GSP/GSP-Agent-Windows
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
GSP-Agent-Windows/OGP64/usr/share/doc/gawk/ChangeLog

4174 lines
144 KiB
Text
Raw Permalink Blame History

2026-03-02 Arnold D. Robbins <arnold@skeeve.com>
* Checklist: Revisions.
2026-03-01 Weixie Cui <cuiweixie@gmail.com>
* array.c (make_aname): Use sizeof(char) instead of
sizeof(char *) when allocating memory for the aname
string buffer. The buffer holds characters, not
pointers, so the previous sizeof(char *) caused
over-allocation.
2026-02-27 Arnold D. Robbins <arnold@skeeve.com>
* buitin.c (do_sub): Pass searchflags to research in
loop over the whole line so that gensub()'s global works
with \N construct. Thanks to Koichi Murase <myoga.murase@gmail.com>
and Finn Steffens <finn.steffens@protonmail.com> for
reporting the bug.
2026-02-22 Arnold D. Robbins <arnold@skeeve.com>
* 5.4.0: Release tar ball made.
2026-02-22 Arnold D. Robbins <arnold@skeeve.com>
* README: Updated.
2026-02-15 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2026-02-13 Arnold D. Robbins <arnold@skeeve.com>
* Checklist: Add more stuff.
2026-02-12 Arnold D. Robbins <arnold@skeeve.com>
* Checklist: Add more stuff.
* Makefile.am, Makefile.in, awk.h, awkgram.y, command.y,
debug.c, eval.c, ext.c, field.c, gawkapi.h, gawkmisc.c,
interpret.h, io.c, mpfr.c, node.c, nonposix.h, printf.c,
profile.c, replace.c, symbol.c: Update copyright year.
2026-02-10 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
* main.c (UPDATE_YEAR): Update to 2026.
2026-02-09 Arnold D. Robbins <arnold@skeeve.com>
Bug fix for untyped elements used in redirections.
Thanks to Ed Morton for the report.
* builtin.c (do_print): Call elem_new_to_scalar() on the redirection.
(do_getline_redir): Ditto.
2026-02-09 Arnold D. Robbins <arnold@skeeve.com>
* main.c (__dso_handle): Remove. PCC no longer needs it.
* printf.c (isinf): Remove macros, PCC now handles it okay.
2026-02-05 Arnold D. Robbins <arnold@skeeve.com>
* main.c (__dso_handle): Define if compiling with PCC.
* printf.c (isinf): For PCC undef it and define it as __builtin_isinf
which PCC supports.
2026-02-04 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: Look for <sys/procctl.h> and <sys/sysctl.h>, and
for procctl() and sysctl(), on FreeBSD.
* NEWS: Updated.
2026-02-03 Arnold D. Robbins <arnold@skeeve.com>
* Makefile.am: For NetBSD 10, turn off ASLR via paxctl.
* NEWS: Updated.
2026-02-02 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
* re.c (regexflags2str): Remove declaration of function that
doesn't exist.
(resetup): Clean up the code a little.
(minrxcompflags2str, minrxexecflags2str): New functions.
2026-02-03 John E. Malmberg <wb8tyw@qsl.net>
* builtin.c (do_dump_node): Honor the HAVE_LONG_LONG_INT macro.
2026-01-28 John E. Malmberg <wb8tyw@qsl.net>
* ChangeLog: Spelling fixes
* custom.h: OpenVMS Posix is long gone.
2026-01-27 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2026-01-24 Arnold D. Robbins <arnold@skeeve.com>
* TODO: Updated.
2026-01-20 Arnold D. Robbins <arnold@skeeve.com>
* TODO: Updated.
2026-01-20 J. <thirdedition@gmx.net>
* configure.ac: Remove \ continuation in lists inside m4
constructs. Needed for MSYS2.
2026-01-17 Arnold D. Robbins <arnold@skeeve.com>
* eval.c (posix_compare): Simplify further to always use
strcoll() everywhere but MinGW.
* NEWS: Updated. Only OpenBSD 7 works with PMA.
2026-01-16 Arnold D. Robbins <arnold@skeeve.com>
* printf.c (format_args): Revert fix of 2026-01-15.
* eval.c (posix_compare): For Cygwin, always uses strcoll().
2026-01-15 Corinna Vinschen <vinschen@redhat.com>
UTF-8 and GB18030 allow 4 byte multibyte sequences which
have to be converted to surrogate pairs on systems with
wchar_t == UTF-16. That only affects Cygwin for now, given
AIX is ignored and native Windows is handled separately.
This affects the printf() functions because the character and byte
counting in mbc_char_count() and mbc_byte_count() can't handle
characters yet which have to be converted to surrogate pairs.
First of all, both functions do not expect the mbstate_t to be
in a shift state after it read the first character. However, this
is essential to keep track of the full character during surrogate
handling. When the code restarts reading the string from the
beginning for counting, it neglects to reset the shift state,
so the next read encounters an invalid character sequence. So make
sure to reset the shift state when reading from the beginning.
For Cygwin:
Add code to mbc_char_count() to count a surrogate pair as a single
character, not as two characters.
Add code to mbc_byte_count() to handle only a full surrogate as a
full character.
2026-01-15 Arnold D. Robbins <arnold@skeeve.com>
* printf.c (format_args): For %c, check for -2 return value
from c32rtomb().
2026-01-14 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_dump_node): Change %zd to %zu.
For MB_CUR_MAX, cast to int; helps on Cygwin.
2026-01-12 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_substr): Add casts to silence clang warnings.
* node.c (r_format_val): Ditto.
* main.c (load_procinfo): Ditto.
(initfds): Change use of newfd to silence set but not used warning.
2026-01-09 Arnold D. Robbins <arnold@skeeve.com>
* mpfr.c (force_mpnum): Use bool instead of int for parameters.
(mpg_force_number): Sync some with r_force_number. Accept
non-decimal data if --posix, as in r_force_number. Thanks to
Jason C. Kwan for the report.
2026-01-06 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Edited.
2026-01-04 Arnold D. Robbins <arnold@skeeve.com>
* bootstrap.sh: No need to touch doc/gawk.texi. Add touch
of pc/Makefile.tst.
2026-01-04 Eli Zaretskii <eliz@gnu.org>
* NEWS: Updated with info about UTF-8 on Windows.
2025-12-30 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2025-12-30 Arnold D. Robbins <arnold@skeeve.com>
* awk.h: Style edit: POSIX in all caps.
* main.c: Style edits: No space before parens in function calls.
* node.c (parse_escape): Move memset() of mbs to outside ifdefs.
Factors out duplicate code.
2025-12-28 Eli Zaretskii <eliz@gnu.org>
* awk.h: Remove MinGW stuff that is no longer needed.
* custom.h [__MINGW32__]: Add redirections for char32_t functions
and their prototypes. Override MB_CUR_MAX with a MinGW-specific
function. Add prototypes and redirections for stdio functions to
work around a bug in MSVCRT.
* gawkmisc.c [__MINGW32__]: Include pc/mbc32.c.
* main.c (main) [__MINGW32__]: Adjust the value of gawk_mb_cur_max
when the locale changes.
* node.c: Remove MinGW-specific code, which is not longer needed.
* nonposix.h: Add prototype of 'mingw_using_utf8'.
* printf.c: Fix a typo: wcrtomb should not be called directly, but,
if needed, via c32rtomb.
* profile.c: Avoid compiler warnings about functions that are
unused when SIGHUP and SIGUSR1 are not defined.
* replace.c [__MINGW32__ && !_UCRT]: Include stdio-consolesafe.c
to work around MSVCRT bugs when writing multibyte characters to
the console.
2025-12-28 Arnold D. Robbins <arnold@skeeve.com>
* TODO: Updated.
2025-12-27 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y (yylex): Add support for hexadecimal floating point.
* builtin.c (nondec2awknum): Improve hex floating point support.
2025-12-24 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: New option, --enable-O3 to use -O3 at compile
time. This restores the default of -O2.
* NEWS: Updated.
Unrelated:
* Checklist: Updated with more items.
2025-12-23 Eli Zaretskii <eliz@gnu.org>
* awk.h [_UCRT]: Use MinGW64-provided <uchar.h> when possible.
(char32_t): Use the same typedef as in uchar.h. Include
<stdint.h> that's required for uint_least32_t.
2025-12-23 Arnold D. Robbins <arnold@skeeve.com>
* bootstrap.sh: Be smarter about touching the doc/*.info files.
2025-12-22 Arnold D. Robbins <arnold@skeeve.com>
* awk.h: Fixes for systems without uchar.h.
2025-12-22 Arnold D. Robbins <arnold@skeeve.com>
Switch to using the char32_t and associated routines.
* configure.ac: Add checks for uchar.h, mbrtoc32(), c32rtomb().
* awk.h: Add declarations and macros, based on configuration.
(NODE, wstrstr, wcasestrstr): Switch to char32_t from wchar_t.
* builtin.c: Ditto, use mbrtoc32(), across the board.
* eval.c: Ditto, use c32slen(), c32scoll() in posix_compare().
* interpret.h: Use char32_t.
* node.c: Use char32_t and associated functions.
* printf.c: Ditto.
2025-12-18 Arnold D. Robbins <arnold@skeeve.com>
* awk.h: Add a pragma for z/OS so that enum sizes are the same
size as int. Prevents a core dump in the randtest test. Much
thanks to Daniel Richard G. <skunk@iskunk.org> for the fix.
2025-12-17 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_match): If $0 was the regex, after unrefing
tre->re_exp, set it to NULL so that it won't get unref'ed
a second time. Thanks to valgrind.
Unrelated:
* node.c (parse_escape): Set type of `i' to long. Avoids
undefined behavior warning from UBSan.
2025-12-16 Arnold D. Robbins <arnold@skeeve.com>
* awk.h, configure.ac, main.c, re.c, version.c: Minor edits
before merging into master.
2025-12-15 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_dump_node): Add additional information, fix
print param_cnt member.
Unrelated:
* configure.ac: Add strftime back to the list of functions
we search for. It's needed on MinGW.
* replace.c: Include missing_d/strftime.c if needed, along
with MinGW stuff. Thanks to Eli Zaretskii <eliz@gnu.org>.
Unrelated:
* TODO: Updated.
2025-12-12 Eli Zaretskii <eliz@gnu.org>
* awk.h: Add declaration for timegm if HAVE_TIMEGM not defined.
Needed for MinGW.
* io.c (wait_any_block_signals): Add ifdefs as needed for MinGW.
(file_can_timeout): Ditto.
2025-12-12 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_dump_node): Add an additional ifdef for MPFR.
2025-12-11 Arnold D. Robbins <arnold@skeeve.com>
* debug.c (CHECK_PROG_RUNNING): Add file, line number and function
names to error message.
(serialize_list): Move history related variables and code into
#ifdefs so that things compile without readline.
2025-12-10 Arnold D. Robbins <arnold@skeeve.com>
Add a function to dump the NODE structure. Should have done
this ages ago.
* awk.h (do_dump_node): Declare the function.
* builtin.c (do_dump_node): New function.
* awkgram.y (tokentab): Add entry for _dump_node(). Note the "_".
2025-12-09 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated with info about MinRX.
2025-12-09 Arnold D. Robbins <arnold@skeeve.com>
* mpfr.c (mpg_mod): Keep only comments relevant to compatibility.
And refer to the test.
* builtin.c (do_intdiv): Fix typos in a comment.
* mpfr.c (do_mpfr_intdiv): Ditto.
Thanks to Nethox <nethox+awk@gmail.com> for the patch.
2025-12-06 Arnold D. Robbins <arnold@skeeve.com>
* awk.h: If using gettimeofday(), include <sys/time.h>.
This is helpful on old macOS, at least.
2025-11-30 Arnold D. Robbins <arnold@skeeve.com>
Improve the timeout check to be made when the IOP is created
instead of on every record read.
* awk.h (IOBUF): Add a can_timeout member.
* io.c (get_a_record): Use it instead of calling file_can_timeout().
(file_can_timeout): Adjust to take fd and mode arguments.
(iop_alloc): Set the can_timeout member by calling file_can_timeout.
2025-11-24 Arnold D. Robbins <arnold@skeeve.com>
* README: Updated some more.
2025-11-22 Arnold D. Robbins <arnold@skeeve.com>
Improve raw I/O speed by not checking for a timeout
when there's no need. Issue noticed during profiling
of something unrelated.
* io.c (file_can_timeout): New routine. Returns true if an
input file is one that can timeout. Regular files cannot.
(get_a_record): Use it.
2025-11-16 Arnold D. Robbins <arnold@skeeve.com>
* io.c (inetfile): Issue a warning if udp is used. It's
now obsolete.
* NEWS: Updated.
2025-11-04 Arnold D. Robbins <arnold@skeeve.com>
Continue the cleanup:
* configure.ac: Don't look for system().
* replace.c: Don't include missing_d/system.c. It was for DJGPP
which is not supported anymore.
* custom.h: Remove the bit for QNX. Improve the declaration of
strsignal().
2025-11-03 Arnold D. Robbins <arnold@skeeve.com>
Considerably simplify use of replacement functions. Many of the
replacement functions are for routines that now widely standard,
so we assume they're available.
* Makefile.am (base_sources): Remove protos.h.
* protos.h: Remove this file.
* awk.h: Don't include protos.h. Don't do gawk_strtod() stuff.
* configure.ac: Remove checks for a whole bunch of standard functions.
Remove AC_FUNC_MKTIME check.
* custom.h: Remove old checks for HP/UX and macOS.
* replace.c: Remove checks and includes of files that are no
longer needed.
2025-10-31 Arnold D. Robbins <arnold@skeeve.com>
After discussion with Nelson Beebe, allow hexadecimal
floating point for strtonum() and --non-decimal-data.
NOTE: NOT in program source code! (At least, not yet. Sigh.)
* NEWS: Updated.
* builtin.c (nondec2awknum): Revise logic.
* mpfr.c (force_mpnum): Ditto.
2025-10-29 Arnold D. Robbins <arnold@skeeve.com>
* re.c (make_regexp): Remove assertion that string is zero
terminated. It isn't always. Instead, save and restore the
character past the value of `len'. This required making the
first parameter not be const anymore. Thanks to Ben Hoyt
<benhoyt@gmail.com> for the report.
* awk.k (make_regexp): Adjust declaration.
2025-10-28 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y: Remove never-documented support for tawk extension
`delete(a)' and lint warnings for regexp modifiers. Tawk is dead,
it's time to bury the corpse.
2025-10-27 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): Op_push_arg_untyped: Don't issue
lint warning for untyped variables, since isarray() and typeof()
don't change the type of the parameter. Thanks to
Greg Minshall <minshall@umich.edu> for the suggestion.
2025-10-25 Arnold D. Robbins <arnold@skeeve.com>
* field.c (do_split, do_patsplit): Use do_lint_extensions before
warning about extensions.
(set_FPAT): Clean up a lint warning.
* io.c (set_RS): Ditto.
2025-10-23 Arnold D. Robbins <arnold@skeeve.com>
* field.c (do_split): If someone passed a Node_elem_new
or subarray as the second argument to split(), issue a
lint warning. Thanks to Hermann Peifer for the suggestion.
(do_patsplit): Add the same warning.
2025-10-22 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): For Op_sub_array, don't warn
about using an uninitialized element, as there is no
way to create the parent of a subarray first. Thanks to
Greg Minshall <minshall@umich.edu> for the report.
Unrelated:
* Checklist: Add an item to run the test suite with MPFR.
2025-10-19 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: Don't disable assertions anymore. They provide
value, and on modern systems don't affect performance.
* NEWS: Updated.
2025-10-19 Arnold D. Robbins <arnold@skeeve.com>
* re.c (re_update): Revert previous fix, it didn't do the trick.
* builtin.c (do_match): If someone passed $0 as the second
argument, deref it. It's a special case, but it keeps the
assertions happy.
Thanks to Eli Zaretskii <eliz@gnu.org> for the report.
2025-10-17 Arnold D. Robbins <arnold@skeeve.com>
Fix a bug when match() is called with $0 as the second argument.
The bug only showed up with assertions enabled. At the same time,
add a warning if a regexp constant is used as the first argument.
Thanks to Daniel David Goren <danieldavidgoren@gmail.com> for
the report.
* awkgram.y (snode): For do_match, add a warning if the first
argument is a regexp constant.
* re.c (re_update): Unref the original text of the regexp
if the copy is different from it.
Unrelated:
* mpfr.c (mpg_mod): Use mpz_tdiv_r() instead of mpz_tdiv_qr()
to simplify the code. Suggested by Jason C. Kwan on the
help-gawk list.
2025-10-04 Arnold D. Robbins <arnold@skeeve.com>
* printf.c (format_args): Fix call to fatal() to not have "fatal:"
in it.
2025-09-27 Arnold D. Robbins <arnold@skeeve.com>
Fix configure --disable-extensions. Thanks to
Frank Busse <f.busse@imperial.ac.uk> for the report.
* configure.ac: When modifying the Makefile for different CFLAGS,
use test -f to make sure that one is there.
* ext.c (init_extensions_list): Add dummy version for when extensions
are disabled.
(load_ext): Update the function argument list for the dummy
version to match the real one and the declaration in awk.h.
* Checklist: Add an item to check `configure --disable-extensions'
so that this won't happen again.
2025-09-01 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Fix spacing / indentation.
* builtin.c (do_typeof): Memory leak fix for Node_elem_new. This
showed up with valgrind in debug build on the test/elemnew4.awk
file.
* builtin.c: Move commented out main() test program for add_thousands()
to ...
* printf.c: ... here from builtin.c.
2025-07-30 Andrew J. Schorr <aschorr@telemetry-investments.com>
Add #ifdef'ed out patch to io.c to use a hashtable to store pipe
pids to fix a problem with a docker container that defaulted to
a ulimit of a billion open fds.
* io.c (write_pipe): If hashing, add a next field.
(save_pipe): New function to add a pipe to the hash table.
(gawk_popen_write): If hashing, allocate index of PIPE_HASH_SIZE
instead of getdtablesize(), and use save_pipe to store the pipe info
instead of sticking it directly into the open_pipes array.
(gawk_popen_write_close): If hashing, search the bucket list
for a matching process. If the match is not at the head of the
linked list, remove it from the list and free it.
2025-07-29 Arnold D. Robbins <arnold@skeeve.com>
Make gawk's system() function smarter. It has to ignore SIGINT
and SIGQUIT and block SIGCHLD (per POSIX for the C system()
function.) Thanks to Finn Magnusson <finnmag@yahoo.com> for
the report. The test case was:
gawk 'BEGIN { system("cat"); print "\n1" }'
Then immediately interrupt the "cat" with a Control-C.
Older versions of gawk that used the C library system() would
still print the "1". Gawk 5.3.1 would not. These changes
restore the old behavior.
Note that different versions of awk do different things
in this test case. BWK awk does print the "1", mawk does not.
* io.c (wait_any): Now has just the code to do the wait and
update the status.
* io.c (wait_any_block_signals): This was originally wait_any().
Simplified it by having just the signal handling here and making
a call to wait_any().
* builtin.c (gawk_system): Added signal handling code to
ignore/block signals in the parent before the fork, restore
things in the child before the exec, and restore them in
the parent after calling wait_any().
2025-07-29 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (emalloc_real, ezalloc_real, erealloc_real): Use %zu
in the error message instead of %ld and casting to long.
2025-07-28 Arnold D. Robbins <arnold@skeeve.com>
Rationalize do_traditional throughout the code base:
1. Bring things into sync with current BWK awk.
2. Change `do_posix || do_traditional' tests to just `do_traditional'
since --posix automatically also turns on do_traditional.
* builtin.c (do_system): Remove old code that divided the
exit status by 256 if do_traditional. BWK awk no longer
does that.
* ext.c (load_ext): Simplify check for do_traditional.
* field.c (do_split): Ditto.
* main.c (is_std_var): Ditto.
* re.c (make_regexp): Ditto.
* symbol.c (lookup): Ditto.
* io.c (set_RS): Change check from do_traditional to do_posix
in case where RS->stlen > 1.
2025-07-24 Andrew J. Schorr <aschorr@telemetry-investments.com>
* io.c (redirect): Don't issue a lint complaint if the
redirect string is a STRNUM, because a filename could be
numeric, and that shouldn't be flagged by lint if it came
from user input.
2025-07-22 Arnold D. Robbins <arnold@skeeve.com>
* main.c (check_pma_security): #ifdef out the whole function if
PMA is not enabled. Also the declaration. Avoids compiler warnings.
(enable_pma): Simplify the code some.
2025-07-22 Nick Alcock <nick.alcock@oracle.com>
* main.c (enable_pma): Return false as default if PMA is
not enabled. Otherwise the test suite breaks if gawk is
built with `configure --disable-pma'.
2025-07-14 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_length): As length(array) is now part of POSIX,
remove checks and warnings.
* NEWS: Updated with this info.
2025-07-10 Arnold D. Robbins <arnold@skeeve.com>
* ext.c (load_ext): Allow a magic environment variable to allow
loading of extensions even if --traditional / --posix.
Mainly for testing purposes.
2025-07-10 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (fixtype): Take Node_elem_new into account.
Bug report from Douglas Nelson, nelson@msu.edu.
2025-06-13 Jim Mellander <jmellander@gmail.com>
* builtin.c (do_srand): Change casts of seed values to
unsigned long. Avoids some overflow problems.
2025-06-03 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated w.r.t. PMA usage changes.
* awk.h (load_ext): Change declaration, new additional parameter.
(init_extension_list): Add declaration.
(extension_list): Add declaration.
(version_string, persist_file): Add declarations.
* awkgram.y (load_library): Add s->src in call to load_ext().
* ext.c (extension_list): Maintain a list of loaded extensions
when using PMA so can check if an extension is loaded from
a different path.
(search_ext_list, init_extension_list): New functions.
(extension_list): Add definition.
(load_ext): Add additional argument `name'. Search the extension
list if PMA.
(make_builtin): If PMA and found the function already installed,
update its function pointer and return true. Returning true
avoids a warning.
* main.c (main): Update call to load_ext().
(persist_file): New global variable.
* symbol.c (struct root_pointers): Add ext_list and version_string
members.
(init_symbol_table): For first time, call init_extension_list,
save it and the version_string. Restore on second time and
check version.
2025-06-01 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (using_pma): Removed. We already have
using_persistent_malloc.
* main.h (using_pma): Ditto. Adjust uses.
* ext.c (load_ext): Use using_persistent_malloc.
2025-05-28 Arnold D. Robbins <arnold@skeeve.com>
* io.c (two_way_open): Fix direction in non-MinGW call
to os_close_on_exec().
2025-05-23 Arnold D. Robbins <arnold@skeeve.com>
* gawkapi.h (gawk_api_minor_version): Increased to 1, because
of the previous change.
* NEWS: Updated.
2025-05-22 Andrew J. Schorr <aschorr@telemetry-investments.com>
* gawkapi.c (api_set_array_element): Ensure that the new index
is indeed a string. Thanks to SBP <seirios@ik.me> for the report.
2025-05-22 Arnold D. Robbins <arnold@skeeve.com>
Persistent memory and extension functions don't mix. Make
it a fatal error that gawk catches instead of weird errors
and core dumps. Issue reported by "Stan Marsh" (sic)
<gazelle@xmission.com>.
* awk.h (using_pma): New variable, add extern declaration.
* ext.c (load_ext): If using_pma, fatal message.
* main.c (using_pma): New variable, add definition.
(enable_pma): Set it.
2025-05-18 Arnold D. Robbins <arnold@skeeve.com>
* eval.c (update_ERRNO_int, update_ERRNO_string, unset_ERRNO):
If do_traditional, return early.
* io.c: As needed, for calls to the above, remove test for
! do_traditional. The code was not being consistent about
always checking. Thanks to Andrew Schorr for pointing out
the issue.
2025-05-16 Arnold D. Robbins <arnold@skeeve.com>
Make sure all non-standard variables are unavailable
with --traditional. Thanks to Andrew Schorr for the push.
* io.c (get_a_record): Don't mess with RT if do_traditional.
* main.c (init_vars): Don't install non-standard variables.
* NEWS: Updated.
2025-05-14 Arnold D. Robbins <arnold@skeeve.com>
* awk.h [RE_NEED_SUB]: New flag for if need submatch info.
* builtin.c (do_sub, do_match): Use it when needed.
* re.c (research): For MinRX, adjust match_count value based
on RE_NEED_SUB and RE_NEED_START. Allows for eventual
optimizations within MinRX.
2025-05-06 Arnold D. Robbins <arnold@skeeve.com>
* re.c (research): Undo previous change. MINRX_REG_NOSUB
is a flag for compiling the regex, not execing it. :-(
In gawk, we don't know at regex compile time if we will
need the subpatterns or not. Oh well.
2025-04-28 Arnold D. Robbins <arnold@skeeve.com>
* re.c (research): If not need_start, add MINRX_REG_NOSUB
to the flags. In preparation for when MinRX implements
that flag.
2025-04-08 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: Check for posix_spawnp() also, for older
macOS systems.
2025-04-06 Arnold D. Robbins <arnold@skeeve.com>
* re.c (make_regexp): Add MINRX_REG_MINDISABLE for traditional awk.
2025-04-02 Arnold D. Robbins <arnold@skeeve.com>
* README: Version adjusted for release.
* configure.ac: Ditto.
* 5.3.2: Release tar made.
2025-03-27 Arnold D. Robbins <arnold@skeeve.com>
* field.c (do_split): Fix subtle issue with " " vs. / /
as separator argument to split(). Reported by Jason Kwan
with assistance from Nethox.
2025-03-25 Arnold D. Robbins <arnold@skeeve.com>
* io.c (find_source): Always append .awk when searching, even if
--traditional. Thanks to Nethox <nethox+awk@gmail.com> for
pointing out the inconsistency.
* NEWS: Updated.
2025-03-19 Collin Funk <collin.funk1@gmail.com>
Fix build on AIX.
* awk.h [_AIX]: Include <sys/cred.h> before defining macros which
cause errors when expanded in function prototypes.
2025-03-18 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (POP_SCALAR): Use DEREF on original value for Node_var_new.
2025-03-17 Arnold D. Robbins <arnold@skeeve.com>
Fix problems calling functions indirectly with Node_var_new.
Thanks to Denis Shirokov <cosmogen@gmail.com> for the report
and test case.
* awk.h (POP_SCALAR): Handle Node_var_new.
(fixtype): Change assertion into a runtime test and failure.
That way it'll happen even if compiled with NDEBUG.
Unrelated cleanup:
* builtin.c (call_match): Remove redundant tests for
Node_var_new inside a test for either Node_var_new or
Node_elem_new.
2025-03-11 Arnold D. Robbins <arnold@skeeve.com>
* re.c (check_bracket_exp): Rewrite the routine to correctly handle
special characters after the opening bracket. Thanks to
Mohamed Akram <mohd.akram@outlook.com> for the report.
(reflags2str): Unrelated: Add RE_DEBUG to the list.
2025-03-09 Arnold D. Robbins <arnold@skeeve.com>
Bug fix:
* eval.c (elem_new_reset): Set vname to NULL. This is a bug fix that
only showed up when compiling for 32 bit.
Memory cleanup:
* eval.c (setup_frame, PUSH_CODE, init_interpret): After call to
getnode(), memset the new node to zero.
* field.c (set_record): Ditto.
* interpret.h (r_interpret): Case Op_array_for_init: Ditto.
* node.c (make_str_node): Ditto.
* profile.c (pp_push): Ditto.
* symbol.c (append_symbol): Ditto.
Compiler warning fix:
* main.c (enable_pma): Add a return true to silence compiler
warnings when USE_PERSISTENT_MALLOC is not defined.
General improvement:
* interpret.h (r_interpret): Case Op_func_call, if do_itrace,
print the function name.
2025-02-24 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
* main.c (UPDATE_YEAR): Update.
* array.c, awk.h, awkgram.y, cint_array.c, command.y, debug.c,
eval.c, ext.c, field.c, floatcomp.c, gawkapi.c, int_array.c,
interpret.h, io.c, mpfr.c, node.c, printf.c, profile.c, re.c,
str_array.c, symbol.c: Update copyright year.
2025-02-14 Paul Eggert <eggert@cs.ucla.edu>
* main.c (main): Omit BSD hack that misbehaved by putting stderr
in O_APPEND mode, which can cause later stderr output to be put in
the wrong place, even after Gawk has exited and even if Gawk
never outputs anything. For example, the shell command
"echo abcdefgh >foo; (gawk 'BEGIN{}'; echo ouch >&2) 2<>foo"
incorrectly appended "ouch" to the end of foo rather than
correctly overwriting the start of "foo" with "ouch".
The BSD issue can be addressed in the test case instead.
2025-02-16 Arnold D. Robbins <arnold@skeeve.com>
* int_array.c (INT_CHAIN_MAX): Bump it up from 2 to 10.
* str_array.c (STR_CHAIN_MAX): Ditto.
2025-02-06 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: Update CXXFLAGS for debug also, not just CFLAGS.
2025-02-05 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: Add checks for spawn.h and _NSGetExecutablePath
function.
* awk.h (os_disable_aslr): Add function declaration.
* main.c (enable_pma): Move OS specific code out of this function
and in posix/gawkmisc.c. Instead, call os_disable_aslr().
2025-02-05 Arnold D. Robbins <arnold@skeeve.com>
* main.c (enable_pma): Remove unused argc parameter and adjust
call. Add call to unsetenv() for magic env var if it was
there, so that the environment is as it was before the exec.
2025-02-05 Arnold D. Robbins <arnold@skeeve.com>
* main.c (enable_pma): New function. Has the entire init flow
for PMA. Also has new linux-specific code to deal with being
an PIE executable.
(main): Call enable_pma().
* configure.ac: Add checks for <sys/personality.h> and the
personality() system call.
* NEWS: Updated.
2025-02-05 Arnold D. Robbins <arnold@skeeve.com>
Bug fixes for indirect calls of match and patsplit.
Thanks to Denis Shirokov <cosmogen@gmail.com> for the report
and test case.
* builtin.c (call_match): Pop the text of the stack. Handle
the case that the regex is Node_var_new or Node_elem_new.
Same for the text. Push text onto the stack first, then
everything else.
Also, free the vname for the regex argument if it's a Node_var_new;
same for the text argument.
* eval.c (setup_frame): Handle Node_regex and Node_dynregex.
* field.c (do_patsplit): Handle the case of the source being
Node_param_list, and then if Node_var_new or Node_elem_new.
2025-02-05 Arnold D. Robbins <arnold@skeeve.com>
* node.c (r_unref): For Node_var_new, free the vname.
2025-02-05 Arnold D. Robbins <arnold@skeeve.com>
Stylistic cleanups and small code cleanups.
* array.c, eval.c, ext.c, interpret.h: Some stylistic cleanups.
* awk.h (eln_pa, eln_vn): Renamed to elemnew_parent and
elemnew_vname. All uses adjusted.
(union worn): Renamed to `z' and only needs the wsp and vn
pointers in it.
2025-02-03 Cristian Ioneci <mekanofox@astropostale.com>
Fix issues triggered by using in certain ways array elements passed to
a function; issues include crash dumps, reads-after-free, internal errors,
incorrect results.
See test cases in test/ar2fn_*.awk.
* awk.h: Add an union around `wsp', `wslen', part of NODE's `sub.val', in
order to overlap a char* over `wsp'; adjust/add #defines to access the
moved/added members.
(force_string_fmt, force_number): Clear Node_elem_new specific members
when the type is changed.
* interpret.h (r_interpret): Make a newborn Node_elem_new remember its
parent and a string representation of the index in the parent; for that
use `typre' and the added char* mentioned above (in awk.h)
(r_interpret): Clear Node_elem_new specific members when the type is
changed.
* array.c (force_array): The node will become a Node_var_array: if
present, use the values stashed in the previously mentioned fields to
properly set its `parent_array' and `vname'.
(adjust_param_node): New function.
(adjust_fcall_stack): Extra handling of Node_elem_new nodes.
* eval.c (elem_new_reset): New function.
(elem_new_to_scalar): Clear Node_elem_new specific members using the newly
introduced function above.
* ext.c (get_actual_argument): Clear Node_elem_new specific members when
the type is changed.
* gawkapi.c (api_sym_update): Clear Node_elem_new specific members when
the type is changed.
* mpfr.c (mpg_force_number): Clear Node_elem_new specific members when
the type is changed.
* node.c (r_force_number): Clear Node_elem_new specific members when
the type is changed.
(r_unref): Add code to free the eln_vn member of a Node_elem_new.
2025-01-26 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): Case Op_store_var; move REGEX
check to ...
* node.c (r_unref): ... here.
* builtin.c (do_sub): Remove REGEX freeing code, simplify what
remained when checking if original target was a regex.
2025-01-24 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): Case Op_store_var; add a
check for valref == 1. Duh.
2025-01-23 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): Case Op_store_var; free up
typed regex stuff to avoid memory leaks. See memleak2 test.
Thanks to Denis Shirokov <cosmogen@gmail.com> for the report
and test case.
2025-01-22 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: Add hack if GCC to use -O3 instead of -O2.
Unrelated:
* builtin.c (call_sub, call_split_func): Handle the case of
the function name being preceded by awk::.
Thanks to Denis Shirokov <cosmogen@gmail.com> for the reports.
2025-01-22 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: Set CXXFLAGS correctly for MinRX: Use -O3
and -std=c++20.
2025-01-20 Arnold D. Robbins <arnold@skeeve.com>
* TODO: Add item about variable line number and filename
definition point.
2025-01-07 Arnold D. Robbins <arnold@skeeve.com>
Thanks to Cristian Ioneci <mekanofox@astropostale.com> for the
report and test case and fixes for these.
* builtin.c (do_isarray): Handle Node_param_list also.
(do_typeof): Additional checking for passed-in array.
2025-01-06 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_typeof): Handle Node_param_list. Thanks to
Thanks to Denis Shirokov <cosmogen@gmail.com> for the report and
test case.
2025-01-02 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated. Copyright year updated, too.
2024-12-24 Arnold D. Robbins <arnold@skeeve.com>
* re.c: Finish the emalloc / erealloc changes.
2024-12-15 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (emalloc, erealloc, ezalloc): Move to using __func__
instead of requiring the function name as a string in the
source code.
* All source files: Updated all uses.
2024-12-11 Arnold D. Robbins <arnold@skeeve.com>
* awk.h: Remove accidentally left-over Tandem bits.
2024-10-24 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_lshift, do_rshift): If shifting by too many
bits, return zero. Adjust lint messages. Thanks to
prospero <prospero@cyber-wizard.com> for the report.
* NEWS: Updated.
2024-10-17 Paul Eggert <eggert@cs.ucla.edu>
* floatcomp.c (adjust_uint): Add commentary.
2024-10-13 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2024-10-10 Arnold D. Robbins <arnold@skeeve.com>
Check if RE match at end of string could be an exact
match. This is true if the maybe_long flag is false.
Thanks to Ronald D. Rechenmacher <ron@fnal.gov> for
the idea.
* io.c (rsrescan): Add an additional check when the
match is exactly at the end.
* re.c (make_regexp): Add backslash to the list of
characters which sets the maybe_long field.
2024-09-25 Arnold D. Robbins <arnold@skeeve.com>
Clean up spurious newlines in pretty printer output.
Thanks to John Devin <john.m.devin@gmail.com> for the
motivation.
* awk.h (close_prof_file): New function.
* main.c (main): All close_prof_file.
* profile.c (close_prof_file): New function.
(at_start): New variable.
(pprint, print_lib_list, print_include_list, print_comment,
pp_func, pp_namespace): Use it.
2024-09-19 Arnold D. Robbins <arnold@skeeve.com>
* array.c (do_delete): Handle case where subscript is Node_elem_new.
Thanks to Denis Shirokov <cosmogen@gmail.com> for the report and
test case.
2024-09-17 Arnold D. Robbins <arnold@skeeve.com>
* 5.3.1: Release tar made.
2024-09-17 Arnold D. Robbins <arnold@skeeve.com>
* Makefile.am, awkgram.y, command.y, debug.c, eval.c,
gawkapi.c, interpret.h, mpfr.c: Update copyright year.
* Checklist: Updated a bit.
* NEWS: Version changed.
* configure.ac: Bump version.
2024-09-04 Arnold D. Robbins <arnold@skeeve.com>
* Checklist: Add two more items to do.
2024-09-01 Arnold D. Robbins <arnold@skeeve.com>
* re.c (make_regexp): Add MINRX_REG_NATIVE1B flag.
2024-08-28 Arnold D. Robbins <arnold@skeeve.com>
* Makefile.am (CCLD): Set to $(CXX) so that builds with clang
can work too. Thanks to Nelson H.F. Beebe.
2024-08-16 Arnold D. Robbins <arnold@skeeve.com>
* printf.c (zero_fill_to_precision): Fix calculation of buflen.
Thanks and a tip of the hat to Address Sanitizer.
2024-08-12 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y (push_ns_onto_namespace_chain): New function.
(mk_function, set_namespace): Use it.
2024-08-12 Arnold D. Robbins <arnold@skeeve.com>
Fix some C23 compilation issues. Thanks to
Jeffrey Cliff <jeffrey.cliff@gmail.com> for the report.
* awkgram.y (get_src_buf): Remove obsolete comment.
* io.c (iop_alloc, get_read_timeout): Fix casts for read.
2024-08-11 John E. Malmberg <wb8tyw@qsl.net>
* printf.c (format_mpg_integer_digits): Improved fix
from Arnold for if !HAVE_MPFR.
2024-08-11 John E. Malmberg <wb8tyw@qsl.net>
* printf.c (format_float): Fix cant_happen macro call.
(format_mpg_integer_digits): Return NULL if !HAVE_MPFR.
2024-08-10 Arnold D. Robbins <arnold@skeeve.com>
Thanks again to Michal Jaegermann for tips on removing
compiler warnings:
* eval.c (cmp_scalars): Default ret to false and add a
default case to the switch.
2024-08-09 Arnold D. Robbins <arnold@skeeve.com>
Thanks to Michal Jaegermann for tips on removing
compiler warnings:
* eval.c (opcode2str): Return "" instead of NULL.
(cmp_doubles): Default ret to false and add a default case
to the switch.
* printf.c (format_signed_integer): Add some additional
parentheses.
2024-08-07 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2024-08-06 Arnold D. Robbins <arnold@skeeve.com>
* printf.c (format_unsigned_integer, format_signed_integer):
Additional fixes. New added tests pass.
2024-07-29 Mike Haertel <mike@ducky.net>
* re.c (re_numsubpats): For MinRX, return re_nsub + 1.
2024-08-01 Arnold D. Robbins <arnold@skeeve.com>
Per Andy's suggestion, @ns_include --> @nsinclude.
* awk.h (SRC_NSINC): Renamed from SRC_NS_INC.
* command.y, debug.c, io.c, main.c, profile.c, awkgram.y: All
uses adjusted.
* awkgram.y (LEX_NSINCLUDE): Renamed from LEX_NS_INCLUDE.
(tokentab): "nsinclude" token renamed from "ns_include".
(yylex): Adjust logic a bit.
2024-07-31 Arnold D. Robbins <arnold@skeeve.com>
Add @ns_include to include a file into the current namespace.
* awk.h (struct srcfile): Add new enum value SRC_NS_INC.
* awkgram.y (use_current_namespace): New boolean.
(LEX_NS_INCLUDE): New token and new production for it.
(tokentab): Add "ns_include" token.
(print_included_from, add_srcfile, include_source, yylex): Adjust
logic.
* command.y (srcfile_generator): Handle SRC_NS_INC.
* debug.c (source_find): Ditto.
(debug_prog): Ditto.
* io.c (srcopen): Ditto.
* main.c (main): Ditto.
* profile.c (print_include_list): Ditto.
2024-07-28 Arnold D. Robbins <arnold@skeeve.com>
* TODO: Updated.
2024-07-26 Arnold D. Robbins <arnold@skeeve.com>
* re.c (check_bracket_exp): Make the code actually work.
2024-07-25 Arnold D. Robbins <arnold@skeeve.com>
* re.c (refree): Call minrx_regfree() on rp->mre_pat.
Thanks to valgrind and Mike Haertel.
2024-07-23 Arnold D. Robbins <arnold@skeeve.com>
Finish cleaning up the printf refactoring. All tests run, both
regular and -M. Test suite is valgrind clean!
Summary of changes:
* printf.c: New file.
(format_args): Moved here and renamed from format_tree. Considerably
simplified by moving code out into new routines.
(struct flags): New struct.
(add_thousands, do_printf, do_sprintf, format_nan_inf, mbc_byte_count,
mbc_char_count, out_of_range, printf_common, reverse): Moved to here.
(add_alt_format, add_plus_or_space_and_fill, adjust_flags, compute_zero_flag,
fill_to_field_width, format_float, format_integer_digits,
format_mpg_integer_digits, format_out_of_range, format_signed_integer,
format_unsigned_integer, zero_fill_to_precision): New routines.
2024-07-21 Arnold D. Robbins <arnold@skeeve.com>
* printf.c: Simplify printing of signed integers; move several
common tasks out into separate routines, and use char * instead
of const char * to avoid a lot of casting.
2024-07-20 Tim Rice <trice@posteo.net>
* io.c (csvscan): Set the sentinel if we found CR-LF.
2024-07-20 Arnold D. Robbins <arnold@skeeve.com>
* version.c: Add "-minrx" to the version, for now.
2024-07-20 Arnold D. Robbins <arnold@skeeve.com>
* printf.c: Continuing fixes and cleanups.
2024-07-17 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
* printf.c (fill_to_field_width): New function.
(format_out_of_range, format_unsigned_integer): Start to use it.
2024-07-16 Arnold D. Robbins <arnold@skeeve.com>
* printf.c: Continue cleaning up code and removing now-dead code.
2024-07-16 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (format_args): Renamed from format_tree. Remove
format_tree_old.
* debug.c, main.c, mpfr.c, node.c: Adjusted.
* printf.c (format_args): Renamed. Removed all old, factored-out
code.
(format_tree_old): Removed.
2024-07-16 Arnold D. Robbins <arnold@skeeve.com>
* printf.c: Use mpfr_sprintf() just to generate digits.
All tests now passing, even with -M.
2024-07-13 Arnold D. Robbins <arnold@skeeve.com>
* printf.c: Continued progress against the printf-corners test.
2024-07-11 Arnold D. Robbins <arnold@skeeve.com>
* printf.c: Signed and unsigned integer printing and float printing
working for current tests. MPFR integer printing also for current
tests. A few tests still to go in printf-corners.
2024-07-09 Arnold D. Robbins <arnold@skeeve.com>
* printf.c: Signed printing I think is working. Unsigned is better.
Checkpoint commit.
2024-07-07 Arnold D. Robbins <arnold@skeeve.com>
* printf.c: Signed and unsigned integer printing mostly working now.
Still to go: Finish that up, get GMP/MPFR integer printing working.
2024-07-05 Arnold D. Robbins <arnold@skeeve.com>
* printf.c: Lots of cleanup. Floating point format pretty
much working. Signed integer printing getting there.
Unsigned integer printing still to do.
2024-07-02 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (format_tree_new, format_tree_old): Add declarations.
(format_tree): Temporarily make into a function.
* main.c (format_tree): Declare it.
(main): Check PRINTF_NEW environment variable and if so,
use the new version. Otherwise use the old one.
* printf.c (struct flags): Declare.
(format_tree_old): The original version.
(format_tree_new): The new one, being refactored.
(add_thousands): Use the global `loc' variable, not a parameter.
static const char *format_integer(NODE *arg, struct flags *flags);
(format_mpg_integer, format_float, format_out_of_range,
compute_zero_flag): New functions.
2024-07-02 Arnold D. Robbins <arnold@skeeve.com>
* re.c (make_regexp): \u escapes also now treated literally
for do_traditional.
2024-06-26 Arnold D. Robbins <arnold@skeeve.com>
Start on printf rework.
* printf.c: New file, with stuff for printf in it.
* builtin.c: Move printf functions into printf.c.
* awk.h (efwrite): Add declaration.
* Makefile.am (base_source): Added printf.c.
2024-06-21 Arnold D. Robbins <arnold@skeeve.com>
* io.c (csvscan): Add a missing buffer overrun check. Restores
`AwKBUFSIZE=42 make check' to passing. :-)
2024-06-16 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): Op_push: When converting a
Node_elem_new into a Node_var, clear the MPFN and MPZN flags.
Allows the test suite to run again with -M.
2024-06-10 Arnold D. Robbins <arnold@skeeve.com>
* main.c (nostalgia): Removed function.
[COPYRIGHT_YEAR]: Bumped to 2024.
(main): Removed --nostalgia from options and from handling.
* NEWS: Updated.
2024-06-08 Arnold D. Robbins <arnold@skeeve.com>
* io.c: Remove check for OSF/1. It's too obsolete now.
2024-06-05 Arnold D. Robbins <arnold@skeeve.com>
* io.c: Fix a comment typo for enum recvalues.
2024-06-03 Arnold D. Robbins <arnold@skeeve.com>
* awk.h [USE_REAL_MALLOC]: Define this so that emalloc() and friends
use the real routines. For memory debugging.
* configure.ac: Enable USE_REAL_MALLOC if 'mallocdebug' is in
.developing. Add debug to .y files if 'yaccdebug' is in
.developing. Use '^...$' in grep test for all tests from
the .developing file.
2024-05-19 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.c: Updated.
* builtin.c (gawk_system): Remove an unused variable.
* io.c (gawk_popen_write): Ditto.
2024-05-18 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y (check_params): Make the message about not allowing
reserved variables as parameters explicit that this is from POSIX.
2024-04-28 Arnold D. Robbins <arnold@skeeve.com>
* re.c: Update copyright year.
2024-04-20 Arnold D. Robbins <arnold@skeeve.com>
* eval.c (cmp_scalars): Call fixtype() on the values before
trying to compare them. Thanks to Ed Morton for the bug
report. See test/strsubscript.awk for the test case.
2024-02-29 Arnold D. Robbins <arnold@skeeve.com>
Document all the changes needed for MinRX integration.
Some of these are temporary while MinRX is still in C++;
it should eventually be rewritten in C.
* Makefile.am (CCLD): Set to g++ in order to get gawk to link.
* awk.h: Include "minrx.h".
(struct Regexp): Add in minrx members.
(restart, re_end, resubpatstart, resubpatend, renumsubpats): Declare
new functions.
(RESTART, RE_END, RESUBPATSTART, RESUBPATEND, RENUMSUBPATS): Change
the macros to call the functions.
(use_gnu_matchers): Declare new variable.
* awkgram.y: Lint warning about /*...*/ comment is only true for
the GNU matchers.
* builtin.c (do_match): `s' is now ssize_t.
(do_sub): Fix value of len parameter in all to research depending upon
the matcher.
* configure.ac: Set CXXFLAGS if we have C++ versions of the files.
* main.c (use_gnu_matchers): Define, set to false.
(main): Check for GAWK_GNU_MATCHERS env var and set use_gnu_matchers
to true if it exists.
(parse_args): Temporarily add 'G' flag for using GNU matchers.
* re.c (get_minrx_regerror): New routine.
(make_regexp): Adjust for the fact that we can use either GNU matchers
or minrx.
(research): Ditto.
(restart, re_end, resubpatstart, resubpatend, renumsubpats): New functions.
Eventually these should be turned back into macros.
2024-02-29 Arnold D. Robbins <arnold@skeeve.com>
* re.c (do_sub): Fix arguments in call to SUBPATSTART().
2024-02-23 Arnold D. Robbins <arnold@skeeve.com>
Another attempt to fix the race conditions on SIGPIPE. See
https://lists.gnu.org/archive/html/bug-gawk/2023-12/msg00011.html.
* awk.h (do_nothing_on_signal): Remove declaration.
(wait_any): Add declaration.
* builtin.c (gawk_system): New function.
(do_system): Call it.
* io.c (wait_any): Make extern, not static.
(gawk_popen_write, gawk_popen_write_close): New functions.
(redirect_string): Call gawk_popen_write.
(close_rp): Call gawk_popen_write_close.
(do_nothing_on_signal): Removed.
(getdtablesize): Define as a macro if needed.
* configure.ac: Add check for getdtablesize.
* awk.h, builtin.c, io.c, node.c, gawkapi.h, configure.ac: Updated
copyright years also.
2024-02-01 Arnold D. Robbins <arnold@skeeve.com>
* .gitignore: Ignore .dirstamp files. Thanks to "M"
<c24p0101@gmail.com> for the suggestion.
2024-01-28 Arnold D. Robbins <arnold@skeeve.com>
* node.c (str2wstr): Always allocate the array for use
by do_match(). Bug reported by Sam James <sam@gentoo.org>.
Thanks to Miguel Pineiro Jr. <mpj@pineiro.cc> for the
simplified test case.
2024-01-20 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: MPFR on 32 bit Power PC is back
to working. Remove the check that disables it.
* NEWS: Updated.
2024-01-16 Andrew J. Schorr <aschorr@telemetry-investments.com>
* gawkapi.h: Add a comment suggesting that do_flags be converted
to a bitmask if we ever need to add another value, since the current
approach breaks ABI compatibility for each new flag.
2024-01-12 Arnold D. Robbins <arnold@skeeve.com>
Use newer interfaces for getting the time if they're available.
Thanks to Andreas Schwab <schwab@suse.de> and Andrew Schorr.
* builtin.c (do_systime): Use clock_gettime() or gettimeofday() or
time(), in that order.
* configure.ac: Add checks for clock_gettime() and gettimeofday().
2023-12-26 Arnold D. Robbins <arnold@skeeve.com>
* gawkapi.c (api_get_argument): Sync handling of AWK_UNDEFINED
with the doc.
(node_to_awk_value): Similar.
2023-12-25 Arnold D. Robbins <arnold@skeeve.com>
* debug.c (serialize_list): Fix a compiler warning about
an unused variable.
2024-12-25 Andrew J. Schorr <aschorr@telemetry-investments.com>
* io.c (get_read_timeout): Make gawk match the doc and have
GAWK_READ_TIMEOUT work if no entry in PROCINFO. Thanks
to Ed Morton for the report.
2023-12-24 Arnold D. Robbins <arnold@skeeve.com>
* debug.c (print_array): Use the subscript as the
subarray name instead of the `vname' field. This field
is not there for arrays created with NUMINT subscripts.
Thanks again to Hermann Peifer for the report.
Unrelated:
* debug.c (find_subscript): Fail early on Node_var_new
and Node_elem_new. Thanks yet again to Hermann Peifer
for the report.
2023-12-21 Arnold D. Robbins <arnold@skeeve.com>
Fix multidimensional array printing in the debugger, again.
Thanks again to Hermann Peifer for the report.
* debug.c (print_array): Maintain a stack of array names
and indexes, and print it when printing the value.
(print_array_names): New function.
2023-12-21 Arnold D. Robbins <arnold@skeeve.com>
Try to close a race condition window on SIGPIPE. See the thread at
https://lists.gnu.org/archive/html/bug-gawk/2023-12/msg00011.html.
Thanks to Andreas Schwab <schwab@suse.de>.
* awk.h (ignore_sigpipe, set_sigpipe_to_default): Remove decls,
they're not functions.
(do_nothing_on_signal): Add declaration.
(silent_catch_sigpipe): New macro.
* builtin.c (do_system): Replace call of set_sigpipe_to_default()
with call to silent_catch_sigpipe().
* io.c (redirect_string, gawk_popen): Ditto.
(do_nothing_on_signal): New function.
2023-12-12 Eli Zaretskii <eliz@gnu.org>
* io.c (redirect_string): Check registered parsers before failing
due to EISDIR.
2023-12-02 Arnold D. Robbins <arnold@skeeve.com>
* debug.c (print_array): Fix printing of multidimensional arrays.
Thanks to Hermann Peifer for the report.
2023-11-28 Arnold D. Robbins <arnold@skeeve.com>
Straighten out Node_elem_new some more. Thanks to "M"
<c24p0101@gmail.com> for the report and to Andrew Schorr for
the reduced test case.
* awk.h (force_string_fmt): Don't clear NUMBER from the flags.
(force_number): Handle Node_elem_new case. (Ooops.)
* builtin.c (do_typeof): Make unassigned case check smarter.
* mpfr.c (mpg_force_number): Don't clear STRING and don't
change the string value.
* node.c (r_force_number): Ditto.
2023-11-20 Arnold D. Robbins <arnold@skeeve.com>
* command.y (yylex): Change _("%s") into "%s". Sheesh.
2023-11-16 Arnold D. Robbins <arnold@skeeve.com>
* eval.c (cmp_scalars): Check for REGEX as well as STRING
in order to do a string compare. Thanks to
"*" <cl2ap0101@gmail.com> for the report.
2023-11-04 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Fix the spelling of "Ukrainian". Thanks to
Bruce Horrocks.
2023-11-02 Arnold D. Robbins <arnold@skeeve.com>
* 5.3.0: Release tar ball made.
2023-10-31 Corinna Vinschen <vinschen@redhat.com>
* awk.h (wcitomb): Add declaration.
* node.c (parse_escape): Use it on Cygwin.
2023-10-24 Arnold D. Robbins <arnold@skeeve.com>
* eval.c (setup_frame): Fix a potential use after free
error. Thanks to Michal Jaegermann for the report.
2023-10-23 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_sub): Fix a use-after-free issue reported
by Miguel Pineiro Jr. <mpj@pineiro.cc>.
2023-10-22 Arnold D. Robbins <arnold@skeeve.com>
* nonposix.h, re.c: Update copyright year.
2023-10-21 Arnold D. Robbins <arnold@skeeve.com>
* main.c (parse_args): Deal better with things if -p
given twice. Thanks to Hermann Peifer for the report.
* Checklist: Revised a bit.
* awkgram.y (isnoeffect): Add Op_exec_count to the list so
that `gawk --lint --profile BEGIN{0}' also produces a lint
warning. Thanks yet again to Hermann Peifer.
2023-10-19 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
* README: Version bumped to 5.3.0. Updated info about makeinfo
and texinfo.tex versions.
2023-10-17 Arnold D. Robbins <arnold@skeeve.com>
* re.c (make_regexp): Fix setting ok_to_escape.
2023-10-15 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Editing, in preparation for release.
2023-10-12 Eli Zaretskii <eliz@gnu.org>
* node.c (parse_escape) [__MINGW32__]: Use w32_wc_to_lc to convert
Unicode codepoints.
* nonposix.h (w32_wc_to_lc) [__MINGW32__]: Add prototype.
2023-10-08 Arnold D. Robbins <arnold@skeeve.com>
* node.c (parse_escape): Fix warning message for \u escape.
* interpret.h (Op_lint_plus): Improve the test, especially
for if dealing with user input. Thanks to Hermann Peifer
for the report.
2023-10-03 Arnold D. Robbins <arnold@skeeve.com>
* profile.c (pprint): Add case for Op_lint_plus. Thanks to
Hermann Peifer for the report.
2023-09-19 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2023-09-13 Arnold D. Robbins <arnold@skeeve.com>
Fix handling of zero length matching in multibyte locales
with sub/gsub. Thanks again to Ed Morton for the report.
* builtin.c (do_sub): When the match is empty and we copy in
the subsequent character, check gawk_mb_cur_max to know whether
to copy one byte or multiple bytes.
2023-09-11 Arnold D. Robbins <arnold@skeeve.com>
* field.c (do_split): Fix some whitespace.
Unrelated: Add do_csv variable to the API.
* gawkapi.h: Bump major version to 4, minor version back to zero.
(DO_FLAGS_SIZE): Increase to 7.
(gawk_do_csv, do_csv): New macros.
* gawkapi.c (init_ext_api): Init the do_csv element. Replace
numeric indices with symbolic ones.
* NEWS: Updated for do_csv in the API, and the major and
minor API versions.
2023-09-01 Miguel Pineiro Jr <mpj@pineiro.cc>
Fix the handling of zero-length matches in multibyte locales.
Thanks to Ed Morton <mortoneccc@comcast.net> for the report.
* builtin.c (do_match): Translate rstart (byte idx to char idx)
even when rlength is zero. For this we tweak the conversion of
rlength to keep it in bounds when rstart and rlength are both 0.
* node.c (str2wstr): Add an entry to the indices array for the
terminating null. It facilitates the tweak above and is needed
to translate the idx of a zero-width match at the end of the
string.
2023-08-27 Arnold D. Robbins <arnold@skeeve.com>
* re.c (make_regexp): When do_traditional and looking to see
if an escape created a metacharacter, use `metas' instead of
a separate string, since BWK awk now also handles interval
expressions. Found by inspection.
2023-08-25 Miguel Pineiro Jr <mpj@pineiro.cc>
* awk.h (enum escape_results): Fix ESCAPE_LINE_CONINUATION typo.
* node.c (make_str_node): Ditto.
* (parse_escape): Ditto.
* re.c (make_regexp): Ditto.
Unrelated: Use size_t for multibyte character lengths (aside
from it being the type which the interfaces return, using it
avoids implementation-defined behavior when -1 or -2 is cast
to size_t then stored in an int).
* node.c (make_str_node): Change mblen from int to size_t.
Unrelated: Change the length of a translated escape sequence
from int to size_t (to match the underlying interfaces).
* awk.h (parse_escape): Change prototype.
* node.c (make_str_node): Adjust to accommodate new type.
* (parse_escape): Ditto.
* re.c (make_regexp): Ditto.
Unrelated: Clarify make_regexp, inspired by make_str_node.
* re.c (make_regexp): Copy multibyte characters at once, move
the escape sequence post-processing into the switch statement
to explicitly connect a return value with its handling, add a
default case to catch unrecognized values, and move ok_to_escape
next to where it's used (and only set it once).
2023-08-21 Arnold D. Robbins <arnold@skeeve.com>
* node.c (make_str_node): Don't use N_() in cant_happen message.
* re.c (make_regexp): Ditto.
2023-08-20 Arnold D. Robbins <arnold@skeeve.com>
Change parse_escape() to return an enum instead of relying
on manifest constants.
* awk.h (enum escape_results): New enum.
(parse_escape): Now returns a value from it, add nbytes pointer
parameter.
* node.c (make_str_node): Revised to use switch/case on
parse_escape return value.
(parse_escape): Update code.
* re.c (make_regexp): Revised to use switch/case on parse_escape
return value.
2023-08-20 Miguel Pineiro Jr <mpj@pineiro.cc>
* awk.h (ELIDE_BACK_NL): Removed.
(parse_escape): Update declaration.
* main.c (cmdline_fs): Don't use ELIDE_BACK_NL.
(arg_assign): Ditto.
* node.c (make_str_node): Rework handling of return value
from parse_escape.
(parse_escape): Change return code to convey more info,
fully parse escapes and handle conversion from unicode to
UTF-8.
* re.c (make_regexp): Rework for new interface to parse_escape.
2023-07-09 Arnold D. Robbins <arnold@skeeve.com>
* re.c (make_regexp): In error message, use the original text
directly as a string and not with calculated length. Thanks
to Neil R. Ormos <ormos-gnulists17@ormos.org> for the report.
2023-06-27 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2023-06-22 Arnold D. Robbins <arnold@skeeve.com>
* field.c (comma_parse_field): Don't set the field if len
is zero. Kudos to BWK's awk for finding this gawk bug.
2023-06-12 Arnold D. Robbins <arnold@skeeve.com>
* main.c (cmdline_fs): Add ELIDE_BACK_NL to flags in call
to make_str_node(). Fixes weirdness with -F. Thanks to
Denys Vlasenko <dvlasenk@redhat.com> for the report.
2023-06-04 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_sub): Fix a Day 1 gensub bug, reported by
Denys Vlasenko <dvlasenk@redhat.com>.
2023-06-02 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2023-05-07 Arnold D. Robbins <arnold@skeeve.com>
* 5.2.2: Release tar ball made.
2023-05-07 Arnold D. Robbins <arnold@skeeve.com>
* README: Update version.
* debug.c: Update copyright year.
2023-04-22 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y, builtin.c, debug.c, eval.c, interpret.h,
profile.c: Fix some implicit fall through warnings.
Thanks to "Jannick" <thirdedition@gmx.net> for the report.
2023-04-21 Arnold D. Robbins <arnold@skeeve.com>
* array.c (asort_actual): Initialize `value' to avoid
compiler complaints. Thanks to Michal Jaegermann
for the report.
* main.c (usage): Cast array subscript to int to silence
a compiler warning. Thanks to "Jannick" <thirdedition@gmx.net>
for the report.
2023-04-14 Arnold D. Robbins <arnold@skeeve.com>
* array.c (asort_actual): Handle Node_elem_new. Add braces
for scoping, fixes tcc compiler error (GCC is too permissive!).
Update copyright year.
* Makefile.am, configure.ac, awk.h, awkgram.y, builtin.c,
command.y, eval.c, field.c, io.c, msg.c, node.c, profile.c,
symbol.c: Update copyright year.
2023-04-13 Arnold D. Robbins <arnold@skeeve.com>
* array.c (asort_actual): Handle Node_var_new. Can happen
if you sort SYMTAB. Thanks to zhou shuiqing
<zhoushuiqing321@outlook.com> for the report.
2023-04-13 Andrew J. Schorr <aschorr@telemetry-investments.com>
* io.c (csvscan): Convert CR-LF pairs to plain LF, both at the
end of the line and when embedded. Plain CRs are not touched.
2023-04-07 Andrew J. Schorr <aschorr@telemetry-investments.com>
* io.c (csvscan): Instead of stripping all carriage returns in the
input, simply include the CR in the RT if it occurs just before
the linefeed character.
2023-04-07 Arnold D. Robbins <arnold@skeeve.com>
* array.c (assoc_info): Update to handle additional cases so
that one may dump SYMTAB or FUNCTAB. Issue reported by
zhou shuiqing <zhoushuiqing321@outlook.com>.
2023-03-26 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2023-03-24 Arnold D. Robbins <arnold@skeeve.com>
Add support for \u escape sequence.
* awk.h (parse_escape): New bool* parameter for unicode.
* node.c (make_str_node): Deal with a \u escape.
(parse_escape): Handle \u.
* re.c (make_regexp): Deal with a \u escape.
2023-03-24 Arnold D. Robbins <arnold@skeeve.com>
* io.c (set_RS): Fix first time and warning logic for csv mode.
2023-03-23 Arnold D. Robbins <arnold@skeeve.com>
* io.c (csvscan): Strip CRs.
2023-03-21 Manuel Collado <mcollado2011@gmail.com>
* io.c (csvscan): Real implementation.
2023-03-18 Miguel Pineiro Jr. <mpj@pineiro.cc>
Allow records longer than INT_MAX. For test cases,
see https://lists.gnu.org/archive/html/bug-gawk/2021-05/msg00003.html.
* awk.h (set_record): `cnt' is now size_t.
* field.c (set_record): `cnt' is now size_t. Adjust databuf_size
and MAX_SIZE macro.
* io.c (get_a_record): Add new second parameter, size_t *len,
which holds the size. Adjust return value meaning. Deal with
return from an extension's get_a_record() function.
(do_getline_redir): Adjust calling get_a_record().
(do_getline): Ditto.
(inrec): Ditto.
2023-03-17 Arnold D. Robbins <arnold@skeeve.com>
* field.c (do_split): Modify behavior for --csv.
2023-03-16 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (enum do_flag_values): Add DO_CSV.
(do_csv): New macro.
(init_csv_fields, init_csv_records): Add declarations.
* field.c (init_csv_fields): New function.
(set_parser): Don't set the parser if doing CSV. Add warnings.
* io.c (csvscan): New function (placeholder for now).
(init_csv_records): New function.
(set_RS): Don't set the parser if doing CSV. Add warnings.
* main.c (optab): Add new options -k/--csv.
(main): Fatal out if --posix and --csv. Call init_csv_records()
and init_csv_fields().
(usage): Add a line for the new options.
(load_procinfo): Install PROCINFO["CSV"] if doing CSV.
(parse_args): Update for new options.
2023-03-09 Arnold D. Robbins <arnold@skeeve.com>
* gawkapi.h: Update copyright year. Small edit in leading comment.
Unrelated:
* NEWS: Updated. Update copyright year.
2023-03-07 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y (optimize_assignment): Turn Op_store_field_exp
into Op_store_field. Add some explanatory comments.
2023-03-02 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): Revert this change of 2023-01-23,
it's handled by the fix of 2023-02-27.
| * interpret.h (r_interpret): Op_assign case, don't unref the
| old lhs NODE until after the assignment. See test case in
| the comment in the code and test/fieldassign.awk. Thanks to
| Yasuhiro Yamada <yamada@gr3.ie> for the report.
* gawkapi.h: Add a note at the top pointing to the online
documentation, warning that the header file's comments aren't
enough. Thanks to Eli Zaretskii <eliz@gnu.org> for the suggestion.
2023-02-27 Arnold D. Robbins <arnold@skeeve.com>
* gawkmisc.c: Use #elif. Update copyright year.
Unrelated: Fix handling of $a = $b so that it's done OK also in
expressions, and not just in statements. This fixes a memory
leak seen with valgrind in test/fieldassign.awk.
* awk.h (enum opcodeval): Add new member, Op_store_field_exp.
* awkgram.y (mk_assignment): Generate Op_store_field_exp under the
right circumstances.
(optimize_assignment): Remove case for Op_field_spec_lhs.
* eval.c (optypes): Add entry for Op_store_field.
* interpret.h (r_interpret): Handle Op_store_field in the switch.
* profile.c (pprint): Same.
Unrelated: Improve how information is passed to input parsers:
* io.c (iop_alloc): lstat() or stat() the file if could not be opened.
See the doc for explanation as to why.
* NEWS: Updated.
2023-02-26 Arnold D. Robbins <arnold@skeeve.com>
* Multiple files: Remove trailing whitespace.
2023-02-24 Arnold D. Robbins <arnold@skeeve.com>
* field.c (set_comma_field): Free the string node. Fixes a
memory leak. Thanks to Andrew Schorr for the report.
2023-02-24 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (boolval): Handle Node_var. Thanks to
Daniel Pouzzner <douzzer@mega.nu> for the report, relayed
by Andrew Schorr.
* builtin.c (do_typeof): Handle Node_var, could come from a
converted Node_elem_new. Thanks to Andrew Schorr for the
report.
2023-02-23 Arnold D. Robbins <arnold@skeeve.com>
* io.c (find_input_parser): Improve check to see if iop is
already being handled by an input parser. Found by
code inspection.
Unrelated:
* awk.h (struct redirect): Change RED_NOBUF to RED_FLUSH.
* builtin.c (efwrite): Ditto
* io.c (redflags2str, check_duplicated_redirections,
redirect_string): Ditto.
2023-02-12 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2023-02-10 Arnold D. Robbins <arnold@skeeve.com>
* io.c (avoid_flush): New function.
(redirect_string): Use it when setting up an output pipeline.
Thanks to Alexandre Ferrieux <alexandre.ferrieux@orange.com>
for pointing out the performance issue.
2023-02-09 Arnold D. Robbins <arnold@skeeve.com>
* main.c (usage): Include URL for source code in the message.
2023-02-05 Arnold D. Robbins <arnold@skeeve.com>
* io.c (do_getline_redir): If `s' from get_a_record is NULL,
pass "" to make_string() instead. Avoids undefined behavior
complaint from GCC -fsanitize=undefined. Thanks to Sam James
<sam@gentoo.org> for the report.
(do_getline): Ditto.
2023-02-03 John E. Malmberg <wb8tyw@qsl.net>
* custom.h (bool) OpenVMS compilers may have a bool type
2023-02-02 John E. Malmberg <wb8tyw@qsl.net>
* custom.h (bool) OpenVMS compilers do not have a bool type
2023-01-31 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): For op_subscript, upref the variable
if it's Node_var, also. Thanks to Volodymyr Gubarkov <xonixx@gmail.com>
for the initial report and to Vincent Forest
<vincent.forest@meso-star.com> for help in getting a small
reproducer.
Also, update the copyright year.
* interpret.h (r_interpret): Op_assign case, don't unref the
old lhs NODE until after the assignment. See test case in
the comment in the code and test/fieldassign.awk. Thanks to
Yasuhiro Yamada <yamada@gr3.ie> for the report.
2023-01-22 Arnold D. Robbins <arnold@skeeve.com>
* main.c: Update copyright year.
(check_pma_security): New routine.
(main): Call it.
(UPDATE_YEAR): Bump to 2023.
* NEWS: Updated.
2023-01-03 Arnold D. Robbins <arnold@skeeve.com>
* README: Updated with texinfo.tex and makeinfo versions.
Update copyright year.
2022-12-27 Arnold D. Robbins <arnold@skeeve.com>
* main.c (extra_stack): Remove variable and its uses. Left over
from libsigsegv cleanup.
2022-12-23 Arnold D. Robbins <arnold@skeeve.com>
* README: Updated with texinfo.tex version.
2022-12-22 Arnold D. Robbins <arnold@skeeve.com>
* README: Updated with texinfo.tex version.
* NEWS: Updated.
2022-12-11 Arnold D. Robbins <arnold@skeeve.com>
* main.c (load_procinfo): Remove "mb_cur_max" and "utf8"
entries, they're not of any real use.
* NEWS: Updated.
2022-12-11 Arnold D. Robbins <arnold@skeeve.com>
* main.c (get_pma_version): Add declaration.
(load_procinfo): Use it to set PROCINFO["pma"] if PMA is in use.
Thanks to J Naman for the feature suggestion.
* NEWS: Updated.
2022-12-04 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
* awk.h (nextfree): Include dimension so we know how big it is.
(pma_save_free_lists): Declare new function.
* main.c (main): Call pma_save_free_lists() via atexit() upon exit.
(parse_args): Call pma_mpfr_check() unconditionally.
* symbol.c (struct root pointer): Add copy of nextfree[] array.
(init_symbol_table): On first alloc, zero out the root_pointers
struct. On subsequent uses, copy saved nextfree into real nextfree.
(pma_mpfr_check): Check using_persistent_malloc and return early if
not using it.
(pma_save_free_lists): New function to save nextfree.
2022-12-01 Arnold D. Robbins <arnold@skeeve.com>
Disallow using / not using -M across invocations of gawk with
a persistent heap.
* awk.h (pma_mpfr_check): Add declaration.
* symbol.c (pma_mpfr_check): New function. Enhance root pointer
structure.
* main.c (parse_args): Call pma_mpfr_check() if using persistent
malloc.
2022-12-01 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: Removed use of libsigsegv.
* main.c: Removed code related to libsigsegv.
* NEWS: Updated.
2022-11-25 Arnold D. Robbins <arnold@skeeve.com>
* field.c (comma_parse_field): Remove unused variable `field'.
2022-11-23 Arnold D. Robbins <arnold@skeeve.com>
* msg.c: Update to GPL 3. Thanks to Corinna Vinschen for
noticing the GPL version issue.
* Makefile.am (install-exec-hook): Check for existence of the
gawk-x.y.z file before hard linking. Otherwise make in POSIX
mode is unhappy. Thanks to Tzvetelin Katchov <katchov@gnu.org>
for the report.
2022-11-17 Arnold D. Robbins <arnold@skeeve.com>
* 5.2.1: Release tar ball made.
2022-11-17 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: Update version.
* README: Updated.
2022-11-01 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2022-10-28 Arnold D. Robbins <arnold@skeeve.com>
* Makefile.am (DEBUG): New variable with debug compilation flags.
* configure.ac: Rework debug compilation to use DEBUG in Makefile.am
and to add %no-lines to grammar files. Works around GDB issue
with macros and #line directives. Thanks to Ulrich Drepper
for the help and for the suggestions.
2022-10-26 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: Try to get macros to work in the debugger.
2022-10-25 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): For Op_push*, simplify the code for
Node_var_new and Node_elem_new.
2022-10-23 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
* debug.c (execute_code): Comment out declaration and assignment of
save_stack_size variable. It's not used anymore.
* configure.ac: Sync code for .developing with extension/configure.ac.
2022-10-23 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y (mk_binary): Check that divisor is a number before
checking it's zero. Thanks to Ben Hoyt <benhoyt@gmail.com> for
the report.
2022-10-20 Paul Eggert <eggert@cs.ucla.edu>
Port to current Gnulib
Adjust to current Gnulib, which uses C23 style keywords for bool,
true, false and static_assert, and assumes that <config.h>
supplies compatibility macros on older systems.
* awk.h: Do not include stdbool.h.
* configure.ac: Use gl_C_BOOL, not AC_HEADER_STDBOOL.
2022-10-20 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2022-10-14 Andrew J. Schorr <aschorr@telemetry-investments.com>
* awkgram.y: Use Node_var_new instead of Node_var_array for
a subscript expression.
* gawkapi.c (api_sym_update): Make things smarter for dealing
with an untyped variable that needs to become an array.
2022-10-14 Arnold D. Robbins <arnold@skeeve.com>
* symbol.c (make_symbol): Save and restore symbol_table in
the root_pointers struct instead of using lookup.
(func_count, var_count): Removed the variables and their uses.
(get_symbols): Getting functions, use the_table->table_size for
the count, not func_count.
2022-10-11 Arnold D. Robbins <arnold@skeeve.com>
* symbol.c (get_symbols): Getting variables, use
the_table->table_size for the count, not var_count. Bug
was tickled when using --dump-var together with persistent
memory. (Why didn't valgrind catch this?) Thanks to Hermann
Peifer for the report. Thanks to Terence Kelly for detective work
that helped pinpoint the search area for the bug.
2022-09-30 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): For the `push' opcodes with
Node_elem_new, deref `m' before assigning m = dupnode(Nnull_string).
Thanks to Andrew Schorr and valgrind for the report.
2022-09-23 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): For the `push' opcodes with
Node_elem_new, put dupnode(Nnull_string) on the stack as
the local value. Thanks to Jeff Chua <jeff.chua.linux@gmail.com>
for the report.
2022-09-22 Arnold D. Robbins <arnold@skeeve.com>
* gawkbug.in: Add attestation about not modifying the sources
before building.
2022-09-19 Arnold D. Robbins <arnold@skeeve.com>
Fix negative NaN issues on RiscV. See bug list citation
in the code. Thanks to Andreas Schwab <schwab@suse.de>
for the report.
* eval.c (fix_nan_sign): New function.
* interpret.h (r_interpret): Use it for plus and minus cases.
2022-09-16 Arnold D. Robbins <arnold@skeeve.com>
* debug.c (do_eval): If a fatal error occurs, don't attempt to
recover, just restart. Thanks to Pascal Maugis
<pascal.maugis@lsce.ipsl.fr> for the report.
2022-09-14 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): For the `push' opcodes, don't upref
the Node_elem_new. Instead, convert the very original value to
Node_var and use elem_new_to_scalar for the local variable.
See test case elemnew3.awk. Thanks to Pascal Terjan <pterjan@gmail.com>
for the report and test case.
2022-09-14 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): For the `push' opcodes, upref
the Node_elem_new. See test case elemnew1. Thanks to
Emanuel Attila Czirai <corre.a.buscar@gmail.com> for the report,
and to Jan Alexander Steffens (heftig)" <heftig@archlinux.org>
for the fix.
2022-09-04 Arnold D. Robbins <arnold@skeeve.com>
* 5.2.0: Release tar ball made.
2022-09-04 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated with real URL of "MPFR is on parole."
* configure.ac: Bump to release version.
2022-09-02 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.c, command.c: Regenerated, using Bison 3.8.2.
2022-09-01 Arnold D. Robbins <arnold@skeeve.com>
* field.c (do_split): Check for not do_posix if separator is comma.
2022-08-31 Arnold D. Robbins <arnold@skeeve.com>
* field.c (do_split): Make "," as third arg do CSV parsing.
2022-08-30 Arnold D. Robbins <arnold@skeeve.com>
* field.c (comma_parse_field): Add resizing of new field buffer;
No Arbitrary Limits!
2022-08-29 Arnold D. Robbins <arnold@skeeve.com>
* field.c (comma_parse_field, set_comma_field): New functions.
(set_FS): If FS is "," and not posix mode, use CSV parsing.
Code follows what BWK's new code does.
2022-08-25 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y (yyerror): Exit at the end, to make syntax errors
immediately fatal. This prevents problems with fuzzer-generated
programs at the cost of not continuing to diagnose subsequent
problems. Oh well, too bad. Hooray for the fuzzers. Bleah.
* NEWS: Updated.
2022-08-17 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Mention that VAX/VMS is no longer supported.
* main.c (deprecate_mpfr): Removed the function for now. It
was causing grief for the translators.
* builtin.c (call_sub): Fix message text for indirect call
to gensub. Thanks to Benno Schulenberg
<coordinator@translationproject.org> for the report.
2022-08-16 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: Remove checks for memset_ulong and memcpy_ulong.
* awk.h: Remove checks and macro definitions for those two
functions. They used to be used in the PC code but were
removed.
2022-08-14 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y (mk_binary): Check for division by constant zero in
`/' and `%' when numerator is not a constant. Thanks to Ed Morton
<mortoneccc@comcast.net> for the report.
2022-08-12 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Small update: mention pm-gawk.1 man page.
2022-08-11 Arnold D. Robbins <arnold@skeeve.com>
* main.c (get_pma_version): Check strchr() return value for
close parenthesis. Ooops.
2022-08-10 Arnold D. Robbins <arnold@skeeve.com>
* symbol.c (init_the_tables): New function.
(init_symbol_table): Rework to make functions persistent also
by saving in the root pointer a struct pointing to both the
global table and the function table, and restoring this
on subsequent runs. The code ended up being nicely cleaner.
2022-08-08 Arnold D. Robbins <arnold@skeeve.com>
* awk.h: Include "floatmagic.h" here so that we no longer have to
worry about it on an individual file basis.
* builtin.c, eval.c, node.c: Remove the include in these files.
* configure.ac: Check for isnan in libm also. This may not
be the right way to do this...
2022-08-08 Eli Zaretskii <eliz@gnu.org>
* nonposix.h [__MINGW32__]: Add prototype for strsignal.
* awk.h (flags): Rename BOOL => BOOLVAL, to avoid conflicts with
MS-Windows headers. All users changed.
2022-08-07 Arnold D. Robbins <arnold@skeeve.com>
* main.c (get_pma_version): New function.
(version): Call it.
Unrelated:
* eval.c: Include "floatmagic.h" as it also calls isnan().
Thanks to John Malmberg for the report.
Unrelated:
* main.c (load_environ): Force ENVIRON_node to point to
str_array_func before loading it. Fixes issues on MinGW
while still solving the original `environ[0] = "0=foo"'
problem. Move the call to init_env_array back to after the
loop.
2022-08-06 John E. Malmberg <wb8tyw@qsl.net>
* custom.h: Update for new VMS Versions
2022-08-03 Arnold D. Robbins <arnold@skeeve.com>
* mpfr.c (mpfr_mem_alloc, mpfr_mem_realloc, mpfr_mem_free):
New functions that just call malloc/realloc/free. Those are
redefined to the pma_* functions if using pma.
* awk.h (mpfr_mem_alloc, mpfr_mem_realloc, mpfr_mem_free):
Add declarations.
* main.c (main): Call mp_set_memory_functions with the new
functions as arguments.
2022-08-03 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (format_tree): When collecting positional field widths or
precisions, check for wrap around to negative values. Thanks to
YU Jiongchi <jcyu.2022@phdcs.smu.edu.sg> for the report.
2022-07-29 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (efwrite): Check ferror() also, fixes some weird
cases. Reported by Nikita Zlobin <cook60020tmp@mail.ru>.
* io.c (gawk_fwrite): Remove static from definition.
* awk.h (gawk_fwrite): Add declaration.
2022-07-27 Arnold D. Robbins <arnold@skeeve.com>
* gawkapi.c (awk_value_to_node): Put tval declaration inside
MPFR ifdef. Avoids warning if no MPFR when compiling.
2022-07-26 Arnold D. Robbins <arnold@skeeve.com>
* io.c (iopflags2str): Remove 'static' from declaration, causes
the compiler to stop complaining. Thanks to Pat Rankin for the
suggestion.
* re.c (reflags2str): Ditto.
* debug.c (debug_prog): Translate the "Restarting ..." message.
Thanks to Antonio Colombo for pointing this out.
2022-07-22 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2022-07-21 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2022-07-19 Arnold D. Robbins <arnold@skeeve.com>
* main.c (catchsig): Adjust error message in cant_happen() call.
* configure.ac: Add check for strsignal function. It doesn't
exist on z/OS.
* replace.c: Include missing_d/strsignal.c if needed.
2022-07-18 Arnold D. Robbins <arnold@skeeve.com>
* main.c (main): Don't use fatal() if pma_init() fails, apparently
it tries deep down to allocate memory. Print return value and
pma_errno value. Thanks to Terence Kelly for the analysis and
suggestion.
* custom.h (pma_errno): Define to 0 if not using persistent malloc.
2022-07-17 Arnold D. Robbins <arnold@skeeve.com>
* main.c (parse_args): Handle --persist with no argument to issue a
better error message. Thanks to Terence Kelly for the suggestion.
2022-07-14 Arnold D. Robbins <arnold@skeeve.com>
* gawkmisc.c, replace.c: Update copyright.
2022-07-11 Arnold D. Robbins <arnold@skeeve.com>
* Makefile.am (base_sources): Remove mbsupport.h.
2022-07-07 Arnold D. Robbins <arnold@skeeve.com>
* mbsupport.h: Removed.
2022-07-06 Arnold D. Robbins <arnold@skeeve.com>
* README: Updated. Remove mention of DJGPP.
* awk.h, custom.h, eval.c, gawkmisc.c, io.c, main.c, nonposix.h,
re.c, replace.c: Remove all code for __DJGPP__.
2022-07-03 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (TOP_SCALAR): Put converted value back on the stack
for any subsequent use.
2022-06-30 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (POP_SCALAR): Convert Node_elem_new into a scalar.
(TOP_SCALAR): Ditto.
Unrelated:
* README: Updated. Remove mention of OS/2.
* awk.h, awkgram.y. gawkmisc.c, io.c, main.c, nonposix.h:
Remove all code for __EMX__.
2022-06-20 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: MPFR is now "on parole" instead of deprecated.
* main.c (deprecate_mpfr): ifdef'ed out.
(parse_args): Comment out call to deprecate_mpfr().
2022-06-20 Arnold D. Robbins <arnold@skeeve.com>
* main.c (optab): Add --persist.
(parse_args): For --persist, fatal message to use the
environment variable. Thanks to Terence Kelly for suggesting
the option, since it's mentioned in his paper.
2022-06-17 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Typo fix.
* main.c (main): Correct the return value check in the call
to pma_init(). Thanks to Terence Kelly for the report.
2022-06-15 Andrew J. Schorr <aschorr@telemetry-investments.com>
* main.c (load_environ): Init ENVIRON_node before filling
it. Fixes a crash if something like "0=foo" is the first thing
in the environment. Thanks to David Arroyo <david@aqwari.net>
for the report.
2022-06-14 Arnold D. Robbins <arnold@skeeve.com>
* main.c (main): Set gawk_name first thing, in case we fatal out
on pma_init.
* configure.ac: Add --disable-pma option.
2022-06-13 Arnold D. Robbins <arnold@skeeve.com>
* main.c (main): If persistent memory is not available and
GAWK_PERSIST_FILE is supplied, issue a warning. Similarly, don't
do mtrace if using persistent malloc.
(version): Include PMA version in the output.
2022-06-10 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (using_persistent_malloc): Declare new variable
* main.c (main): Check for error from pma_init. Set
using_persistent_malloc.
* symbol.c (init_symbol_table): Check using_persistent_malloc
appropriately. Adjust the flow for it.
* NEWS: Updated.
* Makefile.am (LDADD): Add $(LDFLAGS).
* custom.h (pma_init): Define to 0.
2022-06-09 Arnold D. Robbins <arnold@skeeve.com>
* custom.h: Deal with use/non-use of persistent malloc.
* main.c (main): Bracket mtrace call in #ifndef. Call pma_init()
first thing.
* configure.ac (GAWK_USE_PERSISTENT_MALLOC): Add call.
2022-06-06 Andrew J. Schorr <aschorr@telemetry-investments.com>
* gawkapi.h (gawk_api_t): Add new api_destroy_array hook to clear
and free an array.
(destroy_array): New macro wrapper for api_destroy_array.
* gawkapi.c (api_destroy_array): Add new function to clear and free
an array.
(api_impl): Add hook for api_destroy_array.
2022-06-03 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (elem_new_to_scalar): Add declaration.
* eval.c (elem_new_to_scalar): No longer static. Make it handle
things when valref is > 1. Returns a NODE * instead of being void.
(cmp_scalars): Change usage of elem_new_to_scalar().
* interpret.h (r_interpret): Op_push*: Make Node_elem_new into
a separate case, using elem_new_to_scalar().
* array.c (force_array): For Node_elem_new, clear symbol->stptr.
* eval.r (r_get_lhs): Use efree() instead of free() in Node_elem_new
case.
2022-06-02 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_sub): Fix memory corruption when substituting
into a value that is a typed regex. Thanks to valgrind.
2022-06-02 Arnold D. Robbins <arnold@skeeve.com>
Fix a nasty memory corruption problem. Thanks to
Andrew Schorr for the report.
* eval.c (r_get_lhs): Handle Node_elem_new separately
to free the string value. This was not the problem.
* interpret.h (r_interpret): For Op_assign, when turning
Node_elem_new into the null string, DEREF the old value
and dupnode the new one. THIS was the problem.
2022-05-24 Adam Van Scyoc <avanscy@g.clemson.edu>
* awkgram.y (make_assignable): Handle Op_field_assign.
Avoids a core dump upon a certain kind of syntax error;
see test/getlnfa.awk.
2022-05-22 Arnold D. Robbins <arnold@skeeve.com>
* array.c (force_array): When converting Node_elem_new to an array,
release the string storage. Thanks to valgrind, by way of Andrew
Schorr, for the report.
2022-05-11 Arnold D. Robbins <arnold@skeeve.com>
* profile.c (pp_list): Typo fix in a comment.
2022-05-10 Arnold D. Robbins <arnold@skeeve.com>
* gawkbug.in: Add attestation for having read bug reporting
section in the manual. (Yes, I'm pretty paranoid.)
2022-05-03 Arnold D. Robbins <arnold@skeeve.com>
* main.c (deprecate_mpfr): New function.
(parse_args): Call it for -M.
2022-04-28 Arnold D. Robbins <arnold@skeeve.com>
Disallow SYMTAB["x"]["y"]. Reported by Jason C. Kwan.
* interpret.h (r_interpret): Add checks at Op_sub_array case.
Unrelated. Fix MPFR mode unary minus applied to a variable with
a zero MPG value. Also reported by Jason C. Kwan.
* mpfr.c (mpg_interpret): Fix Op_unary_minus to create an
MPFR -0 value. MPG doesn't distinguish negative zero from
positive zero.
2022-04-21 Arnold D. Robbins <arnold@skeeve.com>
Fix some profiling issues related to comments in switch / case
statements.
* awkgram.y (merge_comments): Remove newline between comments being
merged.
* profile.c (pprint): Improve code for Op_K_switch, Op_K_case.
2022-04-20 Arnold D. Robbins <arnold@skeeve.com>
* mpfr.c (mpg_force_number): Small whitespace fix.
* node.c (r_force_number): Ditto.
2022-04-18 Arnold D. Robbins <arnold@skeeve.com>
* gawkbug.in: Small changes.
2022-04-06 Arnold D. Robbins <arnold@skeeve.com>
Further cleanups.
* array.c (new_array_element): Set s->stfmt to STFMT_UNUSED.
* awk.h (force_string_fmt): No need to set s->stfmt anymore.
* builtin.c (do_typeof): Collapse Node_var_new and Node_elem_new
cases.
* symbol.c (print_vars): Node_elem_new can't be in a top level
variable. Remove check and add a call to cant_happen if some
other node type is received.
* NEWS: Updated.
2022-04-01 Arnold D. Robbins <arnold@skeeve.com>
Small fix from the persistent-gawk guys.
* profile.c (pprint): Use estrdup instead of strdup.
2022-03-31 Arnold D. Robbins <arnold@skeeve.com>
* array.c (new_array_element): Make a full null string instead of
an empty node; it can be more easily converted to scalar that way.
* awk.h (force_string_fmt): Change the node type, remove NUMBER flag,
set STFMT_UNUSED.
* cint_array.c (lear_clear): Remove check for not equal Node_elem_new.
* int_array.c (int_clear): Ditto.
* str_array.c (str_clear): Ditto.
* eval.c (elem_new_to_scalar): Now only needs to change the node type.
* interpret.h (r_interpret): Treat Node_elem_new like Node_val, and
UPREF it when pushed. Remove test before DEREFing when popping.
* mpfr.c (mpg_force_number): Simplify converting Node_elem_new to
Node_val. Fix the string value.
* node.c (r_force_number): Ditto.
2022-03-28 Arnold D. Robbins <arnold@skeeve.com>
Allow unreferenced array elements to turn into subarrays.
* awk.h (Node_elem_new): New node type.
(new_array_element): Add declaration.
(POP_ARRAY): Handle Node_elem_new.
(force_string_fmt): Ditto. Make sure valref has the right value.
* array.c (force_array): Handle Node_elem_new.
(do_sort_up_value_type): Ditto.
(new_array_element): New function, creates Node_elem_new node.
* builtin.c (do_length): Handle Node_elem_new.
(do_typeof): Ditto.
* cint_array.c (leaf_lookup): Use new_array_element() when creating
a new element.
(left_clear): Don't unref Node_elem_new.
* debug.c (print_symbol, do_set_var, watchpoint_triggered,
initialize_watch_item, print_memory, do_print_f): Handle Node_elem_new.
* eval.c (nodetypes): Add Node_elem_new.
(r_get_lhs, setup_frame): Handle Node_elem_new.
(elem_new_to_scalar): New function.
(cmp_scalars): Use elem_new_to_scalar.
* ext.c (get_actual_argument): Handle Node_elem_new.
* gawkapi.c (api_get_argument, api_set_argument, node_to_awk_value,
api_sym_update): Handle Node_elem_new.
* int_array.c (int_clear): Don't unref Node_elem_new.
(int_insert): Use new_array_element() when creating a new element.
* interpret.h (r_interpret): Handle Node_elem_new as appropriate.
* mpfr.c (mpg_force_number): Handle Node_elem_new.Make sure valref has
the right value.
* node.c (r_force_number): Ditto.
* str_array.c (str_lookup): Use new_array_element() when creating a
new element.
(str_clear): Don't unref Node_elem_new.
* symbol.c (print_vars): Handle Node_elem_new.
2022-03-27 Arnold D. Robbins <arnold@skeeve.com>
Allow nested indirect function calls. Thanks to
Kaz Kylheku <kaz@kylheku.com> for the report.
* awkgram.y (at_seen): Turn into an int that is incremented
and decremented as appropriate.
2022-03-22 Arnold D. Robbins <arnold@skeeve.com>
Make lint checks for builtin functions smarter. Initial
report from Timothy Sample <samplet@ngyro.com>.
* builtin.c (do_fflush): Require string type.
(printf_common): Ditto.
(do_index): Allow strnum for both arguments.
(do_length): Ditto.
(do_strftime): Ditto.
(do_substr): Require numbers for arguments 2 and 3, string or
strnum for argument 1.
(do_match): Require string or strnum for argument 1.
(do_dcgettext): Require strings for all arguments.
(do_bindtextdomain): Ditto.
(do_dcngettext): Require strings for arguments 5, 4, 2, and 1, and
number for argument 3.
2022-03-03 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2022-02-27 Arnold D. Robbins <arnold@skeeve.com>
* main.c (usage): Update the message for the gawkbug program.
2022-02-27 Arnold D. Robbins <arnold@skeeve.com>
Add `gawkbug' reporting program.
* gawkbug.in: New file.
* Makefile.am (bin_SCRIPTS): New, for gawkbug.
* configure.ac: Add gawkbug to AC_CONFIG_FILES list.
* NEWS: Updated.
* .gitignore: Updated.
2022-02-27 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (cant_happen): Change to allow a format string and arguments,
in order to provide a clue as to what went wrong, and not just
print "internal error".
* awkgram.y, debug.c, eval.c, interpret.h, io.c, main.c, mpfr.c,
profile.c, re.c, symbol.c: Fix all uses of cant_happen().
2022-02-25 Arnold D. Robbins <arnold@skeeve.com>
Memory issues with MPFR, fix part 2. (Part 1 is in the
extension directory.)
* gawkapi.c (awk_value_to_node): Clear the GMP/MPFR values
returned from the C extension after setting the internal
variable.
2022-02-22 Arnold D. Robbins <arnold@skeeve.com>
Enable interval expressions even for --traditional, as BWK
awk has supported them since 2019.
* NEWS: Updated.
* main.c (main): Add lint warning for --re-interval.
(parse_args): Add explanatory comment.
* re.c (resetup): Update the comment w.r.t. interval expressions.
Change the test for enabled them to if do_traditional.
2022-02-22 Arnold D. Robbins <arnold@skeeve.com>
Fix resource leaks found by Coverity. Thanks to
Jakub Martisko <jamartis@redhat.com> for the report.
* ext.c (make_builtin): Free install_name before returning.
* io.c (remap_std_file): Unconditionally close newfd.
* symbol.c (load_symbols): Unref built_in also.
2022-12-10 Andrew J. Schorr <aschorr@telemetry-investments.com>
* io.c (wait_any): When saving a saved exit status returned by
_cwait or waitpid or wait in struct redirect's status field, we
should actually save the value returned by sanitize_exit_status
instead of the raw status. This fixes a bug whereby a saved status
for a previously exited process was being returned by gawk_pclose
without first being sanitized. Thanks to Jakub Martisko
<jamartis@redhat.com> for reporting the bug.
2022-02-10 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (sanitize_exit_status): Fix formatting.
2022-02-07 Arnold D. Robbins <arnold@skeeve.com>
Continue fixing indirect calls of builtins.
* awk.h (check_exact_args, check_args_min_max): Add declarations.
* builtin.c (check_exact_args, check_args_min_max): New functions.
(do_exp, do_fflush, do_index, do_int, do_isarray, do_length, do_log,
do_sqrt, do_strftime, do_systime, do_mktime, do_system, do_tolower,
do_toupper, do_atan2, do_sin, do_cos, do_rand, do_srand, do_match,
do_sub, do_lshift, do_rshift, do_compl, do_strtonum, do_dcgettext,
do_dcngettext, do_bindtextdomain, do_intdiv, do_typeof): Call
the argument checking functions.
(call_sub, call_match): Manually check argument count.
* field.c (do_split, do_patsplit): Call the argument checking
functions.
* interpret.h (r_interpret): For indirect call of extension functions,
pop the function name off the stack when done.
* mpfr.c (do_atan2, do_mpfr_func, do_mpfr_int, do_mpfr_compl,
do_mpfr_lshift, do_mpfr_rshift, do_mpfr_strtonum, do_mpfr_rand,
do_mpfr_srand, do_mpfr_intdiv): Call the argument checking functions.
2022-02-04 Arnold D. Robbins <arnold@skeeve.com>
Start fixing issues with indirect calls of builtins.
Thanks to Denis Shirokov <cosmogen@gmail.com> for the initial report.
Much more remains to be done.
* builtin.c (do_length): Check number of arguments, fatal if not one.
If passed Node_var_new, turn it into the null string.
* interpret.h (r_interpret): For Op_indirect_call, pop the function
name off the stack.
2022-01-05 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y (change_namespace): New function. Extracted from
set_namespace.
(want_namespace): New variable.
[grammar]: Add explanatory comment about the change.
(yylex): Change the namespace immediately from the lexer.
(set_namespace): Adjust code.
* symbol.c (load_symbols): Strip off leading awk:: so that
PROCINFO["identifiers"] is like SYMTAB and FUNCTAB. This makes
the code match the documentation. Thanks to John Naman
<gawker@703n.com> for the bug report.
2021-12-17 Yuri Gribov <tetra2005@gmail.com>
Make a number of functions and variables static.
* cint_array.c (argv_array_func): Make static.
* debug.c (cur_rule, check_until): Ditto.
* field.c (default_FS, FS_re_yes_case, FS_re_no_case, FS_regexp,
FPAT_re_yes_case, FPAT_re_no_case, invalidate_field0: Ditto.
* io.c (iopflags2str): Ditto.
* main.c (ENVIRON_node): Ditto.
* re.c (reflags2str): Ditto.
* str_array.c (env_array_func): Ditto.
2021-12-08 Andrew J. Schorr <aschorr@telemetry-investments.com>
* gawkapi.c (api_sym_update): Allow undefined Node_var_new variables
to be converted to arrays, otherwise API functions have no way
of creating an array that is referenced in an undefined fashion in
the program. This is somewhat comparable to how the set_argument
API allows an undefined argument variable to be converted to an array.
2021-12-02 Andrew J. Schorr <aschorr@telemetry-investments.com>
* builtin.c (efwrite): Don't use return in call of function
returning void. Some compilers might not like it.
2021-12-01 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (efflush): Don't use return in call of function
returning void. It works, but is funky, and I think some
compilers will complain.
Unrelated. Clean up the calls to w32_maybe_set_errno.
* awk.h (os_maybe_set_errno): Add declaration.
* builtin.c (wrerror): Replaced ifdef'ed code calling
w32_maybe_set_errno() with simple call to os_maybe_set_errno().
* io.c (non_fatal_flush_std_file, close_io): Ditto.
* main.c (usage, copyleft): Ditto.
* nonposix.h (w32_maybe_set_errno): Remove declaration.
2021-11-30 Andrew J. Schorr <aschorr@telemetry-investments.com>
Improve output redirection error handling for problems not detected
until the final flush or close. Thanks to Miguel Pineiro Jr.
<mpj@pineiro.cc> for the bug report and suggesting a fix.
* awk.h (efflush): Add declaration.
* builtin.c (efwrite): Break up into 3 functions by moving the
flushing logic into efflush and the error handling logic into
wrerror.
(wrerror): New function containing the error-handling logic extracted
from efwrite.
(efflush): New function containing the fflush logic extracted from
efwrite.
* io.c (close_redir): Call efflush prior to closing the redirection
to identify any problems with flushing output and to take advantage
of the error-handling logic used for print and printf.
2021-11-21 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_typeof): Make Node_array_ref handling smarter.
See test/stupid5.awk.
2021-11-18 Arnold D. Robbins <arnold@skeeve.com>
* gawkapi.c (awk_value_to_node): Fix handling of MPFR values
coming back from an extension.
2021-11-07 Arnold D. Robbins <arnold@skeeve.com>
* mkinstalldirs: Removed, it wasn't used.
* Makefile.in: Regenerated.
2021-11-03 Sam James <sam@gentoo.org>
* configure.ac: Fix AR_FLAGS assignment typo.
* configure.ac: Rename AR_FLAGS to ARFLAGS.
2021-10-27 Arnold D. Robbins <arnold@skeeve.com>
* 5.1.1: Release tar ball made.
2021-10-24 Arnold D. Robbins <arnold@skeeve.com>
* README: Updated.
2021-10-20 Andrew J. Schorr <aschorr@telemetry-investments.com>
* msg.c (err): Need to check that FNR_node->var_value is non-NULL
before testing with is_mpg_number.
2021-10-18 Arnold D. Robbins <arnold@skeeve.com>
Factor duplicated code out into a separate function.
D.R.Y. principle.
* symbol.c (get_name_from_awk_ns): New function.
(lookup): Use it.
(install): Use it.
* awkgram.y (check_qualified_special): Use awk_namespace instead of a
string constant.
* profile.c (pprint): Ditto.
(adjust_namespace): Ditto. Also, fix the check for the qualified
name to look for the "::" in case one namespace is a prefix of
another.
2021-10-13 Arnold D. Robbins <arnold@skeeve.com>
* README: Update version, copyright year.
* custom.h: Update copyright year.
2021-09-26 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_mktime): Update that ISO 8601 does not allow
hours > 23. Thanks to Nethox <nethox+awk@gmail.com> for the info.
2021-09-23 Arnold D. Robbins <arnold@skeeve.com>
* gawkapi.h: Some clean-up of comments in the header file.
2021-09-12 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h: Remove subscript_in_array variable.
Unrelated:
* builtin.c (add_thousands): Fix insertion of thousands_sep
string. Do it in reverse order.
2021-09-10 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h: For Op_subscript, don't allow references into
FUNCTAB and SYMTAB to create new elements. Thanks to
Jason C. Kwan <jasonckwan@yahoo.com> for the bug report.
Unrelated:
* aclocal.m4: Regenerated: Automake 1.16.4.
* NEWS: Updated.
Unrelated:
* awkgram.y (YYERROR_IS_DECLARED): Added for Bison 3.8.
* command.y (YYERROR_IS_DECLARED): Ditto.
* NEWS: Updated.
2021-09-09 Arnold D. Robbins <arnold@skeeve.com>
Move to Autoconf 2.71 (finally!)
* configure.ac: Updated.
* aclocal.m4, configh.in configure: Regenerated.
* NEWS: Updated.
2021-09-06 Arnold D. Robbins <arnold@skeeve.com>
Have the ' flag work for %d in MPFR mode also. Thanks to
Dan Nielsen <catskill549@yahoo.com> for the report.
* builtin.c (add_thousands): New function.
(reverse): New helper function.
(format_tree): Use add_thousands if MPFR and an integer
format.
2021-09-03 Arnold D. Robbins <arnold@skeeve.com>
* main.c (UPDATE_YEAR): Update to 2021.
2021-09-03 Arnold D. Robbins <arnold@skeeve.com>
* mpfr.c (mpg_cmp_as_numbers): Avoid compiler warning about possibly
used before set. Thanks to Michal Jaegermann for the report.
2021-08-18 Arnold D. Robbins <arnold@skeeve.com>
* main.c (main): Change have_srcfile from int to bool.
2021-08-15 Arnold D. Robbins <arnold@skeeve.com>
Allow setting AR and ARFLAGS on the configure command line.
Thanks to Jacob Burkholder <jake.burkholder2@gmail.com> for
the report.
* configure.ac (AR_FLAGS): Provide default value.
(AR): Call AC_SUBST on it.
2021-08-13 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (do_sub): Rationalize handling of strongly typed
regex as argument to sub/gsub, as well as rationalize the return
value from gensub to always be string. Thanks to John Naman
<gawker@703n.com> for the bug report.
2021-08-05 Andrew J. Schorr <aschorr@telemetry-investments.com>
* mpfr.c (do_mpfr_func): New argument, warn_negative. If true,
print a warning message about a negative argument. Bring -M
code into line with regular code. Bug report from Neil Ormos
in the help-gawk list.
Unrelated:
* mpfr.c (do_mpfr_int_div, mpg_div, mpg_mod): If dividing by
zero, print a fatal error messages. Brings MPFR code into line
with the regular code. Thanks to Peng Yu for noticing the bug, in
the help-gawk list.
2021-07-23 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (exec_count_t): Add typedef for Vax VMS C and everyone else.
(EXEC_COUNT_FMT, EXEC_COUNT_PROFILE_FMT): Corresponding format
strings for printf.
* debug.c (print_instruction): Use EXEC_COUNT_FMT.
* profile.c (indent): Ditto and use EXEC_COUNT_PROFILE_FMT.
2021-07-07 Arnold D. Robbins <arnold@skeeve.com>
* array.c (asort_actual): Add a lint warning for passing the
same array as the first two arguments without a third.
Thanks to Peng Yu <pengyu.ut@gmail.com> for the report.
2021-06-30 Arnold D. Robbins <arnold@skeeve.com>
* custom.h: Add a bunch of defines to work around the continuous,
needless churn in support/dfa.c.
2021-06-04 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (format_tree): In MPFR prints, cast values to
int for proper int vs. unsigned comparison. Fixes things
for HP-UX. Thanks to Daniel Richard G. for the report.
2021-05-21 Arnold D. Robbins <arnold@skeeve.com>
* debug.c: For z/OS, use %#p to print pointers, %p everything else.
2021-05-13 Arnold D. Robbins <arnold@skeeve.com>
* custom.h: For z/OS, define _REGEX_INCLUDE_LIMITS_H.
* profile.c (pprint): Add a diagnostic print for an
unexpected node type.
2021-05-05 Arnold D. Robbins <arnold@skeeve.com>
* CMakeLists.txt: Removed.
* cmake: Removed directory and its contents.
2021-05-05 Arnold D. Robbins <arnold@skeeve.com>
Move to Automake 1.16.3.
* configure.ac (AM_INIT_AUTOMAKE): Add subdir-objects option.
* NEWS: Updated.
Get `make distcheck' working again:
* Makefile.am (EXTRA_DIST): Remove files that are now in build-aux.
Unrelated:
* builtin.c (do_sub): Remove check for BOOL as target. It's not
correct anymore.
2021-05-03 Arnold D. Robbins <arnold@skeeve.com>
* eval.c (flags2str): Move BOOL entry into the right place in the
list.
* NEWS: Update with info on mkbool.
2021-05-02 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (do_bool): Renamed do_mkbool.
* awkgram.y (tokentab): Rename bool to mkbool.
* builtin.c (do_bool): Renamed do_mkbool.
2021-04-30 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (boolval): Remove check for BOOL flag. It was incorrect.
Thanks to Andrew Schorr for the catch.
For Node_val, update the comment to describe the BOOL flag, and
move BOOL up to be alongside the flags for Node_val.
2021-04-28 Arnold D. Robbins <arnold@skeeve.com>
Make bools plain numbers that have bool flag instead of
being weird string values.
* array.c (do_sort_up_value_type): Remove special ordering
for booleans.
* awk.h (warn_bool): Remove declaration.
* builtin.c (warn_bool): Remove function and all calls.
* gawkapi.c (node_to_awk_value): Update switches to look for
NUMBER|BOOL.
* node.c (make_bool_node): Revise string values and flag bits.
2021-04-16 Arnold D. Robbins <arnold@skeeve.com>
* main.c (arg_assign): For -v '@/...' make sure that there are
at least three characters there. Thanks to Ed Morton
<mortoneccc@comcast.net> for the report.
2021-04-14 Arnold D. Robbins <arnold@skeeve.com>
Fix up lint warnings for "no effect" to avoid false warnings.
Straighten out the messages, and remove the run-time warning,
since a parse-time warning is already issued. Thanks to
Arkadiusz Drabczyk <arkadiusz@drabczyk.org> for the initial
bug report.
* awkgram.y (isnoeffect): Add more opcodes that are "no effect".
(add_lint): For LINT_no_effect, check that all the opcodes in the
list are "no effect". Only if so, issue the warning. Remove the
addition of the run-time warning.
* interpret.h (r_interpret): Remove LINT_no_effect from Op_lint
case.
2021-04-14 Arnold D. Robbins <arnold@skeeve.com>
* array.c (do_sort_up_value_type): Fix order with bools.
2021-03-30 Arnold D. Robbins <arnold@skeeve.com>
* gawk_api.h (gawk_api_minor_version): Increase to 2.
* gawk_api.c (assign_bool): New function.
(node_to_awk_value): Finish updating for bool types and values.
2021-03-22 Arnold D. Robbins <arnold@skeeve.com>
* gawkapi.h (make_bool): New inline function.
Update table of request/return types.
* gawkapi.c (awk_value_to_node): Add support for AWK_BOOL.
(node_to_awk_value): Start on same. Not yet complete.
2021-03-21 Arnold D. Robbins <arnold@skeeve.com>
* str_array.c (fnv1a_hash_string): New function.
(str_array_init): Use fnv1a_hash_string if AWK_HASH env var
set to "fnv1a".
2021-03-20 Arnold D. Robbins <arnold@skeeve.com>
* array.c (do_sort_up_value_type): Add logic for handling bools.
2021-03-08 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (warn_bool, do_bool): Add function declarations.
* awkgram.y (tokentab): Add entry for "bool" builtin.
* builtin.c (warn_bool): New function.
(do_index, do_substr, do_toupper, do_tolower, do_match,
do_length): Call it.
(do_sub): If first arg to sub/gsub is bool, fatal error.
(do_bool): New function.
* field.c (do_split, do_patsplit): Call warn_bool.
* main.c (load_procinfo_bools): Removed function and call.
2021-03-05 Arnold D. Robbins <arnold@skeeve.com>
Start on a bool type for gawk.
* awk.h (BOOL): New flag value.
(make_bool_node): Add declaration of new function.
(bool_val): Check for BOOL along with NUMBER.
* builtin.c (do_typeof): Add support for BOOL.
* eval.c (flags2str): Ditto.
* gawkapi.h (awk_val_type): Add AWK_BOOL.
(awk_value_t): Add awk_bool_t element to the union and macro for it.
(struct gawk_api): Update the table of request/return values.
* main.c (load_procinfo_bools): New function.
(load_procinfo): Call it.
* node.c (make_bool_node): New function.
2021-02-13 Arnold D. Robbins <arnold@skeeve.com>
* io.c (nextfile): Use the value of ARGC directly in the for
loop, in case it gets changed by code or by a command line
assignment. Thanks to the discussion in comp.lang.awk for
the report.
2021-01-28 Arnold D. Robbins <arnold@skeeve.com>
Fix a Day One Bug!
* main.c (main): For -b option, only set LC_ALL to C if
not --posix. Ooops.
2021-01-09 Arnold D. Robbins <arnold@skeeve.com>
Fix problems turning something like 018 into decimal.
Thanks to Arkadiusz Drabczyk <arkadiusz@drabczyk.org> and to
Jean-Philippe Guérard <jean-philippe.guerard@xn--tigreray-i1a.org>
for the reports.
* builtin.c (nondec2awknum): Use a copy of len, in case we detect
8 or 9 and have to restart as decimal.
* mpfr.c (mpg_strtoui): For 8 or 9, set base to 10.
Unrelated:
* array.c, awk.h, awkgram., builtin.c, cmd.h, command.y, debug.c,
eval.c, gawk.api.c, gawkapi.h, interpret.h, io.c, main.c, mfpr.c,
node.c, profile.c, re.c: Update copyright year.
2021-01-08 Arnold D. Robbins <arnold@skeeve.com>
General tightening up use of const and types. Thanks to
the C++ porting experiment.
* array.c (null_array): Separate out multiple assignment of zero.
* awk.h: Add <math.h> here and pull it out of individual files.
(CONVFMT): Make const char *.
(quote): Make const char *.
(defpath): Make const char *.
(deflibpath): Make const char *.
(envsep): Make const char *.
(DO_FLAG_NONE): New zero enum value.
(getblock): Add a cast to the ty argument.
(make_regnode): First argument is now NODETYPE.
(gawk_name): Returns const char *.
* awkgram.y: Include argument types in read_func pointer.
(make_regnode): First argument is now NODETYPE.
* builtin.c (do_typeof): Use const char *.
(format_nan_inf): Move declaration of val to top of function.
* cmd.h (dbg_prompt, commands_prompt, eval_prompt, dgawk_prompt):
Make const char *.
(struct cmd_token): Use const char *. Rename `class' to `lex_class'.
* command.y (lex_class): Update all uses.
(find_command): Use const char *.
* debug.c (dbg_prompt, commands_prompt, eval_prompt, dgawk_prompt,
output_file): Make const char *.
(struct dbg_option): Use const char *.
(do_info): Use const char *. Cast stop.command to enum argtype.
(do_finish): Cast stop.command to enum argtype.
(do_return, do_until): Ditto.
(print_instruction, set_gawk_output, set_prompt): Add casts to
some function parameters.
* eval.c (update_ERRNO_int): Use const char *.
* gawk.api.c (valtype2str): Return const char *.
(api_get_argument): Use awk_true instead of true in return statements.
(awk_value_to_node): Move declaration of tval to top of function.
(api_release): Cast a_cookie to NODE *.
* gawkapi.h (struct awk_input): Include argument types in
read_func pointer.
* interpret.h (r_interpret): Add cast to argument in call to
do_getline_redir.
* io.c (read_with_timeout): Change second param to void * from char *.
(redirect_string): Add casts to function call parameters.
(path_info): Use const char *.
(init_awkpath): Ditto.
* main.c (CONVFMT, locale_dir): Now const char *.
(do_flags): Init to DO_FLAG_NONE instead of zero.
* mfpr.c (get_rnd_mode): Add cast to return value.
(set_round_mode): Add cast to assignment of rndm.
(do_mpfr_func): Include argument types in type of second parameter.
* node.c (r_make_number): Remove declaration of strtod.
* profile.c (pp_push): Add cast in assignment to n->type. Use
const char * for `pre' at Op_func_call.
* re.c (check_bracket_exp): Add cast to return from memchr.
2021-01-07 Arnold D. Robbins <arnold@skeeve.com>
* builtin.c (format_tree): Generalize handling of invalid
format modifiers. Add support for three more letters. Thanks to
Arkadiusz Drabczyk <arkadiusz@drabczyk.org> for the report.
Unrelated:
* io.c (rsnullscan): Adjust loop test to get to end of string
when re match is null string. Thanks to Ed Morton
<mortoneccc@comcast.net> for the report.
2020-12-26 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2020-12-20 Arnold D. Robbins <arnold@skeeve.com>
Second steps fixing +inform, +nancy, for MPFR.
* builtin.c (format_nan_inf): Use mpfr_signbit instead of mpfr_sgn.
* mpfr.c (force_mpnum): Check for NaN and leading minus and if so
set the signbit with mpfr_setsign.
(mpg_force_number): Copy in code from f_force_number to check
properly for +inf, +nan.
2020-12-19 Arnold D. Robbins <arnold@skeeve.com>
First steps fixing +inform, +nancy. Sigh.
* awk.h (is_ieee_magic_val): Add declaration.
* node.c (is_ieee_magic_val): Make extern, not static.
(r_force_number): Make the check smarter, catch -nan.
2020-11-02 Arnold D. Robbins <arnold@skeeve.com>
Make gawk numeric comparisons act like C doubles.
MPFR differs from doubles w.r.t. NaN, not sure why yet.
* awk.h (scalar_cmp_t): New enum.
* builtin.c (format_nan_inf): Use mpfr_signbit, not mpfr_sgn.
* eval.c (cmp_doubles): New routine.
(cmp_scalars): Change type to bool, rework logic.
* interpret.h (r_interpret): Rework scalar comparisons.
* mpfr.c (mpg_cmp_as_numbers): New routine.
* node.c: Use <math.h>, not "math.h", minor comment edits.
2020-11-02 Arnold D. Robbins <arnold@skeeve.com>
* re.c (make_regexp): Cast len parameter to int to avoid
compiler warnings.
2020-10-31 Arnold D. Robbins <arnold@skeeve.com>
* re.c (make_regexp): When re_compile_pattern fails, use the
original text of the regexp in the error message. Thanks to Neil
R. Ormos <ormos-gnulists17@ormos.org> for the report.
2020-10-29 Arnold D. Robbins <arnold@skeeve.com>
* eval.c (setup_frame): Handle the case of values from FUNCTAB.
Thanks to Denis Shirokov <cosmogen@gmail.com> for the report.
2020-10-12 Arnold D. Robbins <arnold@skeeve.com>
* gawkapi.c (api_lintwarn): Fix comparison, should be to check
lintfunc, not lintwarn. Thanks to OLLIER Jean-François
<lalanne.51@wanadoo.fr>, reported 23 March, 2014.
2020-10-11 Arnold D. Robbins <arnold@skeeve.com>
* re.c (make_regexp): Add {} to characters that cause rp->maybe_long
to be true.
* io.c (rsrescan): Update comment for step 4.
2020-09-21 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (enum redirect_flags): Add RED_NONE.
(redirect_flags_t): New typedef.
* io.c (redflags2str): Handle RED_NONE.
(check_duplicated_redirections): New function.
(redirect_string): Use new typedef. Call new function if do_lint
instead of using inline code.
(close_redir): Add error message for failure on close of two-way pipe.
2020-09-04 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y [GRAMMAR]: Install arrays as Node_var_array. Improves
PROCINFO["identifiers"]. Thanks to anoncoward@nym.hush.com
for the report.
2020-08-25 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): At Op_rule, check if in BEGINFILE. If so,
clear the record. Fixes an issue reported by Pat Rankin in May, 2011.
2020-08-24 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y, builtin.c, command.y, debug.c, gawkapi.c,
io.c, node.c: Make all messages consistent: Remove final periods
from messages that had them.
* TODO: Updated.
2020-08-20 Arnold D. Robbins <arnold@skeeve.com>
* awk.h (NODE): Move valref completely out of the union. Fixes
core dumps on old GCC on CentOS 7.
2020-08-13 Arnold D. Robbins <arnold@skeeve.com>
Fix TIDYMEM / use of mtrace.
* configure.ac: Add mtrace to list of functions searched for.
* symbol.c (make_symbol): Set valref to 1. Allows the
release_all_vars function to work.
Unrelated:
* array.c (value_info): Spell ROUNDMODE correctly in output.
* mpfr.c (set_ROUNDMODE): Ditto.
2020-08-12 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.c, command.c: Regenerated with Bison 3.7.1.
* NEWS: Updated.
Unrelated:
* profile.c (pp_namespace): Always print a newline before the
@namespace, in case it comes after a comment. Thanks to J Naman
for the report.
* symbol.c (comp_symbol): Adjust comparison so that names in the
awk namespace come out first. Makes the order of functions in
pretty-printed output more sane. Thanks also to J Naman for
getting me to explore this issue.
2020-07-29 Arnold D. Robbins <arnold@skeeve.com>
* custom.h (FLEXIBLE_ARRAY_MEMBER): Define to 1 for dfa.h.
2020-07-28 Arnold D. Robbins <arnold@skeeve.com>
Eliminate 'iszero: redefined' warning from tcc.
* awk.h (is_zero): Renamed from `iszero'.
(boolval): Update usage.
* awkgram.y [GRAMMAR]: Ditto.
(negate_num): Ditto.
* debug.c (condition_triggered): Ditto.
* eval.c (set_LINT): Ditto.
2020-07-26 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.c, command.c: Regenerated with Bison 3.7.
* NEWS: Updated.
Unrelated:
* gawkapi.h (api_get_mpfr, api_get_mpz): Mark as obsolete.
2020-07-22 Arnold D. Robbins <arnold@skeeve.com>
* Makefile.in: Regenerated.
2020-07-21 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac (AM_INIT_AUTOMAKE): Update to 1.16.
* ar-lib, compile, config.guess, config.rpath, config.sub,
depcomp, install-sh, missing, ylwrap: Moved to build-aux.
2020-07-20 gettextize <bug-gnu-gettext@gnu.org>
* configure.ac (AM_GNU_GETTEXT_VERSION): Bump to 0.20.2.
2020-07-19 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Fix title on leading set of changes. Thanks to
Marcus <marcuscruse@ya.ru> for the report.
Unrelated: Bump minor API version since semantics have changed.
* gawkapi.h (api_minor_version): Increase to 1.
Unrelated: Fix a spelling error in a comment.
* mpfr.c (force_mpnum): Fix a typo.
2020-07-17 Arnold D. Robbins <arnold@skeeve.com>
* gawkapi.h (gawk_api_minor_version): Increase, since the
MPFR/MPZ stuff changed some. Remains source and binary
compatible, but semantics are different.
2020-07-15 Arnold D. Robbins <arnold@skeeve.com>
Finish cleaning up MPFR valgrind issues.
* awk.h (struct block_header): Remove MPFR and GMP enums.
* gawkapi.c (getmpfr, freempfr, getmpz, freempz): Removed
(api_get_mpfr): Use emalloc instead of getmpfr.
(api_get_mpz): Use emalloc instead of getmpz.
* gawkapi.h: Add comments on MPFR and MPZ
(make_number_mpfr, make_number_mpz): Update comments.
* node.c (nextfree): Remove mpfr and mpz entries.
* NEWS: Updated.
Unrelated:
* config.guess, config.sub: Updated from GNULIB.
2020-07-13 Arnold D. Robbins <arnold@skeeve.com>
* mpfr.c (mpg_interpret): Small speedup. If do_itrace and opcode is
not one that's handled, return early.
2020-07-12 Arnold D. Robbins <arnold@skeeve.com>
* mpfr.c (mpg_interpret): Make the instruction tracing more sane.
Fix the testext test for MPFR.
* gawkapi.c (awk_value_to_node): Use correct code to copy
MPFR/MPN numbers and don't free the original values.
* interpret.h (r_interpret): At Op_ext_builtin, separate out calls
to the external function and awk_value_to_node, for ease in a
debugger.
* NEWS: Updated.
2020-07-10 Arnold D. Robbins <arnold@skeeve.com>
Fix printf issues. Thanks to Michal Jaegermann for the report.
* cint_array.c (cint_dump): Fix printf formats.
* debug.c (print_symbol): Ditto.
* mpfr.c (mpg_interpret): Ditto.
2020-07-09 Arnold D. Robbins <arnold@skeeve.com>
Fix dbugeval2 test for MPFR.
* debug.c (do_eval): Preserve DO_MPFR flag in do_flags before
parsing the string to be evaluated.
Fix indirect call tests for MPFR:
* awkgram.y (lookup_builtin): Only return the MPFR builtin
function if the pointer is not NULL.
Unrelated:
* config.guess, config.sub: Updated from GNULIB.
2020-07-08 Arnold D. Robbins <arnold@skeeve.com>
* str_array.c (str_lookup): If a pure number, copy to a string. Fixes
the printfchar test for MPFR.
Fix typedregex2 and typedregex3 tests for MPFR:
* mpfr.c (mpg_zero): No longer static inline.
(force_mpnum): Handle the case of a strongly typed regex.
(mpg_interpret): Add tracing for do_itrace.
* awk.h (mpg_zero): Add declaration.
* node.c (make_typed_regex): Init to zero for MPFR via mpg_zero.
2020-07-07 Arnold D. Robbins <arnold@skeeve.com>
* awk.h: Turn all the flag defines into enums. GDB can then show
the bit maps directly.
Unrelated: Fix another MPFR bug, reported by Hyunho Cho
<mug896@naver.com>.
* mpfr.c (force_mpnum): Restore the character at the end of
the string if only saw + or -.
2020-07-07 Andrew J. Schorr <aschorr@telemetry-investments.com>
* node.c (r_dupnode): Sanitize the code, particularly for MPFR.
2020-07-05 Arnold D. Robbins <arnold@skeeve.com>
Bug fixes in MPFR, reported by Hyunho Cho <mug896@naver.com>.
* node.c (r_dupnode): Set strndmode and also set stlen of new
string. (How did that one get by us for so long?)
* mpfr.c (force_mpnum): Check for a + or - sign with nothing
following it.
Unrelated. Make do_itrace a real variable so that it's easier to
change the value from a debugger.
* awk.h (do_itrace): Declare.
(DO_ITRACE, do_itrace): Remove macros.
* main.c (do_itrace): Define.
(parse_args): Set do_itrace.
2020-07-03 Arnold D. Robbins <arnold@skeeve.com>
Fix a double free error with -M. Thanks to
"NIDE, Naoyuki" <nide@ics.nara-wu.ac.jp> for the report.
* field.c (clear_mpfr): New macro, clears MPFR and NUMCUR bits.
(reset_record): Use it on the newly copied field values.
(purge_record): Add introductory comment since we're in the file.
2020-07-01 Arnold D. Robbins <arnold@skeeve.com>
* eval.c (posix_compare): Rewrite contributed by
Michael Builov <mbuilov@gmail.com>.
Unrelated:
* awk.h: New option: do_itrace.
* interpret.h (r_interpret): Print instruction trace if enabled.
* main.c (optab): New option: -I/--trace.
(parse_args): Update optlist and the code to handle it.
* NEWS: Updated.
2020-06-29 Arnold D. Robbins <arnold@skeeve.com>
* debug.c: Cleanup messages. Error messages start with lower
case letter.
* TODO: Updated.
2020-06-14 Arnold D. Robbins <arnold@skeeve.com>
Disallow SYMTAB and FUNCTAB as destination arguments to builtin
functions that clear arrays:
* awk.h (check_symtab_functab): Add declaration.
* array.c (asort_actual): Call it in check for second argument.
* builtin.c (check_symtab_functab): New function.
(do_match): Call it in check for third argument.
* field.c (do_patsplit, do_split): Call it in checks for fourth and
second arguments.
2020-06-12 Arnold D. Robbins <arnold@skeeve.com>
* array.c (asort_actual): If SYMTAB or FUNCTAB is the second argument,
fatal out. If they are the first argument, make it work when a
second argument is supplied.
* TODO, NEWS: Updated.
2020-06-10 Arnold D. Robbins <arnold@skeeve.com>
More miscellaneous fixes from Michael Builov <mbuilov@gmail.com>.
* node.c (r_unref): Handle GAWKDEBUG same as in awk.h:unref.
(parse_escape): isxdigit/isdigit/isupper called with possible negative
int: fixed. Dead code under "if (do_lint && j > 2)" eliminated,
j cannot be > 2.
(init_btowc_cache): Also initialize wide-char at index 255.
And so that `make check' passes with GAWKDEBUG defined:
* msg.c (err): Remove abort() call if fatal. Didn't really do
anything special for us.
2020-06-10 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.c, command.c: Regenerated with bison 3.6.3.
* NEWS: Updated.
2020-06-10 Andrew J. Schorr <aschorr@telemetry-investments.com>
Convert exec_count from long long to unsigned long long.
* awk.h (INSTRUCTION): Change ldl type from long long to unsigned
long long.
* debug.c (print_instruction): Fix printf format for exec_count.
* profile.c (indent): The argument is now an unsigned long long,
and fix the printf format to match.
2020-06-10 Arnold D. Robbins <arnold@skeeve.com>
Miscellaneous fixes from Michael Builov <mbuilov@gmail.com>.
* awkgram.y (rule_block): Fix size declaration
(parse_program): Adjust memset call.
(get_comment): Check for end of file before checking for space.
* io.c (gawk_popen): Move os_close_on_exec call inside #endif.
* builtin.c (do_bindtextdomain): Don't pass NULL to strlen.
* debug.c (set_breakpoint): Don't dereference NULL pointer in D_func.
(serialize_list): Get the buffer management right to avoid overrun.
2020-06-10 Arnold D. Robbins <arnold@skeeve.com>
Thanks to Andrew Schorr for suggesting a better way to
handle exec_count being a long long.
* awk.h (INSTRUCTION): Make new ldl member be the long long.
(exec_count): Is now ldl.
* profile.c (pprint): Fix printf of exec_count.
2020-06-09 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y (tokexpand): Use size_t instead of int for size of
buffer.
2020-06-08 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y: Fix `print $"2"' case. Thanks to Ed Morton
<mortoneccc@comcast.net> for the report.
Unrelated:
* awk.h (INSTRUCTION): Make dl a long long for profiling
counts on 32 bit systems. Thanks to Peter Lindgren
for the report.
* profile.c (indent): Take a long long parameter and fix
the printf format.
2020-06-05 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y: Allow @/foo/ as return expression in a function.
Oops.
2020-06-05 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: Add call to AM_PROG_AR for use in
cross compiles. Thanks to Sergei Trofimovich <slyfox@gentoo.org>
for the patch.
2020-05-15 Arnold D. Robbins <arnold@skeeve.com>
* config.guess, config.sub: Updated from GNULIB.
2020-05-14 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.c, command.c: Rebuild with Bison 3.6.
* NEWS: Updated.
2020-05-06 Arnold D. Robbins <arnold@skeeve.com>
Add lint check for string + string. It's not concatenation
in awk, but is in most other languages. Thanks to Roland Illig
<roland.illig@gmx.de> for the suggestion.
* awk.h (Op_lint_plus): New opcode.
* awkgram.y (mk_binary): Add it into the list in the right place
if do_lint.
* eval.c (optypes): Add case for Op_lint_plus.
* interpret.h (r_interpret): Ditto.
2020-04-14 Arnold D. Robbins <arnold@skeeve.com>
* 5.1.0: Release tar ball made.
2020-04-10 Arnold D. Robbins <arnold@skeeve.com>
* field.c (save_FPAT): New global variable.
(get_field): Check if using FPAT or not in order to pass in
the right variables to parse_field function. Fixes bug in
delayed field parsing when using FPAT. Thanks to
luciole75w <luciole75w@gmail.com> for the bug report.
(set_FPAT): Use global save_FPAT instead of local save_fpat.
Unrelated:
* awkgram.c, command.c: Rebuild with Bison 3.5.4.
* NEWS: Updated.
Unrelated:
* awk.h (do_lint_extensions): Define to zero if NO_LINT.
2020-04-02 Arnold D. Robbins <arnold@skeeve.com>
* field.c (re_parse_field): If default parsing and trailing
whitespace, don't set a null field at the end. Thanks to
Ed Morton <mortoneccc@comcast.net> for the report.
2020-03-18 Arnold D. Robbins <arnold@skeeve.com>
* TODO: Add more stuff.
2020-03-15 Arnold D. Robbins <arnold@skeeve.com>
* Makefile.am (zos-diffout): New target.
* awkgram.y, command.y: Upgrade to Bison 3.5.3.
* NEWS: Updated.
2020-03-11 Arnold D. Robbins <arnold@skeeve.com>
* main.c: Further message update. Thanks again to
Roland Illig <roland.illig@gmx.de>.
2020-03-09 Andrew J. Schorr <aschorr@telemetry-investments.com>
* array.c (sort_up_value_string): If either arg is not a Node_val,
call out to sort_up_value_type instead. If cmp_strings returns zero,
fall back to sort_up_index_string as a tie-breaker.
(sort_up_value_number): If either arg is not a Node_val, call out to
sort_up_value_type instead. If cmp_strings returns zero, fall back
to sort_up_index_string as a tie-breaker.
(do_sort_up_value_type): Renamed from sort_up_value_type with one
change: if both arguments have type Node_var, we compare the
var_value NODEs instead.
(sort_up_value_type): New wrapper function around do_sort_up_value_type
to fall back to sort_up_index_string as a tie-breaker.
2020-03-09 Arnold D. Robbins <arnold@skeeve.com>
* array.c, awk.h, awkgram.y, builtin.c, command.y, debug.c,
ext.c, int_array.c, io.c, main.c, symbol.c: Messages refined
based on suggestions from Roland Illig <roland.illig@gmx.de>.
2020-03-06 Jannick <thirdedition@gmx.net>
* Makefile.am: Replace AM_MAKEFLAGS with AM_CFLAGS and AM_LDFLAGS.
Put pc/Makefile.tst into BUILT_SOURCES, instead of in dist-hook.
Add $(srcdir) to rules for the yacc files. Use AM_LDFLAGS in
the efence rule.
* configure.ac: Use GAWK_CANONICAL_HOST instead of AC_CANONICAL_HOST;
this supplies MSYS2 support. Remove use of EXEEXT.
2020-03-04 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2020-02-13 John E. Malmberg <wb8tyw@qsl.net>
* custom.h: OpenVMS needs _REGEX_INCLUDE_LIMITS_H defined.
2020-02-09 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y: Add lint check for assignment in condition to
if and for statements. Thanks to Jannick <thirdedition@gmx.net>
for the suggestion.
* debug.c (do_eval): Remove @eval from the symbol table if there
was a syntax error in the statement being evaluated. Thanks again
to Jannick, for the report.
2020-02-04 John E. Malmberg <wb8tyw@qsl.net>
* custom.h: Fix OpenVMS SIZE_MAX value.
2020-02-01 Arnold D. Robbins <arnold@skeeve.com>
* awk.h, dfa.c: Move include of mbsupport.h to ...
* custom.h: ... here.
* configure.ac: Add check for isblank.
2020-01-27 Arnold D. Robbins <arnold@skeeve.com>
* custom.h: Fix non-VMS compilation.
2020-01-26 John E. Malmberg <wb8tyw@qsl.net>
* custom.h: Additional OpenVMS typedefs needed.
* gettext.h: Fix macro for implementations without gettext
to not force the result to be a const char *.
2020-01-26 Andrew J. Schorr <aschorr@telemetry-investments.com>
* array.c (sort_up_value_type): Small efficiency tweak. Only
run through the list of other types if both values are not scalars.
2020-01-24 Arnold D. Robbins <arnold@skeeve.com>
* array.c, profile.c, cint_array.c, builtin.c, interpret.h,
main.c: Update copyright years.
2020-01-23 Arnold D. Robbins <arnold@skeeve.com>
* array.c (sort_up_value_type): Handle other types of nodes,
in case someone applies PROCINFO["sorted_in"] to SYMTAB or
FUNCTAB. This threw an assertion error if assertions were
enabled. Thanks to John Naman <gawker@703n.com> for the report.
(asort_actual): Disallow SYMTAB and FUNCTAB as first arguments.
* TODO: Note that we should should eventually allow SYMTAB/FUNCTAB
passed to asort/asorti if a destination array is provided. (The
current code breaks right now.)
2020-01-19 Arnold D. Robbins <arnold@skeeve.com>
* profile.c (pp_top): New macro.
(pprint): Be smarter for print[f] with redirection that was
parenthesized, to not print `printf(("hello\n")) > "..."'.
Unrelated:
* compile, config.guess, config.rpath, config.sub, depcomp: Updated
from GNULIB.
2020-01-14 Andrew J. Schorr <aschorr@telemetry-investments.com>
* cint_array.c (cint_array_init): Fix off-by-one error in array
bounds overflow check for an NHAT value set in the environment.
Thanks to Michael Builov <mbuilov@gmail.com> for the report.
2020-01-08 Arnold D. Robbins <arnold@skeeve.com>
Fix a number of subtle memory leaks. Thanks to the
combination of MEMDEBUG and valgrind for finding them.
* builtin.c (do_print): Unref tmp if force_string_ofmt
returned something different.
(call_sub, call_match, call_split_func): Free any
created regexp.
* debug.c (do_eval): Free the node holding the @eval function
when done with it.
* field.c (init_fields): Create fields_arr[0] and Nnull_field
in a rational fashion.
* interpret.h (Op_assign_concat): Handle newly created node
better.
Unrelated:
* interpret.h (Op_subscript): When retrieving from SYMTAB, check
for Node_var_new; variables can exist but have not been assigned
a value. Thanks to Denis Shirokov <cosmogen@gmail.com> for the report.
Unrelated:
* builtin.c (do_typeof): Handle Node_array_ref also. Thanks to
Denis Shirokov <cosmogen@gmail.com> for the report.
2019-12-22 Arnold D. Robbins <arnold@skeeve.com>
* config.guess: Updated from GNULIB.
2019-12-22 Jannick <thirdedition@gmx.net>
* configure.ac: Hardcode the shared object extension to dll
on Windows platforms, i.e.whenever EXEEXT is '.exe'.
2019-12-22 Arnold D. Robbins <arnold@skeeve.com>
* gettext.h: Update from gettext 0.20.1.
* NEWS: Updated.
2019-12-18 Paul Eggert <eggert@Penguin.CS.UCLA.EDU>
Fix memcpy issue found by -fsanitize=undefined
* field.c (set_record): Don't memcpy (databuf, NULL, 0),
as the C standard says the resulting behavior is undefined.
2019-12-16 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y, command.y: Upgrade to Bison 3.5.
* NEWS: Updated.
2019-11-21 Arnold D. Robbins <arnold@skeeve.com>
* config.rpath: Update from GNULIB.
2019-11-10 Arnold D. Robbins <arnold@skeeve.com>
* compile, config.rpath: Update from GNULIB. Adds msys support.
2019-11-03 Arnold D. Robbins <arnold@skeeve.com>
* field.c (get_field): Remove special case code for fpat_parse_field.
Fixes a bug reported by Ed Morton <mortoneccc@comcast.net>.
2019-10-19 Arnold D. Robbins <arnold@skeeve.com>
* compile, depcomp: Update from GNULIB.
2019-10-13 Arnold D. Robbins <arnold@skeeve.com>
Fix handling of assigning @/.../ to FS and RS. Thanks
to Stephane Chazelas <stephane.chazelas@gmail.com>
for the report.
* field.c (set_FS): Check for REGEX in the flags.
* io.c (set_RS): Ditto.
2019-10-04 Arnold D. Robbins <arnold@skeeve.com>
* config.guess, config.sub, install-sh: Updated from GNULIB.
2019-09-30 John E. Malmberg <wb8tyw@qsl.net>
* custom.h: OpenVMS does not provide ULLONG_MAX
2019-09-20 Florian Weimer <fweimer@redhat.com>
* configure.ac (PRINTF_HAS_F_FORMAT): Include <string.h> for
strcmp.
(PRINTF_HAS_A_FORMAT): Likewise.
* configure: Regenerate.
2019-09-17 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.y, commands.y: Upgrade to Bison 3.4.2.
* NEWS: Updated.
2019-09-10 Andrew J. Schorr <aschorr@telemetry-investments.com>
* awk.h (block_header): Remove cnt field and replace it with
highwater. If MEMDEBUG is defined, add active counter.
* node.c (r_getblock): The cnt field was replaced by active. Update
highwater as appropriate.
(r_freeblock): Decrement the active counter (renamed from cnt).
(more_blocks): Bump the highwater counter (renamed from cnt).
* builtin.c (do_typeof): When providing PROCINFO debug memory stats,
replace count_<blah> with <blah>_active and <blah>_highwater. When
MEMDEBUG is not set, we calculate the active value by subtracting the
number of elements in the free list from the highwater value.
2019-09-01 Arnold D. Robbins <arnold@skeeve.com>
* profile.c (pp_list, pp_concat): Fix use-after-free errors
in both routines. Thanks to valgrind and Andy's MEMDEBUG code.
2019-09-01 Andrew J. Schorr <aschorr@telemetry-investments.com>
* awk.h (block_header): Add cnt field even when MEMDEBUG is not
defined, and add a name field.
* node.c (nextfree): Initialize new name field.
(more_blocks): Bump nextfree[id].cnt by BLOCKCHUNK.
* builtin.c (do_typeof): When the 1st argument is PROCINFO and the
2nd arg is provided, return new "count_<blah>" fields containing
the memory allocation accounts of the various block types.
2019-08-30 Andrew J. Schorr <aschorr@telemetry-investments.com>
* configure.ac (.developing): Add -DMEMDEBUG to CFLAGS.
* awk.h (block_header): If MEMDEBUG is defined, add cnt field
to track the number of allocations.
(getblock, freeblock): When MEMDEBUG is defined, replace these
macros with calls to new functions r_getblock and r_freeblock.
* node.c (r_getblock, r_freeblock): New functions that simply
use malloc and free when MEMDEBUG is defined.
2019-08-30 Andrew J. Schorr <aschorr@telemetry-investments.com>
* interpret.h (r_interpret): For Op_match_rec, unref if a
dynamic regexp. Fixes another memory issue. See the thread starting
at https://lists.gnu.org/archive/html/bug-gawk/2019-08/msg00023.html.
2019-08-29 Arnold D. Robbins <arnold@skeeve.com>
Restore a modified version of the change of
2016-05-03, which used realloc to shrink the buffer.
* builtin.c (format_tree): Reduce INITIAL_OUT_SIZE to 64.
At the end, call realloc only if there is more
than (2 * INITIAL_OUT_SIZE) bytes to give back.
2019-08-23 Arnold D. Robbins <arnold@skeeve.com>
* main.c (main): Fix whitespace issues.
* re.c (re_update): Small code cleanup in handling t->re_cnt.
Thanks to Andrew Schorr.
2019-08-21 Arnold D. Robbins <arnold@skeeve.com>
* config.guess: Updated from GNULIB.
2019-08-15 Arnold D. Robbins <arnold@skeeve.com>
Revert a6df7afc605079df7d85318846a522ef64aaa44d, change of
2016-05-03, which used realloc to shrink the buffer, in an
attempt to save memory. In actuality, it could cause increased
memory usage, even though there was no memory leak. See
https://lists.gnu.org/archive/html/bug-gawk/2019-08/msg00003.html
and the rest of thread for more detail.
* builtin.c (format_tree): Don't use realloc, just call
make_str_node with the original buffer. Remove `olen_final'
variable and its use.
2019-08-15 Andrew J. Schorr <aschorr@telemetry-investments.com>
Reduce memory usage by only compiling the IGNORECASE version of
regexp when it's actually needed.
* awkgram.y (make_regnode): Only compile the regular version.
* re.c (re_cache_get): New function.
(re_update): Use it as appropriate.
2019-07-23 Koichi Murase <myoga.murase@gmail.com>
* builtin.c (do_xor): Remove unneeded local variable `i'. Simplify
the loop's computation.
(do_and): Improve the initial value for the result.
2019-07-23 Andrew J. Schorr <aschorr@telemetry-investments.com>
Fix reporting of negative arguments for and(), or() and xor().
Thanks to Koichi Murase <myoga.murase@gmail.com> for the
report.
* builtin.c (do_and): Use nargs instead of i in the loop to
check for non-numeric and negative arguments.
(do_or): Ditto.
(do_xor): Ditto.
2019-07-23 Arnold D. Robbins <arnold@skeeve.com>
* Checklist: Updated.
2019-07-21 Arnold D. Robbins <arnold@skeeve.com>
* command.y, custom.h, eval.c, ext.c, field.c, io.c,
msg.c, node.c, nonposix.h, re.c: Update copyright year.
* NEWS: Updated.
* README: Updated.
2019-07-16 Arnold D. Robbins <arnold@skeeve.com>
* TODO: Updated.
* awkgram.y [GRAMMAR]: Change some lint warnings to use lintwarn.
(yylex): Ditto.
* field.c (choose_fs_function): Ditto.
* main.c (main): Ditto.
* node.c (parse_escape): Ditto.
2019-07-09 Arnold D. Robbins <arnold@skeeve.com>
* gawkapi.h: Update copyright year, fix some spelling errors.
2019-07-08 Andrew J. Schorr <aschorr@telemetry-investments.com>
* gawkapi.h (gawk_api_major_version): Bump from 2 to 3 because the
namespace changes altered the function signatures in gawk_api_t.
And add a comment at the top of the structure with a reminder that
the version number should be bumped whenever the struct is altered
in any way.
2019-06-30 Arnold D. Robbins <arnold@skeeve.com>
* interpret.h (r_interpret): Fix two more cases of warning
about gawk extensions.
* awkgram.y [GRAMMAR]: And two more cases here. Thanks to
Mark Krauze <daburashka@ya.ru> for making me look for them.
Unrelated:
* config.sub: Updated from GNULIB.
2019-06-26 Arnold D. Robbins <arnold@skeeve.com>
* symbol.c (install): Strip off any leading `awk::' before
installing a symbol. Thanks to Andrew Schorr for the
report.
2019-06-21 Arnold D. Robbins <arnold@skeeve.com>
* config.guess, config.sub: Updated from GNULIB.
* awkgram.y (get_src_buf): Set buffers to zero to avoid
possible issues with accessing uninitialized memory.
Thanks to valgrind, as pointed out by Nadezhda Vyukova
<qniva@yandex.ru>.
* NEWS, README.git: Fix some typos.
2019-06-19 Arnold D. Robbins <arnold@skeeve.com>
* re.c (make_regexp): If do_posix, have {} in ok_to_escape.
Thanks to Ed Morton <mortoneccc@comcast.net> for the report.
2019-06-18 Arnold D. Robbins <arnold@skeeve.com>
* 5.0.1: Release tar ball made.
2019-06-06 Arnold D. Robbins <arnold@skeeve.com>
* main.c (usage): Update comment for translators.
2019-06-02 Arnold D. Robbins <arnold@skeeve.com>
* awkgram.c, command.c: Update to Bison 3.4.
* NEWS: Updated.
2019-05-22 Arnold D. Robbins <arnold@skeeve.com>
Add --lint=no-ext. Suggest by Mark Krauze <daburashka@ya.ru>.
* NEWS: Updated.
* awk.h (DO_LINT_EXTENSIONS): New enum.
(do_lint_extensions): New macro.
* awkgram.y (yylex, snode): Use do_lint_extensions instead of
do_lint where appropriate.
* builtin.c (do_length): Ditto.
* eval.c (set_IGNORECASE, set_BINMODE): Ditto.
(set_LINT): Revise logic.
* field.c (do_split, set_FIELDWIDTHS, chose_fs_function, set_FPAT):
Ditto.
* io.c (set_RS): Ditto.
* main.c (usage): Updated.
(parse_args): Revise the code to handle --lint=no-ext.
2019-05-10 Arnold D. Robbins <arnold@skeeve.com>
* NEWS: Updated.
2019-05-06 Arnold D. Robbins <arnold@skeeve.com>
In sandbox mode, disallow assigning filenames that weren't
there initially. Thanks to Nolan Woods <nolan_w@sfu.ca> for
pointing out the gap.
* awk.h (init_argv_array): Add declaration.
* cint_array.c (argv_store): New vtable for ARGV.
(argv_shadow_array): New file static variable
(argv_store, init_argv_array): New functions.
* main.c (init_args): If in sandbox mode, build shadow array of
initial argv values. Call init_argv_array.
2019-05-05 Arnold D. Robbins <arnold@skeeve.com>
* ext.c (load_ext): Fix the message in the version for when
extensions are not available.
2019-04-24 Arnold D. Robbins <arnold@skeeve.com>
* msg.c (msg): Use %ld for the line number value. Thanks to
Michal Jaegermann for the report.
2019-04-23 Arnold D. Robbins <arnold@skeeve.com>
* config.sub: Updated from GNULIB.
2019-02-21 Andrew J. Schorr <aschorr@telemetry-investments.com>
* interpret.h (Op_store_field): Move call to force_string to
here from unfield. Speeds up work with fields that are numeric
only. Thanks to Tom Gray <tom_gray@keysight.com> for the report.
2019-04-21 Arnold D. Robbins <arnold@skeeve.com>
* POSIX.STD: Updated.
* field.c (get_field): If NF == -1, check parse high water to
set in_middle correctly. Thanks to <oguzismailuysal@gmail.com>
for the report.
2019-04-18 Arnold D. Robbins <arnold@skeeve.com>
* msg.c (msg): Add an undocumented feature. "Use the Source, Luke."
* Makefile.am (EXTRA_DIST): Add ChangeLog.1 to the list. Ooops.
* CheckList: Updated.
Fix core dump reported by Steve Kemp <steve@steve.org.uk>:
* awk.h (errcount): Declare.
* awkgram.y (errcount): No longer static.
* command.y (dbg_errcount): Renamed from errcount.
* main.c (catchsig, catchsegv): If errcount > 0, just exit,
don't abort.
2019-04-12 Arnold D. Robbins <arnold@skeeve.com>
* configure.ac: Update version to 5.0.0.
* configure: Regenerated.
* ChangeLog.1: Rotated ChangeLog into this file.
* ChangeLog: Created anew for gawk 5.0.0 and on.
* README: Bump version to 5.0.0.
* 5.0.0: Release tar ball made.
Reference in a new issue View git blame Copy permalink