| Commit message (Collapse) | Author | Age |
... | |
|\ \ \ \
| |_|_|/
|/| | | |
Small fixes.
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Problem : Operands don't affect results (CONSTANT_EXPRESSION_RESULT).
Diagnostic : Harmless issue.
Rationale : n >= LONG_MIN, n being intmax_t, is always true for
architectures where sizeof(intmax_t) == sizeof(long).
Resolution : Add sizes check.
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Problem : Assert can fail for legal values. Modulo-arithmetic of
unsigned types can make so that n * 100 > n, but n has
overflowed.
Solution : Use alternative form of expression.
|
|\ \ \ \
| | | | |
| | | | | |
Macro cleanup
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
These were found with -Wunused-macros. There are many more macros which
triggered that warning, but they were primarily part of larger sets of
macros so leave them alone.
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Neovim always ships with all features[0], so this serves no purpose.
Besides, this always evaluated to true.
[0]: https://github.com/neovim/neovim/wiki/Differences-from-vim
|
| |/ / /
| | | |
| | | |
| | | |
| | | |
| | | | |
Regarding dict_lookup() in eval.c: both definitions are the same, the
only difference being the spacing between the indirection operator and
the indentation level.
|
| | | | |
|
| | | |
| | | |
| | | |
| | | | |
Must check for EOF which will result in row/col being uninitialized.
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
- Set 't_Co' to 256 at startup. The value can be changed by the user for
compatibility with terminals that are less capable.
- `has('gui_running')` will return 1 if at least one rgb UI is attached.
Even though these changes are hacky, they are necessary to make the transition
to the new UI architecture smoother.
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Before sending a resize command to the UIs, flush the current output buffer to
ensure no redraw commands for a screen with invalid size are processed.
|
| | | |
| | | |
| | | |
| | | |
| | | | |
By calling ui_write directly, the internal screen isn't updated and invalid
bytes aren't handled, which breaks the abstract UI model.
|
| | | |
| | | |
| | | |
| | | |
| | | | |
ui_linefeed will scroll the screen when it becomes full. This can happen when
executing external commands.
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | | |
This is required to send redraw notifications while a msgpack-rpc call is being
performed by another channel.
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Forward the command to the remote UI and flush immediately. The
semantics/handling is UI-specific.
|
| | | | |
|
| | | | |
|
| | | |
| | | |
| | | |
| | | | |
When set to false, nvim will send cterm color numbers with `highlight_set`.
|
| | | | |
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | | |
Ignoring invalid key sequences simplifies input handling in UIs. The only
downside is having to use "<lt>" everytime a "<" is needed on functional tests.
|
| | | | |
|
| | | |
| | | |
| | | |
| | | | |
Also disable deferral of attach_ui
|
| | | |
| | | |
| | | |
| | | |
| | | | |
This code is required for screen.c to set a vertical scroll region, which is a
great optimization when scrolling vertically split windows.
|
| | | | |
|
|/ / /
| | |
| | |
| | |
| | | |
It is necessary to notify the UI when the default background/foreground colors
change in order to render correctly.
|
|\ \ \
| | | |
| | | | |
vim-patch: Multiple small patches
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Problem: No test for replacing on a tab in Virtual replace mode.
Solution: Add a test. (Elias Diem)
https://code.google.com/p/vim/source/detail?r=v7-4-563
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Problem: Non-ascii vertical separater characters are always redrawn.
Solution: Compare only the one byte that's stored. (Thiago Padilha)
https://code.google.com/p/vim/source/detail?r=v7-4-567
|
| | | | |
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Problem: CTRL-W } does not open preview window. (Erik Falor)
Solution: Don't set g_do_tagpreview for CTRL-W }.
https://code.google.com/p/vim/source/detail?r=v7-4-466
|
|\ \ \ \
| | | | |
| | | | | |
Remove QNX/pterm remnants
|
| | |/ /
| |/| |
| | | |
| | | |
| | | | |
The function qnx_init() (wrapped in an ifdef in main.c) doesn't even
exist.
|
|\ \ \ \
| | | | |
| | | | | |
Linting: Recommend os_* instead of POSIX functions.
|
| | | | | |
|
| | | | | |
|
|\ \ \ \ \
| |_|/ / /
|/| | | | |
[WIP] "abstract_ui" fixes and improvements
|
| | | | | |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Before sending a resize command to the UIs, flush the current output buffer to
ensure no redraw commands for a screen with invalid size are processed.
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
By calling ui_write directly, the internal screen isn't updated and invalid
bytes aren't handled, which breaks the abstract UI model.
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
ui_linefeed will scroll the screen when it becomes full. This can happen when
executing external commands.
|
| | | | | |
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This is required to send redraw notifications while a msgpack-rpc call is being
performed by another channel.
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
- Remove suspend method from the UI protocol
- Handle `:suspend` by disconnecting the last channel that sent a request to
nvim.
|