diagnostics: give m4 precise locations

Currently we pass only the columns based on the screen-width, which is
important for the carets.  But we don't pass the bytes-based columns,
which is important for the colors.  Pass both.

* src/muscle-tab.c (muscle_boundary_grow): Also pass the byte-based column.
* src/location.c (location_caret): Clarify.
(boundary_set_from_string): Adjust to the new format.
* tests/diagnostics.at (Tabulations and multibyte characters from M4): New.

src/location.c

@@ -238,8 +238,8 @@ location_caret (location loc, const char *style, FILE *out)
     int c = getc (caret_info.source);
     if (c != EOF)
       {
-        /* Quote the file (at most first line in the case of multiline
-           location).  Indent by a single column.  */
+        /* Quote the file (at most the first line in the case of
+           multiline locations).  */
         fprintf (out, "%5d | ", loc.start.line);
         bool single_line = loc.start.line == loc.end.line;
         /* Consider that single point location (with equal boundaries)
@@ -269,9 +269,9 @@ location_caret (location loc, const char *style, FILE *out)
           putc ('^', out);
           /* Underlining a multiline location ends with the first
              line.  */
-          int len = loc.start.line != loc.end.line
-            ? ftell (caret_info.source) - caret_info.offset
-            : loc.end.column;
+          int len = single_line
+            ? loc.end.column
+            : ftell (caret_info.source) - caret_info.offset;
           for (int i = loc.start.column + 1; i < len; ++i)
             putc ('~', out);
           end_use_class (style, out);
@@ -289,17 +289,29 @@ location_empty (location loc)
 }
 
 void
-boundary_set_from_string (boundary *bound, char *loc_str)
+boundary_set_from_string (boundary *bound, char *str)
 {
-  /* Must search in reverse since the file name field may
-   * contain '.' or ':'.  */
-  char *delim = strrchr (loc_str, '.');
-  aver (delim);
-  *delim = '\0';
-  bound->byte = bound->column = atoi (delim+1);
-  delim = strrchr (loc_str, ':');
-  aver (delim);
-  *delim = '\0';
-  bound->line = atoi (delim+1);
-  bound->file = uniqstr_new (loc_str);
+  /* Must search in reverse since the file name field may contain '.'
+     or ':'.  */
+  char *at = strrchr (str, '@');
+  if (at)
+    {
+      *at = '\0';
+      bound->byte = atoi (at+1);
+    }
+  {
+    char *dot = strrchr (str, '.');
+    aver (dot);
+    *dot = '\0';
+    bound->column = atoi (dot+1);
+    if (!at)
+      bound->byte = bound->column;
+  }
+  {
+    char *colon = strrchr (str, ':');
+    aver (colon);
+    *colon = '\0';
+    bound->line = atoi (colon+1);
+  }
+  bound->file = uniqstr_new (str);
 }

src/location.h

@@ -135,7 +135,8 @@ location_cmp (location a, location b)
 /* Whether this is the empty location.  */
 bool location_empty (location loc);
 
-/* LOC_STR must be formatted as 'file:line.column', it will be modified.  */
-void boundary_set_from_string (boundary *bound, char *loc_str);
+/* STR must be formatted as 'file:line.column@byte' or 'file:line.column',
+   it will be modified.  */
+void boundary_set_from_string (boundary *bound, char *str);
 
 #endif /* ! defined LOCATION_H_ */

src/muscle-tab.c

@@ -275,7 +275,7 @@ muscle_boundary_grow (char const *key, boundary bound)
 {
   obstack_sgrow  (&muscle_obstack, "[[");
   obstack_escape (&muscle_obstack, bound.file);
-  obstack_printf (&muscle_obstack, ":%d.%d]]", bound.line, bound.column);
+  obstack_printf (&muscle_obstack, ":%d.%d@@%d]]", bound.line, bound.column, bound.byte);
   char const *extension = obstack_finish0 (&muscle_obstack);
   muscle_grow (key, extension, "", "");
   obstack_free (&muscle_obstack, extension);

tests/diagnostics.at

@@ -33,16 +33,18 @@ AT_BISON_OPTION_PUSHDEFS
 
 AT_DATA_GRAMMAR([[input.y]], [$2])
 
-AT_DATA([experr], [$4])
+AT_DATA([experr.orig], [$4])
+
+# When no style, same messages, but without style.
+AT_CHECK([perl -p -e 's{</?\w+>}{}g' <experr.orig >experr])
 # Cannot use AT_BISON_CHECK easily as we need to change the
 # environment.
 # FIXME: Enhance AT_BISON_CHECK.
-AT_CHECK([LC_ALL=en_US.UTF-8 bison -fcaret --style=debug -Wall input.y], [$3], [], [experr])
-
-# When no style, same messages, but without style.
-AT_CHECK([perl -pi -e 's{</?\w+>}{}g' experr])
 AT_CHECK([LC_ALL=en_US.UTF-8 bison -fcaret -Wall input.y], [$3], [], [experr])
 
+AT_CHECK([cp experr.orig experr])
+AT_CHECK([LC_ALL=en_US.UTF-8 bison -fcaret --style=debug -Wall input.y], [$3], [], [experr])
+
 AT_BISON_OPTION_POPDEFS
 
 AT_CLEANUP
@@ -212,4 +214,29 @@ input.y:9.1-24:      previous definition
 input.y: <warning>warning:</warning> fix-its can be applied.  Rerun with option '--update'. [<warning>-Wother</warning>]
 ]])
 
+
+## ---------------------------------------------- ##
+## Tabulations and multibyte characters from M4.  ##
+## ---------------------------------------------- ##
+
+# Locations coming from m4 need the byte-column for diagnostics.
+
+AT_TEST([[Locations from M4]],
+[[%define api.prefix		{sun}
+%define api.prefix      	{🌞}
+%%
+exp:;
+]],
+[1],
+[[input.y:10.1-35: <error>error:</error> %define variable 'api.prefix' redefined
+   10 | <error>%define api.prefix      	{🌞}</error>
+      | <error>^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~</error>
+input.y:9.1-37:      previous definition
+    9 | <note>%define api.prefix		{sun}</note>
+      | <note>^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~</note>
+input.y: <warning>warning:</warning> fix-its can be applied.  Rerun with option '--update'. [<warning>-Wother</warning>]
+]])
+
+
+
 m4_popdef([AT_TEST])