m4-commit
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[SCM] GNU M4 source repository annotated tag, v1.9a, created. v1.9a


From: Eric Blake
Subject: [SCM] GNU M4 source repository annotated tag, v1.9a, created. v1.9a
Date: Tue, 08 Apr 2008 17:14:57 +0000

This is an automated email from the git hooks/post-receive script. It was
generated because a ref change was pushed to the repository containing
the project "GNU M4 source repository".

http://git.sv.gnu.org/gitweb/?p=m4.git;a=commitdiff;h=b6c4ff3e039c2181d0ffae56c98a87b5486a8030

The annotated tag, v1.9a has been created
        at  b6c4ff3e039c2181d0ffae56c98a87b5486a8030 (tag)
   tagging  0bed0cfbcc944a93edf719af0db3c8097fc8122f (commit)
 tagged by  Eric Blake
        on  Wed May 4 15:46:00 2005 +0000

- Log -----------------------------------------------------------------
use 1.9a, not 1.4q, as version; not a true release

Akim Demaille (112):
      Update.
      Update.
      * config/Makefile.am, config/stackovf.m4: New.
      * config/gnu-obstacks.m4: New.
      More ignores, and fix some typos.
      * m4/evalparse.c: A better introductory comment.
      Perl out the trailing blanks.
      * m4/evalparse.c: A better introductory comment.
      Improve `make distcheck'.
      * configure.in: Require Autoconf 2.52.
      Regen.
      * config/gnu-obstack.m4 (M4_AC_FUNC_OBSTACK): Be sure `m4/' exists
      * m4/m4module.h (m4/list.h): Don't include it.
      * m4/builtin.c (m4_macro_define): Do not reset SYMBOL_TRACED.
      * m4/symtab.c (m4_hack_all_symbols, m4_lookup_symbol): Formatting
      * modules/Makefile.am (ETAGS_ARGS): New.
      Drop Autoconf 2.13 compatibility.
      Start using Autotest.  atconfig creation is still lacking.
      * doc/m4.texinfo (Esyscmd): Fix the paths in the Vice example.
      others.at no longer depends on other-tests/.
      Remove the non Autotest tests.
      * doc/m4.texinfo: Promote proper quotation.
      Fix a serious problem with a patch application.
      More ignores.
      Really apply the patch ``Drop Autoconf 2.13 compatibility.''
      Add.
      * tests/Makefile.am (EXTRA_DIST): Ship generate.awk.
      Regen.
      * tests/testsuite.at (AT_CHECK_M4): Anchor M4PATH in $top_srcdir.
      * tests/others.at (iso8859): Specify the path to the M4 test file.
      Adjust.
      Remove.
      Use sprintf, not ecft and friends since it is standard, portable,
      * examples/include.m4: Typo.
      * m4/utility.c (m4_numeric_arg): Spell out the culprit.
      * config/atconfig.m4: New, until part of Autoconf per se.
      Remove.
      * config/atconfig.m4: s/EOF/ATEOF/, so that configure can be
      * Makefile.am (AUTOMAKE_OPTIONS): 1.5.
      * tests/Makefile.am (CLEANFILES): Remove.
      Let the test suite use a wrapper around the not installed m4 to
      * tests/others.at (capitalize, changeword, comments, ddivert)
      * tests/Makefile.am (installcheck-local): New.
      * tests/others.at (indir): Formatting change.
      * tests/Makefile.am (installcheck-local): Don't use
      Regen.
      Typo.
      * examples/indir.m4, tests/others.at (indir): s/nonsens/nonsense/.
      * tests/generate.awk (normalize): s/@__@/@&t@/.
      * tests/Makefile.am ($(srcdir)/$(TESTSUITE)): Use autom4te's
      * tests/testsuite.at (AT_CHECK_M4): Don't pass -I, let the tests
      * tests/others.at (Freezing modules)
      * tests/others.at (changeword, ddivert, debug, esyscmd, exp, gmp)
      * bootstrap.sh: Create aclocal.m4 instead of modifying it.
      Improve the test suite's selfcontainedness.
      * src/main.c (MODULE_SHORTOPT, MODULEPATH_SHORTOPT): Remove, there
      * m4/symtab.c (m4_lookup_symbol): Give more details when reporting
      * m4/m4module.h (m4_symbol_lookup): Remove M4_SYMBOL_IGNORE,
      * m4/m4module.h (m4_symbol_lookup): Rename as...
      * m4/symtab.c (m4_symbol_popdef, m4_symbol_delete): Rename as...
      * m4/m4module.h (m4_symbol_lookup): Remove M4_SYMBOL_LOOKUP.
      * m4/symtab.c (m4_symbol_insert): Don't use `foo () || bar ()'
      Rationalize warnings, reporting the macro name as if it were a
      * modules/format.c (format): Have a bigger `str'.
      Add dist-bzip2 to Automake options.
      * src/main.c (main): Standardize --version.
      * configure.in: We now need Autoconf 2.52e.
      * modules/stdlib.c, modules/times.c: Misc cleanups.
      * tests/Makefile.am (package.m4): New.
      Regen.
      * config/gmp.m4 (AM_WITH_GMP): Rename as...
      Adjust.
      Update.
      Autotest has changed again.
      tests/m4 shall be position independent.
      * tests/m4.in: Don't use short options.
      Add configure signature.
      * tests/Makefile.am (EXTRA_DIST): `m4' is not to be shipped.
      * modules/gnu.c (RE_SYNTAX_BRE, RE_SYNTAX_ERE): New.
      * tests/generate.awk: Remove debugging code.
      * config/gmp.m4: Consult the user before using GMP: use
      * tests/generate.awk: Tag the tests with `documentation'.
      * m4/utility.c (m4_bad_argc): Detail the mismatches.
      * modules/modtest.c (init_handler): Consistently output to stderr.
      Adjust.
      * modules/gnu.c (m4_regexp_do, m4_patsubst_do): Extracted from
      typo.
      * tests/Makefile.am: Adjust for gnuprog2.
      Typo.
      * m4/output.c (m4_make_diversion, m4_insert_file)
      * m4/utility.c (m4_bad_argc): Display user argument counts, i.e.,
      * m4/macro.c (expand_macro): Let m4_bad_argc handle the cases
      * modules/evalparse.c: Save translators' time: don't translate
      * m4/utility.c (m4_numeric_arg): Use the usual (argc, argv, ...)
      * tests/builtins.at (pushdef/popdef, trace, trace2, trace3): Move
      * tests/macros.at (Arity and defn): New failing test.
      * src/main.c, src/freeze.c, m4/debug.c, m4/input.c, m4/macro.c:
      * m4/ltdl.c: Update.
      * m4/input.c, modules/m4.c: Formatting changes.
      * m4/macro.c (expand_argument): Display where was opened what is
      * m4/input.c (m4_next_token): Display where was opened what is
      * m4/module.c, m4/output.c, src/main.c: Normalize error messages.
      Regen.
      Regen.
      * bootstrap (aclocal_apilibdir): New, to cope with Automake's new
      Regen.
      Fix the headers.
      Regen.
      * configure.ac: Automake 1.7.1, Autoconf 2.54, Gettext 0.11.5,
      * config/Makefile.am (ACLOCAL_MACROS): Ship the Gettext macros.
      Ship regex.m4.
      * doc/m4.texinfo (Esyscmd): Don't grep, use something easier:

Gary V. Vaughan (194):
      Resynchronise with my local tree.
      Modularization of all builtins by extenstion of libm4.
      * config/ltmain.sh:  This needs to be checked in to match the
      mv m4.c main.c
      typo.
      * po/POTFILES.in: List files in their new positions.
      These files are generated, and needn't be in CVS.
      * configure.in: Figure out whether to use the installed
      * m4/Makefile.am (libm4_la_LIBADD): New magic -- we are
      Regenerated.
      * src/main.c (main):  It is now safe to
      * modules/perl.c (builtin_perleval):  Use PL_na for
      * po/POTFILES.in: Take account of these file movements.
      Should have been cvs rm'ed a while back...
      * m4/ltdl.c: Upgraded to latest bleeding edge version again.
      Clear up the file droppings from a cvs -nq up.
      * m4/Makefile.am (libm4_la_SOURCES): Removed obstack.c, since it
      * configure.in (--with-modules): Changed semantics.  This
      * src/main.c (main): Fixed a particularly nasty bug is
      Remove generated files from CVS.
      * tests/other-tests/changeword.test: New test.  Rudimentary
      Move the implementation of GMP support for the mpeval() builtin
      Consolidate the myriad of dispersed test scripts into the
      * TODO: Updated.  Removed some cruft that has since been fixed
      * TODO:  Removed m4exit bug.
      * modules/format.c (format):  Formatting for %f was not
      * ChangeLog:  Cleaned.  Gettextize has a stutter
      * src/main.c (main): Fix improper uses of error (missing format
      * doc/m4.texinfo: Fix some typos, and apply some small
      * src/main.c (main): Add `-b' (for `batch') mode switch to force
      * src/main.c (main): Use lt_dlinsertsearchdir to prepend -M
      * bootstrap: Rewritten to play nicely with Autoconf 2.5x.
      * acconfig.h: Removed.  No longer required.
      * m4/m4module.h (obstack.h): Choose between installed version
      * config/ltmain.sh: Doh!  Import again for a quoting fix that
      m4_modules are no more, we use lt_dlhandles directly and let
      * m4/module.c: Fixup some errors in the description comment.
      * tests/generated-tests/.cvsignore: Deleted, so that 'cvs up -Pd'
      * tests/generated-tests/.cvsignore: Deleted, so that 'cvs up -Pd'
      * bootstrap: revert my last patch.  commited from the wrong directory.
      * config/debug.m4 (M4_AC_CHECK_DEBUGGING): Cleaned up.  Don't test
      * c-boxes.el: Deleted.
      Ansify the source.  Previously we had a mix, where my code was
      * m4/system.h [cygwin*]:  Thanks to Paul Sokolovsky and
      * bootstrap (generate.awk): Generate test cases with new script.
      * tests/generate.awk: When making substitutions to "@&t@", either
      In an attempt to moderate my egomania...
      The experimental `changeword' feature never took off, and has
      * tests/m4: Don't call the m4 wrapper script with libtool, it
      * src/Makefile.am (m4_SOURCES): Don't list $(M4OBJS) here, since
      * m4/m4private.h (SYMBOL_NEXT): Move into m4's name space; renamed
      The shadowing mechanism and, infact, the whole symbol table
      * m4/hash.c, m4/hash.h: forgot to cvs add these files.
      * config/debug.m4 (AC_LIBTOOL_COMPILER_OPTION): Provide a fallback
      Bootstrap will now work with Libtool 1.4.1 or higher:
      * TODO: Remove the items that have been addressed already.
      Fix the obstack.h problem once and for all!  At configure time we
      * m4/symtab.c (m4_symtab_remove_module_references):  Renamed as
      Revert my last patch... too fast on the commit button, will try again in 
small
      * m4/symtab.c (m4_symtab_remove_module_references):  Renamed as
      Get rid of m4_symbol_lookup_t entirely.  With Akim's earlier
      * m4/m4module.h, m4/builtin.c, m4/symtab.c:
      * m4/hash.c (m4_hash_bucket_insert): Symbol shadowing is no longer
      * m4/m4private.h (struct m4_token_data): Simplified by removing
      The `traced' flag needs to be attached to the definition of a
      More cleanup.  After the last patch, m4_symbol holds nothing but
      * m4/evalparse.c: Moved to...
      * config/gmp.m4 (AC_ARG_WITH): Removed.
      * TODO: stackovf.c is basically broken.
      * configure.in: Moved to...
      Reverted my large patch for removing the old m4_symbol
      Reinstate the memory handling improvements from the patch I just
      * configure.ac (TIMESTAMP): Display with AS_BOX at configure
      * configure.ac: `rm -f $m4_getopt_h' was lost.
      * bootstrap: Remove aclocal.m4t when it is no longer required.
      * configure.ac (--with-modules): Forgot a comma in the
      * modules/gnu.c (m4_macro_table): Display the timetamp when
      The trace semantics now attach the trace bit to a symbol name.
      Rather than forcing each builtin definition to manage its own
      * modules/gnu.c (m4_regexp_compile): s/%0/%s/
      * m4/input.c (init_builtin_token):  Renamed from init_macro_token,
      * m4/symtab.c (m4_symbol_destroy): This function calls
      * bootstrap: s/configure.in/configure.ac/
      * m4/symtab.c (m4_symtab_hash, m4_symtab_cmp): Moved from here...
      * m4/hash.c (m4_hash_resize): New function.
      * bootstrap (aclocal_apiversion): The aclocal apiversion is
      Preparations for refactoring syntax tables to allow reverse
      These changes were necessary to get m4 to build on my iBook
      * m4/input.c: Updated various bitrotted comments.
      * m4/macro.c: Format changes.
      Begin preparations for supporting formal parameters in m4 macros.
      First cut at formal parameters in macros.
      * m4/symtab.c (m4_symbol_popdef): Need to pass the hash address to
      Getting rid of the annoying bug with configure --enable-debug,
      * README: Remove references to --enable-changeword, which has been
      Refactoring modules to rationalise the API into an external
      More refactoring to stabilise the module api, this time for
      * m4/hash.c (m4_hash_apply): Pass an initial hash table parameter
      Begin work on lifting the curse of the global variables.  To start
      Further refactoring to stabilise the module API.  Renaming some
      Still refactoring furiously.  This delta represents a change in
      Tidy up style of hash.[ch] in accordance with doc/STYLE.
      Renamed some of the types and their accessors to make more sense.
      Tie down the interface to libm4 some more.  Make more structures
      Two related changes, and a huge knockon effect throughout the
      * m4/hash.c (m4_hash_resize): ifdefed out.  This function is
      * m4/module.c (m4_set_module_macro_table)
      * doc/STYLE: Added notes on callback naming schemes.
      Move the global variables that pertain to syntax to a new `struct
      * doc/m4.texinfo (Eregexp and Regexp): Typo.
      * doc/m4.texinfo (Changesyntax): Document Dollar syntax class.
      Conform to POSIX if the POSIXLY_CORRECT environment is set.
      * NEWS: Note `$' syntax class.
      * m4/input.c (CHAR_EOF, CHAR_BUILTIN, CHAR_RETRY): These token
      Change the macros for checking syntax so that matching against
      More refactoring of the m4module.h API.  Replace #defines with an
      * m4/m4module.h (program_name): Not used by modules at all, so
      * m4/m4module.h: Reformatting and rearranging lines a bit.
      An experiment: There is loads of code in libm4 which should be in
      Create an export table in modules/m4.c in readiness for beefing up
      Sometimes exporting the address of a symbol from a module isn't
      Aside from auditing path.c for m4module.h api obfuscation, this
      * m4/m4module.h (M4_DEFAULT_NESTING_LIMIT): This value is
      * m4/utility.c (dumpdef_cmp): Removed stale declaration.
      * m4/m4module.c (m4_string): Moved from here...
      Libltdl already has an excellent mechanism for accessing C symbols
      * tests/Makefile.am (TESTS_ENVIRONMENT): Now we can pass make
      * config/debug.m4 (M4_AC_CHECK_DEBUGGING): Don't AC_REQUIRE a
      * config/gnu-obstack.m4 (M4_AC_FUNC_OBSTACK): Use the interrim new
      Still cleaning up the users' module API, by taking out the stuff
      * configure.ac (AC_HEADER_STDBOOL, INCLUDE_STDBOOL_H): Check for
      * configure.ac (AC_CONFIG_FILES): Add tests/atlocal.
      * config/gmp.m4 (M4_AC_LIB_GMP): Use AC_INCLUDES_DEFAULT:
      * config/debug.m4: `perl -pi.bak -e 's/(Copyright) (d)/$1 (C) $2/g'`
      * m4/stdbool_.h: New file from gnulib for systems without their
      * m4/system-h.in: For consistency with the gnulib imports, renamed
      * m4/gnu-obstack.h: Updated from CVS gnulib.  For consistency with
      * bootstrap: CVS autoreconf leaves file droppings.  Remove
      * config/Makefile.am (MAINTAINERCLEANFILES, ACLOCAL_MACROS)
      GNU M4 currently builds with a number of files adapted from
      forgot to cvs add this file in the last commit
      Migrate the obstack support into the gnulib directories for easy
      * config/Makefile.am: Reverting yesterdays patch for VPATH builds.
      We can't mirror the gnulib directory structure here, since we need
      The gnulib obstack module requires the gnulib gettext module for
      * config/m4-obstack.m4 (m4_FUNC_OBSTACK): D'oh! Now that we wrap
      * bootstrap: Revert 2003-09-04 patch now that CVS autoconf and
      * gnulib/m4/Makefile.am (MOSTLYCLEANFILES): Typo
      Import the error and progname modules from CVS gnulib.  Our old
      Import the xalloc module from CVS gnulib, along with its
      Reimport the latest xalloc module from CVS gnulib, and adjust the
      Replace the xstrzdup code by importing the xstrndup module from
      * gnulib/m4/Makefile.am (EXTRA_DIST): Removed malloc.c and
      * gnulib/m4/regex.c, gnulib/m4/regex_.h, gnulib/config/regex.m4:
      * configure.ac (pkglibexecdir): Don't try and set it here, PACKAGE
      * configure.ac (AC_CONFIG_FILES): Remove config/Makefile.
      Forgot to commit this file with the last patch
      new file
      reduce noise
      Replace the getopt code with CVS gnulib getopt.  The source files
      * gnulib/m4/strtol.c, gnulib/config/strtol.m4: Import strtol
      * doc/STYLE: Document the gnulib header naming conventions, and
      * gnulib/m4/regex.c:  s/<regex.h>/"regex.h"/ or else the compiler
      * bootstrap: Separate options.
      * gnulib/m4/exit.h: Import exit module from CVS gnulib.
      * gnulib/m4/version-etc.c, gnulib/m4/version-etc.h: Import
      forgot this file in the last commit
      * Makefile.am (ACLOCAL_AMFLAGS): Search in the new acm4 and
      Don't use XMALLOC and XCALLOC.  Once we install the
      Tru64UNIX perl is sloppy with namespace pollution.  This patch is
      * tests/modules.at (modules: importing): Edit the generated stderr
      * tests/modules.at (modules: importing): Apply some lateral
      * tests/module.at: Comment typo.
      The import script was not written properly, and the last gnulib
      * configure.ac (AC_LIBTOOL_TAGS): Don't include shell code for
      * m4/symtab.c (m4_symtab_create): Fix a careless use of sizeof.
      * gnulib/acm4/libtool.m4, m4/ltdl.c, m4/ltdl.h: Reimported from
      * modules/Makefile.am (pkglibexec_LTLIBRARIES): Remove perl.la
      * configure.ac (AM_INIT_AUTOMAKE): Require 1.7g...
      * config/mkstamp: Updated from CVS libtool.
      * config/mailnotify: New file for mailing commit notifications,
      * gnulib/import: Now updates makefile fragments, and configure.ac.
      Tweaking to enable compilation with latest CVS libtool and
      Latest CVS libtool can preload modules in libraries, including
      * gnulib/*: Don't store any of the gnulib files in arch, as they
      * src/main.c: Include gnulib files from the correct directory.
      * m4/m4.c (m4_context_field_table, m4_context_opt_bit_table):
      * configure.ac (INCLUDE_STDBOOL_H): Account for gnulib's move
      * TODO: Add qindir requirement, and defn bug.
      * tests/Makefile (clean-local): Only run the testsuite cleanup
      * tests/Makefile (clean-local): Ignore testsuite cleanup
      The FSF are moving offices today.  Changed their contact address
      http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=84416
      Gnulib has changed again.  Reimport.  Adjust.  Rinse.  Repeat.

Jeff Bailey (1):
      Initial revision

Stepan Kasal (3):
      wish execution stack; fix typos
      clarify the qindir example
      clarify the qindir example

-----------------------------------------------------------------------


hooks/post-receive
--
GNU M4 source repository




reply via email to

[Prev in Thread] Current Thread [Next in Thread]