RFA: add macro commands

RFA: add macro commands

[Jim Blandy]

2002-05-09  Jim Blandy  <jimb@redhat.com>

	Add commands for manually expanding macros and showing their
	definitions.
	* macrocmd.c, macroscope.c, macroscope.h: New files.
	* Makefile.in (SFILES): Add macrocmd.c, macroscope.c.
	(macroscope_h): New variable.
	(HFILES_NO_SRCDIR): Add macroscope.h.
	(COMMON_OBS): Add macrocmd.o, macroscope.o.
	(macroscope.o, macrocmd.o): New rules.

Index: gdb/Makefile.in
===================================================================
RCS file: /cvs/src/src/gdb/Makefile.in,v
retrieving revision 1.180.2.6
diff -c -r1.180.2.6 Makefile.in
*** gdb/Makefile.in	9 May 2002 20:03:35 -0000	1.180.2.6
--- gdb/Makefile.in	9 May 2002 23:49:14 -0000
***************
*** 539,545 ****
  	m2-exp.y m2-lang.c m2-typeprint.c m2-valprint.c main.c maint.c \
  	memattr.c mem-break.c minsyms.c mipsread.c nlmread.c objfiles.c \
  	p-exp.y p-lang.c p-typeprint.c p-valprint.c parse.c \
! 	macrotab.c macroexp.c \
  	printcmd.c remote.c remote-nrom.c scm-exp.c scm-lang.c \
  	scm-valprint.c source.c stabsread.c stack.c symfile.c \
  	symmisc.c symtab.c linespec.c target.c thread.c top.c tracepoint.c \
--- 539,545 ----
  	m2-exp.y m2-lang.c m2-typeprint.c m2-valprint.c main.c maint.c \
  	memattr.c mem-break.c minsyms.c mipsread.c nlmread.c objfiles.c \
  	p-exp.y p-lang.c p-typeprint.c p-valprint.c parse.c \
! 	macrotab.c macroexp.c macrocmd.c macroscope.c \
  	printcmd.c remote.c remote-nrom.c scm-exp.c scm-lang.c \
  	scm-valprint.c source.c stabsread.c stack.c symfile.c \
  	symmisc.c symtab.c linespec.c target.c thread.c top.c tracepoint.c \
***************
*** 635,640 ****
--- 635,641 ----
  linespec_h =	linespec.h
  macroexp_h =	macroexp.h
  macrotab_h =    macrotab.h $(obstack_h) $(bcache_h)
+ macroscope_h =  macroscope.h $(macrotab_h) $(symtab_h)
  memattr_h =     memattr.h
  monitor_h =	monitor.h
  objfiles_h =	objfiles.h
***************
*** 673,679 ****
  	gdb-stabs.h $(inferior_h) language.h minimon.h monitor.h \
  	objfiles.h parser-defs.h serial.h solib.h \
  	symfile.h stabsread.h target.h terminal.h typeprint.h xcoffsolib.h \
! 	macrotab.h macroexp.h \
  	c-lang.h ch-lang.h f-lang.h \
  	jv-lang.h \
  	m2-lang.h  p-lang.h \
--- 674,680 ----
  	gdb-stabs.h $(inferior_h) language.h minimon.h monitor.h \
  	objfiles.h parser-defs.h serial.h solib.h \
  	symfile.h stabsread.h target.h terminal.h typeprint.h xcoffsolib.h \
! 	macrotab.h macroexp.h macroscope.h \
  	c-lang.h ch-lang.h f-lang.h \
  	jv-lang.h \
  	m2-lang.h  p-lang.h \
***************
*** 716,722 ****
  	source.o values.o eval.o valops.o valarith.o valprint.o printcmd.o \
  	symtab.o symfile.o symmisc.o linespec.o infcmd.o infrun.o \
  	expprint.o environ.o stack.o thread.o \
! 	macrotab.o macroexp.o \
  	event-loop.o event-top.o inf-loop.o completer.o \
  	gdbarch.o arch-utils.o gdbtypes.o copying.o $(DEPFILES) \
  	memattr.o mem-break.o target.o parse.o language.o $(YYOBJ) buildsym.o \
--- 717,723 ----
  	source.o values.o eval.o valops.o valarith.o valprint.o printcmd.o \
  	symtab.o symfile.o symmisc.o linespec.o infcmd.o infrun.o \
  	expprint.o environ.o stack.o thread.o \
! 	macrotab.o macrocmd.o macroexp.o macroscope.o \
  	event-loop.o event-top.o inf-loop.o completer.o \
  	gdbarch.o arch-utils.o gdbtypes.o copying.o $(DEPFILES) \
  	memattr.o mem-break.o target.o parse.o language.o $(YYOBJ) buildsym.o \
***************
*** 2123,2128 ****
--- 2124,2134 ----
  
  macrotab.o: macrotab.c $(defs_h) $(obstack_h) $(objfiles_h) $(symtab_h) \
  	$(macrotab_h) $(splay_tree_h) gdb_assert.h $(bcache_h)
+ 
+ macroscope.o: macroscope.c $(defs_h) $(macroscope_h)
+ 
+ macrocmd.o: macrocmd.c $(defs_h) $(macrotab_h) $(macroexp_h) \
+ 	$(macroscope_h) $(command_h) $(gdbcmd_h)
  
  target.o: target.c $(bfd_h) $(defs_h) $(gdbcmd_h) $(inferior_h) \
  	$(objfiles_h) $(symfile_h) $(target_h) $(gdb_string_h) $(regcache_h)
Index: gdb/macrocmd.c
===================================================================
RCS file: gdb/macrocmd.c
diff -N gdb/macrocmd.c
*** gdb/macrocmd.c	1 Jan 1970 00:00:00 -0000
--- gdb/macrocmd.c	9 May 2002 23:49:15 -0000
***************
*** 0 ****
--- 1,287 ----
+ /* C preprocessor macro expansion commands for GDB.
+    Copyright 2002 Free Software Foundation, Inc.
+    Contributed by Red Hat, Inc.
+ 
+    This file is part of GDB.
+ 
+    This program is free software; you can redistribute it and/or modify
+    it under the terms of the GNU General Public License as published by
+    the Free Software Foundation; either version 2 of the License, or
+    (at your option) any later version.
+ 
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU General Public License for more details.
+ 
+    You should have received a copy of the GNU General Public License
+    along with this program; if not, write to the Free Software
+    Foundation, Inc., 59 Temple Place - Suite 330,
+    Boston, MA 02111-1307, USA.  */
+ 
+ 
+ #include "defs.h"
+ #include "macrotab.h"
+ #include "macroexp.h"
+ #include "macroscope.h"
+ #include "command.h"
+ #include "gdbcmd.h"
+ 
+ \f
+ /* The `macro' prefix command.  */
+ 
+ static struct cmd_list_element *macrolist;
+ 
+ static void
+ macro_command (char *arg, int from_tty)
+ {
+   printf_unfiltered
+     ("\"macro\" must be followed by the name of a macro command.\n");
+   help_list (macrolist, "macro ", -1, gdb_stdout);
+ }
+ 
+ 
+ \f
+ /* Macro expansion commands.  */
+ 
+ 
+ static void
+ macro_expand_command (char *exp, int from_tty)
+ {
+   struct macro_scope *ms = 0;
+   char *expanded = 0;
+   struct cleanup *cleanup_chain = make_cleanup (free_current_contents, &ms);
+   make_cleanup (free_current_contents, &expanded);
+ 
+   /* You know, when the user doesn't specify any expression, it would be
+      really cool if this defaulted to the last expression evaluated.
+      Then it would be easy to ask, "Hey, what did I just evaluate?"  But
+      at the moment, the `print' commands don't save the last expression
+      evaluated, just its value.  */
+   if (! exp || ! *exp)
+     error ("You must follow the `macro expand' command with the"
+            " expression you\n"
+            "want to expand.");
+ 
+   ms = default_macro_scope ();
+   if (ms)
+     {
+       expanded = macro_expand (exp, standard_macro_lookup, ms);
+       fputs_filtered ("expands to: ", gdb_stdout);
+       fputs_filtered (expanded, gdb_stdout);
+       fputs_filtered ("\n", gdb_stdout);
+     }
+   else
+     fputs_filtered ("GDB has no preprocessor macro information for "
+                     "that code.\n",
+                     gdb_stdout);
+ 
+   do_cleanups (cleanup_chain);
+   return;
+ }
+ 
+ 
+ static void
+ macro_expand_once_command (char *exp, int from_tty)
+ {
+   struct macro_scope *ms = 0;
+   char *expanded = 0;
+   struct cleanup *cleanup_chain = make_cleanup (free_current_contents, &ms);
+   make_cleanup (free_current_contents, &expanded);
+ 
+   /* You know, when the user doesn't specify any expression, it would be
+      really cool if this defaulted to the last expression evaluated.
+      And it should set the once-expanded text as the new `last
+      expression'.  That way, you could just hit return over and over and
+      see the expression expanded one level at a time.  */
+   if (! exp || ! *exp)
+     error ("You must follow the `macro expand-once' command with"
+            " the expression\n"
+            "you want to expand.");
+ 
+   ms = default_macro_scope ();
+   if (ms)
+     {
+       expanded = macro_expand_once (exp, standard_macro_lookup, ms);
+       fputs_filtered ("expands to: ", gdb_stdout);
+       fputs_filtered (expanded, gdb_stdout);
+       fputs_filtered ("\n", gdb_stdout);
+     }
+   else
+     fputs_filtered ("GDB has no preprocessor macro information for "
+                     "that code.\n",
+                     gdb_stdout);
+ 
+   do_cleanups (cleanup_chain);
+   return;
+ }
+ 
+ 
+ static void
+ show_pp_source_pos (struct ui_file *stream,
+                     struct macro_source_file *file,
+                     int line)
+ {
+   fprintf_filtered (stream, "%s:%d\n", file->filename, line);
+ 
+   while (file->included_by)
+     {
+       fprintf_filtered (gdb_stdout, "  included at %s:%d\n",
+                         file->included_by->filename,
+                         file->included_at_line);
+       file = file->included_by;
+     }
+ }
+ 
+ 
+ static void
+ show_macro_command (char *name, int from_tty)
+ {
+   struct macro_scope *ms = 0;
+   struct cleanup *cleanup_chain = make_cleanup (free_current_contents, &ms);
+   struct macro_definition *d;
+   
+   if (! name || ! *name)
+     error ("You must follow the `show macro' command with the name"
+            " of the macro\n"
+            "whose definition you want to see.");
+ 
+   ms = default_macro_scope ();
+   if (! ms)
+     error ("GDB has no preprocessor macro information for that code.\n");
+ 
+   d = macro_lookup_definition (ms->file, ms->line, name);
+   if (d)
+     {
+       int line;
+       struct macro_source_file *file
+         = macro_definition_location (ms->file, ms->line, name, &line);
+ 
+       fprintf_filtered (gdb_stdout, "Defined at ");
+       show_pp_source_pos (gdb_stdout, file, line);
+       fprintf_filtered (gdb_stdout, "#define %s", name);
+       if (d->kind == macro_function_like)
+         {
+           int i;
+ 
+           fputs_filtered ("(", gdb_stdout);
+           for (i = 0; i < d->argc; i++)
+             {
+               fputs_filtered (d->argv[i], gdb_stdout);
+               if (i + 1 < d->argc)
+                 fputs_filtered (", ", gdb_stdout);
+             }
+           fputs_filtered (")", gdb_stdout);
+         }
+       fprintf_filtered (gdb_stdout, " %s\n", d->replacement);
+     }
+   else
+     {
+       fprintf_filtered (gdb_stdout,
+                         "The symbol `%s' has no definition as a C/C++"
+                         " preprocessor macro\n"
+                         "at ", name);
+       show_pp_source_pos (gdb_stdout, ms->file, ms->line);
+     }
+ 
+   do_cleanups (cleanup_chain);
+ }
+ 
+ 
+ \f
+ /* User-defined macros.  */
+ 
+ /* A table of user-defined macros.  Unlike the macro tables used for
+    symtabs, this one uses xmalloc for all its allocation, not an
+    obstack, and it doesn't bcache anything; it just xmallocs things.  So
+    it's perfectly possible to remove things from this, or redefine
+    things.  */
+ static struct macro_table *user_macros;
+ 
+ static void
+ macro_define_command (char *exp, int from_tty)
+ {
+   error ("Command not implemented yet.");
+ }
+ 
+ 
+ static void
+ macro_undef_command (char *exp, int from_tty)
+ {
+   error ("Command not implemented yet.");
+ }
+ 
+ 
+ static void
+ macro_list_command (char *exp, int from_tty)
+ {
+   error ("Command not implemented yet.");
+ }
+ 
+ 
+ \f
+ /* Initializing the `macrocmd' module.  */
+ 
+ void
+ _initialize_macrocmd (void)
+ {
+   struct cmd_list_element *c;
+ 
+   /* We introduce a new command prefix, `macro', under which we'll put
+      the various commands for working with preprocessor macros.  */
+   add_prefix_cmd
+     ("macro", class_info, macro_command,
+      "Prefix for commands dealing with C preprocessor macros.",
+      &macrolist, "macro ", 0, &cmdlist);
+ 
+   add_cmd
+     ("expand", no_class, macro_expand_command,
+      "Fully expand any C/C++ preprocessor macro invocations in EXPRESSION.\n"
+      "Show the expanded expression.",
+      &macrolist);
+   add_alias_cmd ("exp", "expand", no_class, 1, &macrolist);
+   add_cmd
+     ("expand-once", no_class, macro_expand_once_command,
+      "Expand C/C++ preprocessor macro invocations appearing directly in"
+      " EXPRESSION.\n"
+      "Show the expanded expression.\n"
+      "\n"
+      "This command differs from `macro expand' in that it only expands macro\n"
+      "invocations that appear directly in EXPRESSION; if expanding a macro\n"
+      "introduces further macro invocations, those are left unexpanded.\n"
+      "\n"
+      "`macro expand-once' helps you see how a particular macro expands,\n"
+      "whereas `macro expand' shows you how all the macros involved in an\n"
+      "expression work together to yield a pre-processed expression.",
+      &macrolist);
+   add_alias_cmd ("exp1", "expand-once", no_class, 1, &macrolist);
+ 
+   add_cmd
+     ("macro", no_class, show_macro_command,
+      "Show the definition of MACRO, and its source location.",
+      &showlist);
+ 
+   add_cmd
+     ("define", no_class, macro_define_command,
+      "Define a new C/C++ preprocessor macro.\n"
+      "The GDB command `macro define DEFINITION' is equivalent to placing a\n"
+      "preprocessor directive of the form `#define DEFINITION' such that the\n"
+      "definition is visible in all the inferior's source files.\n"
+      "For example:\n"
+      "  (gdb) macro define PI (3.1415926)\n"
+      "  (gdb) macro define MIN(x,y) ((x) < (y) ? (x) : (y))",
+      &macrolist);
+ 
+   add_cmd
+     ("undef", no_class, macro_undef_command,
+      "Remove the definition of the C/C++ preprocessor macro with the"
+      " given name.",
+      &macrolist);
+ 
+   add_cmd
+     ("list", no_class, macro_list_command,
+      "List all the macros defined using the `macro define' command.",
+      &macrolist);
+ 
+   user_macros = new_macro_table (0, 0);
+ }
Index: gdb/macroscope.c
===================================================================
RCS file: gdb/macroscope.c
diff -N gdb/macroscope.c
*** gdb/macroscope.c	1 Jan 1970 00:00:00 -0000
--- gdb/macroscope.c	9 May 2002 23:49:15 -0000
***************
*** 0 ****
--- 1,107 ----
+ /* Functions for deciding which macros are currently in scope.
+    Copyright 2002 Free Software Foundation, Inc.
+    Contributed by Red Hat, Inc.
+ 
+    This file is part of GDB.
+ 
+    This program is free software; you can redistribute it and/or modify
+    it under the terms of the GNU General Public License as published by
+    the Free Software Foundation; either version 2 of the License, or
+    (at your option) any later version.
+ 
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU General Public License for more details.
+ 
+    You should have received a copy of the GNU General Public License
+    along with this program; if not, write to the Free Software
+    Foundation, Inc., 59 Temple Place - Suite 330,
+    Boston, MA 02111-1307, USA.  */
+ 
+ #include "defs.h"
+ 
+ #include "macroscope.h"
+ #include "symtab.h"
+ #include "target.h"
+ #include "frame.h"
+ #include "inferior.h"
+ 
+ 
+ struct macro_scope *
+ sal_macro_scope (struct symtab_and_line sal)
+ {
+   struct macro_source_file *main;
+   struct macro_scope *ms;
+ 
+   if (! sal.symtab
+       || ! sal.symtab->macro_table)
+     return 0;
+ 
+   ms = (struct macro_scope *) xmalloc (sizeof (*ms));
+ 
+   main = macro_main (sal.symtab->macro_table);
+   ms->file = macro_lookup_inclusion (main, sal.symtab->filename);
+ 
+   if (! ms->file)
+     internal_error
+       (__FILE__, __LINE__,
+        "\n"
+        "the symtab `%s' refers to a preprocessor macro table which doesn't\n"
+        "have any record of processing a file by that name.\n",
+        sal.symtab->filename);
+ 
+   ms->line = sal.line;
+ 
+   return ms;
+ }
+ 
+ 
+ struct macro_scope *
+ default_macro_scope ()
+ {
+   struct symtab_and_line sal;
+   struct macro_source_file *main;
+   struct macro_scope *ms;
+ 
+   /* If there's a selected frame, use its PC.  */ 
+   if (selected_frame)
+     sal = find_pc_line (selected_frame->pc, 0);
+   
+   /* If the target has any registers at all, then use its PC.  Why we
+      would have registers but no stack, I'm not sure.  */
+   else if (target_has_registers)
+     sal = find_pc_line (read_pc (), 0);
+ 
+   /* If all else fails, fall back to the current listing position.  */
+   else
+     {
+       /* Don't call select_source_symtab here.  That can raise an
+          error if symbols aren't loaded, but GDB calls the expression
+          evaluator in all sorts of contexts.
+ 
+          For example, commands like `set width' call the expression
+          evaluator to evaluate their numeric arguments.  If the
+          current language is C, then that may call this function to
+          choose a scope for macro expansion.  If you don't have any
+          symbol files loaded, then select_source_symtab will raise an
+          error.  But `set width' shouldn't raise an error just because
+          it can't decide which scope to macro-expand its argument in.  */
+       sal.symtab = current_source_symtab;
+       sal.line = current_source_line;
+     }
+ 
+   return sal_macro_scope (sal);
+ }
+ 
+ 
+ /* Look up the definition of the macro named NAME in scope at the source
+    location given by BATON, which must be a pointer to a `struct
+    macro_scope' structure.  */
+ struct macro_definition *
+ standard_macro_lookup (const char *name, void *baton)
+ {
+   struct macro_scope *ms = (struct macro_scope *) baton;
+ 
+   return macro_lookup_definition (ms->file, ms->line, name);
+ }
Index: gdb/macroscope.h
===================================================================
RCS file: gdb/macroscope.h
diff -N gdb/macroscope.h
*** gdb/macroscope.h	1 Jan 1970 00:00:00 -0000
--- gdb/macroscope.h	9 May 2002 23:49:15 -0000
***************
*** 0 ****
--- 1,63 ----
+ /* Interface to functions for deciding which macros are currently in scope.
+    Copyright 2002 Free Software Foundation, Inc.
+    Contributed by Red Hat, Inc.
+ 
+    This file is part of GDB.
+ 
+    This program is free software; you can redistribute it and/or modify
+    it under the terms of the GNU General Public License as published by
+    the Free Software Foundation; either version 2 of the License, or
+    (at your option) any later version.
+ 
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU General Public License for more details.
+ 
+    You should have received a copy of the GNU General Public License
+    along with this program; if not, write to the Free Software
+    Foundation, Inc., 59 Temple Place - Suite 330,
+    Boston, MA 02111-1307, USA.  */
+ 
+ #ifndef MACROSCOPE_H
+ #define MACROSCOPE_H
+ 
+ #include "macrotab.h"
+ #include "symtab.h"
+ 
+ 
+ /* All the information we need to decide which macro definitions are
+    in scope: a source file (either a main source file or an
+    #inclusion), and a line number in that file.  */
+ struct macro_scope {
+   struct macro_source_file *file;
+   int line;
+ };
+ 
+ 
+ /* Return a `struct macro_scope' object corresponding to the symtab
+    and line given in SAL.  If we have no macro information for that
+    location, or if SAL's pc is zero, return zero.  */
+ struct macro_scope *sal_macro_scope (struct symtab_and_line sal);
+ 
+ 
+ /* Return a `struct macro_scope' object describing the scope the `macro
+    expand' and `macro expand-once' commands should use for looking up
+    macros.  If we have a selected frame, this is the source location of
+    its PC; otherwise, this is the last listing position.
+ 
+    If we have no macro information for the current location, return zero.
+ 
+    The object returned is allocated using xmalloc; the caller is
+    responsible for freeing it.  */
+ struct macro_scope *default_macro_scope (void);
+ 
+ 
+ /* Look up the definition of the macro named NAME in scope at the source
+    location given by BATON, which must be a pointer to a `struct
+    macro_scope' structure.  This function is suitable for use as
+    a macro_lookup_ftype function.  */
+ struct macro_definition *standard_macro_lookup (const char *name, void *baton);
+ 
+ 
+ #endif /* MACROSCOPE_H */

Re: RFA: add macro commands

[Daniel Jacobowitz]

On Thu, May 09, 2002 at 06:58:22PM -0500, Jim Blandy wrote:
> 
> 2002-05-09  Jim Blandy  <jimb@redhat.com>
> 
> 	Add commands for manually expanding macros and showing their
> 	definitions.
> 	* macrocmd.c, macroscope.c, macroscope.h: New files.
> 	* Makefile.in (SFILES): Add macrocmd.c, macroscope.c.
> 	(macroscope_h): New variable.
> 	(HFILES_NO_SRCDIR): Add macroscope.h.
> 	(COMMON_OBS): Add macrocmd.o, macroscope.o.
> 	(macroscope.o, macrocmd.o): New rules.

This came up recently... Jim, could you please use PATCH: rather than
RFA:?  The 'R' is request, which is obviously inappropriate since
you're committing them at the same time.

Meanwhile, glad to see this work being finished!

-- 
Daniel Jacobowitz                           Carnegie Mellon University
MontaVista Software                         Debian GNU/Linux Developer

Re: RFA: add macro commands

[Andrew Cagney]

doco?

Re: RFA: add macro commands

[Jim Blandy]

Daniel Jacobowitz <drow@mvista.com> writes:

> On Thu, May 09, 2002 at 06:58:22PM -0500, Jim Blandy wrote:
> > 
> > 2002-05-09  Jim Blandy  <jimb@redhat.com>
> > 
> > 	Add commands for manually expanding macros and showing their
> > 	definitions.
> > 	* macrocmd.c, macroscope.c, macroscope.h: New files.
> > 	* Makefile.in (SFILES): Add macrocmd.c, macroscope.c.
> > 	(macroscope_h): New variable.
> > 	(HFILES_NO_SRCDIR): Add macroscope.h.
> > 	(COMMON_OBS): Add macrocmd.o, macroscope.o.
> > 	(macroscope.o, macrocmd.o): New rules.
> 
> This came up recently... Jim, could you please use PATCH: rather than
> RFA:?  The 'R' is request, which is obviously inappropriate since
> you're committing them at the same time.

I'm committing them to jimb-macro-020506-branch, not to the trunk.

Re: RFA: add macro commands

[Jim Blandy]

Andrew Cagney <ac131313@cygnus.com> writes:
> doco?

Coming soon.

Re: RFA: add macro commands

[Eli Zaretskii]

> From: Jim Blandy <jimb@redhat.com>
> Date: Thu,  9 May 2002 18:58:22 -0500 (EST)
> 
> +   add_cmd
> +     ("expand", no_class, macro_expand_command,
> +      "Fully expand any C/C++ preprocessor macro invocations in EXPRESSION.\n"
> +      "Show the expanded expression.",
> +      &macrolist);

Do I understand correctly that this command definition, as well as
others you added in this patch, makes the command use the default
completion function?  If so, I'm not sure that's a good idea, unless
the tables scanned by make_symbol_completion_list store macro names.

It would be nice to have completion only on macro names in these
commands.  But if that cannot be done easily, at least let's disable
completion entirely until something appropriate is coded.

Re: RFA: add macro commands

[Jim Blandy]

"Eli Zaretskii" <eliz@is.elta.co.il> writes:
> > From: Jim Blandy <jimb@redhat.com>
> > Date: Thu,  9 May 2002 18:58:22 -0500 (EST)
> > 
> > +   add_cmd
> > +     ("expand", no_class, macro_expand_command,
> > +      "Fully expand any C/C++ preprocessor macro invocations in EXPRESSION.\n"
> > +      "Show the expanded expression.",
> > +      &macrolist);
> 
> Do I understand correctly that this command definition, as well as
> others you added in this patch, makes the command use the default
> completion function?  If so, I'm not sure that's a good idea, unless
> the tables scanned by make_symbol_completion_list store macro names.
> 
> It would be nice to have completion only on macro names in these
> commands.  But if that cannot be done easily, at least let's disable
> completion entirely until something appropriate is coded.

I hadn't thought about completion at all.

Given that the macros are stored in a splay tree, they're all sorted
in a way that would make efficient completion straightforward.  So it
wouldn't be too hard to write a macro-only completion function.

But the argument is an expression.  People may want to use real
variable names in the expressions they expand: they may not want to
substitute short dummy names for the long variable names in the
expression they have inmind.  So why should it complete differently
from other expressions?  Why shouldn't GDB help people use long
variable names in their expressions to expand?

I think a better change would be to simply include macro names in the
completion process.

Re: RFA: add macro commands

[Eli Zaretskii]

> From: Jim Blandy <jimb@redhat.com>
> Date: 10 May 2002 10:55:45 -0500
> 
> I think a better change would be to simply include macro names in the
> completion process.

Yes, if you can make make_symbol_completion_list complete on macros as
well as on other symbols, it would be great.

Re: RFA: add macro commands

[Neil Booth]

Jim Blandy wrote:-

> > It would be nice to have completion only on macro names in these
> > commands.  But if that cannot be done easily, at least let's disable
> > completion entirely until something appropriate is coded.
> 
> I hadn't thought about completion at all.
> 
> Given that the macros are stored in a splay tree, they're all sorted
> in a way that would make efficient completion straightforward.  So it
> wouldn't be too hard to write a macro-only completion function.

Just in case you'd not realized, macro expansions are frequently
different, depending on location in the source files.

Neil.

Re: RFA: add macro commands

[Eli Zaretskii]

> Date: Sat, 11 May 2002 00:08:13 +0100
> From: Neil Booth <neil@daikokuya.demon.co.uk>
> 
> Jim Blandy wrote:-
> 
> > Given that the macros are stored in a splay tree, they're all sorted
> > in a way that would make efficient completion straightforward.  So it
> > wouldn't be too hard to write a macro-only completion function.
> 
> Just in case you'd not realized, macro expansions are frequently
> different, depending on location in the source files.

I'm not sure I understand what you mean.  I was talking about
completing on macro _names_, not trhe results of their _expansion_.
That is, if I type "print regnum < NUM_" and hit TAB, I'd expect to
see NUM_REGS and NUM_PSEUDO_REGS as possible completions.  But I
don't care about the values these two expand to.

So this kind of completion does not depend on the location in the
source files, only on the macros recorded in the debug info.

Am I missing something?

Re: RFA: add macro commands

[Neil Booth]

Eli Zaretskii wrote:-

> Am I missing something?

Sorry, I'd misunderstood you.

Neil.

Re: RFA: add macro commands

[Jim Blandy]

Neil Booth <neil@daikokuya.demon.co.uk> writes:
> Jim Blandy wrote:-
> 
> > > It would be nice to have completion only on macro names in these
> > > commands.  But if that cannot be done easily, at least let's disable
> > > completion entirely until something appropriate is coded.
> > 
> > I hadn't thought about completion at all.
> > 
> > Given that the macros are stored in a splay tree, they're all sorted
> > in a way that would make efficient completion straightforward.  So it
> > wouldn't be too hard to write a macro-only completion function.
> 
> Just in case you'd not realized, macro expansions are frequently
> different, depending on location in the source files.

GDB's macro data structures are meant to take into account macros
being #defined, #undefined, and re-#defined, and source files being
#included multiple times, and perhaps #defining different symbols each
time (e.g., under the influence of #if directives).

In order to look up a macro's definition, you need to provide the
macro name, the line number, and a structure representing, not a
source file, but a particular #inclusion of a source file.  The lookup
function returns you the #definition that was in force at that line of
that #inclusion.  (The main source file is treated as the "root"
#inclusion.)

I believe GDB's data structures are as precise as libcpp's linemap
structures and logical line numbers.  I think the linemap structure is
more graceful, though.

For the details, see the data structures in macrotab.h and macrotab.c: 
http://sources.redhat.com/ml/gdb-patches/2002-05/msg00135.html

Re: RFA: add macro commands

[Jim Blandy]

I've committed this patch.  This is the first user-visible change, so
tests will be coming soon.

Jim Blandy <jimb@redhat.com> writes:

> 2002-05-09  Jim Blandy  <jimb@redhat.com>
> 
> 	Add commands for manually expanding macros and showing their
> 	definitions.
> 	* macrocmd.c, macroscope.c, macroscope.h: New files.
> 	* Makefile.in (SFILES): Add macrocmd.c, macroscope.c.
> 	(macroscope_h): New variable.
> 	(HFILES_NO_SRCDIR): Add macroscope.h.
> 	(COMMON_OBS): Add macrocmd.o, macroscope.o.
> 	(macroscope.o, macrocmd.o): New rules.