Removing newlines in rc.xml and menu.xml caused parser error with
following content:
<!--
-
- Some comments
-
-->
...though it is a valid XML.
Let's not do that. I moved `grab_file()` to `buf.c` and renamed it to
`buf_from_file()`, because it now directly touches `struct buf` and
I don't like having a source file only for one function.
...to enable configuration of the action prompt command.
Also set some better defaults for labnag.
The new default command is:
labnag \
--message '%m' \
--button-dismiss '%n' \
--button-dismiss '%y' \
--background '%b' \
--text '%t' \
--border '%t' \
--border-bottom '%t' \
--button-background '%b' \
--button-text '%t' \
--border-bottom-size 1 \
--button-border-size 3 \
--timeout 0
...where the conversion specifiers are defined as follows:
%m: the `<prompt>` message option
%n: _("No")
%y: _("Yes")
%b: osd.bg.color
%t: osd.label.text.color
This config options also enables the use of a different dialog client, for
example like this:
<core>
<promptCommand>zenity --question --text="%m"</promptCommand>
</core>
This is a common practice in C projects, which simply enforces that
each header must compile cleanly without implicit dependencies on
other headers (see also the previous commit).
This works around a wrong truncation warning in older GCC versions:
```
../src/common/buf.c:110:10: error: null destination pointer [-Werror=format-truncation=]
110 | int n = vsnprintf(NULL, size, fmt, ap)
```
Add a BUF_INIT macro, which makes it easier to initialize a struct buf
to an empty string (without a heap allocation).
Add buf_move() to move the contents of one struct buf to another (the
source is reset to BUF_INIT, analogous to C++ move-assignment).
Use buf_reset() instead of directly calling `free(s->buf)` since the
internal buf may not always be allocated by malloc() now.
Before this patch, the OSD would repeat the last buffer
content in case the new buffer content would be empty.
This was mostly happening for the `title` OSD field that is intended
to be empty when it matches the app_id / WM_CLASS of the application.
Due to only buffer.len being reset but its internal allocations being
untouched, buffer.buf would still carry the old data.
This patch fixes it by also overwriting the first byte in the buffer
allocation with '\0' via the new `buf_clear()` function.
Do the same for buf_expand_shell_variables() although that one should
have been fine before as it always writes new data to the buffer.
There is at least one user of the buffer API that reuse a single
buffer by just resetting `buf.len` to `0`. This works as long as
the new user of the buffer actually adds something to the buffer.
However, if we don't add anything but still provide `buf.buf` to
a consumer, the old content will be re-used.
This patch thus adds two new clearing variants to the buffer API:
- `buf_clear()` which doesn't reset the internal allocations
- `buf_reset()` which does free the internal allocations
Additionally, this patch makes `buffer_add_char()` public which
allows adding single characters to an existing buffer. This will
be used in a future PR which implements custom format strings for
the OSD.
...<command> argument (but still resolve tilde).
This makes it easier to write sh -c '' constructs without turning labwc
into a shell parser in order to expand environment variables, whilst
respecting single quotes and escaped characters as well as ignoring
subshells syntax like $(foo) and backticks.
Also, fix bug where buffer length+alloc get out-of-sync
Expanding shell variables, including tilde, enables the following type
of keybind:
<keyboard>
<keybind key="XF86AudioMute">
<action name="Execute">
<command>bash ~/mute-script.sh</command>
</action>
</keybind>
</keyboard>
Fixes issue #32
