In some cases it was possible that the `roster_pending_presence` list was
not free'd correctly.
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
Sometimes the server is too busy sending other stanzas or our connection is
saturated because of something else, so the pong arrives too late.
Prevent the autoping disconnect event by extending the timeout
each time a stanza is received.
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
We have to lazy load the keyfiles as they will maybe be written at one
point and should be initialized by that time.
Fixup of ca2df180d8
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
If `n` is the number of `substr` replace operations, the old
implementation did `n+1` allocations and always first copied start,
then replacement, then the remaining (unsearched) part of the original
string.
Now calculate the number of occurences of `substr` before allocating the
resulting string.
Change the algorithm to parse through the original string and only
copy the parts that must be kept into the resulting string plus the
replacements where they belong at.
Now we also only have to go through the original string twice plus we only
do a single allocation.
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
Previously we printed a format string into a `GString` to then print that
result again into a `GString` and display that.
This patch removes one level of `GString` printing by forwarding the
`va_arg` to the newly added API `win_println_va()`.
Introducing `static win_println_va_internal()` also allowed refactoring
most of the `win_print*()` and `win_append*()` implementations and removing
all the redundant code.
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
When we are not connected and run `/plugins install` we crash because we
get the account struct to check for the (xmpp) tls setting.
To apply that to the http (etc) connection to download the plugin from a
server.
This got introduced in 3a86b8c29 to fix#1624.
There are several ways to handle this (some described in 1880) in this
patch I took the route that it will use secure connection when we are
nto connected and will only check the tls.trust account setting if we
are connected.
Fix https://github.com/profanity-im/profanity/issues/1880
Thanks to @alarixnia fot the patch and bugreport.
```
Forwarding a bug reported using the NetBSD bug tracker ("Profanity crashes at login"): http://gnats.netbsd.org/57050
The code in question here handles messages by people not in the user's contact list, and what is probably happening is that the XMPP server is sending such a message before profanity has set up its window layout.
Setting roster.unsubscribed=false indeed makes profanity start without crashing, and the attached patch bails out early in the problematic code path if layout->subwin is NULL (Adrian verified for me that either of the two makes profanity work for him again). But I am not familiar enough with the profanity code to say if this has any issues.
Profanity compiled with ncurses doesn't crash because ncurses's wattron() has a guard against the user passing null pointers, but the X/Open curses library standard doesn't have this guarantee and there's no guard in the NetBSD curses implementation.
```
Fix https://github.com/profanity-im/profanity/issues/1769
Add ability to completely turn logs off,
Needed since `/logging` and `/history` didn't stop history logging
in the DB, only in files.
Command might break something on usage,
hence it was primarily introduced for privacy geeks.
Privacy command discussion #1836
Refactor `p_gpg_list_keys`
Add `/pgp autoimport` command,
it's not described in XEP-0027, but used in some clients,
such as PSI, Pidgin. It will autoimport keys received with
`/pgp sendpub`, in plain text as a message, or using features,
provided in other clients. It doesn't autoassign them, but shows
command to assign, letting user to decide.
Improve documentation for some preexisting functions
Add contact argument to `/pgp sendpub`
Improve documentation
Update code formatting and indentation for better readability
Show state if no argument provided
Show if argument wasn't changed
Reduce amount of arguments
Command allows to share your PGP pub key with ease,
it's not described in XEP-0027, but used in some clients,
such as PSI, Pidgin.
Fix typos
Minor improvements
The strings have to be dup'ed, otherwise we'll free them twice.
Fixes#1855
Introduced by b6bb50ceb2
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
Tell the user that a MAM request will be issued delayed instead of showing
a warning that MAM isn't supported by the server.
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
When compiling with `CFLAGS="-O2 -fexceptions -Wp,-D_FORTIFY_SOURCE=2"`, as
done per default when creating an AUR package e.g. via `makepkg`, this
error is produced.
Refactor `stanza_create_mam_iq()` in order to fix said warning-turned-into-
an-error.
Edit by jubalh:
I still think the error message is a false positive. But this commit improves the
code a lot.
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
`MB_CUR_MAX` looks like a macro, but it's a function call and therefore
creates a VLA. We don't want that.
Also this array being uninitialized created the following Valgrind error
```
==503529== Conditional jump or move depends on uninitialised value(s)
==503529== at 0x619F15E: waddnstr (lib_addstr.c:67)
==503529== by 0x1929B7: _inp_write (inputwin.c:353)
==503529== by 0x1937D5: _inp_redisplay (inputwin.c:619)
==503529== by 0x61511B1: rl_forced_update_display (display.c:2693)
==503529== by 0x193F9D: _inp_rl_send_to_editor (inputwin.c:957)
==503529== by 0x614642F: _rl_dispatch_subseq (readline.c:916)
==503529== by 0x6146C85: _rl_dispatch_callback (readline.c:823)
==503529== by 0x616739F: rl_callback_read_char (callback.c:241)
==503529== by 0x1923DB: inp_readline (inputwin.c:188)
==503529== by 0x149860: prof_run (profanity.c:117)
==503529== by 0x2283E8: main (main.c:186)
==503529== Uninitialised value was created by a stack allocation
==503529== at 0x1928B1: _inp_write (inputwin.c:334)
```
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
`data_dir` would have been leaked if directory creation failed.
`editor_argv` was leaked at some point, no idea why.
```
==1244734== 118 (32 direct, 86 indirect) bytes in 1 blocks are definitely lost in loss record 6,299 of 7,824
==1244734== at 0x4846CC3: realloc (vg_replace_malloc.c:1451)
==1244734== by 0x5E85AD0: g_realloc (in /usr/lib/libglib-2.0.so.0.7600.1)
==1244734== by 0x5E4A004: ??? (in /usr/lib/libglib-2.0.so.0.7600.1)
==1244734== by 0x5E4A7B1: g_ptr_array_add (in /usr/lib/libglib-2.0.so.0.7600.1)
==1244734== by 0x5EA4235: g_strsplit (in /usr/lib/libglib-2.0.so.0.7600.1)
==1244734== by 0x1F143C: get_message_from_editor (editor.c:92)
==1244734== by 0x193F6B: _inp_rl_send_to_editor (inputwin.c:950)
==1244734== by 0x614642F: _rl_dispatch_subseq (readline.c:916)
==1244734== by 0x6146C85: _rl_dispatch_callback (readline.c:823)
==1244734== by 0x616739F: rl_callback_read_char (callback.c:241)
==1244734== by 0x1923DB: inp_readline (inputwin.c:188)
==1244734== by 0x149860: prof_run (profanity.c:117)
==1244734== by 0x2283E8: main (main.c:186)
```
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
There were multiple reports where after a reconnect the window of the
MUC that was last opened, was empty.
`muc_join()` creates an instance of a MUC, `presence_join_room()` works
with this instance. Therefore the instance has to exist before working on
it.
I'm not sure if this really fixes the issue, but at least it didn't
happen anymore after I applied this modification.
I can't remember how I stumbled over this, either while looking at debug
logs or while looking at Valgrind output while a reconnect happened, but
something went wrong. Then I came to the conclusion that this may fix
the issue and for now it did ... maybe it comes back, then my RCA was
wrong.
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
Each of the cmd_executable_editor(), cmd_executable_urlopen() etc calls
this helper function since they were all doing the same just for
different configurations.
This is done in `/executable avatar set` since d7848e38b.
The command help also didn't mention this anymore.
Seems like it was forgotten to remove the actual parsing of this.
Add "/avatar disable" to comply with point "3.5 Publisher Disables
Avatar Publishing" in XEP-0084.
src/command/cmd_defs.c:2416
Add "disable" argument. Reword the "/avatar" command description
so it flows better.
src/command/cmd_ac.c:1101
Add "disable" to the "/avatar" autocomplete dictionary.
src/command/cmd_funcs.c:9277
Split "/avatar" commands into two groups with an if statement; those
with a parameter and those without. "cons_bad_cmd_usage()" is in both
groups, which is messy. "disable" has similar logic to "set", but it
includes a failure message.
src/xmpp/avatar.c:152
"avatar_publishing_disable()" uses the same logic to publish metadata
as in "avatar_set()".
src/xmpp/avatar.c:238
Add a message to inform users when they do not receive an avatar after
using "/avatar get" and "/avatar open". In case of a failure, the user
will be subscribed to future avatar updates as long as they continue
to use their current instance of profanity. Adding
"caps_remove_feature()" after "cons_show()" will unsubscribe the user
from avatar updates and prevent the user from downloading an avatar
unexpectedly hours later without issuing an "/avatar" command.
src/xmpp/stanza.c:2698
The new "disable" function follows the same logic as
"stanza_create_avatar_metadata_publish_iq()".
Before it tried to unload the plugin first and check the output.
But if broken plugin was loaded, then it couldn't unload it,
so before it require uninstall and install after it,
making update useless for plugin development purposes.
Unload is part of the uninstall so no unload is needed inside of the cmd function.
Refactoring of cmd_plugins_update.
Introduce new feature: sessions_alarm.
Added new account setting: max_connections. On exceeding this number,
user will get an alert. If number is less than 1, no alert will happen.
Tests altered to fit new feature.
Expected behaviour
When you type
/command
/command
it should be just 1 entry in the history.
Behaviour
All the entries were saved.
Behaviour is changed by introducing check. Before adding to history,
entry now is compared to the last history entry.
Under certain circumstances setting plain_str[len] to 0 might lead to crash
and it does not follow the best practices as well.
This change allows better handling of buffer copying and prevents crash.
Let's set this to NULL.
Someone was using `picaur` and got an error about potential
uninitialization. Even though the code is fine let's set this to NULL to
make the compiler happy.
The strange thing was that building manually on the same system worked.
Related to 2e43b0ae62.
If custom client is not set and include_os is true,
stanzas might be released without initializiting.
This commit fixes it by introducing an additional check
on usage of custom client.
Related to commit e52ca2fbaa
While testing https://github.com/profanity-im/profanity/pull/1817
we found that:
```
/statusbar show name (I have this off by default
/msg someone where I chose someone in my roster that has a nick assigned
/statusbar chat jid
```
Will only take effect upon creation or redraw of that tab.
Reason is that we do this in create_tab to limit building this
dynamically.
Let's leave it like it is but add a note to users.
Not worth build that string all the time.
- Fixes statusbar tabs.
Bug: Statusbar used nickname if it was set in roster,
irrelevant to /statusbar chat setting.
Expected behaviour would be using this setting set as "user" to show nickname,
and to show jid with "jid" setting.
Other solution is to give a user control over it with another settings.
- _status_bar_draw_maintext cleaned up, no changes to behaviour
Before this change, only nickname or JID (if no nickname set) is shown in the titlebar.
Change allows to alter nickname/JID preference
using already implemented /titlebar show jid|name setting.
If both are set to "show", then it is displayed in "nickname <JID>" format.
The existing way how active tabs are displayed didn't allow showing more
than 10 tabs. This patch adds a mode where the statusbar shows a
comma-separated list of tabs which were active since the last time viewed.
This view is inspired by how `irssi` shows the active tabs, therefore
it is also called `actlist`.
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
It is possible, that a server sends a form with a field as follows, which
has no (default) value assigned.
```
<field label="foo" type="list-single" var="bar">
<option label="a"><value>a</value></option>
<option label="b"><value>b</value></option>
<option label="c"><value>c</value></option>
</field>
```
This patch fixes profanity to show that list. Before this patch
profanity showed nothing.
I stumbled over this while running the `/room config` command inside
a newly created muc on the `sure.im` XMPP server.
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
Before this change the same hashtable was re-created each time one used the
auto-completion feature.
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
* Fix typos.
* Add words that are not typos to codespell's ignore words list and
ignore regex.
* Make codespell ignore URIs.
* Make `make doublecheck` throw no error.
Signed-off-by: Daniel Santos <dacs.git@brilhante.top>
Sometimes after a reconnect the current window would get cleared. This
was a deliberate change to fix the profanity window looking all garbled
up after providing the passphrase for a gpg key using pinentry-curses.
Fixes https://github.com/profanity-im/profanity/issues/1556
* Make editor executable into a string to be able to support
(multiple) flags.
* Change /help executable to suit this new feature
Signed-off-by: Daniel Santos <dacs.git@brilhante.top>
Rewrite `win_get_title()` to using switch without a default case. So the compiler warns us
(`enumeration value ‘WIN_XXX’ not handled in switch`)
in case we add a new window type (WIN_CHAT, WIN_PRIV etc) and forget to
adapt this function.
Add assert() in the end to make compiler happy when he realizes there
the function might have no return value (`control reaches end of
non-void function`). This should ever be reached.
Replace the default value of `win_to_string()`, `win_get_tab_identifier()` as
well, and replace it with an assert.
See discussion started at https://github.com/profanity-im/profanity/pull/1799#discussion_r1142444684
Fixes https://github.com/profanity-im/profanity/issues/1794
Explanation
The problem is the download's identifier. Downloads are given an ID so they can be referenced later when their progress changes. Currently, the download's ID is the download's URL. When you download the same file twice, you have two downloads with the same ID. Download progress updates are shown on the first of both downloads with the same ID.
Solution
Change the download's ID from its URL to a random number. A random ID is generated when get_random_string() is called from cmd_funcs.c. Several other functions are updated to cope with the new ID format.
After adding MAM quote and url autocompletion wouldn't pick up messages
from MAM or history(from DB) or would have them in the wrong order. This
commit fixes that.
Fixes https://github.com/profanity-im/profanity/issues/1770
The timer will be auto-removed if the callback returns `0`. The way it was
done before breaks some internals of libstrophe.
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
This also fixes a memory leak from within `_handle_groupchat()` in [0].
[0] src/xmpp/message.c
Signed-off-by: Steffen Jaeckel <jaeckel-floss@eyet-services.de>
Previously after the url if you pressed tab, even if you typed out a
filepath, profanity would erase that and cycle through url autocomplete
results.
This patch solves that and autocompletes filepaths after the url.
Fixes https://github.com/profanity-im/profanity/issues/1783
Locally I have clang-format 15.0.2.
Our CI runs ubuntu-20.04 so it has 10.0.
Seems like a default value changed but I can't find it.
When I still had clang-format 14.x everything was fine and checking:
https://releases.llvm.org/15.0.0/tools/clang/docs/ReleaseNotes.html#clang-format
I only see irrelevant things.
Someone on the llvm IRC channel sais "there is no guarantee" and "maybe
a regression or intended change".
But seems like noone knows which setting could be the one we need.
Since I don't have a better solution for now I will just edit this by
hand to apply to the clang-format version we have on the CI.
Will look into updating this one.
But in any case it would be best if the configuration file could have
this setting so formatting works the same for all contributors.
When loading messages from MAM profanity would segfault. Reason was that
we were freeing the timestamp of messages when displaying them and we
needed it for loading MAM.
