GNU bug report logs - #77336
[PATCH] Fix mouse highlighting for compact mode lines

Previous Next

Package: emacs;

Reported by: Pengji Zhang <me <at> pengjiz.com>

Date: Fri, 28 Mar 2025 10:56:01 UTC

Severity: normal

Tags: patch

Done: Eli Zaretskii <eliz <at> gnu.org>

Bug is archived. No further changes may be made.

Full log


Message #23 received at 77336 <at> debbugs.gnu.org (full text, mbox):

From: Pengji Zhang <me <at> pengjiz.com>
To: Eli Zaretskii <eliz <at> gnu.org>
Cc: 77336 <at> debbugs.gnu.org
Subject: Re: bug#77336: [PATCH] Fix mouse highlighting for compact mode lines
Date: Wed, 02 Apr 2025 21:02:19 +0800
[Message part 1 (text/plain, inline)]
Eli Zaretskii <eliz <at> gnu.org> writes:

>> From: Pengji Zhang <me <at> pengjiz.com>
>> Cc: 77336 <at> debbugs.gnu.org
>> Date: Mon, 31 Mar 2025 11:31:36 +0800
>> 
>> Eli Zaretskii <eliz <at> gnu.org> writes:
>> 
>> >> Is casting the 'Lisp_Object' to a number acceptable here?
>> >
>> > No, because Lisp_Object could be a struct (if Emacs was configured with
>> > "--enable-check-lisp-object-type").
>> 
>> Sorry I did not make it clear. I meant 'make_fixnum(XLI(elt))'. I
>> suppose 'XLI' will handle that case correctly?
>
> I'd prefer not to do that unless there's no other reasonable solution.
>
>> I am actually concerned that there could be a collision between two
>> different 'elt's, e.g., a symbol and a string, because 'make_fixnum'
>> alters the tag of 'elt'. I think it might be fine because 'elt' should
>> always be a 'Lisp_String' when this is called.
>
> Sorry, I don't understand: what do you mean by "alters the tag of
> 'elt'", and how could make_fixnum affect 'elt'?

It does not affect 'elt'. I meant that 'make_fixnum' may change the
value of the number from the tagged pointer:

    INLINE Lisp_Object
    make_fixnum (EMACS_INT n)
    {
      ...
        {
          n &= INTMASK;
          n += (int0 << VALBITS);
        }
      return XIL (n);
    }

So if I understand it correctly, that means 'make_fixnum(XLI(x))' could
give the same result for, e.g., a fixnum and a string.

>> > A global variable is what I had in mind.  I don't see why it would be
>> > less clean.
>> 
>> It is just that maintaining a global state is not as tasteful to me. If
>> 'make_fixnum(XLI(elt))' is not good here, I am happy to go this way.
>
> Yes, a global variable is cleaner than using XLI for this purpose.

Thanks. Please see the attached patch.

[0001-Fix-mouse-highlighting-for-compact-mode-lines-bug-77.patch (text/x-patch, inline)]
From e6ca2165f2e49364f57c53c5c1472a127a7b76a4 Mon Sep 17 00:00:00 2001
From: Pengji Zhang <me <at> pengjiz.com>
Date: Wed, 2 Apr 2025 20:52:30 +0800
Subject: [PATCH] Fix mouse highlighting for compact mode lines (bug#77336)

When 'mode-line-compact' is non-nil, the mode line string is
displayed as a whole.  That confuses the computation of ranges
of mouse highlighting on the mode line because all the glyphs
have the same Lisp object source.  As such, in this commit we
instead split the mode line string by sources, and display those
elements one by one, so the boundaries of each element could be
correctly detected for the purpose of mouse highlighting.

* src/xdisp.c (display_mode_line): Display mode line elements
one by one when 'mode-line-compact' is non-nil.
(display_mode_element): Record source element number of the
stored string via a text property.
(Fformat_mode_line): Initialize 'mode_line_elt_no' to 0.
(syms_of_xdisp): New symbol for the text property.
---
 src/xdisp.c | 145 ++++++++++++++++++++++++++++++++++++++--------------
 1 file changed, 106 insertions(+), 39 deletions(-)

diff --git a/src/xdisp.c b/src/xdisp.c
index 886f070753f..673b5c19c20 100644
--- a/src/xdisp.c
+++ b/src/xdisp.c
@@ -13674,6 +13674,9 @@ #define MODE_LINE_NOPROP_LEN(start) \
 /* List of strings making up the mode-line.  */
 static Lisp_Object mode_line_string_list;
 
+/* Element number for the stored mode line string. */
+static ptrdiff_t mode_line_elt_no;
+
 /* Base face property when building propertized mode line string.  */
 static Lisp_Object mode_line_string_face;
 static Lisp_Object mode_line_string_face_prop;
@@ -27754,57 +27757,94 @@ display_mode_line (struct window *w, enum face_id face_id, Lisp_Object format)
 	{
 	  /* The window is wide enough; just display the mode line we
 	     just computed. */
-	  display_string (NULL, mode_string, Qnil,
-			  0, 0, &it, 0, 0, 0,
-			  STRING_MULTIBYTE (mode_string));
+	  Lisp_Object start = make_fixnum (0), end;
+	  Lisp_Object elt;
+
+	  /* Display the mode line string one element by one element.
+	     This is to make the ranges of mouse highlighting
+	     correct. */
+	  do
+	    {
+	      end = Fnext_single_property_change (start, Qmode_line_elt_no,
+						  mode_string, Qnil);
+	      elt = Fsubstring (mode_string, start, end);
+	      display_string (NULL, elt, Qnil, 0, 0, &it, 0, 0, 0,
+			      STRING_MULTIBYTE (elt));
+	      start = end;
+	    }
+	  while (!NILP (end));
 	}
       else
 	{
 	  /* Compress the mode line. */
-	  ptrdiff_t i = 0, i_byte = 0, start = 0;
+	  ptrdiff_t i = 0, i_byte = 0;
 	  int prev = 0;
+	  Lisp_Object start = make_fixnum (0), end;
+	  Lisp_Object elt = empty_unibyte_string;
 
-	  while (i < SCHARS (mode_string))
+	  /* Display the mode line string one element by one element.
+	     This is to make the ranges of mouse highlighting
+	     correct. */
+	  do
 	    {
-	      int c = fetch_string_char_advance (mode_string, &i, &i_byte);
-	      if (c == ' ' && prev == ' ')
+	      end = Fnext_single_property_change (start,
+						  Qmode_line_elt_no,
+						  mode_string,
+						  make_fixnum (SCHARS (mode_string)));
+	      while (i < XFIXNUM (end))
 		{
-		  Lisp_Object prev_pos = make_fixnum (i - 1);
-
-		  /* SPC characters with 'display' properties are not
-                     really "empty", since they have non-trivial visual
-                     effects on the mode line.  */
-		  if (NILP (Fget_text_property (prev_pos, Qdisplay,
-						mode_string)))
+		  int c = fetch_string_char_advance (mode_string, &i, &i_byte);
+		  if (c == ' ' && prev == ' ')
 		    {
-		      display_string (NULL,
-				      Fsubstring (mode_string,
-						  make_fixnum (start),
-						  prev_pos),
-				      Qnil, 0, 0, &it, 0, 0, 0,
-				      STRING_MULTIBYTE (mode_string));
-		      /* Skip past the rest of the space characters. */
-		      while (c == ' ' && i < SCHARS (mode_string)
-			     && NILP (Fget_text_property (make_fixnum (i),
-							  Qdisplay,
-							  mode_string)))
+		      Lisp_Object prev_pos = make_fixnum (i - 1);
+		      Lisp_Object display = Fget_text_property (prev_pos,
+								Qdisplay,
+								mode_string);
+
+		      /* SPC characters with 'display' properties are not
+			 really "empty", since they have non-trivial visual
+			 effects on the mode line.  */
+		      if (NILP (display))
 			{
-			  c = fetch_string_char_advance (mode_string,
-							 &i, &i_byte);
+			  elt = concat2 (elt, Fsubstring (mode_string,
+							  start,
+							  prev_pos));
+
+			  /* Skip past the rest of the space characters. */
+			  Lisp_Object display = Fget_text_property (make_fixnum (i),
+								    Qdisplay,
+								    mode_string);
+			  while (c == ' ' && i < XFIXNUM (end)
+				 && NILP (display))
+			    {
+			      c = fetch_string_char_advance (mode_string,
+							     &i, &i_byte);
+			      display = Fget_text_property (make_fixnum (i),
+							    Qdisplay,
+							    mode_string);
+			    }
+
+			  /* Skip the final space no matter how the loop
+			     above ends. */
+			  if (c == ' ' && NILP (display))
+			    start = end;
+			  else
+			    start = make_fixnum (i - 1);
 			}
-		      start = i - 1;
 		    }
+		  prev = c;
 		}
-	      prev = c;
-	    }
 
-	  /* Display the final bit. */
-	  if (start < i)
-	    display_string (NULL,
-			    Fsubstring (mode_string, make_fixnum (start),
-					make_fixnum (i)),
-			    Qnil, 0, 0, &it, 0, 0, 0,
-			    STRING_MULTIBYTE (mode_string));
+	      /* Append the final bit. */
+	      if (XFIXNUM (start) < XFIXNUM (end))
+		elt = concat2 (elt, Fsubstring (mode_string, start, end));
+
+	      display_string (NULL, elt, Qnil, 0, 0, &it, 0, 0, 0,
+			      STRING_MULTIBYTE (elt));
+	      elt = empty_unibyte_string;
+	      start = end;
+	    }
+	  while (XFIXNUM (end) < SCHARS (mode_string));
 	}
     }
   pop_kboard ();
@@ -27926,6 +27966,20 @@ display_mode_element (struct it *it, int depth, int field_width, int precision,
   if (depth > 100)
     elt = build_string ("*too-deep*");
 
+  /* NOTE: Increment the number only for 'MODE_LINE_STRING', because for
+     other cases this variable is not used.  So we do not need to reset
+     the variable in every callers of this function.
+
+     NOTE: This might result in gaps in the stored element numbers
+     because some elements are processed recursively.  However, we
+     cannot increment this number when the number is stored because an
+     element could be stored by parts.  For example, "a%b" is stored as
+     two elements in the 'mode_line_string_list', but they should be
+     considered as one element, so we cannot increment this number when
+     "a" is stored. */
+  if (mode_line_target == MODE_LINE_STRING)
+    mode_line_elt_no++;
+
   depth++;
 
   switch (XTYPE (elt))
@@ -28023,7 +28077,11 @@ display_mode_element (struct it *it, int depth, int field_width, int precision,
 		n += store_mode_line_noprop (SSDATA (elt), -1, prec);
 		break;
 	      case MODE_LINE_STRING:
-		n += store_mode_line_string (NULL, elt, true, 0, prec, Qnil);
+		{
+		  AUTO_LIST2 (src, Qmode_line_elt_no,
+			      make_fixnum (mode_line_elt_no));
+		  n += store_mode_line_string (NULL, elt, true, 0, prec, src);
+		}
 		break;
 	      case MODE_LINE_DISPLAY:
 		n += display_string (NULL, elt, Qnil, 0, 0, it,
@@ -28076,8 +28134,10 @@ display_mode_element (struct it *it, int depth, int field_width, int precision,
 		      Lisp_Object mode_string
 			= Fsubstring (elt, make_fixnum (charpos),
 				      make_fixnum (endpos));
+		      AUTO_LIST2 (src, Qmode_line_elt_no,
+				  make_fixnum (mode_line_elt_no));
 		      n += store_mode_line_string (NULL, mode_string, false,
-						   0, 0, Qnil);
+						   0, 0, src);
 		    }
 		    break;
 		  case MODE_LINE_DISPLAY:
@@ -28157,6 +28217,8 @@ display_mode_element (struct it *it, int depth, int field_width, int precision,
 			{
 			  Lisp_Object tem = build_string (spec);
 			  props = Ftext_properties_at (make_fixnum (charpos), elt);
+			  props = plist_put (props, Qmode_line_elt_no,
+					     make_fixnum (mode_line_elt_no));
 			  /* Should only keep face property in props */
 			  n += store_mode_line_string (NULL, tem, false,
 						       field, prec, props);
@@ -28372,6 +28434,9 @@ display_mode_element (struct it *it, int depth, int field_width, int precision,
 	  n += store_mode_line_noprop ("", field_width - n, 0);
 	  break;
 	case MODE_LINE_STRING:
+	  /* NOTE: Padding implicitly has 'mode-line-elt-no' property
+	     to be nil.  Normally padding spaces are between two real
+	     elements, so this should be good. */
 	  n += store_mode_line_string ("", Qnil, false, field_width - n, 0,
 				       Qnil);
 	  break;
@@ -28579,6 +28644,7 @@ DEFUN ("format-mode-line", Fformat_mode_line, Sformat_mode_line,
     }
 
   push_kboard (FRAME_KBOARD (it.f));
+  mode_line_elt_no = 0;
   display_mode_element (&it, 0, 0, 0, format, Qnil, false);
   pop_kboard ();
 
@@ -37675,6 +37741,7 @@ syms_of_xdisp (void)
   DEFSYM (Qfontification_functions, "fontification-functions");
   DEFSYM (Qlong_line_optimizations_in_fontification_functions,
 	  "long-line-optimizations-in-fontification-functions");
+  DEFSYM (Qmode_line_elt_no, "mode-line-elt-no");
 
   /* Name of the symbol which disables Lisp evaluation in 'display'
      properties.  This is used by enriched.el.  */
-- 
2.49.0


This bug report was last modified 47 days ago.

Previous Next


GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.