CVS log for src/lib/libedit/history.c

Up to [cvs.NetBSD.org] / src / lib / libedit

Request diff between arbitrary revisions

Keyword substitution: kv
Default branch: MAIN

Don't fchmod(fileno(fp), ...) in history_save_fp().

There are two reasons for this, first, the permissions of the history
file should be able to be set by the user, not forced to 0600 every
time the history file is overwritten (or appended to).

And more importantly, the fp used for fileno(fp) might have come
from fmemopen() or funopen() (etc) - none of which put a file
descriptor in the "fd" field (ie: fileno(fp) == -1).

To compensate for that, when a history file is opened (in history_save())
set the default permissions then - if the file is actually created.
As fopen() cannot do that (it simply uses 0666&~umask) create the
(normal type) of fp using (approximately) fdopen(open(...), ...)
where the open supplies the 0600 default permissions that are
desired here (which might still be restricted even more by the
umask).   Callers using history(...,H_SAVE_FP,...) or
history(...,H_NSAVE_FP,...) now need to look after any permission
setting required themselves (but as the doc says absolutely
nothing about that, one way or the other, what happens in this
area has always been unspecified, and still is)

One "feature" of the fchmod() method is lost here - apart from forcing
the 0600 permissions (which isn't really desirable) that fchmod()
would also have failed if the current (effective) uid is not the owner of
the history file (or root).    If that is required, a test for it could
be added later - the effect would be as it has always been, the file
named must have been writable (or its directory writable if the file
did not exist) the open would occur (potentially truncating the file)
after which the fchmod() would be attempted, possibly failing, and if
so, never writing anything.   Any new uid test would work the same way.

OK christos@

Mostly merge changes from HEAD upto 20200411

Change strncpy to either memcpy (when we know the len), or strlcpy (when
we used to NUL terminate explicitly.

Sync with HEAD

Ssync with HEAD

Fix editing mistake, remove ) from func call that is now gone.

more efficient to use decode_result :-) Pointed out by kre@

PR/53597: Yasuhiro Horimoto: Avoid segmentation fault in bad history file.

The order in which the arguments to a function are evaluated is
undefined, so don't use va_arg() twice.

PR lib/52849

PR/51517: Jay West: Tty settings not restored on exit
PR/51518: Jay West: prompt is interleaved with client output

Both these issues are caused by rl_restore_handler not DTRT; fix
it so that it kills the internal libedit state completely. This is
inefficient, but it works.

Also fix:
1. add append_history()/H_NSAVE_FP
2. call the rl_startup_hook before printing the first prompt as documented.
   callint it from rl_initialize breaks python, because the callback ends
   up being invoked before the readline module is installed, and we end up
   dereferencing a NULL pointer.
3. add el_resize_terminal.

With those changes, s/lreadline/ledit/g in python works.

Get rid of private/public; keep protected (Ingo Schwarze)

chartype cleanups from Ingo Schwarze:

 - The file tokenizer.c no longer uses chartype.h,
   so don't include the header.

 - The dummy definitions of ct_{de,en}code_string() for the
   NARROWCHAR case are only used in history.c, so move them there.

 - Now the whole content of chartype.h is for the wide character
   case only.  So remove the NARROWCHAR ifdef and include the
   header only in the wide character case.

 - In chartype.h, move ct_encode_char() below the comment explaining it.

 - No more need for underscores before ct_{de,en}code_string().

 - Make the conversion buffer resize functions private.
   They are only called from the decoding and encoding functions
   inside chartype.c, and no need can possibly arise to call them
   from anywhere else.

Char -> wchar_t from Ingo Schwarze.

more macro WIDECHAR undoing from Ingo Schwarze.

Start removing the WIDECHAR ifdefs; building without it has stopped working
anyway. (Ingo Schwarze)

whitespace and header sorting changes (Ingo Schwarze). No functional changes.

Don't free getline memory (Ingo Schwarze).

Use getline for better portability.

Change the test for the size of encoded buffer to include the NULL, from
OpenBSD; no functional change.

Use fparseln to avoid newline hacks.

Rebase to HEAD as of a few days ago.

Rebase.

sync with head.

for a reference, the tree before this commit was tagged
as yamt-pagecache-tag8.

this commit was splitted into small chunks to avoid
a limitation of cvs.  ("Protocol error: too many arguments")

Add a history function that takes a FILE pointer; needed for Capsicum.
From Eitan Adler

sync with head

Add coverity annotations about unreachable code (Kamil Dudka)

pass -Wconversion

- fix unused params
- unconditionalize vis.h

KNF return (\1); -> return \1;

kill ptr_t and ioctl_t, add  * sizeof(*foo) to all allocations.

include vis.h from dot.

fix narrow compilation

- don't leave cursor dangling on memory failure or after clearing the list
- compute the string length to be strvis'ed after the string is encoded

off by one in fetching history data. From: Gerry Swislow

rename historyw -> history_w for consistency.
add wide tst code and make it the default.

Fix wide build, test it, but don't turn it on yet.

Wide character support (UTF-8) from Johny Mattsson; currently disabled.

apply apple patches from:
http://opensource.apple.com/source/libedit/libedit-11/patches/

Plug memory leak, from MySQL.

Fix memory leak found by valgrind (Julien Torres)

Don't forget to initialize h_del; from  Julien Torres.

PR/30747: David N. Williams: libedit is missing remove_history()
Added, please test.

Fix memory leaks found by valgrind. From Julien Torres

PR/26785: Jess Thrysoee: libedit - H_NEXT and H_PREV shifts cursor on failure

make sure that we round up to 1K.

PR/26725: Sergey S. Kostyliov: Typo in libedit, possible buffer overflow in src/lib/libedit/history.c:history_save()

change allocation policy in el_push to allocate the string itself.
fix issues with strdup.

Move UCB-licensed code from 4-clause to 3-clause licence.

Patches provided by Joel Baker in PR 22280, verified by myself.

From michael@moria.de:
- use __attribute__((__unused__)) in arguments where appropriate.
- some int -> size_t and char * to const char * conversions.

Add a uniquefier for the history function.

don't crash in memory shortage conditions.

write the vis(3) converted string into the file, not the original one.
Noted by Tim Robbins. Hi luke!

- constify; passes all gcc and lint strict checks.
- add config.h [Jason Evans], to create a portable version of libedit that
  can be easily compiled on other OS's.

history_def_enter: fix off-by-one mistake in delete condition (the behaviour
to keep at least one entry on the history list is retained)
This fixes lib/9704 by Phil Nelson.

chmod the history file to 600 so that only the owner can read it.
[inspired by the openbsd fix to readline]

convert to new style guide, which includes:
- ansi prototypes & features (such as stdargs)
- 8 space indents

history_def_set has a `const int' as a third arg, not an `int'.
picked up by the ultrix compiler, reported by simonb@ ...

More trailing white space.

Add missing , in he_errlist[] initializer.  Yay lint.

Encode the history lines using strvis() before saving it. Fixes gdb history
problem with missing newlines.

- missing error message from array could cause core dump
- delint

per e-mail with christos: history_next_event() should return -1, not NULL,
in case of error, just like history_prev_event() does.

Add H_APPEND to simplify the interface.

Add a function to be able to set the cursor to a given event number.

PR/4257: Jaromir Dolecek: history() has no generic error handling and isn't
reentrant. This changes the interface of the history function, so we need
a major number bump.

Fix compiler warnings.

Portability fixes:
__const		-> const
BADSIG		-> SIG_ERR
int flags	-> u_int flags
#if __STDC__	-> #ifdef __STDC__

- convert unsafe strcpy(), strcat() and sprintf() to the `n' versions.
- some KNF.

RCSid police
editline first appeared in 4.4BSD not NetBSD1.0

Added history load and save to file functions.

libedit!

Initial revision