view src/rangetab.c @ 826:6728e641994e

[xemacs-hg @ 2002-05-05 11:30:15 by ben] syntax cache, 8-bit-format, lots of code cleanup README.packages: Update info about --package-path. i.c: Create an inheritable event and pass it on to XEmacs, so that ^C can be handled properly. Intercept ^C and signal the event. "Stop Build" in VC++ now works. bytecomp-runtime.el: Doc string changes. compat.el: Some attempts to redo this to make it truly useful and fix the "multiple versions interacting with each other" problem. Not yet done. Currently doesn't work. files.el: Use with-obsolete-variable to avoid warnings in new revert-buffer code. xemacs.mak: Split up CFLAGS into a version without flags specifying the C library. The problem seems to be that minitar depends on zlib, which depends specifically on libc.lib, not on any of the other C libraries. Unless you compile with libc.lib, you get errors -- specifically, no _errno in the other libraries, which must make it something other than an int. (#### But this doesn't seem to obtain in XEmacs, which also uses zlib, and can be linked with any of the C libraries. Maybe zlib is used differently and doesn't need errno, or maybe XEmacs provides an int errno; ... I don't understand. Makefile.in.in: Fix so that packages are around when testing. abbrev.c, alloc.c, buffer.c, buffer.h, bytecode.c, callint.c, casefiddle.c, casetab.c, casetab.h, charset.h, chartab.c, chartab.h, cmds.c, console-msw.h, console-stream.c, console-x.c, console.c, console.h, data.c, device-msw.c, device.c, device.h, dialog-msw.c, dialog-x.c, dired-msw.c, dired.c, doc.c, doprnt.c, dumper.c, editfns.c, elhash.c, emacs.c, eval.c, event-Xt.c, event-gtk.c, event-msw.c, event-stream.c, events.c, events.h, extents.c, extents.h, faces.c, file-coding.c, file-coding.h, fileio.c, fns.c, font-lock.c, frame-gtk.c, frame-msw.c, frame-x.c, frame.c, frame.h, glade.c, glyphs-gtk.c, glyphs-msw.c, glyphs-msw.h, glyphs-x.c, glyphs.c, glyphs.h, gui-msw.c, gui-x.c, gui.h, gutter.h, hash.h, indent.c, insdel.c, intl-win32.c, intl.c, keymap.c, lisp-disunion.h, lisp-union.h, lisp.h, lread.c, lrecord.h, lstream.c, lstream.h, marker.c, menubar-gtk.c, menubar-msw.c, menubar-x.c, menubar.c, minibuf.c, mule-ccl.c, mule-charset.c, mule-coding.c, mule-wnnfns.c, nas.c, objects-msw.c, objects-x.c, opaque.c, postgresql.c, print.c, process-nt.c, process-unix.c, process.c, process.h, profile.c, rangetab.c, redisplay-gtk.c, redisplay-msw.c, redisplay-output.c, redisplay-x.c, redisplay.c, redisplay.h, regex.c, regex.h, scrollbar-msw.c, search.c, select-x.c, specifier.c, specifier.h, symbols.c, symsinit.h, syntax.c, syntax.h, syswindows.h, tests.c, text.c, text.h, tooltalk.c, ui-byhand.c, ui-gtk.c, unicode.c, win32.c, window.c: Another big Ben patch. -- FUNCTIONALITY CHANGES: add partial support for 8-bit-fixed, 16-bit-fixed, and 32-bit-fixed formats. not quite done yet. (in particular, needs functions to actually convert the buffer.) NOTE: lots of changes to regex.c here. also, many new *_fmt() inline funs that take an Internal_Format argument. redo syntax cache code. make the cache per-buffer; keep the cache valid across calls to functions that use it. also keep it valid across insertions/deletions and extent changes, as much as is possible. eliminate the junky regex-reentrancy code by passing in the relevant lisp info to the regex routines as local vars. add general mechanism in extents code for signalling extent changes. fix numerous problems with the case-table implementation; yoshiki never properly transferred many algorithms from old-style to new-style case tables. redo char tables to support a default argument, so that mapping only occurs over changed args. change many chartab functions to accept Lisp_Object instead of Lisp_Char_Table *. comment out the code in font-lock.c by default, because font-lock.el no longer uses it. we should consider eliminating it entirely. Don't output bell as ^G in console-stream when not a TTY. add -mswindows-termination-handle to interface with i.c, so we can properly kill a build. add more error-checking to buffer/string macros. add some additional buffer_or_string_() funs. -- INTERFACE CHANGES AFFECTING MORE CODE: switch the arguments of write_c_string and friends to be consistent with write_fmt_string, which must have printcharfun first. change BI_* macros to BYTE_* for increased clarity; similarly for bi_* local vars. change VOID_TO_LISP to be a one-argument function. eliminate no-longer-needed CVOID_TO_LISP. -- char/string macro changes: rename MAKE_CHAR() to make_emchar() for slightly less confusion with make_char(). (The former generates an Emchar, the latter a Lisp object. Conceivably we should rename make_char() -> wrap_char() and similarly for make_int(), make_float().) Similar changes for other *CHAR* macros -- we now consistently use names with `emchar' whenever we are working with Emchars. Any remaining name with just `char' always refers to a Lisp object. rename macros with XSTRING_* to string_* except for those that reference actual fields in the Lisp_String object, following conventions used elsewhere. rename set_string_{data,length} macros (the only ones to work with a Lisp_String_* instead of a Lisp_Object) to set_lispstringp_* to make the difference clear. try to be consistent about caps vs. lowercase in macro/inline-fun names for chars and such, which wasn't the case before. we now reserve caps either for XFOO_ macros that reference object fields (e.g. XSTRING_DATA) or for things that have non-function semantics, e.g. directly modifying an arg (BREAKUP_EMCHAR) or evaluating an arg (any arg) more than once. otherwise, use lowercase. here is a summary of most of the macros/inline funs changed by all of the above changes: BYTE_*_P -> byte_*_p XSTRING_BYTE -> string_byte set_string_data/length -> set_lispstringp_data/length XSTRING_CHAR_LENGTH -> string_char_length XSTRING_CHAR -> string_emchar INTBYTE_FIRST_BYTE_P -> intbyte_first_byte_p INTBYTE_LEADING_BYTE_P -> intbyte_leading_byte_p charptr_copy_char -> charptr_copy_emchar LEADING_BYTE_* -> leading_byte_* CHAR_* -> EMCHAR_* *_CHAR_* -> *_EMCHAR_* *_CHAR -> *_EMCHAR CHARSET_BY_ -> charset_by_* BYTE_SHIFT_JIS* -> byte_shift_jis* BYTE_BIG5* -> byte_big5* REP_BYTES_BY_FIRST_BYTE -> rep_bytes_by_first_byte char_to_unicode -> emchar_to_unicode valid_char_p -> valid_emchar_p Change intbyte_strcmp -> qxestrcmp_c (duplicated functionality). -- INTERFACE CHANGES AFFECTING LESS CODE: use DECLARE_INLINE_HEADER in various places. remove '#ifdef emacs' from XEmacs-only files. eliminate CHAR_TABLE_VALUE(), which duplicated the functionality of get_char_table(). add BUFFER_TEXT_LOOP to simplify iterations over buffer text. define typedefs for signed and unsigned types of fixed sizes (INT_32_BIT, UINT_32_BIT, etc.). create ALIGN_FOR_TYPE as a higher-level interface onto ALIGN_SIZE; fix code to use it. add charptr_emchar_len to return the text length of the character pointed to by a ptr; use it in place of charcount_to_bytecount(..., 1). add emchar_len to return the text length of a given character. add types Bytexpos and Charxpos to generalize Bytebpos/Bytecount and Charbpos/Charcount, in code (particularly, the extents code and redisplay code) that works with either kind of index. rename redisplay struct params with names such as `charbpos' to e.g. `charpos' when they are e.g. a Charxpos, not a Charbpos. eliminate xxDEFUN in place of DEFUN; no longer necessary with changes awhile back to doc.c. split up big ugly combined list of EXFUNs in lisp.h on a file-by-file basis, since other prototypes are similarly split. rewrite some "*_UNSAFE" macros as inline funs and eliminate the _UNSAFE suffix. move most string code from lisp.h to text.h; the string code and text.h code is now intertwined in such a fashion that they need to be in the same place and partially interleaved. (you can't create forward references for inline funs) automated/lisp-tests.el, automated/symbol-tests.el, automated/test-harness.el: Fix test harness to output FAIL messages to stderr when in batch mode. Fix up some problems in lisp-tests/symbol-tests that were causing spurious failures.
author ben
date Sun, 05 May 2002 11:33:57 +0000
parents a5954632b187
children c925bacdda60
line wrap: on
line source

/* XEmacs routines to deal with range tables.
   Copyright (C) 1995 Sun Microsystems, Inc.
   Copyright (C) 1995, 2002 Ben Wing.

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. */

/* Written by Ben Wing, August 1995. */

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

Lisp_Object Qrange_tablep;
Lisp_Object Qrange_table;


/************************************************************************/
/*                            Range table object                        */
/************************************************************************/

/* We use a sorted array of ranges.

   #### We should be using the gap array stuff from extents.c.  This
   is not hard but just requires moving that stuff out of that file. */

static Lisp_Object
mark_range_table (Lisp_Object obj)
{
  Lisp_Range_Table *rt = XRANGE_TABLE (obj);
  int i;

  for (i = 0; i < Dynarr_length (rt->entries); i++)
    mark_object (Dynarr_at (rt->entries, i).val);
  return Qnil;
}

static void
print_range_table (Lisp_Object obj, Lisp_Object printcharfun, int escapeflag)
{
  Lisp_Range_Table *rt = XRANGE_TABLE (obj);
  int i;

  write_c_string (printcharfun, "#s(range-table data (");
  for (i = 0; i < Dynarr_length (rt->entries); i++)
    {
      struct range_table_entry *rte = Dynarr_atp (rt->entries, i);
      if (i > 0)
	write_c_string (printcharfun, " ");
      if (rte->first == rte->last)
	write_fmt_string (printcharfun, "%ld ", (long) (rte->first));
      else
	write_fmt_string (printcharfun, "(%ld %ld) ", (long) (rte->first),
			  (long) (rte->last));
      print_internal (rte->val, printcharfun, 1);
    }
  write_c_string (printcharfun, "))");
}

static int
range_table_equal (Lisp_Object obj1, Lisp_Object obj2, int depth)
{
  Lisp_Range_Table *rt1 = XRANGE_TABLE (obj1);
  Lisp_Range_Table *rt2 = XRANGE_TABLE (obj2);
  int i;

  if (Dynarr_length (rt1->entries) != Dynarr_length (rt2->entries))
    return 0;

  for (i = 0; i < Dynarr_length (rt1->entries); i++)
    {
      struct range_table_entry *rte1 = Dynarr_atp (rt1->entries, i);
      struct range_table_entry *rte2 = Dynarr_atp (rt2->entries, i);

      if (rte1->first != rte2->first
	  || rte1->last != rte2->last
	  || !internal_equal (rte1->val, rte2->val, depth + 1))
	return 0;
    }

  return 1;
}

static unsigned long
range_table_entry_hash (struct range_table_entry *rte, int depth)
{
  return HASH3 (rte->first, rte->last, internal_hash (rte->val, depth + 1));
}

static unsigned long
range_table_hash (Lisp_Object obj, int depth)
{
  Lisp_Range_Table *rt = XRANGE_TABLE (obj);
  int i;
  int size = Dynarr_length (rt->entries);
  unsigned long hash = size;

  /* approach based on internal_array_hash(). */
  if (size <= 5)
    {
      for (i = 0; i < size; i++)
	hash = HASH2 (hash,
		      range_table_entry_hash (Dynarr_atp (rt->entries, i),
					      depth));
      return hash;
    }

  /* just pick five elements scattered throughout the array.
     A slightly better approach would be to offset by some
     noise factor from the points chosen below. */
  for (i = 0; i < 5; i++)
    hash = HASH2 (hash, range_table_entry_hash (Dynarr_atp (rt->entries,
							    i*size/5),
						depth));
  return hash;
}

static const struct lrecord_description rte_description_1[] = {
  { XD_LISP_OBJECT, offsetof (range_table_entry, val) },
  { XD_END }
};

static const struct struct_description rte_description = {
  sizeof (range_table_entry),
  rte_description_1
};

static const struct lrecord_description rted_description_1[] = {
  XD_DYNARR_DESC (range_table_entry_dynarr, &rte_description),
  { XD_END }
};

static const struct struct_description rted_description = {
  sizeof (range_table_entry_dynarr),
  rted_description_1
};

static const struct lrecord_description range_table_description[] = {
  { XD_STRUCT_PTR,  offsetof (Lisp_Range_Table, entries),  1, &rted_description },
  { XD_END }
};

DEFINE_LRECORD_IMPLEMENTATION ("range-table", range_table,
                               mark_range_table, print_range_table, 0,
			       range_table_equal, range_table_hash,
			       range_table_description,
			       Lisp_Range_Table);

/************************************************************************/
/*                        Range table operations                        */
/************************************************************************/

#ifdef ERROR_CHECK_STRUCTURES

static void
verify_range_table (Lisp_Range_Table *rt)
{
  int i;

  for (i = 0; i < Dynarr_length (rt->entries); i++)
    {
      struct range_table_entry *rte = Dynarr_atp (rt->entries, i);
      assert (rte->last >= rte->first);
      if (i > 0)
	assert (Dynarr_at (rt->entries, i - 1).last < rte->first);
    }
}

#else

#define verify_range_table(rt)

#endif

/* Look up in a range table without the Dynarr wrapper.
   Used also by the unified range table format. */

static Lisp_Object
get_range_table (EMACS_INT pos, int nentries, struct range_table_entry *tab,
		 Lisp_Object default_)
{
  int left = 0, right = nentries;

  /* binary search for the entry.  Based on similar code in
     extent_list_locate(). */
  while (left != right)
    {
      /* RIGHT might not point to a valid entry (i.e. it's at the end
	 of the list), so NEWPOS must round down. */
      int newpos = (left + right) >> 1;
      struct range_table_entry *entry = tab + newpos;
      if (pos > entry->last)
	left = newpos+1;
      else if (pos < entry->first)
	right = newpos;
      else
	return entry->val;
    }

  return default_;
}

DEFUN ("range-table-p", Frange_table_p, 1, 1, 0, /*
Return non-nil if OBJECT is a range table.
*/
       (object))
{
  return RANGE_TABLEP (object) ? Qt : Qnil;
}

DEFUN ("make-range-table", Fmake_range_table, 0, 0, 0, /*
Return a new, empty range table.
You can manipulate it using `put-range-table', `get-range-table',
`remove-range-table', and `clear-range-table'.
*/
       ())
{
  Lisp_Range_Table *rt = alloc_lcrecord_type (Lisp_Range_Table,
					      &lrecord_range_table);
  rt->entries = Dynarr_new (range_table_entry);
  return wrap_range_table (rt);
}

DEFUN ("copy-range-table", Fcopy_range_table, 1, 1, 0, /*
Return a new range table which is a copy of RANGE-TABLE.
It will contain the same values for the same ranges as RANGE-TABLE.
The values will not themselves be copied.
*/
       (range_table))
{
  Lisp_Range_Table *rt, *rtnew;

  CHECK_RANGE_TABLE (range_table);
  rt = XRANGE_TABLE (range_table);

  rtnew = alloc_lcrecord_type (Lisp_Range_Table, &lrecord_range_table);
  rtnew->entries = Dynarr_new (range_table_entry);

  Dynarr_add_many (rtnew->entries, Dynarr_atp (rt->entries, 0),
		   Dynarr_length (rt->entries));
  return wrap_range_table (rtnew);
}

DEFUN ("get-range-table", Fget_range_table, 2, 3, 0, /*
Find value for position POS in RANGE-TABLE.
If there is no corresponding value, return DEFAULT (defaults to nil).
*/
       (pos, range_table, default_))
{
  Lisp_Range_Table *rt;

  CHECK_RANGE_TABLE (range_table);
  rt = XRANGE_TABLE (range_table);

  CHECK_INT_COERCE_CHAR (pos);

  return get_range_table (XINT (pos), Dynarr_length (rt->entries),
			  Dynarr_atp (rt->entries, 0), default_);
}

void
put_range_table (Lisp_Object table, EMACS_INT first,
		 EMACS_INT last, Lisp_Object val)
{
  int i;
  int insert_me_here = -1;
  Lisp_Range_Table *rt = XRANGE_TABLE (table);

  /* Now insert in the proper place.  This gets tricky because
     we may be overlapping one or more existing ranges and need
     to fix them up. */

  /* First delete all sections of any existing ranges that overlap
     the new range. */
  for (i = 0; i < Dynarr_length (rt->entries); i++)
    {
      struct range_table_entry *entry = Dynarr_atp (rt->entries, i);
      /* We insert before the first range that begins at or after the
	 new range. */
      if (entry->first >= first && insert_me_here < 0)
	insert_me_here = i;
      if (entry->last < first)
	/* completely before the new range. */
	continue;
      if (entry->first > last)
	/* completely after the new range.  No more possibilities of
	   finding overlapping ranges. */
	break;
      if (entry->first < first && entry->last <= last)
	{
	  /* looks like:

	                 [ NEW ]
	         [ EXISTING ]

	   */
	  /* truncate the end off of it. */
	  entry->last = first - 1;
	}
      else if (entry->first < first && entry->last > last)
	/* looks like:

	         [ NEW ]
	       [ EXISTING ]

	 */
	/* need to split this one in two. */
	{
	  struct range_table_entry insert_me_too;

	  insert_me_too.first = last + 1;
	  insert_me_too.last = entry->last;
	  insert_me_too.val = entry->val;
	  entry->last = first - 1;
	  Dynarr_insert_many (rt->entries, &insert_me_too, 1, i + 1);
	}
      else if (entry->last > last)
	{
	  /* looks like:

	       [ NEW ]
	         [ EXISTING ]

	   */
	  /* truncate the start off of it. */
	  entry->first = last + 1;
	}
      else
	{
	  /* existing is entirely within new. */
	  Dynarr_delete_many (rt->entries, i, 1);
	  i--; /* back up since everything shifted one to the left. */
	}
    }

  /* Someone asked us to delete the range, not insert it. */
  if (UNBOUNDP (val))
    return;

  /* Now insert the new entry, maybe at the end. */

  if (insert_me_here < 0)
    insert_me_here = i;

  {
    struct range_table_entry insert_me;

    insert_me.first = first;
    insert_me.last = last;
    insert_me.val = val;

    Dynarr_insert_many (rt->entries, &insert_me, 1, insert_me_here);
  }

  /* Now see if we can combine this entry with adjacent ones just
     before or after. */

  if (insert_me_here > 0)
    {
      struct range_table_entry *entry = Dynarr_atp (rt->entries,
						    insert_me_here - 1);
      if (EQ (val, entry->val) && entry->last == first - 1)
	{
	  entry->last = last;
	  Dynarr_delete_many (rt->entries, insert_me_here, 1);
	  insert_me_here--;
	  /* We have morphed into a larger range.  Update our records
	     in case we also combine with the one after. */
	  first = entry->first;
	}
    }

  if (insert_me_here < Dynarr_length (rt->entries) - 1)
    {
      struct range_table_entry *entry = Dynarr_atp (rt->entries,
						    insert_me_here + 1);
      if (EQ (val, entry->val) && entry->first == last + 1)
	{
	  entry->first = first;
	  Dynarr_delete_many (rt->entries, insert_me_here, 1);
	}
    }
}

DEFUN ("put-range-table", Fput_range_table, 4, 4, 0, /*
Set the value for range (START, END) to be VALUE in RANGE-TABLE.
*/
       (start, end, value, range_table))
{
  EMACS_INT first, last;

  CHECK_RANGE_TABLE (range_table);
  CHECK_INT_COERCE_CHAR (start);
  first = XINT (start);
  CHECK_INT_COERCE_CHAR (end);
  last = XINT (end);
  if (first > last)
    invalid_argument_2 ("start must be <= end", start, end);

  put_range_table (range_table, first, last, value);
  verify_range_table (XRANGE_TABLE (range_table));
  return Qnil;
}

DEFUN ("remove-range-table", Fremove_range_table, 3, 3, 0, /*
Remove the value for range (START, END) in RANGE-TABLE.
*/
       (start, end, range_table))
{
  return Fput_range_table (start, end, Qunbound, range_table);
}

DEFUN ("clear-range-table", Fclear_range_table, 1, 1, 0, /*
Flush RANGE-TABLE.
*/
       (range_table))
{
  CHECK_RANGE_TABLE (range_table);
  Dynarr_reset (XRANGE_TABLE (range_table)->entries);
  return Qnil;
}

DEFUN ("map-range-table", Fmap_range_table, 2, 2, 0, /*
Map FUNCTION over entries in RANGE-TABLE, calling it with three args,
the beginning and end of the range and the corresponding value.

Results are guaranteed to be correct (i.e. each entry processed
exactly once) if FUNCTION modifies or deletes the current entry
\(i.e. passes the current range to `put-range-table' or
`remove-range-table'), but not otherwise.
*/
       (function, range_table))
{
  Lisp_Range_Table *rt;
  int i;

  CHECK_RANGE_TABLE (range_table);
  CHECK_FUNCTION (function);

  rt = XRANGE_TABLE (range_table);

  /* Do not "optimize" by pulling out the length computation below!
     FUNCTION may have changed the table. */
  for (i = 0; i < Dynarr_length (rt->entries); i++)
    {
      struct range_table_entry *entry = Dynarr_atp (rt->entries, i);
      EMACS_INT first, last;
      Lisp_Object args[4];
      int oldlen;

    again:
      first = entry->first;
      last = entry->last;
      oldlen = Dynarr_length (rt->entries);
      args[0] = function;
      args[1] = make_int (first);
      args[2] = make_int (last);
      args[3] = entry->val;
      Ffuncall (countof (args), args);
      /* Has FUNCTION removed the entry? */
      if (oldlen > Dynarr_length (rt->entries)
	  && i < Dynarr_length (rt->entries)
	  && (first != entry->first || last != entry->last))
	goto again;
      }

  return Qnil;
}


/************************************************************************/
/*                         Range table read syntax                      */
/************************************************************************/

static int
rangetab_data_validate (Lisp_Object keyword, Lisp_Object value,
			Error_Behavior errb)
{
  Lisp_Object rest;

  /* #### should deal with errb */
  EXTERNAL_LIST_LOOP (rest, value)
    {
      Lisp_Object range = XCAR (rest);
      rest = XCDR (rest);
      if (!CONSP (rest))
	sferror ("Invalid list format", value);
      if (!INTP (range) && !CHARP (range)
	  && !(CONSP (range) && CONSP (XCDR (range))
	       && NILP (XCDR (XCDR (range)))
	       && (INTP (XCAR (range)) || CHARP (XCAR (range)))
	       && (INTP (XCAR (XCDR (range))) || CHARP (XCAR (XCDR (range))))))
	sferror ("Invalid range format", range);
    }

  return 1;
}

static Lisp_Object
rangetab_instantiate (Lisp_Object data)
{
  Lisp_Object rangetab = Fmake_range_table ();

  if (!NILP (data))
    {
      data = Fcar (Fcdr (data)); /* skip over 'data keyword */
      while (!NILP (data))
	{
	  Lisp_Object range = Fcar (data);
	  Lisp_Object val = Fcar (Fcdr (data));

	  data = Fcdr (Fcdr (data));
	  if (CONSP (range))
	    Fput_range_table (Fcar (range), Fcar (Fcdr (range)), val,
			      rangetab);
	  else
	    Fput_range_table (range, range, val, rangetab);
	}
    }

  return rangetab;
}


/************************************************************************/
/*                         Unified range tables                         */
/************************************************************************/

/* A "unified range table" is a format for storing range tables
   as contiguous blocks of memory.  This is used by the regexp
   code, which needs to use range tables to properly handle []
   constructs in the presence of extended characters but wants to
   store an entire compiled pattern as a contiguous block of memory.

   Unified range tables are designed so that they can be placed
   at an arbitrary (possibly mis-aligned) place in memory.
   (Dealing with alignment is a pain in the ass.)

   WARNING: No provisions for garbage collection are currently made.
   This means that there must not be any Lisp objects in a unified
   range table that need to be marked for garbage collection.
   Good candidates for objects that can go into a range table are

   -- numbers and characters (do not need to be marked)
   -- nil, t (marked elsewhere)
   -- charsets and coding systems (automatically marked because
				   they are in a marked list,
				   and can't be removed)

   Good but slightly less so:

   -- symbols (could be uninterned, but that is not likely)

   Somewhat less good:

   -- buffers, frames, devices (could get deleted)


   It is expected that you work with range tables in the normal
   format and then convert to unified format when you are done
   making modifications.  As such, no functions are provided
   for modifying a unified range table.  The only operations
   you can do to unified range tables are

   -- look up a value
   -- retrieve all the ranges in an iterative fashion

*/

/* The format of a unified range table is as follows:

   -- The first byte contains the number of bytes to skip to find the
      actual start of the table.  This deals with alignment constraints,
      since the table might want to go at any arbitrary place in memory.
   -- The next three bytes contain the number of bytes to skip (from the
      *first* byte) to find the stuff after the table.  It's stored in
      little-endian format because that's how God intended things.  We don't
      necessarily start the stuff at the very end of the table because
      we want to have at least ALIGNOF (EMACS_INT) extra space in case
      we have to move the range table around. (It appears that some
      architectures don't maintain alignment when reallocing.)
   -- At the prescribed offset is a struct unified_range_table, containing
      some number of `struct range_table_entry' entries. */

struct unified_range_table
{
  int nentries;
  struct range_table_entry first;
};

/* Return size in bytes needed to store the data in a range table. */

int
unified_range_table_bytes_needed (Lisp_Object rangetab)
{
  return (sizeof (struct range_table_entry) *
	  (Dynarr_length (XRANGE_TABLE (rangetab)->entries) - 1) +
	  sizeof (struct unified_range_table) +
	  /* ALIGNOF a struct may be too big. */
	  /* We have four bytes for the size numbers, and an extra
	     four or eight bytes for making sure we get the alignment
	     OK. */
	  ALIGNOF (EMACS_INT) + 4);
}

/* Convert a range table into unified format and store in DEST,
   which must be able to hold the number of bytes returned by
   range_table_bytes_needed(). */

void
unified_range_table_copy_data (Lisp_Object rangetab, void *dest)
{
  /* We cast to the above structure rather than just casting to
     char * and adding sizeof(int), because that will lead to
     mis-aligned data on the Alpha machines. */
  struct unified_range_table *un;
  range_table_entry_dynarr *rted = XRANGE_TABLE (rangetab)->entries;
  int total_needed = unified_range_table_bytes_needed (rangetab);
  void *new_dest = ALIGN_PTR ((char *) dest + 4, EMACS_INT);

  * (char *) dest = (char) ((char *) new_dest - (char *) dest);
  * ((unsigned char *) dest + 1) = total_needed & 0xFF;
  total_needed >>= 8;
  * ((unsigned char *) dest + 2) = total_needed & 0xFF;
  total_needed >>= 8;
  * ((unsigned char *) dest + 3) = total_needed & 0xFF;
  un = (struct unified_range_table *) new_dest;
  un->nentries = Dynarr_length (rted);
  memcpy (&un->first, Dynarr_atp (rted, 0),
	  sizeof (struct range_table_entry) * Dynarr_length (rted));
}

/* Return number of bytes actually used by a unified range table. */

int
unified_range_table_bytes_used (void *unrangetab)
{
  return ((* ((unsigned char *) unrangetab + 1))
	  + ((* ((unsigned char *) unrangetab + 2)) << 8)
	  + ((* ((unsigned char *) unrangetab + 3)) << 16));
}

/* Make sure the table is aligned, and move it around if it's not. */
static void
align_the_damn_table (void *unrangetab)
{
  void *cur_dest = (char *) unrangetab + * (char *) unrangetab;
  if (cur_dest != ALIGN_PTR (cur_dest, EMACS_INT))
    {
      int count = (unified_range_table_bytes_used (unrangetab) - 4
		   - ALIGNOF (EMACS_INT));
      /* Find the proper location, just like above. */
      void *new_dest = ALIGN_PTR ((char *) unrangetab + 4, EMACS_INT);
      /* memmove() works in the presence of overlapping data. */
      memmove (new_dest, cur_dest, count);
      * (char *) unrangetab = (char) ((char *) new_dest - (char *) unrangetab);
    }
}

/* Look up a value in a unified range table. */

Lisp_Object
unified_range_table_lookup (void *unrangetab, EMACS_INT pos,
			    Lisp_Object default_)
{
  void *new_dest;
  struct unified_range_table *un;

  align_the_damn_table (unrangetab);
  new_dest = (char *) unrangetab + * (char *) unrangetab;
  un = (struct unified_range_table *) new_dest;

  return get_range_table (pos, un->nentries, &un->first, default_);
}

/* Return number of entries in a unified range table. */

int
unified_range_table_nentries (void *unrangetab)
{
  void *new_dest;
  struct unified_range_table *un;

  align_the_damn_table (unrangetab);
  new_dest = (char *) unrangetab + * (char *) unrangetab;
  un = (struct unified_range_table *) new_dest;
  return un->nentries;
}

/* Return the OFFSETth range (counting from 0) in UNRANGETAB. */
void
unified_range_table_get_range (void *unrangetab, int offset,
			       EMACS_INT *min, EMACS_INT *max,
			       Lisp_Object *val)
{
  void *new_dest;
  struct unified_range_table *un;
  struct range_table_entry *tab;

  align_the_damn_table (unrangetab);
  new_dest = (char *) unrangetab + * (char *) unrangetab;
  un = (struct unified_range_table *) new_dest;

  assert (offset >= 0 && offset < un->nentries);
  tab = (&un->first) + offset;
  *min = tab->first;
  *max = tab->last;
  *val = tab->val;
}


/************************************************************************/
/*                            Initialization                            */
/************************************************************************/

void
syms_of_rangetab (void)
{
  INIT_LRECORD_IMPLEMENTATION (range_table);

  DEFSYMBOL_MULTIWORD_PREDICATE (Qrange_tablep);
  DEFSYMBOL (Qrange_table);

  DEFSUBR (Frange_table_p);
  DEFSUBR (Fmake_range_table);
  DEFSUBR (Fcopy_range_table);
  DEFSUBR (Fget_range_table);
  DEFSUBR (Fput_range_table);
  DEFSUBR (Fremove_range_table);
  DEFSUBR (Fclear_range_table);
  DEFSUBR (Fmap_range_table);
}

void
structure_type_create_rangetab (void)
{
  struct structure_type *st;

  st = define_structure_type (Qrange_table, 0, rangetab_instantiate);

  define_structure_type_keyword (st, Qdata, rangetab_data_validate);
}