aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAge
* I/O: ignore ENOTSUP for failed fsync()Justin M. Keyes2019-02-21
| | | | | | | | | Suggested by ZyX in https://github.com/neovim/neovim/issues/6725#issuecomment-312197691 : > There already is an exception if writing to a “device” (e.g. FIFO). > It makes sense to ignore certain errors like ENOTSUP or EOPNOTSUPP > since it is not something we or user can do anything about. ref #6725
* deps: update to libuv v1.26.0Justin M. Keyes2019-02-21
| | | | | | | | | | | | | | | | | | | | | | | Notable changes since v1.23.2: - v1.26.0 - uv_os_uname() - unix: don't attempt to invalidate invalid fd https://github.com/libuv/libuv/commit/1ce6393a5780538ad8601cae00c5bd079b9415a9 - v1.25.0 - unix: better handling of unsupported F_FULLFSYNC (fixes #6725) https://github.com/libuv/libuv/commit/6fc797c3fe18d8df71b36ecf2184f085c0283251 - tty,win: fix Alt+key under WSL https://github.com/libuv/libuv/commit/d2e59bb6003d707bdebd7a381f5a7e1d0cc3fd3b - fsevents: really watch files with fsevents on macos 10.7+ https://github.com/libuv/libuv/commit/2d2af382ce84b91d6ee7a185af32fca7f0acd84b - win: fix duplicate tty vt100 fn key - v1.24.0 - win,fs: retry if uv_fs_rename fails https://github.com/libuv/libuv/commit/e94c184c7c4a18f3de569c97caeb83f4ff98a4b2 - later [reverted](https://github.com/libuv/libuv/issues/2098) but may be useful reference - win: support more fine-grained windows hiding https://github.com/libuv/libuv/commit/4c2dcca27b80945d6b7063f0ea031b8a75a46a52
* fix "E667: Fsync failed" on macOSJustin M. Keyes2019-02-21
| | | | | | | | | | | | | macOS: Try direct fsync() if F_FULLFSYNC fails. closes #6725 ref https://github.com/vim/vim/pull/4016 vim-patch:8.1.0957 > on macOS F_FULLFSYNC fails with ENOTSUP for unsupported storage systems > (e.g. SMB), though this is not documented in the Apple fcntl man page. libuv fixed this in v1.25.0: https://github.com/libuv/libuv/commit/6fc797c3fe18d8df71b36ecf2184f085c0283251
* man.vim: g:man_hardwrap #9633Curtis McEnroe2019-02-21
| | | | | | This option allows restoring the behaviour prior to #9023. Fixes #9583 Closes #9633
* Merge #9626 from janlazo/vim-8.1.0926Justin M. Keyes2019-02-18
|\
| * vim-patch:8.1.0276: no test for 'incsearch' highlighting with :sJan Edmund Lazo2019-02-17
| | | | | | | | | | | | Problem: No test for 'incsearch' highlighting with :s. Solution: Add a screendump test. https://github.com/vim/vim/commit/164251ff805e89a3d9a850a77e3139e28908f44a
| * vim-patch:8.1.0387: no test for 'ambiwidth' detectionJan Edmund Lazo2019-02-17
| | | | | | | | | | | | Problem: No test for 'ambiwidth' detection. Solution: Add a test. https://github.com/vim/vim/commit/24839edc54e985ed88d063513226922a6f5b7554
| * vim-patch:8.1.0668: no test for overstrike mode in the command lineJan Edmund Lazo2019-02-17
| | | | | | | | | | | | Problem: No test for overstrike mode in the command line. Solution: Add a test. (Dominique Pelle, closes vim/vim#3742) https://github.com/vim/vim/commit/c0676bab92bd6488e17d05a4a70dfff993ef524e
| * vim-patch:8.1.0926: no test for :wnext, :wNext and :wpreviousJan Edmund Lazo2019-02-17
| | | | | | | | | | | | Problem: No test for :wnext, :wNext and :wprevious. Solution: Add a test. (Dominique Pelle, closes vim/vim#3963) https://github.com/vim/vim/commit/e93e5a504f481bd0dad9c504d5fcf0e5f0dfc6e6
* | Merge #9625 from janlazo/vim-8.1.0945Justin M. Keyes2019-02-18
|\ \ | |/ |/|
| * vim-patch:8.1.0937: invalid memory access in search patternJan Edmund Lazo2019-02-17
| | | | | | | | | | | | | | Problem: Invalid memory access in search pattern. (Kuang-che Wu) Solution: Check for incomplete collation element. (Dominique Pelle, closes vim/vim#3985) https://github.com/vim/vim/commit/f1b57ab2ab18032d19f64bff7d22f3adb3fe93f7
| * vim-patch:8.1.0934: invalid memory access in search patternJan Edmund Lazo2019-02-17
| | | | | | | | | | | | Problem: Invalid memory access in search pattern. (Kuang-che Wu) Solution: Check for incomplete equivalence class. (closes vim/vim#3970) https://github.com/vim/vim/commit/985079c514e9ab85598b7bca019c77d3e42526f5
| * lintJan Edmund Lazo2019-02-17
| |
| * vim-patch:8.1.0945: internal error when using pattern with NL in the rangeJan Edmund Lazo2019-02-17
|/ | | | | | | Problem: Internal error when using pattern with NL in the range. Solution: Use an actual newline for the range. (closes vim/vim#3989) Also fix error message. (Dominique Pelle) https://github.com/vim/vim/commit/a5483448cba6997517003a22a8029f0de1007d0e
* build: checkprefix: skip if empty #9624Daniel Hahler2019-02-17
| | | | | | | | | | | | | | | This regressed in 3b473bb14f9f452461016c7042949a23499df629: ``` % make 'CMAKE_BUILD_TYPE=RelWithDebInfo' 'CMAKE_EXTRA_FLAGS=-DCMAKE_INSTALL_PREFIX=/vim-build/neovim/neovim/master -DENABLE_JEMALLOC=OFF' 'DEPS_CMAKE_FLAGS=-DUSE_BUNDLED=OFF' error: CMAKE_INSTALL_PREFIX '/vim-build/neovim/neovim/master' does not match cached value '' Run this command, then try again: cmake build -DCMAKE_INSTALL_PREFIX=/vim-build/neovim/neovim/master make: *** [Makefile:169: checkprefix] Error 1 ``` It was checking before for non-empty also [1]. 1: https://github.com/neovim/neovim/pull/9621/files#diff-b67911656ef5d18c4ae36cb6741b7965L22
* vim-patch:8.1.0932: remove Farsi support (#9622)Justin M. Keyes2019-02-17
| | | | | Problem: Farsi support is outdated and unused. Solution: Delete the Farsi support. https://github.com/vim/vim/commit/14184a3133b9a6ee5f711d493c04e41ba4fa7c2f
* Merge #9593 'API: nvim_create_buf: "scratch" param'Justin M. Keyes2019-02-17
|\
| * API: nvim_create_buf: add `scratch` parameterJustin M. Keyes2019-02-17
| | | | | | | | | | Creating a scratch buffer is a chore/ritual, and would be more useful/common if formally exposed.
* | terminal: Fix potential invalid local 'scrollback' (#9605)Justin M. Keyes2019-02-17
| | | | | | | | | | | | TermOpen autocmd may set local 'scrollback' to -1, this needs to be adjusted as in on_scrollback_option_changed(). fixes #9588 (OOM, out of memory)
* | build/Makefile: validate prefix for specific targets (#9621)Justin M. Keyes2019-02-17
|/ | | | | | `clean` and `distclean` should not check CMAKE_INSTALL_PREFIX. ref #9615 Helped-by: bruce-hill
* vim-patch:8.1.0803: session restore: handle single quotes #9620Marco Hinz2019-02-17
| | | | | | | | | Problem: Session file has problem with single quote in file name. (Jon Crowe) Solution: Use a double quoted string. Add a test. https://github.com/vim/vim/commit/ad36a3588d32985ee27bd11aa97e5195ef623158 Fixes https://github.com/neovim/neovim/issues/9618
* Merge #9617Justin M. Keyes2019-02-16
|\
| * build/Makefile: check CMAKE_INSTALL_PREFIXJustin M. Keyes2019-02-16
| | | | | | | | closes #9615
| * build/CMakeLists.txt: group related logicJustin M. Keyes2019-02-16
|/ | | | | | - Group CMAKE_BUILD_TYPE-related logic together. - Print the MIN_LOG_LEVEL message (in the build-log) next to the CMAKE_BUILD_TYPE message.
* Merge #9614Justin M. Keyes2019-02-16
|\
| * clang/"null pointer dereference": win_rotateJustin M. Keyes2019-02-14
| |
| * PVS/V1028: cast operands, not the resultJustin M. Keyes2019-02-14
|/
* vim-patch:8.1.0852: findfile() and finddir() are not properly tested (#9609)Jan Edmund Lazo2019-02-13
| | | | | Problem: findfile() and finddir() are not properly tested. Solution: Extend the test and add more. (Dominique Pelle, closes vim/vim#3880) https://github.com/vim/vim/commit/ed71ed37bcd44bcbfcf59beee1248799de3315ec
* Merge #9589 from justinmk/cterm-pumblendJustin M. Keyes2019-02-12
|\ | | | | UI: 'pumblend' for cterm (256-color TUI)
| * UI: change implementation of hl_rgb2cterm_color()Justin M. Keyes2019-02-12
| | | | | | | | | | Replace the implementation cargo-culted from Vim's source with something simpler which "seems to look better" with 'pumblend'.
| * UI: 'pumblend' for cterm (256-color TUI)Justin M. Keyes2019-02-12
|/ | | | | | hl_rgb2cterm_color, hl_cterm2rgb_color were adapted from Vim 8.1 (color2index, cterm_color2rgb). ref: https://github.com/vim/vim/commit/c5cd88554f1e0b2e9ff08d9a0748238dd8340ce1
* Merge #9601 from justinmk/tui-italicJustin M. Keyes2019-02-12
|\
| * TUI: assume italics support in all xterm-likesJustin M. Keyes2019-02-12
| | | | | | | | | | | | https://github.com/neovim/neovim/issues/9598#issuecomment-462421302 > On systems other than macOS, terminfo always has them set. It's part > of the ANSI standard SGR codes
| * TUI: sniff nsterm (Terminal.app) from $TERM_PROGRAMJustin M. Keyes2019-02-11
| |
| * TUI: force italics in tmuxJustin M. Keyes2019-02-11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | tmux users are encouraged to use "tmux-256color" instead of the old "screen-256color". Put this in your .tmux.conf : if-shell 'infocmp tmux-256color' 'set -g default-terminal "tmux-256color"' 'set -g default-terminal "screen-256color"' If default-terminal is "tmux-256color" AND tmux finds its terminfo, THEN it will enable italics (regardless of whether the terminfo contains sitm/ritm). OTHERWISE tmux "emulates screen", which means it emits italics as "reverse". That's unfortunate, but it's an acceptable compromise that allows Nvim to support italics in tmux by default in compatible (and increasingly-common) configurations. Test case: tmux nvim -u NORC +'hi Comment cterm=italic' CMakeLists.txt ref #9598
| * TUI: italics in Terminal.app (nsterm)Justin M. Keyes2019-02-11
| | | | | | | | | | | | | | | | | | | | | | | | This is the conservative approach, requiring TERM=nsterm as recommended at `:help $TERM`. It won't work with Terminal.app's default of TERM=xterm-256color, nor in tmux. Test case: TERM=nsterm nvim -u NORC +'hi Comment cterm=italic' CMakeLists.txt closes #9598
* | Merge pull request #9603 from bfredl/reversepumBjörn Linse2019-02-11
|\ \ | |/ |/| highlight: handle blending with gui=reverse and guisp attribute
| * highlight: handle blending with gui=reverse and guisp attributesBjörn Linse2019-02-11
|/
* Merge pull request #7466 from bfredl/ext_messagesBjörn Linse2019-02-10
|\ | | | | ui: implement ext_messages to externalize all messages in msg area
| * ui: implement ext_messagesBjörn Linse2019-02-10
|/ | | | Co-Author: Dongdong Zhou <dzhou121@gmail.com>
* Merge pull request #9272 from bfredl/createbufBjörn Linse2019-02-09
|\ | | | | api: add nvim_create_buf to create a new empty buffer.
| * api: add nvim_create_buf to create a new empty buffer.Björn Linse2019-02-09
| | | | | | | | | | | | | | | | | | | | | | | | Loading existing files into a buffer is non-trivial and requires a window. Creating an unnamed emtpy buffer is trivial and safe though, thus worth a special case. Change nvim_buf_set_option to use aucmd_prepbuf. This is necessary to allow some options to be set on a not yet displayed buffer, such as 'buftype' option. vim-patch:7.4.1858: Add BLN_NEW to enforce buflist_new creating new buffer
* | doc: 'fillchars' is local to windowMarco Hinz2019-02-09
|/ | | | | This was forgotten in 352811fe5 since the commit was originally meant to only change 'listchars'.
* Merge pull request #9586 from bfredl/screencleanupBjörn Linse2019-02-07
|\ | | | | screen: cleanup allocation and scrolling
| * screen: cleanup allocation, clearing and validationBjörn Linse2019-02-07
| | | | | | | | | | | | grid_xx functions shouldn't randomly be called when the screen is not valid. Nvim deterministically initializes a default screen early in main(), so the default grid can be assumed to be allocated.
| * screen: simplify scrolling codeBjörn Linse2019-02-07
|/ | | | | Since the separation between internal screen and TUI, internal scroll cannot FAIL. Delete the conditions for this.
* Merge pull request #9571 from bfredl/pum_transparentBjörn Linse2019-02-07
|\ | | | | UI: implement 'pumblend' option for semi-transparent popupmenu
| * UI: implement 'pumblend' option for semi-transparent popupmenuBjörn Linse2019-02-07
|/ | | | | | | | | | | | | Why? - Because we can. - Because the TUI is just another GUI™ - Because it looks kinda nice, and provides useful context like 1 out of 100 times Complies with "don't pay for what you don't use". Some crashes for resizing were unfolded, add tests for those.
* man.vim: set 'linebreak'Justin M. Keyes2019-02-06
| | | | closes #9583
* Merge pull request #9581 from bfredl/alwaysrgbBjörn Linse2019-02-05
|\ | | | | UI: always use contrete colors for default_colors_set