various frame-geometry fixes -------------------- ChangeLog entries follow: -------------------- src/ChangeLog addition: 2010-02-15 Ben Wing <ben@xemacs.org> * EmacsFrame.c: * EmacsFrame.c (EmacsFrameResize): * console-msw-impl.h: * console-msw-impl.h (struct mswindows_frame): * console-msw-impl.h (FRAME_MSWINDOWS_TARGET_RECT): * device-tty.c: * device-tty.c (tty_asynch_device_change): * event-msw.c: * event-msw.c (mswindows_wnd_proc): * faces.c (Fface_list): * faces.h: * frame-gtk.c: * frame-gtk.c (gtk_set_initial_frame_size): * frame-gtk.c (gtk_set_frame_size): * frame-msw.c: * frame-msw.c (mswindows_init_frame_1): * frame-msw.c (mswindows_set_frame_size): * frame-msw.c (mswindows_size_frame_internal): * frame-msw.c (msprinter_init_frame_3): * frame.c: * frame.c (enum): * frame.c (Fmake_frame): * frame.c (adjust_frame_size): * frame.c (store_minibuf_frame_prop): * frame.c (Fframe_property): * frame.c (Fframe_properties): * frame.c (Fframe_displayable_pixel_height): * frame.c (Fframe_displayable_pixel_width): * frame.c (internal_set_frame_size): * frame.c (Fset_frame_height): * frame.c (Fset_frame_pixel_height): * frame.c (Fset_frame_displayable_pixel_height): * frame.c (Fset_frame_width): * frame.c (Fset_frame_pixel_width): * frame.c (Fset_frame_displayable_pixel_width): * frame.c (Fset_frame_size): * frame.c (Fset_frame_pixel_size): * frame.c (Fset_frame_displayable_pixel_size): * frame.c (frame_conversion_internal_1): * frame.c (get_frame_displayable_pixel_size): * frame.c (change_frame_size_1): * frame.c (change_frame_size): * frame.c (generate_title_string): * frame.h: * gtk-xemacs.c: * gtk-xemacs.c (gtk_xemacs_size_request): * gtk-xemacs.c (gtk_xemacs_size_allocate): * gtk-xemacs.c (gtk_xemacs_paint): * gutter.c: * gutter.c (update_gutter_geometry): * redisplay.c (end_hold_frame_size_changes): * redisplay.c (redisplay_frame): * toolbar.c: * toolbar.c (update_frame_toolbars_geometry): * window.c: * window.c (frame_pixsize_valid_p): * window.c (check_frame_size): Various fixes to frame geometry to make it a bit easier to understand and fix some bugs. 1. IMPORTANT: Some renamings. Will need to be applied carefully to the carbon repository, in the following order: -- pixel_to_char_size -> pixel_to_frame_unit_size -- char_to_pixel_size -> frame_unit_to_pixel_size -- pixel_to_real_char_size -> pixel_to_char_size -- char_to_real_pixel_size -> char_to_pixel_size -- Reverse second and third arguments of change_frame_size() and change_frame_size_1() to try to make functions consistent in putting width before height. -- Eliminate old round_size_to_char, because it didn't really do anything differently from round_size_to_real_char() -- round_size_to_real_char -> round_size_to_char; any places that called the old round_size_to_char should just call the new one. 2. IMPORTANT FOR CARBON: The set_frame_size() method is now passed sizes in "frame units", like all other frame-sizing functions, rather than some hacked-up combination of char-cell units and total pixel size. This only affects window systems that use "pixelated geometry", and I'm not sure if Carbon is one of them. MS Windows is pixelated, X and GTK are not. For pixelated-geometry systems, the size in set_frame_size() is in displayable pixels rather than total pixels and needs to be converted appropriately; take a look at the changes made to mswindows_set_frame_size() method if necessary. 3. Add a big long comment in frame.c describing how frame geometry works. 4. Remove MS Windows-specific character height and width fields, duplicative and unused. 5. frame-displayable-pixel-* and set-frame-displayable-pixel-* didn't use to work on MS Windows, but they do now. 6. In general, clean up the handling of "pixelated geometry" so that fewer functions have to worry about this. This is really an abomination that should be removed entirely but that will have to happen later. Fix some buggy code in frame_conversion_internal() that happened to "work" because it was countered by oppositely buggy code in change_frame_size(). 7. Clean up some frame-size code in toolbar.c and use functions already provided in frame.c instead of rolling its own. 8. Fix check_frame_size() in window.c, which formerly didn't take pixelated geometry into account.

/* Repository for inline functions
   Copyright (C) 1995 Sun Microsystems, Inc.

This file is part of XEmacs.

XEmacs 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, or (at your option) any
later version.

XEmacs 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 XEmacs; see the file COPYING.  If not, write to
the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
Boston, MA 02111-1307, USA.  */

/* Synched up with: Not in FSF. */

/* The purpose of this file is so that there is at least one actual
   definition of each inline function.  This is needed under GCC.  The
   reason is that under GCC we declare our inline functions `inline
   extern', which causes the inlined version to get used only for
   inlining, and in other cases to generate an external reference to
   the function.  This is more efficient than declaring our inline
   functions `inline static', which (in many cases) would cause a separate
   version of the function to get inserted into every source file that
   included the corresponding header file.  See internals.texi.

   Some compilers that recognize `inline' may not do the same
   `inline extern' business, so on those we just do `inline static'.
   */

/* Note to maintainers: This file contains a list of all header files
   that use the INLINE macro, either directly, or by using DECLARE_LRECORD.
   i.e. the output of ``grep -l -w 'DECLARE_LRECORD|INLINE_HEADER' *.h'' */

#define DONT_EXTERN_INLINE_HEADER_FUNCTIONS

#include <config.h>
#include "lisp.h"

#include "sysfile.h"

#include "buffer.h"
#include "bytecode.h"
#include "casetab.h"
#include "chartab.h"
#include "device-impl.h"
#include "elhash.h"
#include "events.h"
#include "extents-impl.h"
#include "faces.h"
#include "frame-impl.h"
#include "glyphs.h"
#include "gui.h"
#include "keymap.h"
#include "lstream.h"
#include "objects-impl.h"
#include "opaque.h"
#include "process.h"
#include "rangetab.h"
#include "specifier.h"
#include "symeval.h"
#include "syntax.h"
#include "window.h"

/* If we demand !defined (HAVE_SHLIB) the INLINE_HEADERS aren't instantiated.
   This only shows up in --with-error-checking=types builds AFAIK.
   On Mac OS X 10.3.9 with the Apple toolchain (GCC 3.3) gives a buildtime
   link error (the lrecord error_check functions are undefined).
   Debian GNU/Linux `sid' with GCC 4.0.3 prerelease & binutils 2.16.91 gives
   a runtime link error (the lrecord error_check functions are undefined).
   It is possible that this can be fixed trickily by appropriately defining
   INLINE, or that it should be done in the module itself somehow.  If you
   can do it better or more elegantly, please feel free to consult me.
   --stephen 2005-11-07 */
#if defined (HAVE_LDAP)
#include "../modules/ldap/eldap.h"
#endif

/* We can't ask for !defined (HAVE_SHLIB).  See HAVE_LDAP, above. */
#if defined (HAVE_POSTGRESQL)
#include "../modules/postgresql/postgresql.h"
#endif

#ifdef HAVE_TOOLBARS
#include "toolbar.h"
#endif

#ifdef HAVE_SCROLLBARS
#include "scrollbar.h"
#endif

#ifdef HAVE_DATABASE
#include "database.h"
#endif

#ifdef HAVE_X_WINDOWS
#include "glyphs-x.h"
#ifdef HAVE_XFT
#include "font-mgr.h"
#endif
#endif

#ifdef HAVE_MS_WINDOWS
#include "console-msw.h"
#endif

#ifdef HAVE_GTK
#include "console-gtk.h"
#include "ui-gtk.h"
#endif

#include "file-coding.h"

#ifdef TOOLTALK
#include "tooltalk.h"
#endif