Mercurial > hg > xemacs-beta

comparison src/ChangeLog @ 734:8bd30fae1bce

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
[xemacs-hg @ 2002-01-25 16:46:24 by stephent] Per patch <87665q9yfh.fsf@tleepslib.sk.tsukuba.ac.jp>.
author stephent
date Fri, 25 Jan 2002 16:46:26 +0000
parents b1f74adcc1ff
children 2e5e2ccbeed2
comparison
equal deleted inserted replaced
733:b1f74adcc1ff 734:8bd30fae1bce
1 2002-01-26 Stephen J. Turnbull <stephen@xemacs.org>
2
3 * ChangeLog (2001-09-19 Ben Wing): Entry dosed with Dramamine.
4 * README.integral-types: New file excised from ChangeLog.
5
1 2002-01-19 Jan Vroonhof <jan@xemacs.org> 6 2002-01-19 Jan Vroonhof <jan@xemacs.org>
2 7
3 * cmdloop.c: New variable Qdisabled_command_loop. 8 * cmdloop.c: New variable Qdisabled_command_loop.
4 Move Vdisabled_command_loop declaration here. 9 Move Vdisabled_command_loop declaration here.
5 * cmdloop.c (syms_of_cmdloop): Declare Qdisabled_command_loop. 10 * cmdloop.c (syms_of_cmdloop): Declare Qdisabled_command_loop.
1178 * window.c (Fcurrent_pixel_column): 1183 * window.c (Fcurrent_pixel_column):
1179 * window.h (struct window): 1184 * window.h (struct window):
1180 1185
1181 The great integral types renaming. 1186 The great integral types renaming.
1182 1187
1183 The purpose of this is to rationalize the names used for various 1188 Rationalize the names used for various integral types, so that
1184 integral types, so that they match their intended uses and follow 1189 they match their intended uses and follow consist conventions, and
1185 consist conventions, and eliminate types that were not semantically 1190 eliminate types that were not semantically different from each
1186 different from each other. 1191 other. The following global substitutions were made:
1187 1192
1188 The conventions are: 1193 Memory_Count Bytecount
1189 1194 Lstream_Data_Count Bytecount
1190 -- All integral types that measure quantities of anything are 1195 Element_Count Elemcount
1191 signed. Some people disagree vociferously with this, but their 1196 Hash_Code Hashcode
1192 arguments are mostly theoretical, and are vastly outweighed by 1197 extcount bytecount
1193 the practical headaches of mixing signed and unsigned values, 1198 bufpos charbpos
1194 and more importantly by the far increased likelihood of 1199 bytind bytebpos
1195 inadvertent bugs: Because of the broken "viral" nature of 1200 memind membpos
1196 unsigned quantities in C (operations involving mixed 1201 bufbyte intbyte
1197 signed/unsigned are done unsigned, when exactly the opposite is 1202 Extcount Bytecount
1198 nearly always wanted), even a single error in declaring a 1203 Bufpos Charbpos
1199 quantity unsigned that should be signed, or even the even more 1204 Bytind Bytebpos
1200 subtle error of comparing signed and unsigned values and 1205 Memind Membpos
1201 forgetting the necessary cast, can be catastrophic, as 1206 Bufbyte Intbyte
1202 comparisons will yield wrong results. -Wsign-compare is turned 1207 EXTCOUNT BYTECOUNT
1203 on specifically to catch this, but this tends to result in a 1208 BUFPOS CHARBPOS
1204 great number of warnings when mixing signed and unsigned, and 1209 BYTIND BYTEBPOS
1205 the casts are annoying. More has been written on this 1210 MEMIND MEMBPOS
1206 elsewhere. 1211 BUFBYTE INTBYTE
1207 1212 MEMORY_COUNT BYTECOUNT
1208 -- All such quantity types just mentioned boil down to EMACS_INT, 1213 LSTREAM_DATA_COUNT BYTECOUNT
1209 which is 32 bits on 32-bit machines and 64 bits on 64-bit 1214 ELEMENT_COUNT ELEMCOUNT
1210 machines. This is guaranteed to be the same size as Lisp 1215 HASH_CODE HASHCODE
1211 objects of type `int', and (as far as I can tell) of size_t 1216
1212 (unsigned!) and ssize_t. The only type below that is not an 1217 This resulted in some duplications.
1213 EMACS_INT is Hashcode, which is an unsigned value of the same 1218
1214 size as EMACS_INT. 1219 * lisp.h: removed duplicate declarations of Bytecount.
1215 1220
1216 -- Type names should be relatively short (no more than 10 1221 * lstream.h: removed duplicate declaration of Bytecount.
1217 characters or so), with the first letter capitalized and no 1222 Rewrote the comment about this type.
1218 underscores if they can at all be avoided. 1223
1219 1224 * dumper.c: remove duplicate case tag XD_BYTECOUNT, and the
1220 -- "count" == a zero-based measurement of some quantity. Includes 1225 accompanying duplicate code, from 4 switchs tatements.
1221 sizes, offsets, and indexes. 1226
1222 1227 See README.integral-types in this directory for more details.
1223 -- "bpos" == a one-based measurement of a position in a buffer.
1224 "Charbpos" and "Bytebpos" count text in the buffer, rather than
1225 bytes in memory; thus Bytebpos does not directly correspond to
1226 the memory representation. Use "Membpos" for this.
1227
1228 -- "Char" refers to internal-format characters, not to the C type
1229 "char", which is really a byte.
1230
1231 -- For the actual name changes, see the script below.
1232
1233 I ran the following script to do the conversion. (NOTE: This script
1234 is idempotent. You can safely run it multiple times and it will
1235 not screw up previous results -- in fact, it will do nothing if
1236 nothing has changed. Thus, it can be run repeatedly as necessary
1237 to handle patches coming in from old workspaces, or old branches.)
1238 There are two tags, just before and just after the change:
1239 `pre-integral-type-rename' and `post-integral-type-rename'. When
1240 merging code from the main trunk into a branch, the best thing to
1241 do is first merge up to `pre-integral-type-rename', then apply the
1242 script and associated changes, then merge from
1243 `post-integral-type-change' to the present. (Alternatively, just do
1244 the merging in one operation; but you may then have a lot of
1245 conflicts needing to be resolved by hand.)
1246
1247 Script `fixtypes.sh' follows:
1248
1249
1250 ----------------------------------- cut ------------------------------------
1251 files="*.[ch] s/*.h m/*.h config.h.in ../configure.in Makefile.in.in ../lib-src/*.[ch] ../lwlib/*.[ch]"
1252 gr Memory_Count Bytecount $files
1253 gr Lstream_Data_Count Bytecount $files
1254 gr Element_Count Elemcount $files
1255 gr Hash_Code Hashcode $files
1256 gr extcount bytecount $files
1257 gr bufpos charbpos $files
1258 gr bytind bytebpos $files
1259 gr memind membpos $files
1260 gr bufbyte intbyte $files
1261 gr Extcount Bytecount $files
1262 gr Bufpos Charbpos $files
1263 gr Bytind Bytebpos $files
1264 gr Memind Membpos $files
1265 gr Bufbyte Intbyte $files
1266 gr EXTCOUNT BYTECOUNT $files
1267 gr BUFPOS CHARBPOS $files
1268 gr BYTIND BYTEBPOS $files
1269 gr MEMIND MEMBPOS $files
1270 gr BUFBYTE INTBYTE $files
1271 gr MEMORY_COUNT BYTECOUNT $files
1272 gr LSTREAM_DATA_COUNT BYTECOUNT $files
1273 gr ELEMENT_COUNT ELEMCOUNT $files
1274 gr HASH_CODE HASHCODE $files
1275 ----------------------------------- cut ------------------------------------
1276
1277
1278 `fixtypes.sh' is a Bourne-shell script; it uses 'gr':
1279
1280
1281 ----------------------------------- cut ------------------------------------
1282 #!/bin/sh
1283
1284 # Usage is like this:
1285
1286 # gr FROM TO FILES ...
1287
1288 # globally replace FROM with TO in FILES. FROM and TO are regular expressions.
1289 # backup files are stored in the `backup' directory.
1290 from="$1"
1291 to="$2"
1292 shift 2
1293 echo ${1+"$@"} | xargs global-replace "s/$from/$to/g"
1294 ----------------------------------- cut ------------------------------------
1295
1296
1297 `gr' in turn uses a Perl script to do its real work,
1298 `global-replace', which follows:
1299
1300
1301 ----------------------------------- cut ------------------------------------
1302 : #-*- Perl -*-
1303
1304 ### global-modify --- modify the contents of a file by a Perl expression
1305
1306 ## Copyright (C) 1999 Martin Buchholz.
1307 ## Copyright (C) 2001 Ben Wing.
1308
1309 ## Authors: Martin Buchholz <martin@xemacs.org>, Ben Wing <ben@xemacs.org>
1310 ## Maintainer: Ben Wing <ben@xemacs.org>
1311 ## Current Version: 1.0, May 5, 2001
1312
1313 # This program is free software; you can redistribute it and/or modify
1314 # it under the terms of the GNU General Public License as published by
1315 # the Free Software Foundation; either version 2, or (at your option)
1316 # any later version.
1317 #
1318 # This program is distributed in the hope that it will be useful, but
1319 # WITHOUT ANY WARRANTY; without even the implied warranty of
1320 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
1321 # General Public License for more details.
1322 #
1323 # You should have received a copy of the GNU General Public License
1324 # along with XEmacs; see the file COPYING. If not, write to the Free
1325 # Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA
1326 # 02111-1307, USA.
1327
1328 eval 'exec perl -w -S $0 ${1+"$@"}'
1329 if 0;
1330
1331 use strict;
1332 use FileHandle;
1333 use Carp;
1334 use Getopt::Long;
1335 use File::Basename;
1336
1337 (my $myName = $0) =~ s@.*/@@; my $usage="
1338 Usage: $myName [--help] [--backup-dir=DIR] [--line-mode] [--hunk-mode]
1339 PERLEXPR FILE ...
1340
1341 Globally modify a file, either line by line or in one big hunk.
1342
1343 Typical usage is like this:
1344
1345 [with GNU print, GNU xargs: guaranteed to handle spaces, quotes, etc.
1346 in file names]
1347
1348 find . -name '*.[ch]' -print0 | xargs -0 $0 's/\bCONST\b/const/g'\n
1349
1350 [with non-GNU print, xargs]
1351
1352 find . -name '*.[ch]' -print | xargs $0 's/\bCONST\b/const/g'\n
1353
1354
1355 The file is read in, either line by line (with --line-mode specified)
1356 or in one big hunk (with --hunk-mode specified; it's the default), and
1357 the Perl expression is then evalled with \$_ set to the line or hunk of
1358 text, including the terminating newline if there is one. It should
1359 destructively modify the value there, storing the changed result in \$_.
1360
1361 Files in which any modifications are made are backed up to the directory
1362 specified using --backup-dir, or to `backup' by default. To disable this,
1363 use --backup-dir= with no argument.
1364
1365 Hunk mode is the default because it is MUCH MUCH faster than line-by-line.
1366 Use line-by-line only when it matters, e.g. you want to do a replacement
1367 only once per line (the default without the `g' argument). Conversely,
1368 when using hunk mode, *ALWAYS* use `g'; otherwise, you will only make one
1369 replacement in the entire file!
1370 ";
1371
1372 my %options = ();
1373 $Getopt::Long::ignorecase = 0;
1374 &GetOptions (
1375 \%options,
1376 'help', 'backup-dir=s', 'line-mode', 'hunk-mode',
1377 );
1378
1379
1380 die $usage if $options{"help"} or @ARGV <= 1;
1381 my $code = shift;
1382
1383 die $usage if grep (-d || ! -w, @ARGV);
1384
1385 sub SafeOpen {
1386 open ((my $fh = new FileHandle), $_[0]);
1387 confess "Can't open $_[0]: $!" if ! defined $fh;
1388 return $fh;
1389 }
1390
1391 sub SafeClose {
1392 close $_[0] or confess "Can't close $_[0]: $!";
1393 }
1394
1395 sub FileContents {
1396 my $fh = SafeOpen ("< $_[0]");
1397 my $olddollarslash = $/;
1398 local $/ = undef;
1399 my $contents = <$fh>;
1400 $/ = $olddollarslash;
1401 return $contents;
1402 }
1403
1404 sub WriteStringToFile {
1405 my $fh = SafeOpen ("> $_[0]");
1406 binmode $fh;
1407 print $fh $_[1] or confess "$_[0]: $!\n";
1408 SafeClose $fh;
1409 }
1410
1411 foreach my $file (@ARGV) {
1412 my $changed_p = 0;
1413 my $new_contents = "";
1414 if ($options{"line-mode"}) {
1415 my $fh = SafeOpen $file;
1416 while (<$fh>) {
1417 my $save_line = $_;
1418 eval $code;
1419 $changed_p = 1 if $save_line ne $_;
1420 $new_contents .= $_;
1421 }
1422 } else {
1423 my $orig_contents = $_ = FileContents $file;
1424 eval $code;
1425 if ($_ ne $orig_contents) {
1426 $changed_p = 1;
1427 $new_contents = $_;
1428 }
1429 }
1430
1431 if ($changed_p) {
1432 my $backdir = $options{"backup-dir"};
1433 $backdir = "backup" if !defined ($backdir);
1434 if ($backdir) {
1435 my ($name, $path, $suffix) = fileparse ($file, "");
1436 my $backfulldir = $path . $backdir;
1437 my $backfile = "$backfulldir/$name";
1438 mkdir $backfulldir, 0755 unless -d $backfulldir;
1439 print "modifying $file (original saved in $backfile)\n";
1440 rename $file, $backfile;
1441 }
1442 WriteStringToFile ($file, $new_contents);
1443 }
1444 }
1445 ----------------------------------- cut ------------------------------------
1446
1447
1448 In addition to those programs, I needed to fix up a few other
1449 things, particularly relating to the duplicate definitions of
1450 types, now that some types merged with others. Specifically:
1451
1452 1. in lisp.h, removed duplicate declarations of Bytecount. The
1453 changed code should now look like this: (In each code snippet
1454 below, the first and last lines are the same as the original, as
1455 are all lines outside of those lines. That allows you to locate
1456 the section to be replaced, and replace the stuff in that
1457 section, verifying that there isn't anything new added that
1458 would need to be kept.)
1459
1460 --------------------------------- snip -------------------------------------
1461 /* Counts of bytes or chars */
1462 typedef EMACS_INT Bytecount;
1463 typedef EMACS_INT Charcount;
1464
1465 /* Counts of elements */
1466 typedef EMACS_INT Elemcount;
1467
1468 /* Hash codes */
1469 typedef unsigned long Hashcode;
1470
1471 /* ------------------------ dynamic arrays ------------------- */
1472 --------------------------------- snip -------------------------------------
1473
1474 2. in lstream.h, removed duplicate declaration of Bytecount.
1475 Rewrote the comment about this type. The changed code should
1476 now look like this:
1477
1478
1479 --------------------------------- snip -------------------------------------
1480 #endif
1481
1482 /* The have been some arguments over the what the type should be that
1483 specifies a count of bytes in a data block to be written out or read in,
1484 using Lstream_read(), Lstream_write(), and related functions.
1485 Originally it was long, which worked fine; Martin "corrected" these to
1486 size_t and ssize_t on the grounds that this is theoretically cleaner and
1487 is in keeping with the C standards. Unfortunately, this practice is
1488 horribly error-prone due to design flaws in the way that mixed
1489 signed/unsigned arithmetic happens. In fact, by doing this change,
1490 Martin introduced a subtle but fatal error that caused the operation of
1491 sending large mail messages to the SMTP server under Windows to fail.
1492 By putting all values back to be signed, avoiding any signed/unsigned
1493 mixing, the bug immediately went away. The type then in use was
1494 Lstream_Data_Count, so that it be reverted cleanly if a vote came to
1495 that. Now it is Bytecount.
1496
1497 Some earlier comments about why the type must be signed: This MUST BE
1498 SIGNED, since it also is used in functions that return the number of
1499 bytes actually read to or written from in an operation, and these
1500 functions can return -1 to signal error.
1501
1502 Note that the standard Unix read() and write() functions define the
1503 count going in as a size_t, which is UNSIGNED, and the count going
1504 out as an ssize_t, which is SIGNED. This is a horrible design
1505 flaw. Not only is it highly likely to lead to logic errors when a
1506 -1 gets interpreted as a large positive number, but operations are
1507 bound to fail in all sorts of horrible ways when a number in the
1508 upper-half of the size_t range is passed in -- this number is
1509 unrepresentable as an ssize_t, so code that checks to see how many
1510 bytes are actually written (which is mandatory if you are dealing
1511 with certain types of devices) will get completely screwed up.
1512
1513 --ben
1514 */
1515
1516 typedef enum lstream_buffering
1517 --------------------------------- snip -------------------------------------
1518
1519
1520 3. in dumper.c, there are four places, all inside of switch()
1521 statements, where XD_BYTECOUNT appears twice as a case tag. In
1522 each case, the two case blocks contain identical code, and you
1523 should *REMOVE THE SECOND* and leave the first.
1524 1228
1525 2001-09-17 Ben Wing <ben@xemacs.org> 1229 2001-09-17 Ben Wing <ben@xemacs.org>
1526 1230
1527 * fileio.c (normalize_filename): 1231 * fileio.c (normalize_filename):
1528 * fileio.c (Fexpand_file_name): 1232 * fileio.c (Fexpand_file_name):