name : getX ()) {
>         }
>     }
> }
>
> In current trunk (compiled five minutes ago)
> 'this_is_not_a_function_name' above is fontified as a function
> name. This seems to depend on whether I have '()' after 'getX' or
> not.
>
> Paul

Jordon Biondo <jordonbiondo@gmail.com> writes:

> I have a potential solution for this bug, as it has impacted me as well.
>
> Not sure what the best solution would be but I have included a patch that does
> solve the problem. From what I can tell there are no tests for java
> fortification but I have not seen any issues arising in my own java
> projects. The issue was that in java, a “:” can mark the end of a declaration in
> a foreach, not just a “;” or “,”.
>
> === modified file 'lisp/ChangeLog'
> *** lisp/ChangeLog	2014-09-01 15:03:45 +0000
> --- lisp/ChangeLog	2014-09-01 19:11:46 +0000
> ***************
> *** 1,3 ****
> --- 1,8 ----
> + 2014-09-01  Jordon Biondo  <jordonbiondo@gmail.com>
> + 
> + 	* progmodes/cc-fonts.el (c-font-lock-declarators): Special
> + 	case for declarations inside Java foreach loops (Bug#18378)
> + 
>   2014-09-01  Eli Zaretskii  <eliz@gnu.org>
>   
>   	* ls-lisp.el (ls-lisp-use-string-collate)
>
> === modified file 'lisp/progmodes/cc-fonts.el'
> *** lisp/progmodes/cc-fonts.el	2014-08-24 20:50:11 +0000
> --- lisp/progmodes/cc-fonts.el	2014-09-01 19:11:59 +0000
> *************** casts and declarations are fontified.  U
> *** 1116,1123 ****
>   	    ;; initializing brace lists.
>   	    (let (found)
>   	      (while
> ! 		  (and (setq found (c-syntactic-re-search-forward
> ! 			     "[;,]\\|\\s)\\|\\'\\|\\(=\\|\\s(\\)" limit t t))
>   		       (eq (char-before) ?\[)
>   		       (c-go-up-list-forward))
>   		     (setq brackets-after-id t))
> --- 1116,1128 ----
>   	    ;; initializing brace lists.
>   	    (let (found)
>   	      (while
> !                   (and (setq found (c-syntactic-re-search-forward
> !                                     (concat "[;,]\\|\\s)\\|\\'\\|\\(=\\|"
> !                                             (if (c-major-mode-is 'java-mode)
> !                                                 ":\\|"
> !                                               "")
> !                                             "\\s(\\)")
> !                                     limit t t))
>   		       (eq (char-before) ?\[)
>   		       (c-go-up-list-forward))
>   		     (setq brackets-after-id t))

bug#18378: [Patch] wrong fontification of Java code

[Alan Mackenzie]

Hello, Stefan.

On Mon, Jan 20, 2020 at 22:16:46 +0100, Stefan Kangas wrote:
> Hi Alan,

> Could you please help review the below patch for cc-fonts.el?

I missed this bug report back in 2014.  The code has moved around so
much since then that Jordon's patch no longer applies.  But a simple
patch with the same idea appears to fix this bug:



diff -r 3d883b66c245 cc-engine.el
--- a/cc-engine.el	Tue Jan 07 19:24:00 2020 +0000
+++ b/cc-engine.el	Wed Jan 22 20:49:53 2020 +0000
@@ -9025,7 +9025,7 @@
 			   (if (looking-at c-:-op-cont-regexp)
 			       (progn (goto-char (match-end 0)) t)
 			     (not
-			      (and (c-major-mode-is 'c++-mode)
+			      (and (c-major-mode-is '(c++-mode java-mode))
 				   (save-excursion
 				     (and
 				      (c-go-up-list-backward)

This patch should apply cleanly to the emacs-27 branch, and most
probably to master, too.


> Thanks in advance.

> Best regards,
> Stefan Kangas

-- 
Alan Mackenzie (Nuremberg, Germany).


> Paul Pogonyshev <pogonyshev@gmail.com> writes:

> > class X
> > {
> >     public void x ()
> >     {
> >         for (int this_is_not_a_function_name : getX ()) {
> >         }
> >     }
> > }
> >
> > In current trunk (compiled five minutes ago)
> > 'this_is_not_a_function_name' above is fontified as a function
> > name. This seems to depend on whether I have '()' after 'getX' or
> > not.
> >
> > Paul

> Jordon Biondo <jordonbiondo@gmail.com> writes:

> > I have a potential solution for this bug, as it has impacted me as well.
> >
> > Not sure what the best solution would be but I have included a patch that does
> > solve the problem. From what I can tell there are no tests for java
> > fortification but I have not seen any issues arising in my own java
> > projects. The issue was that in java, a “:” can mark the end of a declaration in
> > a foreach, not just a “;” or “,”.
> >
> > === modified file 'lisp/ChangeLog'
> > *** lisp/ChangeLog	2014-09-01 15:03:45 +0000
> > --- lisp/ChangeLog	2014-09-01 19:11:46 +0000
> > ***************
> > *** 1,3 ****
> > --- 1,8 ----
> > + 2014-09-01  Jordon Biondo  <jordonbiondo@gmail.com>
> > + 
> > + 	* progmodes/cc-fonts.el (c-font-lock-declarators): Special
> > + 	case for declarations inside Java foreach loops (Bug#18378)
> > + 
> >   2014-09-01  Eli Zaretskii  <eliz@gnu.org>
> >   
> >   	* ls-lisp.el (ls-lisp-use-string-collate)
> >
> > === modified file 'lisp/progmodes/cc-fonts.el'
> > *** lisp/progmodes/cc-fonts.el	2014-08-24 20:50:11 +0000
> > --- lisp/progmodes/cc-fonts.el	2014-09-01 19:11:59 +0000
> > *************** casts and declarations are fontified.  U
> > *** 1116,1123 ****
> >   	    ;; initializing brace lists.
> >   	    (let (found)
> >   	      (while
> > ! 		  (and (setq found (c-syntactic-re-search-forward
> > ! 			     "[;,]\\|\\s)\\|\\'\\|\\(=\\|\\s(\\)" limit t t))
> >   		       (eq (char-before) ?\[)
> >   		       (c-go-up-list-forward))
> >   		     (setq brackets-after-id t))
> > --- 1116,1128 ----
> >   	    ;; initializing brace lists.
> >   	    (let (found)
> >   	      (while
> > !                   (and (setq found (c-syntactic-re-search-forward
> > !                                     (concat "[;,]\\|\\s)\\|\\'\\|\\(=\\|"
> > !                                             (if (c-major-mode-is 'java-mode)
> > !                                                 ":\\|"
> > !                                               "")
> > !                                             "\\s(\\)")
> > !                                     limit t t))
> >   		       (eq (char-before) ?\[)
> >   		       (c-go-up-list-forward))
> >   		     (setq brackets-after-id t))

bug#18378: [Patch] wrong fontification of Java code

[Stefan Kangas]

Hi Alan,

Alan Mackenzie <acm@muc.de> writes:

> I missed this bug report back in 2014.  The code has moved around so
> much since then that Jordon's patch no longer applies.  But a simple
> patch with the same idea appears to fix this bug:

Thanks for taking a look.

With your fix, applied on the master branch, in the example given:

    class X
    {
        public void x ()
        {
            for (int this_is_not_a_function_name : getX ()) {
            }
        }
    }

The "int" no longer has font-lock-type-face.  Should it?

Best regards,
Stefan Kangas

bug#18378: [Patch] wrong fontification of Java code

[Alan Mackenzie]

Hello, Stefan.

On Thu, Jan 23, 2020 at 15:13:05 +0100, Stefan Kangas wrote:
> Hi Alan,

> Alan Mackenzie <acm@muc.de> writes:

> > I missed this bug report back in 2014.  The code has moved around so
> > much since then that Jordon's patch no longer applies.  But a simple
> > patch with the same idea appears to fix this bug:

> Thanks for taking a look.

> With your fix, applied on the master branch, in the example given:

>     class X
>     {
>         public void x ()
>         {
>             for (int this_is_not_a_function_name : getX ()) {
>             }
>         }
>     }

> The "int" no longer has font-lock-type-face.  Should it?

That's funny.  It has font-lock-type-face for me, in all of emacs-27,
master and standalone CC Mode (in master).  font-lock-type-face is indeed
correct, here.

Are you sure you were ....  (I don't think I need to fill in the dots for
you.  ;-)

> Best regards,
> Stefan Kangas

-- 
Alan Mackenzie (Nuremberg, Germany).

bug#18378: [Patch] wrong fontification of Java code

[Alan Mackenzie]

Hello again, Stefan.

Ping?

-- 
Alan Mackenzie (Nuremberg, Germany).


On Thu, Jan 23, 2020 at 18:09:47 +0000, Alan Mackenzie wrote:
> Hello, Stefan.

> On Thu, Jan 23, 2020 at 15:13:05 +0100, Stefan Kangas wrote:
> > Hi Alan,

> > Alan Mackenzie <acm@muc.de> writes:

> > > I missed this bug report back in 2014.  The code has moved around so
> > > much since then that Jordon's patch no longer applies.  But a simple
> > > patch with the same idea appears to fix this bug:

> > Thanks for taking a look.

> > With your fix, applied on the master branch, in the example given:

> >     class X
> >     {
> >         public void x ()
> >         {
> >             for (int this_is_not_a_function_name : getX ()) {
> >             }
> >         }
> >     }

> > The "int" no longer has font-lock-type-face.  Should it?

> That's funny.  It has font-lock-type-face for me, in all of emacs-27,
> master and standalone CC Mode (in master).  font-lock-type-face is indeed
> correct, here.

> Are you sure you were ....  (I don't think I need to fill in the dots for
> you.  ;-)

> > Best regards,
> > Stefan Kangas

> -- 
> Alan Mackenzie (Nuremberg, Germany).

bug#18378: [Patch] wrong fontification of Java code

[Alan Mackenzie]

Hello, all.

I've committed a fix to this bug to the emacs-27 branch at savannah.
I'm closing the bug with this post.

-- 
Alan Mackenzie (Nuremberg, Germany).


On Mon, Jan 20, 2020 at 22:16:46 +0100, Stefan Kangas wrote:
> Hi Alan,

> Could you please help review the below patch for cc-fonts.el?

> Thanks in advance.

> Best regards,
> Stefan Kangas

> Paul Pogonyshev <pogonyshev@gmail.com> writes:

> > class X
> > {
> >     public void x ()
> >     {
> >         for (int this_is_not_a_function_name : getX ()) {
> >         }
> >     }
> > }

> > In current trunk (compiled five minutes ago)
> > 'this_is_not_a_function_name' above is fontified as a function
> > name. This seems to depend on whether I have '()' after 'getX' or
> > not.

> > Paul

> Jordon Biondo <jordonbiondo@gmail.com> writes:

> > I have a potential solution for this bug, as it has impacted me as well.

> > Not sure what the best solution would be but I have included a patch that does
> > solve the problem. From what I can tell there are no tests for java
> > fortification but I have not seen any issues arising in my own java
> > projects. The issue was that in java, a “:” can mark the end of a declaration in
> > a foreach, not just a “;” or “,”.

> > === modified file 'lisp/ChangeLog'
> > *** lisp/ChangeLog	2014-09-01 15:03:45 +0000
> > --- lisp/ChangeLog	2014-09-01 19:11:46 +0000
> > ***************
> > *** 1,3 ****
> > --- 1,8 ----
> > + 2014-09-01  Jordon Biondo  <jordonbiondo@gmail.com>
> > + 
> > + 	* progmodes/cc-fonts.el (c-font-lock-declarators): Special
> > + 	case for declarations inside Java foreach loops (Bug#18378)
> > + 
> >   2014-09-01  Eli Zaretskii  <eliz@gnu.org>

> >   	* ls-lisp.el (ls-lisp-use-string-collate)

> > === modified file 'lisp/progmodes/cc-fonts.el'
> > *** lisp/progmodes/cc-fonts.el	2014-08-24 20:50:11 +0000
> > --- lisp/progmodes/cc-fonts.el	2014-09-01 19:11:59 +0000
> > *************** casts and declarations are fontified.  U
> > *** 1116,1123 ****
> >   	    ;; initializing brace lists.
> >   	    (let (found)
> >   	      (while
> > ! 		  (and (setq found (c-syntactic-re-search-forward
> > ! 			     "[;,]\\|\\s)\\|\\'\\|\\(=\\|\\s(\\)" limit t t))
> >   		       (eq (char-before) ?\[)
> >   		       (c-go-up-list-forward))
> >   		     (setq brackets-after-id t))
> > --- 1116,1128 ----
> >   	    ;; initializing brace lists.
> >   	    (let (found)
> >   	      (while
> > !                   (and (setq found (c-syntactic-re-search-forward
> > !                                     (concat "[;,]\\|\\s)\\|\\'\\|\\(=\\|"
> > !                                             (if (c-major-mode-is 'java-mode)
> > !                                                 ":\\|"
> > !                                               "")
> > !                                             "\\s(\\)")
> > !                                     limit t t))
> >   		       (eq (char-before) ?\[)
> >   		       (c-go-up-list-forward))
> >   		     (setq brackets-after-id t))