| Age | Commit message (Collapse) | Author |
|---|
|
This further tweaks the hack introduced to ignore the last match of ^
at the end of the file, see 1a158268c7693b00bf43c7e81034816d8d00358c.
|
|
|
|
|
|
|
|
|
|
This is useful when writing "debian/rules" files in packages.
|
|
|
|
|
|
|
|
Fix #867
|
|
These are the permission bits of struct stat's st_mode field at the
time of the most recent load/save.
Close #861
|
|
These are currently not mapped by default but can be enabled by mappings
using their virtual key names.
|
|
This renames the functions and constants implementing the to/till motions.
The new names should indicate that matches are only returned within the
current line (not globally). Apart from the changed virtual key/command
name this contains no functional changes.
|
|
These are either accidentally global or unused variables. In the file
type pluging the mime type variable was wrongly scoped, meaning the
text/plain settings applied to all otherwise unknown files.
|
|
|
|
|
|
|
|
f4f0f5b allowed "<<-EOF" heredocs to be parsed correctly, but it
introduced a bug that made the lexer fail when the beginning of a
here-doc was the last string in a file (optionally followed only
by blanks).
In order to fix this, move everything regarding "delimiter" within
the block that is executed only if "delimiter" is not nil.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
If a here-doc start delimiter begins with "-", then spaces are allowed
to come before the closing delimiter.
This patch fixes what would otherwise be parsed incorrectly:
<<-EOF
....
EOF
|
|
|
|
|
|
|
|
The for loop in selection_free won't run because the next element
will always be NULL, because we are freeing from the end.
Close #852
|
|
|
|
operators
|
|
This test does not seem to be stable on the GitHub macOS CI.
|
|
There are two main ways how the input queue is managed in vis:
- vis_keys_feed(..) appends new input to the queue and immediately
starts processing it. Starting from the position before the call
i.e. ignoring any previously queued input.
This is typically used in key binding handlers where the input
queue still contains the mapping leading to the invocation of the
handler. In that case new input should be interpreted immediately,
before the handler eventually returns and its mapping is consumed.
- vis_keys_push(..) with pos=0, appends new input to the end of the
queue and starts processing it from the start of the queue, taking
the full content into consideration.
This is used by the main loop when new input becomes available.
This patch switches the handling of <C-c> after a SIGINT from the former
to the latter mechanism and fixes mappings using <C-c> in a non-leading
position.
|
|
Gracefully fallback to curses(3) (no libcurses `.pc' file is present
and `-l$libcurses' is used and hence `-lcurses' (not `-llibcurses'!)).
|
|
This makes it thread safe. While it is unlikely that we use threads in
the near future, it is sensible to avoid unnecessary global state.
|
|
|
|
|
|
|
|
Previously these were only used to compile the main project source,
resulting in inconsistencies between the feature detection and actual
usage.
|
|
memrchr() signature is not exposed by default for NetBSD. If one does
not add -D_NETBSD_SOURCE, the compiler uses its own signature for
memrchr() that returns a 32-bit integer, leading to misbehaviour in some
situations. Defining _BSD_SOURCE doesn't work, so we define
_NETBSD_SOURCE.
|
|
There is no .lily file but there is a .ily file in lilypond for
stylesheets. They are like css for lilypond. See here:
http://lilypond.org/doc/v2.18/Documentation/learning/style-sheets
|
|
|
|
This was added in c240368d5da8208c15e0263034384414d938afb3 to work around
a possibly bogus tis-interpreter warning regarding multiple accesses
in the same expression.
|
|
|
|
|
|
While the ACSL specification for the function prototype is present,
the actual implementation is missing. Not sure why?
|
|
|
|
Cygwin does seemingly not support overlapping mmap(2) regions.
|
|
|
