2014-05-01 Andreas Kupries <andreask@activestate.com>
* exp_main_sub.c: Updated EXP_VERSION to 5.45.3
* configure, configure.in: Updated expect to version 5.45.3
* expect.man [http://sourceforge.net/p/expect/bugs/86/]. Report
and fix by Vitezlav Crhonek.
* expect.c [http://sourceforge.net/p/expect/bugs/76/]. Report and
fix by Mutsuhito Iikura. On finding a full buffer during
matching the sliding window mechanism slides too far, truncating
the whole buffer and preventing matches across the boundary. Fix
is shortening the slide distance (slide only one 1/3).
* expect.c: [http://sourceforge.net/p/expect/patches/18/]. Report
and fix both by Nils Carlson <nils.carlson@ludd.ltu.se>.
Replaced a cc==0 check with proper Tcl_Eof() check.
2013-11-04 Andreas Kupries <andreask@activestate.com>
* tclconfig/tcl.m4: [http://sourceforge.net/p/expect/patches/17/]
* configure: Extended Tcl header detection for OS X Mountain Lion.
* exp_main_sub.c: Updated EXP_VERSION to 5.45.2
* configure, configure.in: Updated expect to version 5.45.2
* expect.c: [http://sourceforge.net/p/expect/patches/16/]. Report
and fix both by Per Cederqvist. Replaced a memcpy with memmove
as the latter properly handles overlapping memory, whereas the
original code does not.
2012-08-15 Andreas Kupries <andreask@activestate.com>
* exp_main_sub.c: Updated EXP_VERSION to 5.45.1
* configure, configure.in: Updated expect to version 5.45.1
* exp_chan.c: Applied patch sent in by Ogawa Hirofumi
<hirofumi@mail.parknet.co.jp>. The patch fixes a problem when
talking a tty where the writer has died. Some operating systems
report the condition as EIO with nothing read, while this
actually an EOF. Without the patch the returned data is
incomplete due to the error reported immediately and dropping
data in buffers.
2012-05-16 Andreas Kupries <andreask@activestate.com>
* exp_chan.c: [Bug 3526461]: Applied patch by Michael Cleverly
<cleverly@users.sourceforge.net> fixing a problem with the
iteration over the expect channel list where the loop code may
modify the list, breaking the iterator.
* exp_chan.c: [Bug 3526707]: Applied patch by Michael Cleverly
* exp_command.h: <cleverly@users.sourceforge.net> fixing problem
* expect.c: with an insufficient test for a lost channel in
exp_background_channelhandler.
2010-10-26 Andreas Kupries <andreask@activestate.com>
* expect.c: [Bug 3095935]: Convert #bytes information to #chars to
prevent later code to fail when copying strings around and
miscalculating how much to copy, for strings containing
non-ASCII utf chars.
2010-09-30 Andreas Kupries <andreask@activestate.com>
* example/autopasswd: Updated to use tclsh in PATH, and 'package
* example/chess.exp: require Expect'. Obsoletes fixline1.
* example/cryptdir:
* example/decryptdir:
* example/dislocate:
* example/dvorak:
* example/ftp-inband:
* example/ftp-rfc:
* example/gethostbyaddr:
* example/kibitz:
* example/lpunlock:
* example/mkpasswd:
* example/multixterm:
* example/passmass:
* example/read1char:
* example/rftp:
* example/rlogin-cwd:
* example/robohunt:
* example/rogue.exp:
* example/telnet-cwd:
* example/timed-read:
* example/timed-run:
* example/unbuffer:
* example/virterm:
* example/weather:
* example/xkibitz:
* example/xpstat:
* Makefile.in: Continued work on the 'make dist' target (config.{sub,guess})
* install.sh: Removed unused file.
* mkinstalldirs: Removed unused file.
* tclconfig/README.txt: Removed unused file.
* testsuite/config/: Removed contents of unused directory.
* test/.Sanitize: Removed unused file.
2010-09-17 Jeff Hobbs <jeffh@ActiveState.com>
* Makefile.in: improved make dist target to include the necessary
files, and remove old dist_orig.
Enable Makefile regen target
2010-09-16 Jeff Hobbs <jeffh@ActiveState.com>
* configure: regen with ac-2.59
* tclconfig/tcl.m4: update for correct LDFLAGS on HP-UX
* Makefile.in (expect): use TEA LDFLAGS instead of tclConfig.sh to
build expect executable
2010-09-01 Andreas Kupries <andreask@activestate.com>
* Makefile.in: [Bug 865278] Semi-applied Don Porter's patch adding
code which prevents the package from registering with Tcl
interpreters which cannot load it, i.e. below the version it was
compiled against. Semi because the pkgIndex.in in his patch is
not used, the pkgIndex.tcl is generated by the Makefile(.in).
* pkgIndex.in: Removed. File not used.
2010-08-31 Andreas Kupries <andreask@activestate.com>
* Various cleanups, local patches of ActiveState.
* exp_clib.c: Remove local copy of Tcl_ErrnoMsg().
* exp_inter.c: Hack access to TCL_REG_BOSONLY when not present,
became private with Tcl 8.5 and higher.
* expect.h: Remove the local fiddling with the memory allocation
and panic macros.
* Dbg.c: [Bug 2972727]: Applied the parts of Larry Virden's patch
which were still current. Most of it applied to the partially
ansified sources we had in trunk for a time and where errors in
that set of changes, thus out of date now.
* example/unbuffer: [Bug 2949748]: Applied patch by Vitezslav
Crhonek <vcrhonek@users.sourceforge.net> to prevent unbuffer
from eating the exit code of the command it ran (regular mode
only, not -p). Slightly modified, I removed the superfluous
'eval' used to invoke 'wait', invoking it directly.
2010-08-27 Jeff Hobbs <jeffh@ActiveState.com>
* retoglob.c: Fail if the generated glob contains more than
two asterisks. Fixes [Expect SF Bug 3010684] (cederqvist)
* exp_main_sub.c: add return to silence compiler warning.
Updated EXP_VERSION to 5.45.0
* config.guess, config.sub, expect.m4 (removed):
* configure, configure.in, tclconfig/tcl.m4: Update to TEA 3.9
* Makefile.in, aclocal.m4: Partial cleanup of the build system.
* testsuite/Makefile.in: Remove unused EXP_ from configure.in
* testsuite/aclocal.m4: and no longer use Tk.
* testsuite/configure: Update testsuite bits to use TEA info.
* testsuite/configure.in:
Update expect to version 5.45
* Dbg.c: account for removal of tcldbgcf.h
* DbgMkfl.in, Dbgconfig.in, Dbgconfigure, DbgpkgInd.in (removed):
* tcldbgcf.h.in (removed): removed Dbg debugger subcomponent as it
no longer built and debugger functionality is included in expect
library and binary
* pty_termios.c: add HAVE_STRING_H include string.h
* exp_trap.c: add HAVE_STRING_H include string.h
* expectk.man, exp_main_tk.c (removed): expectk no longer
built. Use tclsh with package require Tk and Expect instead.
* tests/all.tcl: add package require Expect
* example/archie, example/autoexpect: minor code cleanup
* example/tkpasswd, example/tknewsbiff, example/tkterm: use
package require statements in lieu of expectk
2009-11-03 Andreas Kupries <andreask@activestate.com>
* exp_command.c (Exp_SpawnObjCmd): [Expect SF Bug 2891422]. Fixed
error message when the command to spawn could not be executed.
Tried to use a Tcl_Obj* as argument for a %s. Replaced with the
correct char* variable. Thanks to Elchonon Edelson
<eee@users.sourceforge.net> for the report.
[Expect SF Bug 2891563] Ditto for the same type of bug when
emitting error 'usage: -ignore unknown signal name'. The
remainder of the exp_error calls are ok however.
* configure.in: Bumped version to 5.44.1.15.
* configure: Regen'd, autoconf 2.59.
2009-07-14 Andreas Kupries <andreask@activestate.com>
* exp_clib.c (TclRegComp): Fixed memory leak reported by
<ettlmartin@users.sourceforge.net> in [Expect SF Bug 2814263].
2009-06-22 Andreas Kupries <andreask@activestate.com>
* pty_unicos.c (pty_stty): Fixed missing double-quotes for sprint
formatting strings. Thanks to <ettlmartin@users.sourceforge.net>
for the report, i.e. [Expect SF Bug 2809496].
2009-06-18 Andreas Kupries <andreask@activestate.com>
* exp_command.c (Exp_LogFileObjCmd): Fix argument parsing logic
error in the log_file command breaking the use of options -open
and -leaveopen. Applied patch supplied by Andy Belsey
<Andrew.Belsey@Sun.COM>. With thanks for both the analysis of
the problem and the patch.
* configure.in: Bumped version to 5.44.1.14.
* configure: Regen'd, autoconf 2.59.
2009-05-27 Andreas Kupries <andreask@activestate.com>
* exp_tty.c (Exp_SttyCmd, Exp_SystemCmd): Applied patch by
Reinhard Max (max@suse.de) fixing buffer-overflows in the 'stty'
command due to an off-by-one error in the buffer size. See the
bugs https://bugzilla.novell.com/show_bug.cgi?id=506873
and https://bugzilla.novell.com/show_bug.cgi?id=501291
* configure.in: Bumped version to 5.44.1.13.
* configure: Regen'd, autoconf 2.59.
2009-05-06 Andreas Kupries <andreask@activestate.com>
* retoglob.c: Accepted the patch by Mike Cumings
<mcumings@users.sourceforge.net> fixing [Expect SF Bug 13179].
The updated code checks a (?...) sequence that it contains only
ARE options before processing it as such. This prevents the
misinterpretation of non-capturing paren groups as ARe options
with subsequent segmentation fault.
* configure.in: Bumped version to 5.44.1.12.
* configure: Regen'd, autoconf 2.59.
2008-09-30 Andreas Kupries <andreask@activestate.com>
* configure.in: Bumped version to 5.44.1.11.
* configure: Regen'd, autoconf 2.59.
* exp_command.c (Exp_OverlayObjCmd): Fixed [Expect SF Bug 2127456]
reported by <eee@users.sourceforge.net>, with thanks. Changed
retrieval of command from objv[0] (== 'overlay' itself), to
objv[i] containing the actual user command.
* expect.c (string_case_first, string_first, eval_case_string):
Applied patch supplied by Andy Belsey <Andrew.Belsey@Sun.COM>
fixing the seg.fault in 'expect -exact'. With thanks for both
the analysis of the problem and the patch. See also [Expect SF
Bug 2114547].
2008-08-28 Andreas Kupries <andreask@activestate.com>
* exp_trap.c (tophalf): Fixed inverted condition setting the
interpreter used for trap handling to NULL, causing a crash when
trying to handle ^C. This fixes [SF Bug 1757471] reported by
Matthias Kraft <matzek@users.sourceforge.net>.
* configure.in: Bumped version to 5.44.1.10.
* configure: Regen'd, autoconf 2.59.
2008-08-18 Jeff Hobbs <jeffh@ActiveState.com>
* exp_main_sub.c (exp_interpreter): cleaner handling of commandPtr
to prevent crash. (das)
2008-06-03 Andreas Kupries <andreask@activestate.com>
* exp_glob.c (Exp_StringCaseMatch2): Fixed bug in the handling of
glob classes, see [SF Bug 1873404]. The code tried to match the
closing bracket of a class in the input because it was not
properly skipped after the class was matched successfully.
Additional trace output added.
* configure.in: Bumped version to 5.44.1.9.
* configure: Regen'd, autoconf 2.59.
2008-05-05 Andreas Kupries <andreask@activestate.com>
* exp_pty.c: Minimal ansification of function definitions to match
them to their prototypes where promoted types are otherwise
causing a mismatch for some compilers, like AIX in 64bit mode.
* configure.in: Bumped version to 5.44.1.8.
* configure: Regen'd, autoconf 2.59.
2008-04-03 Andreas Kupries <andreask@activestate.com>
* configure.in: Bumped version to 5.44.1.7.
* configure: Regen'd, autoconf 2.59.
* The following set of changes was sent our way by Reinhard Max
<max@tclers.tk>.
* exp_command.c: Fixed more compiler warnings, and started
* exp_command.h: to ansify the code base, beginning with
* exp_inter.c: the introduction of proper function prototypes.
* exp_main_exp.c:
* exp_pty.h:
* exp_tty.c:
* exp_tty.h:
* exp_win.c:
* expect.c:
* pty_termios.c:
* retoglob.c:
2008-04-03 Andreas Kupries <andreask@activestate.com>
* configure.in: Bumped version to 5.44.1.6.
* configure: Regen'd, autoconf 2.59.
* The following set of changes was sent our way by Reinhard Max
<max@tclers.tk>.
* configure.in: Fixed checking of stty on linux, do not restrict
to specific cpu's. Further try with stdin, and stdin redirected
to /dev/tty when determining if stdout is read.
* testsuite/configure.in: Brought up to TEA 3.5.
* testsuite/aclocal.m4: New file, to import the TEA definitions.
* Dbg.c: Added missed CONST in declaration and definition of
'debugger_trap'.
* exp_command.c: Fixed pointer aliasing trouble with
'Tcl_DetachPids', and added the missing initialization of the
command string in the 'overlay' command.
* expect.c: Fixed missing initialization of 'simple_start' element
of 'ecase'.
* exp_inter.c: Fixed bogus use of 'slen' in 'intMatch'. The
relevant string is Tcl_Unichar, not Utf.
* Makefile.in: Replaced bogus INSTALL_ROOT with DESTDIR, and added
missing DESTDIR references to the target for the installation of
the manpages.
2008-02-27 Andreas Kupries <andreask@activestate.com>
* expect.c: Fixed refcounting error when parsing a single expect
* configure.in: argument into a list of patterns and
* configure: actions. Updated the version number to 5.44.1.5.
* Dbg.c: Added missing 'return TCL_OK' to debugger_trap.
2007-12-13 Jeff Hobbs <jeffh@ActiveState.com>
* exp_log.c (expStdoutLogU): correct which buf is output.
* exp_command.c (Exp_SendLogObjCmd): fix '--' handling
(Exp_SendObjCmd): fix '--' handling to expect last argument
2007-09-24 Andreas Kupries <andreask@activestate.com>
* exp_inter.c: Changed inclusion of tcl.h to tclInt.h to get the
* expect.c: definition of TCL_REG_BOSONLY, which was moved to that
header in Tcl 8.5. Ditto for expect.c, for the macro
TclUtfToUniChar (was a function in 8.4). Expect now compiles
again for both Tcl 8.4 and 8.5.
* configure.in: Bumped version to 5.44.1.4.
* configure: Regenerated.
2007-09-19 Andreas Kupries <andreask@activestate.com>
* retoglob.c (EMITC): Keep the characters '^' and '$' quoted as
well, we do not wish to invoke their special interpretation by
the Expect glob matcher.
* configure.in: Bumped version to 5.44.1.3.
* configure: Regenerated.
2007-08-09 Andreas Kupries <andreask@activestate.com>
* retoglob.c: We had ExpChopNested and ExpBackslash locally
ansified (prototypes) to avoid a compiler error on AIX (type
promotion trouble). This now integrated into the mainline
sources, conditional to AIX.
* expect.c (Exp_TimestampObjCmd): Fixed argument processing broken
by objc,objv conversion. Tried to use command name as the
argument for -seconds. Also did not detect when one argument to
many was specified.
* configure.in: Bumped version to 5.44.1.2.
* configure: Regenerated.
2007-07-17 Andreas Kupries <andreask@activestate.com>
* expect.c: Circumvented problems with the C compiler by use of a
temporary variable to hold the unicode pointer of a glob
pattern. The computed pattern length given to
Exp_StringCaseMatch was bogus.
* exp_glob.c: Added tracing of the glob matcher internals (Can be
enabled by define EXP_INTERNAL_TRACE_GLOB). Fixed bug in a guard
condition in the optimized handling of '*'. The bad condition
caused the code to miss possible matches at the beginning of the
input (first char).
* tests/expect.test: Added tests which check the glob matcher and
RE gate keeping.
* configure.in: Bumped to 5.44.1.1 to separate this from the
regular 5.44.1 sources.
* configure: Regenerated.
2007-07-12 Andreas Kupries <andreask@activestate.com>
* expect.c: Found bugs mismanaging input and pattern in the
* exp_glob.c: updated glob matcher. We cannot check for '\0'
anymore to find the end of the string, these are counted
arrays now. Rewritten to use sentinel pointers.
2007-07-11 Andreas Kupries <andreask@activestate.com>
* exp_chan.c: Converted the buffering system from UTF-8 in Tcl_Obj
* exp_command.h: to UTF-16 C-array, to avoid the repeated conversion
* expect.c: of the input from utf-8 to utf-16. Updated the glob
* exp_glob.c: matching code to use the same tricks for speed which
* exp_inter.c: are used by the Tcl core. Extended the regexp
* exp_log.c: matching path with a glob matcher which uses a gate
* exp_log.h: keeper glob pattern to weed out most non-candidates
* retoglob.c (New file): in a speedy manner. Regexp matching now
has to be done only for the small number of candidates
identified by the gate keeper. Overall speed improvement as glob
matching is faster than regexp matching. Added code translating
regular expressions into their gate keeper glob pattern.
* Dbg.c: Converted the lot of argc,argv based command
* exp_command.c: implementations over to objc,objv.
* expect.c:
* exp_main_sub.c:
* Dbg.c: Cleaned up the direct access to interp->result,
* exp_command.c: replaced with the proper functions and
* expect.c: Tcl_Obj's.
* exp_main_exp.c:
* exp_main_sub.c:
* exp_main_tk.c:
* exp_prog.h:
* exp_trap.c:
* exp_tty.c:
* exp_win.c:
* exp_win.h:
* tests/cat.test: Added proper 'package require Expect'
* tests/expect.test: to the test setup code (JeffH).
* tests/logfile.test:
* tests/pid.test:
* tests/send.test:
* tests/spawn.test:
* tests/stty.test:
* exp_command.c: Reformatted overlong lines, whitespace,
* expect.c: comments. Added braces to some if-constructs.
* exp_inter.c: Reworked if-constructs interleaved with
* exp_tty.c: #if for better formatting in emacs.
* Dbg.c: Added note about unhandled cases in a switch.
* exp_chan.c: Added code to suppress unhandled warning for
unreachable code.
* exp_command.c: Removed unused variable.
* expect.c: Removed unused static function, added code to suppress
unhandled warning for unreachable code.
* exp_command.c: Fixed typo in comment.
2007-06-28 Andreas Kupries <andreask@activestate.com>
* Merged changes from the official version 5.44.1 of expect into
the SF sources. See the details below.
--------------------
Marius Schamsula <marius173@mchsi.xcom> reported tclconfig
missing, evidentally for new TEA.
Lots of massaging to fix TEAification of Makefile and configure
including that version numbers will now be full three part.
Daniel Wong <danielwong@berkeley.xedu> noted the home page
should note that Wikipedia has a very readable entry for Expect.
Andre Alves <aalves@escloyalty.xcom> noted passmass needed some
fixes to handle Solaris 9 passwd prompt changes.
Andreas fixed several things: changes to better support TEA, fix
debugger interaction with nonblocking mode, and probably other
things I'm overlooking.
Martin Dietze <di@fh-wedel.xde> noted that autoconf 2.59 is
confused by C comment after undefs in expect_cf.h.in.
Added additional code to unbuffer -p so that if a process
earlier in the pipeline exits, unbuffer attempts to recover any
remaining output from the spawned proc before unbuffer itself
exits.
Jeffrey Hobbs noted that once stty was called, a bg'd script
would be suspended at exit. Turned out to be overaggressive
code in stty that recorded what 'damage' the user might have
caused when calling stty in the first place.
Jens Petersen provided patch to make setpgrp configure better on
some Linux systems.
Added example/getpassck script to test for getpass bug.
multixterm had debugging stuff leftover ("hello").
--------------------
2006-02-27 Andreas Kupries <andreask@activestate.com>
* exp_main_sub.c: Added command 'exp_configure' for magic configuration.
* exp_command.c: Accepts option -strictwrite. Default is 0, ignoring
* exp_chan.c: write errors (compatible to 5.41). Setting to 1 re-
* expect_tcl.h: activates 5.42 behaviour.
2006-01-25 Jeff Hobbs <jeffh@ActiveState.com>
* tclconfig/tcl.m4, configure: Fix LD_SEARCH_FLAGS setting in tcl.m4
* example/unbuffer: whitespace police
* example/beer.exp: brace exprs
* expect.man: use clock instead of exec date, minor nroff fixes.
2006-01-24 Andreas Kupries <andreask@activestate.com>
* tclconfig/tcl.m4: Updated to TEA 3.5
* configure.in: Ditto.
* configure: Regenerated.
2006-01-10 Jeff Hobbs <jeffh@ActiveState.com>
* tests/expect.test: ensure iso8859-1 for /tmp/null (steffen).
2005-09-19 Andreas Kupries <andreask@activestate.com>
* exp_chan.c (ExpOutputProc): Added guard to intercept and ignore
empty write operations, i.e. operations trying to write zero
bytes.
2005-09-09 Andreas Kupries <andreask@activestate.com>
* exp_chan.c (ExpBlockModeProc): No, stdin is not ok (See last
entry). Fixed.
2005-07-07 Andreas Kupries <andreask@activestate.com>
* exp_chan.c (ExpBlockModeProc): [Expect SF Bug 1108551]. Excluded
manipulation of the blocking status for stdin/stderr. This is
handled by the Tcl core itself and we must absolutely not pull
the rug out from under it. The standard setting to non-blocking
will mess with the core which had them set to blocking, and
makes all its decisions based on that assumption. Setting to
non-blocking can cause hangs and crashes. Stdin is ok however,
apparently.
This problem was introduced at '2004-06-14'.
2005-06-22 Andreas Kupries <andreask@activestate.com>
* exp_chan.c: Fixed bug causing crash of expect on exit when a
* exp_command.c: Tcl channel is used with -(leave)open more than
* exp_command.h: once. It tried to close such channels multiple
times, going through already freed memory. Added data structures
to track and refcount all such channels, to close them only when
the last user goes away.
2005-06-09 Andreas Kupries <andreask@activestate.com>
* Makefile.in: Upgraded build system to TEA 3.2.
* configure.in:
* config.guess:
* config.sub
* exp_main_sub.c:
* aclocal.m4:
2005-03-29 Andreas Kupries <andreask@activestate.com>
* exp_chan.c: Fixed problem with debugger introduced at '2004-06-14'.
* tcldbg.h: For a nonblocking stdin the debugger blowed up on the
* Dbg.c: empty reads it could get, exiting the application. I
guess that this was an implicit 'panic'. Fix:
- Split ExpBlockmodeProc into high- and lowlevel
code, the latter for use by the debugger. The
high-level code tells the debugger which state
stdin is in (doing this explicitly because if
FIONBIO is used it is not possible to query the fd
directly, and I saw no reason to #ifdef for fcntl
which can).
- Debugger now exports a function for use by the
blockmode proc, and in each interaction it checks
for nonblocking input, forces blocking if
necessary. At the end of each interaction the true
mode is restored. Both operations use the
low-level blockmode code.
2005-03-07 Jeff Hobbs <jeffh@ActiveState.com>
* exp_tty.c (Exp_SttyCmd): fix from Libes that controlling
terminal check example (book p372) by restricting set of
exp_ioctled_devtty variable.
2005-02-15 Andreas Kupries <andreask@activestate.com>
* Merged changes from the official versions 5.42.1 and 5.43.0 of
expect into the SF sources. See the details below.
--------------------
Martin Forssen <maf@tkrat.xorg> fixed bug in ExpOutputProc
that caused misbehavior during partial writes.
Someone noted that gets stdin behaves differently (returns -1
immediately) from tclsh because with 5.42, stdin is unblocked by
defaults.
Robroy Gregg <robroy@armory.xcom> noted that expect_background
ignores timeouts. Added to documentation.
Jens Peterson <peterson@redhat.xcom> provided patch for
"mkpasswd -vo".
Gary Bliesener <gary.bliesener@nextel.xcom> noted that
multixterm failed on his system which had an old Tk that didn't
support the Tk package.
Removed beta designation.
Daniel A. Steffen <steffen@ics.mq.edu.xau> provided patch for
MacOS to avoid panic-redefinition.
--------------------
2005-01-21 Andreas Kupries <andreask@activestate.com>
* exp_inter.c: Changed all uses of 'time(3)' to Tcl_GetTime. IOW
* expect.c: go through the abstract core API instead of
directly acessing OS time. This makes the code
dependent on Tcl 8.4, as Tcl_GetTime was not public
before. See TIP #73 for its introduction into the
public API. As for the reason behind _this_ change
see TIP #233. Our change here now causes Expect to
be automatically in sync with any virtualization
set up in the core.
2004-08-19 Andreas Kupries <andreask@activestate.com>
* exp_chan.c (ExpOutputProc): Added code posted by Don on
c.l.t. to handle interrupted system calls (EAGAIN).
2004-07-15 Andreas Kupries <andreask@activestate.com>
* Merged changes from the official version 5.42b0 of expect into
the SF sources. See details below.
--------------------
Alexander Doktorovich <alexander.doktorovich@ericsson.xcom>
wanted to use Expect as a filter. This is possible but 'too
hard'. To make it easier, added close_on_eof command to control
whether expect/interact automatically close the channel on eof.
This should simplify/enable other scripts.
Kurt Heberlein <kurth@3pardata.xcom> noted that Expect would
hang. Andreas tracked it down to a change in Tcl such that when
Tcl had data left in its buffers, it would check for more data
rather than returning what it had to Expect first. If no data
was forthcoming then Tcl would hang because the pty driver runs
in blocked mode. Recoded to use nonblocking mode.
Yi Luo <yluo@brocade.xcom> noted that multixterm xterms were
reporting the parent's X window ids (via the WINDOWID env
variable) instead of the new ones.
Dick Van Deun <dirk@dinf.vub.ac.xbe> noted that kibitz expects
to find write in /bin but it is in /usr/bin on Slackware. Seems
safe to drop the prefix.
Steve Lee <steve@tuxsoft.xcom> noted that building Expect failed
on Linux when built from scratch because stty ends up in
/usr/local/bin rather than the assumed /bin. Added code to
support this.
--------------------
2004-06-14 Andreas Kupries <andreask@activestate.com>
* exp_chan.c: Integrated the block mode proc I got by mail from
Don Libes into the channel driver. This fixes an error with
expect hanging on some input if the situation is just
right^Hwrong. Basically if the buffers in driver, Tcl IO core
and Expect itself are aligned just so it can cause Expect to
block in all call to the OS for more data even if all the data
it needs is in ts buffers. Because the driver is blocking and
the Tcl core was told that it can run in non-blocking mode. with
the block mode proc in place the driver knows that it should be
non-blocking and is able to tell this to the OS as well. The
call to the OS still happens, but is not blocking anymore, and
so the problem is gone.
A number of incompat changes in the Tcl IO core to work around
this problem in Expect will be removed now.
2004-06-03 Andreas Kupries <andreask@activestate.com>
* aclocal.m4 (TCLHDIRDASHI): Extended with path to unix headers as
well.
* expect_tcl.h: Added inclusion of <stdio.h>.
Both changes required for Expect to compile against Tcl 8.5
after the header reform.
* configure: Regenerated.
2004-05-19 Andreas Kupries <andreask@activestate.com>
* Merged changes from the official version 5.41 of expect into the
SF sources. See details below.
--------------------
Simon Taylor <simon@unisolve.com.xau> provided fix for interact
-o which was completely broken by 5.40.1.
Added scroll support to official tkterm. Copied all fixes
from/to term_expect to/from tkterm.
Kiran Madabhushi <maskiran@hotmail.xcom> encountered interact
diagnostics incorrectly pointing to expect_background. Also,
found multiple -o flags behaving unexpectedly. Added diag.
Kristoffer Eriksson <ske@pkmab.xse> noted typo in SIMPLE code in
exp_inter.c. However, this is extremely unlikely to affect any
machines.
Reinhard Max <max@suse.xcom> noted that "make test" failed when
run in the background. The log testcase was testing the
send_tty command. Added code in both Expect and in the test to
handle this.
--------------------
2004-02-25 Andreas Kupries <andreask@activestate.com>
* Merged changes from the official version 5.40 of expect into the
SF sources. See details below. Partially already done (Rich
Kennedy's patch).
--------------------
Eric Raymond <esr@snark.thyrsus.xcom> provided troff-related
fixes for the expect, lib, and dislocate man pages.
Rich Kennedy <rickenne@cisco.xcom> noted a bug having to do
with our caching of whether we have registered a filehandler.
This broke when Tcl was setting a handler on the same file.
Ken Pizzini <ken.pizzini@explicate.xorg> provided patch for
leak in spawn error handling.
Pete Lancashire <plancashire@columbia.xcom> noted autopasswd
example broke on Solaris which capitalized prompts.
--------------------
2003-10-20 Andreas Kupries <andreask@activestate.com>
* exp_event.c (exp_get_next_event): Applied a patch made by Don
Libes in response to a bug report posted to news:comp.lang.tcl
by Rich Kennedy <rickenne@cisco.com>. Patch was posted to c.l.t
too.
> Subject: Re: 2nd interact does not grab stdin
> Date: 17 Oct 2003 15:33:38 -0400
> From: Don Libes <libes@nist.gov>
> Organization: National Institute of Standards and Technology
> Newsgroups: comp.lang.tcl
> References: <3F86D6F8.E535CBDE@cisco.com>
> It's a bug - some overaggressive caching regarding stdin. A fix
> appears below. (You should also be able to avoid the problem by
> replacing the gets with expect_user.)
> Don
> Rich Kennedy <rickenne@cisco.com> writes:
> > Hi,
> >
> > The following little expect script gets in trouble with second
> > 'interact' statement. The 'less' program does not get stdin
> > so it doesn't respond to typed characters...
> >
> > Is there something that must be done after the 1st interact to
> > allow the second to work correctly? (Note, you have to run
> > as a script because it works correctly if you type the commands
> > interactively to expect)
> >
> > #!/usr/local/bin/expect
> >
> > gets stdin a
> > spawn less file1
> > interact
> > wait
> > gets stdin junk
> > spawn less file2
> > interact
> > wait
> >
> > Thanks
> >
> > Rich Kennedy
2003-09-05 Andreas Kupries <andreask@activestate.com>
* Merged changes from the official version 5.39 of expect into the
SF sources. See details below. Partially already done.
--------------------
Poorva Gupta <poorva@cup.hp.xcom> noted that grantpt/unlockpt
order was backward. Strange that this was never a prob before!
Eric Raymond <esr@snark.thyrsus.xcom> provided a troff-related
fix for the multixterm man page.
Nicolas Roeser <n-roeser@gmx.xnet> noted confusion with md5 so I
made the Expect page more explicit about which file that hash
was based on.
Josh Purinton noted that earlier fix wasn't quite right. Exit
on INT/TERM should cause Expect to exit with signal embedded in
status. He also requested I obfuscate email addresses in this
file.
Guido Ostkamp <Guido.Ostkamp@t-online.xde> and Igor Sobrado
<sobrado@string1.ciencias.uniovi.xes> noted that fixline1
rewrote scripts to be expect scripts even if they were expectk
scripts.
Dirk Petera <dirkpetera@yahoo.xcom> noted that any_spawn_id used
to work but did no longer. Looks like a bug left over from the
the I18L conversion. Fixed.
Steve Szabo noted exp_log_file -open channel failed. Fixed.
Fixed bug from 5.31 that prevent stty from returning messages
from underlying program.
Thomas Dickey <dickey@herndon4.his.xcom> noted that ncurses
ignores 2-char term names because of, well, poor assumptions and
coding. Changed tkterm to use longer names.
Heath Moore <hmoore@systran.xcom> noted that exp_clib could lock
up if remtime happened to be precisely 0. Recoded to avoid.
At request of Per Otterholm <otterholm@telia.xcom>, wrote script
to read from stdin and echo passwords (exercise 9 in Tk chapter
of Expect book). Added to example directory as passwdprompt.
Josh Purinton <josh@purinton.xorg> pointed out that by default,
SIGINT/TERM should cause expect's return status to be 1, not 0.
Paul Reithmuller <paul.reithmuller@eng.sun.xcom> noted that
unbuffer shouldn't postprocess its output. Added stty_init.
Mordechai T. Abzug <morty@sanctuary.arbutus.md.xus> noted that
log_file wasn't recording -append status.
James Kelly <macubergeek@comcast.xnet> noted weather example
needed new source.
Dimitar Haralanov <mitko@tahoenetworks.xcom> noted that interact
dumped core with interact { timeout 1 }
--------------------
2003-06-16 Andreas Kupries <andreask@activestate.com>
* exp_command.c: Applied patch provided on c.l.t., by Don Libes
<libes@nist.gov> in response to a bug report by Dirk Petera
<dirkpetera@yahoo.com> in same place. See thread reference
below:
http://groups.google.ca/groups?threadm=4791f5a6.0305250619.1a660299%40posting.google.com
2003-05-08 Andreas Kupries <andreask@activestate.com>
* exp_clib.c (expectv): Applied patch provided on c.l.t., by Don
Libes <libes@nist.gov> in response to a bug report by "Heath
Moore" <hmoore@systran.com> in same place:
> Regarding expect 5.38...
>
> I'm using libexpect on RedHat 8.0 to communicate via telnet,
> and am having problems with it locking up instead of timing
> out. Causing traffic during the lockup breaks the lockup. I
> looked at the sources, and think I may have found the reason.
>
> It appears as though i_read can be called with remtime== 0,
> which means do > one read() and return without using alarm(),
> even when exp_timeout is non-zero. This would happen if
> i_read were to return after receiving non-matching data when
> end_time == current_time. The subsequent i_read would then
> wait until it received data.
2003-02-17 Andreas Kupries <andreask@activestate.com>
* Makefile.in:
* configure.in: Removed the check of configure against
configure.in and Makefile.in. It is a hassle to deal with when
trying to build straight from CVS, making unsupervised automatic
builds difficult
2003-02-14 Andreas Kupries <andreask@activestate.com>
* configure.in: Made expect aware of tcl stubs. Derived from the
* exp_command.h: patches to expect done by Steve Landers
* exp_command.c: <stevel@digital-smarties.com>. Modifications:
* exp_main_sub.c No global cmdinfo structures for 'close' and
* exp_main_exp.c: 'return'. Made this per-interpreter information
* exp_main_tk.c: as it should be. Using interp assoc data for this.
NOTE: stubs is not default, but has to be activated via '--enable-stubs'.
* configure: Regenerated.
2003-02-14 Andreas Kupries <andreask@activestate.com>
* exp_chan.c (exp_close_all): Save the nextPtr in a local variable
before calling 'exp_close' as 'expStateFree' can be called from
it under some circumstances, possibly causing the memory
allocator to smash the value in 'esPtr'.
2003-02-03 Andreas Kupries <andreask@activestate.com>
* exp_log.c (expLogChannelOpen): Fixed the bug reported on
comp.lang.tcl by Mordechai T. Abzug
<morty@sanctuary.arbutus.md.us>. The bugfix itself was provided
by Don Libes.
2002-10-09 Andreas Kupries <andreask@activestate.com>
* exp_command.c (Exp_SpawnCmd): Tcl_GetChannelHandle expected a
ClientData*, but got an int*. sizeof(int) != sizeof(ClientData)
on 64bit platforms. Crashed the command on a PA-RISC 2.0 machine
with --enable-64bit set. Fix: Use temp. variables of type
ClientData to retrieve the fd's, and copy this into the actual
variables, with a cast to int.
2002-09-25 Jeff Hobbs <jeffh@ActiveState.com>
* configure: regen'ed
* configure.in: use tcl_libdir in EXP_LIB_SPEC instead of
${INSTALL_ROOT}${exec_prefix}/lib (steffen)
* exp_main_tk.c (Tk_Init2): don't call XSynchronize on OS X.
2002-08-08 Andreas Kupries <andreas_kupries@users.sourceforge.net>
* Merged changes from the official version 5.38 of expect into the
SF sources. See details below.
* Makefile.in: Added generation of MD5 checksum for distributed
archive.
* rftp: Bugfix by Curt Schroeder, see HISTORY
* HISTORY: Updated with new info.
* configure: Updated version info.
* configure.in: Updated version info.
2002-06-26 David Gravereaux <davygrvy@pobox.com>
* example/weather: Updated script to use rainmaker.wunderground.com
instead of cirrus.sprl.umich.edu. The old service is closed.
Added Larry Virden's notes about how rainmaker needs reverse DNS
from the peer making the connection or no data can retrieved.
This appears to be a blind error condition.
2002-06-17 Andreas Kupries <andreas_kupries@users.sourceforge.net>
* exp_main_tk.c: #ifdef'd definition of "matherr". This hack is
not required for 8.4 anymore. But still for 8.3.
2002-03-25 Andreas Kupries <andreas_kupries@users.sourceforge.net>
* exp_main_sub.c: Fixed typo in merge of #459646. Thanks to Hemang Lavana.
* exp_log.c (expStdoutLogU): Merged fix for SF Bug #513382 into
the HEAD (The source of the patch is "expect-sf418892-sf439042-branch").
Use Tcl_WriteChars/Tcl_Flush instead of 'fwrite' for tcl 8.1 and
beyond to enforce correct conversion of the internal UTF/8 into
the external representation.
* Merged fix for SF Bug #459646 into the HEAD (The source of the
patch is "expect-sf418892-sf439042-branch").
* Merged fix for SF Bug #439042 into the HEAD (The source of the
patch is "expect-sf418892-sf439042-branch").
* Merged fix for SF Bug #418892 into the HEAD (The source of the
patch is "expect-sf418892-sf439042-branch").
2002-03-23 Don Libes <libes@users.sourceforge.net>
* Andreas Kupries mods to provide CONST support per TIP 27 (Fixed
SF Patch #525074).
2002-02-25 Andreas Kupries <andreas_kupries@users.sourceforge.net>
* expect.c: Applied patch by Don Libes fixing improper
internationalization.
2002-02-21 Andreas Kupries <andreas_kupries@users.sourceforge.net>
* expect.man: Changed the paragraph about [exp_continue] to
contain information about the flag "-continue_timer". This fixes
the bug SF #520630.
2002-02-08 Andreas Kupries <andreas_kupries@users.sourceforge.net>
* expect.man: Changed abbreviation of "-notransfer" from "-n" to
"-not". "-n" is no longer unique due to the addition of
"-nocase". This fixes the bug SF #514994.
2002-02-07 Andreas Kupries <andreas_kupries@users.sourceforge.net>
* Applied patch for SF #514590 to correct behaviour of expect when
expecting and send from and to bogus spawn id's.
2002-01-16 Andreas Kupries <andreas_kupries@users.sourceforge.net>
* Resynchronization of SourceForge with Don's sources to Expect
version 5.34. The changes are
Don Porter <don.porter@nist.gov> provided package-related fixes
for test suite.
Brian Theado <brian.theado@usa.net> noted that interact's -re
support broke when offsets kicked in. Turned out that the
regexp engine supports them during execution but the results are
delivered RELATIVE to the offset. (I suspect this was done
due to expediency.)
2001-12-05 Andreas Kupries <andreas_kupries@users.sourceforge.net>
* exp_inter.c: Applied patch posted by Don libes to c.l.t. on his
behalf to keep the SF repository in sync with his changes. Don's
notes: I obviously missed the fact that although
"Tcl_RegExpExecObj" supports offsets, they aren't delivered to
"Tcl_RegExpGetInfo".
2001-09-12 David Gravereaux <davygrvy@pobox.com>
* 'telco-tec-win32-branch' branch created.
2001-08-01 Jeff Hobbs <jeffh@ActiveState.com>
* Dbg.c (Dbg_On): fixed handling of stepping. [Bug: #446412]
2000-04-26 Rob Savoye <rob@welcomehome.org>
* pty_termios.h: Only include stropts.h if it exists, rather than
deciding it exists based on HAVE_PTMX.
* configure.in: Make sure libpt exists, rather than blindly using
it for all our configure tests, which then all fail. Also assume
our svr4 style ptys are broken, if /dev/ptmx exists, but stropts.h
doesn't exist.
1999-08-31 Jennifer Hom <jenn@scriptics.com>
* Makefile.in: Changed test target to source tests/all.tcl instead
of tests/all
* tests/README: Modified documentation to reflect the change from
usage of a defs file to the use of package tcltest to run the tests
* tests/all:
* tests/defs:
* tests/all.tcl:
* tests/cat.test:
* tests/expect.test:
* tests/logfile.test:
* tests/pid.test:
* tests/send.test:
* tests/spawn.test
* tests/stty.test: Modified test files to use package tcltest,
removed tests/all and tests/defs, and added tests/all.tcl
1999-06-22 <stanton@scriptics.com>
* expect.c: Fixed bug in token parsing where index was not being
incremented properly.
* configure.in: Changed version number to 5.31.
* aclocal.m4: Fixed CY_AC_LOAD_TKCONFIG so it tests for Tk_Init
instead of Tk_Main (which is only a macro in 8.1 and later). Also
added TCL_BUILD_LIB_SPEC to the set of flags used in this test to
avoid linker errors.
* Dbgconfig.in: move CY_*_TCLCONFIG tests below AC_PROG_CC so it
will work with gcc
Thu Mar 20 14:27:45 1997 Geoffrey Noer <noer@cygnus.com>
* configure.in: don't check if stty reads stdout for
i[[3456]]86-*-sysv4.2MP during config; hard code instead
Tue Nov 19 09:22:08 1996 Tom Tromey <tromey@cygnus.com>
* Makefile.in (install_shared_lib): Put else clause onto each if.
Fri Nov 15 11:23:43 1996 Tom Tromey <tromey@cygnus.com>
* Makefile.in (XCFLAGS): Use EXP_SHLIB_CFLAGS, not
TCL_SHLIB_CFLAGS.
(TCL_SHLIB_CFLAGS): Define.
* configure.in: Allow arguments to --enable-blah to work.
Compute and AC_SUBST EXP_SHLIB_CFLAGS.
Added missing AC_MSG_CHECKING.
Wed Oct 2 10:13:37 1996 Tom Tromey <tromey@cygnus.com>
* configure: Regenerated.
* configure.in (stty_reads_stdout): /bin/stty on DG/UX fails.
Fri Sep 27 10:15:48 1996 Tom Tromey <tromey@creche.cygnus.com>
* expect.c (exp_i_read): Pass interp as first arg to exp_error.
* configure.in (stty_reads_stdout): /bin/stty on OSF2.0, OSF3.2,
HPUX 9.X, HPUX 10.X guesses wrong, so set value explicitly.
Mon Sep 9 10:29:32 1996 Tom Tromey <tromey@creche.cygnus.com>
* configure: Regenerated.
* configure.in: Added code to actually handle --with-x.
* configure: Regenerated.
* configure.in: Don't bother looking for Tk if --with-x=no
specified.
Thu Sep 5 11:01:09 1996 Tom Tromey <tromey@creche.cygnus.com>
* configure: Regenerated.
* configure.in (stty_reads_stdout): AIX fails "stty" test in
background, so set explicitly. Ditto HPUX 9 and 10.
Thu Aug 29 17:04:55 1996 Michael Meissner <meissner@tiktok.cygnus.com>
* configure.in (i[345]86-*-*): Recognize i686 for pentium pro.
* configure: Regenerate.
Mon Aug 5 12:55:06 1996 Tom Tromey <tromey@creche.cygnus.com>
* Makefile.in (XCFLAGS): New macro.
(CFLAGS): Define to just @CFLAGS@.
(CFLAGS_INT): Use $(XCFLAGS).
(expect, expect_installed, expect.tc, expectk, expectk_installed,
expectk.tc): Use $(XCFLAGS).
Mon Feb 12 23:11:38 1996 Rob Savoye <rob@chinadoll>
* aclocal.m4: Fix typo of ac_cv_tkh to be ac_cv_tclh so it works
on all systems.
* configure, DBGconfigure, testsuite/configure: Regenerated with
autoconf 2.7.
Tue Feb 6 11:48:05 1996 Tom Tromey <tromey@creche.cygnus.com>
* exp_clib.c, exp_printify.c, expect_comm.h: For Tcl 7.5 and
greater, use ../compat/stdlib.h, not compat/stdlib.h.
Tue Jan 30 12:21:37 1996 Fred Fish <fnf@kalessin.cygnus.com>
* exp_regexp.c (regmatch, regrepeat): Only declare strchr if it is not
a macro.
Mon Jan 22 11:17:06 1996 Tom Tromey <tromey@creche.cygnus.com>
* configure.in: Check for -lieee, -ldl, and -ldld.
* Makefile.in (OFILES): Include @LIBOBJS@.
(strerror.o): New target.
* strerror.c: New file.
* configure.in: Test for strerror.
Fri Jan 19 11:08:11 1996 Tom Tromey <tromey@creche.cygnus.com>
* Makefile.in (install, ${SCRIPT_LIST}, test): Find new Tcl libraries.
Thu Jan 18 13:43:13 1996 Tom Tromey <tromey@creche.cygnus.com>
* Most files: Update to expect 5.19.
Fri Jan 12 16:22:12 1996 Tom Tromey <tromey@creche.cygnus.com>
* exp_closetcl.c (exp_close_tcl_files): Skip stdin, stdout,
stderr.
* expect_comm.h: Declare exp_close_files_interp.
* exp_command.c (exp_init_most_cmds): Set exp_close_files_interp.
Thu Jan 11 09:43:14 1996 Tom Tromey <tromey@creche.cygnus.com>
* exp_closetcl.c (exp_close_files_interp): New variable for Tcl
7.5.
(exp_close_tcl_files): Updated for Tcl 7.5.
Prototype and varargs changes:
* expect.c: Don't include <varargs.h>.
* Dbg.c: Copied in many defines from expect_comm.h.
(print): Use new varargs defines.
* exp_clib.c (exp_fexpectl): Use EXP_VARARGS_START.
* expect_comm.h: Include "tclInt.h".
* exp_console.c (exp_console_manipulation_failed): First arg to
errorlog is char*, not FILE*.
* exp_log.c (debuglog): Pass name of last argument to
EXP_VARARGS_START.
* expect_cf.h.in (tcl_AsyncReady): Removed define.
* expect.c (Exp_ExpectGlobalCmd): Added cast.
* exp_command.c (exp_i_update): First arg to exp_debuglog is
* exp_poll.c (exp_get_next_event): Likewise.
char*, not Tcl_Interp*.
* exp_log.h: Use prototypes everywhere. Include "expect_comm.h".
* expect_tcl.h: Use EXP_VARARGS, not VARARGS.
(tcl_AsyncReady): New define for Tcl 7.5.
* aclocal.m4 (CY_AC_PATH_TCLH): Handle Tcl 7.5 and greater.
(CY_AC_PATH_TCLLIB): Handle Tcl 7.5 and greater.
(CY_AC_PATH_TKH): Handle Tk 4.1 and greater.
(CY_AC_PATH_TKLIB): Handle Tk 4.1 and greater. Properly quote
argument to AC_REQUIRE.
* configure: Regenerated.
Tue Jan 9 16:26:47 1996 Rob Savoye <rob@chinadoll.cygnus.com>
* Makefile.in: Change SHORT_BINDIR to $prefix, rather than
exec_prefix. This is only used to store the platform independant
expect scripts.
Dec 18 17:22:05 1995 Brendan Kehoe <brendan@lisa.cygnus.com>
* configure.in, configure: For a solaris2 machine doing a static
build, add `-ldl -lw' to avoid unresolved refs using the
OpenWindows libraries.
Wed Nov 22 08:49:01 1995 Rob Savoye <rob@chinadollchinadoll.cygnus.com>
* Most files: Update to expect 5.18.1.
Fri Nov 17 17:31:55 1995 Rob Savoye <rob@chinadoll.cygnus.com>
* configure.in: Add support for SCO OpenServer. It doesn't like
the trap either.
Thu Nov 16 09:28:53 1995 Rob Savoye <rob@chinadoll.cygnus.com>
* configure.in: Use $host to get the OS type rather than trying to
get the host name ourselves. Use the $host to set the
STTY_READS_STDOUT for hosts were the test is known to fail. It
also now configures in the background.
* configure.in, Dbgconfig.in, testsuite/configure.in: Use
AC_PROG_CC again since Cygnus configure now does the sames thing.
Mon Oct 30 18:16:48 1995 Jason Molenda (crash@phydeaux.cygnus.com)
* configure.in (no_tk): zero out X_PROGS if we can't find tk
libraries.
Tue Oct 24 18:25:09 1995 Jason Molenda (crash@phydeaux.cygnus.com)
* Makefile.in (X11HDIR): Changed to X11_CFLAGS.
(X11_LIB_FLAGS): Changed to X11_LDFLAGS.
(X11_LIB): Changed to X11_LIBS.
(CPPFLAGS_SIMPLE): Use X11_CFLAGS.
(expectk, expectk.tc, tk): use X11_LDFLAGS & X11_LIBS.
* configure.in (X11HDIR, X11_LIB_FLAGS, X11_LIB): Use X11_CFLAGS,
X11_LDFLAGS, X11_LIBS. Link X11 statically on Solaris, SunOS and
HPUX.
Thu Oct 19 20:55:54 1995 Fred Fish <fnf@cygnus.com>
* Makefile.in: Remove extraneous tabs and blanks in otherwise
empty lines. That confuses older non-GNU versions of "make".
Mon Oct 9 20:58:50 1995 Jason Molenda (crash@phydeaux.cygnus.com)
* testsuite/aclocal.m4: New file. Include ../aclocal.m4.
Thu Aug 31 00:16:26 1995 Rob Savoye <rob@darkstar.cygnus.com>
* HISTORY, Makefile.in, aclocal.m4, exp_command.h, exp_inter.c,
exp_main_tk.c, exp_pty.c, expect.c, tests/all,
testsuite/Makefile.in. Update to the 5.18.0 release. Minor
changes.
Thu Aug 17 18:47:21 1995 Rob Savoye <rob@darkstar.cygnus.com>
* Most files: Update to the 5.17.7 release.
Thu Aug 3 22:47:36 1995 Jeff Law (law@snake.cs.utah.edu)
* pty_termios.c (HAVE_PTMX): Undefine if both HAVE_PTYM and
HAVE_PTMX are defined (as happens for hpux10).
Thu Jul 27 16:31:23 1995 J.T. Conklin <jtc@cygnus.com>
* Makefile.in (configure): Removed rule that automatically
rebuilds configure script. Users might not have autoconf.
Tue Jul 18 23:15:03 1995 Fred Fish <fnf@fishbowl>
* expect.c (Exp_ExpectGlobalCmd): Cast ckrealloc first arg to char*.
Sun Jun 18 13:02:41 1995 Fred Fish <fnf@amigalib.com>
* configure, configure.in (XLIBS): When adding -lX11, also preserve
the previous libraries that we went to the trouble of finding.
Sun Jun 18 12:15:44 1995 Fred Fish <fnf@amigalib.com>
* Makefile.in (exp_clib.o): Add dependencies.
Mon May 1 16:50:22 1995 Rob Savoye <rob@darkstar.cygnus.com>
* configure.in: Also set XINCLUDES in the Makefile.
Fri Apr 28 18:56:02 1995 Rob Savoye <rob@darkstar.cygnus.com>
* aclocal.m4: Create a clone of AC_C_CROSS called CY_C_CROSS that
has better error handling in case the native compiler is hosed.
* aclocal.m4: Look for tcl and tk directories as just tcl (and tk)
or tcl[0-9] (and tk[0-9)] so it doesn't match the tclX
release. Print an error and exit if any of the --with-{tcl,tk}*
options are used and point to bogus paths. Based Tcl header search
on tclInt./h rather than tcl.h.
* Makefile.in: Add dependancies for back in for configure and
Dbgconfigure targets.
Mon Apr 24 16:46:01 1995 Rob Savoye <rob@darkstar.cygnus.com>
* exp_command.c, exp_event.h, exp_inter.c, exp_main_tk.c,
exp_poll.c, exp_select.c, exp_simple.c, exp_tk.c, exp_trap.c,
exp_tty.c, FAQ, README, HISTORY: Update to expect 5.16.3.
Fri Apr 14 12:00:39 1995 Rob Savoye <rob@darkstar.cygnus.com>
* configure.in: Copy Dbg_cf.h to objdir, not srcdir.
Tue Apr 11 18:52:24 1995 Rob Savoye <rob@darkstar.cygnus.com>
* aclocal.m4: Split the macros so header and library searches are
seperate macros. AC_PATH_{TCL,TK} nows only calls the macros. Lots
of optimization to the AC_PATH_T* macros. Supports the use of
--with-tclinclude, --with-tcllib, --with-tkinclude, --with-tklib
to specify alternative search dirs for tcl and tk stuff.
* Makefile.in, testsuite/Makefile.in: Re-write targets for
configure, Dbgconfigure so they work in an unconfigured srcdir.
* configure.in: Put AC_PATH_X before AC_PATH_TK and make the TK
test conditional. Fix how Dbgconfigure gets passed the Tcl header
dir to use --with-PACKAGE which is much simpler. Removed the test
for user override of X search paths since AC_PATH_X uses.
--x-includes and --x-libraries instead.
* Dbgconfig.in: Use AC_PATH_TCLH to find just the headers, and
test for LynxOS.
* debugger/: Remove directory. Recursive configuring is so much
easier...
* DbgMkfl.in, Dbg_cf.h.in, Dbg.c, Dbg.h, Dbgconfigure,
Dbgconfig.in: Sigh, moved back to the top-level expect directory.
Wed Apr 5 17:25:45 1995 Rob Savoye <rob@darkstar.cygnus.com>
* configure.in: Add a trap so the configure runs in the
background.
Thu Mar 16 16:56:08 1995 Rob Savoye <rob@darkstar.cygnus.com>
* debugger: New directory for the Tcl debugger.
* debugger/Dbg.c, debugger/Dbg.h, debugger/Dbg_cf.h.in: Moved from
the top level expect directory so it builds standalone.
* DbgMkfl.in, debugger/Makefile.in: Moved to debugger dir and
renamed.
* install-sh, mkinstalldirs: New files borrowed from the autoconf
distribution.
* aclocal.m4: New autoconf macros.
* Makefile.in: Tweaked so it's recursive.
* configure.in: Use new macros in aclocal.m4 rather than hunting
for the Tcl and Tk stuff ourseleves.
* debugger/Makefile.in: Build debugger standalone.
* testsuite/Makefile.in, testsuite/configure.in: New files for
autoconf support.
* exp_test.c, testsuite/exp_test.c: Move test file.
Fri Jan 13 15:30:30 1995 Ian Lance Taylor <ian@sanguine.cygnus.com>
* Makefile.in (check): Pass EXPECT correctly to runtest.
Thu Oct 20 18:04:06 1994 Rob Savoye <rob@darkstar.cygnus.com>
* Makefile.in: Add X11_INCLUDE_FLAGS so top level flags get used
too.
Tue Jun 14 12:32:07 1994 David J. Mackenzie (djm@rtl.cygnus.com)
* aclocal.m4: Copy from TCL directory.
* configure.in: Improve checks for installed Tcl and Tk.
* configure: Rebuilt.
Tue Jun 7 13:52:34 1994 Ian Lance Taylor (ian@tweedledumb.cygnus.com)
* Makefile.in (mostlyclean, realclean): New targets.
Wed May 18 12:21:06 1994 Ian Lance Taylor (ian@tweedledumb.cygnus.com)
* Makefile.in (install): Add another ``else true''.
Fri Apr 29 16:49:36 1994 Ian Lance Taylor (ian@tweedledumb.cygnus.com)
* Makefile.in (install): Always use else in if conditional to
avoid Ultrix sh bug.
Mon Apr 11 15:22:12 1994 Rob Savoye (rob@cirdan.cygnus.com)
* Upgrade to the new "official" beta release of expect 5.7.
Wed Mar 30 17:15:28 1994 Rob Savoye (rob@cirdan.cygnus.com)
* testsuite/expect.tests/expect-test.exp: Just run the new expect
tests and format the outout under DejaGnu.
Mon Mar 28 14:33:55 1994 Rob Savoye (rob@cirdan.cygnus.com)
* Upgrade to expect 5.6.3.
Thu Dec 2 16:26:54 1993 Rob Savoye (rob@darkstar.cygnus.com)
* configure.in: Add tests to find Tcl and Tk headers and
libraries.
Thu Aug 19 18:26:49 1993 Rob Savoye (rob@darkstar.cygnus.com)
* upgraded to version 4.7.6, add OSF/1 patches in again.
Wed Aug 18 20:10:16 1993 Rob Savoye (rob@rtl.cygnus.com)
* upgraded to version 4.7.4, add OSF/1 patches in again.
Tue Aug 17 20:17:40 1993 Rob Savoye (rob@darkstar.cygnus.com)
* pty_termios.c, exp_command.c, configure.in: Add support for
using ptmx_bsd's if they exist. Only found on OSF/1. (patch
applied from Gregory Depp <depp@osf.org>
Thu Jun 10 11:36:09 1993 david d `zoo' zuhn (zoo at cirdan.cygnus.com)
* exp_main.h: fix prototype for exp_cook
Fri Jun 4 08:55:22 1993 Ian Lance Taylor (ian@cygnus.com)
* Makefile.in (TCLLIB): If ../tcl/libtcl.a does not exist, use
-ltcl.
Tue May 25 14:45:12 1993 Rob Savoye (rob@darkstar.cygnus.com)
* Makefile.in, configure.in: Add some support for autoconfiguring
for X.
Sun May 23 22:32:09 1993 Rob Savoye (rob at darkstar.cygnus.com)
* exp_command.c: Fix so send_log still works when master is out of
bounds. (ok since it doesn't get used).
Mon May 17 19:51:52 1993 Rob Savoye (rob@cygnus.com)
* configure.in: Change test for ranlib so it kicks out "true"
rather than "@:" if it can't be found.
Thu Apr 15 14:11:50 1993 Rob Savoye (rob@cygnus.com)
* configure.in, Makefile.in: If using ptmx's (SVR4 style pty's)
then check for libpt.a too.
Thu Apr 8 17:13:39 1993 david d `zoo' zuhn (zoo at cirdan.cygnus.com)
* Makefile.in: all doesn't depend on $(SCRIPTS). When building
$(SCRIPTS) using fixline & sources in $(srcdir), not the current
directory. When installing manpages, install from $(srcdir).
Don't install like "install foo $(bindir)" but rather "install foo
$(bindir)/foo".
Mon Mar 22 23:56:29 1993 david d `zoo' zuhn (zoo at cirdan.cygnus.com)
* Makefile.in: add check & installcheck targets
Tue Mar 2 20:28:30 1993 david d `zoo' zuhn (zoo at cirdan.cygnus.com)
* configure.in, configure: declare SETUID to be @: instead of echo
* pty_termios.c: declare ptynum
* Makefile.in: a number of changes, including use of the AR and
ARFLAGS variables, the appropriate variables for X11 (as passed
down from the top level Makefile), clean up some doc lines
Mon Mar 1 15:05:40 1993 Rob Savoye (rob at darkstar.cygnus.com)
* configure.in, defs.h.in: Fixed problem for systems that think
getpty() should be _getpty().
Thu Feb 25 15:34:34 1993 Rob Savoye (rob at darkstar.cygnus.com)
* exp_tty.h: Defines portable tty macros.
* pty_termios.c: New file, slightly based on pty_usg.c. Uses
portable macros and also supports termio.
* pty_sgttyb.c: Was pty_bsd.c.
* configure.in, Makefile.in, configure: autoconf support for
expect.
Sun Feb 21 17:42:28 1993 Rob Savoye (rob at darkstar.cygnus.com)
* exp_tty.h: Removed and renamed the macros to use configure's.
Wed Feb 17 18:56:36 1993 Rob Savoye (rob at darkstar.cygnus.com)
* expect.c, Makefile.in: Changed SIG_FN_RETURN to RETSIGTYPE
since that's what autoconf kicks out.
Thu Dec 24 15:07:32 1992 david d `zoo' zuhn (zoo at cirdan.cygnus.com)
* Makefile.in: added dummy dvi target
Wed Dec 16 11:26:16 1992 Ian Lance Taylor (ian@cygnus.com)
* inter_select.c (init_interact): if SCO is defined, use sysconf
to get maxfds, rather than getdtablesize.
* configure.in (*-*-sco*): Use mh-sco.
* mh-sco: New file; like mh-sysv, but pass -DSCO in HDEFS.
Tue Nov 17 14:28:20 1992 david d `zoo' zuhn (zoo at cirdan.cygnus.com)
* config/mh-{hpux,aix,irix4,sysv*}: updated with appropriate
values for the host machine (HDEFS, RANLIB, etc)
* configure.in: use that
* Makefile.in: use $(HDEFS) in compiling C files
Sun Nov 15 21:46:16 1992 Fred Fish (fnf@cygnus.com)
* Update to base 3.24.0 release, merging back in changes made
by cygnus to 3.22.12 release.
Sat Nov 14 20:16:26 1992 Fred Fish (fnf@cygnus.com)
* Makefile.in (CFLAGS): Rework use of CFLAGS to fit in better with
cygnus configuration standard.
* config/mh-svr4: Removed.
* config/mh-sysv4: New file, renamed from mh-svr4.
* configure.in (*-sysv4): New configuration.
* configure.in (*-sun-solaris2, *-sysv4): Use mh-sysv4.
* expect.c (sigwinch_handler): Fix #if without any condition.
* command.c, expect.c, global.h, lib_exp.c, main.c, term.h:
Test for SYSV4 as well as SYSV3.
* inter_select.c (sys/sysconfig.h): Include when SYSV4 defined.
* inter_select.c (init_interact): Add sysconf call for SYSV4.
* pty_svr4.c (ptsname): Declare for SYSV4.
Thu Oct 22 17:35:07 1992 Rob Savoye (rob@cygnus.com)
* command.c: Added a "send_log" command. It only writes to a log
file if one was opened by the "log_file" command.
* main.c: Added setbuf commands for stdin, stdout, stderr to turn
off buffering.
Generated by dwww version 1.15 on Wed May 22 21:30:12 CEST 2024.