Commit Graph

6801 Commits

Author SHA1 Message Date
Justin M. Keyes
ae9a3d3b99 'inccommand': Simplify cursor placement logic.
Remove the inner for-loop; just use the column of the first in-line
match.
2016-11-11 05:45:06 +01:00
Justin M. Keyes
795f2bb9b9 'inccommand': Restore cursor/view if cancelled.
Restore cursor position and window "view", if command is cancelled.
This is how 'incsearch' works.
2016-11-10 18:09:48 +01:00
Justin M. Keyes
2e1217da46 'inccommand': buftype=nofile, restore cursor/view
- Use a standard scratch buffer instead of a new 'buftype', functions
  like curbufIsChanged() already have special handling for scratch bufs.
- Cleanup some stuff from the previous merge.
- Add support for :smagic, :snomagic. Closes #5578
2016-11-10 15:48:00 +01:00
Justin M. Keyes
0213e99aaf Merge #5561 'inccommand'
Initial work by:
  Robin Elrharbi-Fleury (Robinhola)
  Audrey Rayé (Adrey06)
  Philémon Hullot (DesbyP)
  Aymeric Collange (aym7)
  Clément Guyomard (Clement0)

Major revisions by:
  KillTheMule
  Björn Linse <bjorn.linse@gmail.com>
  Justin M. Keyes <justinkz@gmail.com>
2016-11-09 03:19:22 +01:00
Justin M. Keyes
aa0e09d251 'inccommand': Preserve curbuf->b_u_newhead.
Add tests for undotree().

Helped-by: Björn Linse <bjorn.linse@gmail.com>

When "curhead" points to a valid head, the value of "newhead" is
meaningless (and really should be set to null). The undo state for
a buffer is _logically_ the enum:

  enum UndoState {
    CurrentHead(head),
    NewHead(head),
    EmptyTree
  }

nvim _represents_ this as: whenever `curbuf->b_u_curhead` is nonnull it
should be used as the current head, and `curbuf->b_u_newhead` is
ignored. If the there is a current head, then this will be redoed on the
next redo, and its parent will be undone on next undo. Only if
`b_u_curhead` is NULL, `b_u_newhead` will be used as the head to undo
(and it is not possible to redo). Also both can be NULL, to indicate an
empty undotree. (To be fair, this only strictly true when calling undo.c
from the outside, in some places _within_ a function in undo.c both
values might be meaningful)

Apparently `undotree()` breaks this non-abstraction, this _cosmetic_
issue can easily be fixed by `ex_substitute` also saving and restoring
`b_u_newhead`, but is doesn't reflect any error really how
`u_undo_and_forget` manipulates the _actual_ state of the undo tree.
2016-11-09 01:03:47 +01:00
Justin M. Keyes
21dfbfbb9a perf: do_sub(): avoid work, avoid screen updates
- Don't fill matched_lines if not showing a preview (!eap->is_live).
- Encapsulate: Move cursor placement logic to show_sub().
2016-11-08 21:22:24 +01:00
Justin M. Keyes
f3e8ca3bf5 'inccommand': preserve 'modified'
During the live preview, the buffer-local 'modified' flag
should not be changed.
2016-11-08 21:22:24 +01:00
Justin M. Keyes
ff6ec703d5 'inccommand': Do not trigger during scripts, feedkeys(). 2016-11-08 21:22:24 +01:00
KillTheMule
527ba2b12a 'inccommand': test: scripts/feedkeys() should not trigger preview 2016-11-08 21:22:24 +01:00
Justin M. Keyes
6a3f8d48d0 'inccommand': rename 'incsubstitute'
'inccommand' allows us to expand the feature to other commands, such as:
    :cdo
    :cfdo
    :global

Also rename "IncSubstitute" highlight group to "Substitute".
2016-11-08 21:22:13 +01:00
Justin M. Keyes
1e0e301062 'inccommand': format line numbers as "|123| "
This matches what Quickfix traditionally does.
2016-11-08 21:21:46 +01:00
Justin M. Keyes
3ccf69c1de 'inccommand': set buffer name to [Preview]
[inc_sub] is less obvious for users. Also, in the future we may want to
generalize the idea of a "preview buffer", or "incremental commands"
besides :substitute.
2016-11-08 21:21:46 +01:00
Justin M. Keyes
d0689eb0b2 'inccommand': disable 'cursorline', 'spell' in preview 2016-11-08 21:21:43 +01:00
Justin M. Keyes
e31f9007e4 'inccommand': preserve b:changedtick 2016-11-08 21:21:00 +01:00
Justin M. Keyes
e4e7b2d239 lint 2016-11-08 21:21:00 +01:00
Justin M. Keyes
c04ffe866d 'inccommand': rework
- Eliminate/isolate static/global variables
- Remove special-case parameter from buflist_new()
- Remove special-case ECMD_RESERVED_BUFNR
- To determine when u_undo_and_forget() should be done, check
  b_changedtick instead of a heuristic.
- use mb_string2cells() instead of strlen() to measure the :sub patterns
- call ml_close() before buf_clear_file(). Avoids leaks caught by ASan.

Original patch by:
  Robin Elrharbi-Fleury (Robinhola)
  Audrey Rayé (Adrey06)
  Philémon Hullot (DesbyP)
  Aymeric Collange (aym7)
  Clément Guyomard (Clement0)
2016-11-08 21:20:08 +01:00
Marco Hinz
bd33f1165b CheckHealth: ignore non-existing markdownCodeBlock group (#5570)
People using [1] would experience an error, because their markdown syntax
doesn't define the markdownCodeBlock group.

[1]: https://github.com/plasticboy/vim-markdown

Closes #5569
2016-11-08 19:01:15 +01:00
Tommy Allen
3089c86f4e CheckHealth: timeout system() calls (#5565) 2016-11-08 14:34:03 +01:00
Björn Linse
9147331e21 Merge pull request #2905 from bfredl/utf8
Only allow encoding=utf-8 and simplify multibyte code
2016-11-05 18:12:14 +01:00
Björn Linse
4ab3fe8eaa encoding: update documentation 2016-11-05 14:49:24 +01:00
Björn Linse
b3ece5c81c encoding: update tests
Change shada reencoding tests to check for
correct handling of UTF-8 and binary strings.

Delete enc=latin1 json tests.
2016-11-05 14:49:24 +01:00
Björn Linse
18f56c8e90 encoding: delete non-UTF-8 implementations of multibyte functions
Deleted documentation was duplicated at specific utf_ implementation
2016-11-05 14:49:24 +01:00
Björn Linse
5072ab9e5b encoding: cleanup mbyte.c given fixed encoding=utf-8
Eliminate mb_init():
Set "enc_utf" and "has_mbyte" early. Eliminate "enc_unicode" and "enc_latin1like".
init_chartab() and screenalloc() are already invoked elsewhere
in the initialization process.
The EncodingChanged autocmd cannot be triggered.
At initialization, there is no spellfiles to reload
2016-11-05 14:49:24 +01:00
Björn Linse
4804001aff encoding: only allow encoding=utf-8 2016-11-05 14:49:23 +01:00
Marco Hinz
32d9c19e29 Health: introduce help links (#5557)
We can now use help links like in normal help pages. The bars around them will
be concealed as well.

* Health: link string "SUCCESS" to ModeMsg

So far we linked "SUCCESS" to the Function highlight group.

The newly introduced healthHelp group links to Idenfifier as it does for links
in normal help pages.

Now the problem is that the Function group links again to Identifier as well, so
both, "SUCCESS" and help links, would use the same colors.

* Health: don't use Markdown code blocks

Every line indented by more than 4 spaces would lose its highlighting otherwise.

* Health: add helper health#help_to_link()
2016-11-04 11:03:05 +01:00
James McCoy
01f1028496
Merge pull request #5524 from brcolow/vim-7.4.1967
vim-patch:7.4.1967
2016-11-02 20:06:23 -04:00
Michael Ennen
b8bb1f6729
vim-patch:7.4.1967
Problem: Falling back from NFA to old regexp engine does not work properly.
(fritzophrenic)
Solution: Do not restore nfa_match. (Christian Brabandt, closes vim/vim#867)

6747fabc73

Helped-by: jamessan
2016-11-02 20:05:45 -04:00
James McCoy
9ef4be9aab Merge pull request #5550 from jamessan/big-endian-fixes
Making nvim big-endian friendly
2016-11-02 20:04:10 -04:00
James McCoy
654e92186b
lint 2016-11-02 10:06:27 -04:00
James McCoy
9b545dd3db
nvim_win_get_number: Use int as rv to fix BE issues
rv is int64_t but its address is being passed into win_get_tabwin as if
it were an int.  This breaks on big-endian systems, since win_get_tabwin
will store the data to the "wrong" half of the int64_t, thus returning
invalid data out of nvim_win_get_number.
2016-11-02 10:06:27 -04:00
James McCoy
37e64d79cc
object_to_vim: Fix buffer/window/tabpage conversion on BE systems
Since data.integer is a different (larger) integer type than
data.{buffer,window,tabpage}, we cannot abuse the union by using
data.integer to access the value for all 4 types.  Instead, remove the
{buffer,window,tabpage} members and always use the integer member.

In order to accomodate this, perform distinct validation and coercion
between the Integer type and Buffer/Window/Tabpage types in
object_to_vim, msgpack_rpc helpers, and gendispatch.lua.
2016-11-02 10:06:27 -04:00
James McCoy
87ff2682d7
Use int as the standard type for boolean options.
All options are accessed by passing char_u pointers around, casting the
pointer to the right pointer type for the specific option, and then
dereferencing that pointer.

This dance works fine on little-endian systems when some bool options
are int types (as in Vim) and some are bool types (as would make more
sense), but on big-endian systems *(int *)varp when varp is pointing to
a bool will read random memory.

Therefore, all boolean options must remain a consistent type and int is
currently the easiest to choose.
2016-11-02 10:06:26 -04:00
Justin M. Keyes
349fa0048b Merge #5542 from mhinz/health/clipboard
CheckHealth: add clipboard check
2016-11-02 12:17:32 +01:00
Marco Hinz
3a802e3c16 provider/clipboard.vim: never show a warning on sourcing
Never throw an error when provider/clipboard.vim is sourced for the first time.
Save the error instead and expose it via `provider#clipboard#Error()`, mimicking
provider/python.vim.
2016-11-02 12:06:57 +01:00
Marco Hinz
797d72a997 health/provider.vim: add clipboard check 2016-11-02 12:06:54 +01:00
Marco Hinz
0a1b852cd1 provider/clipboard.vim: refactor 2016-11-02 12:04:43 +01:00
Justin M. Keyes
36c0ec6dd4 tui/suspend_event(): set STDIN to "blocking"
Set STDIN to blocking on "suspend". This was missed in #2598.
See 8a782f1699.

Also on "continue"/SIGCONT explicitly set STDIN back to NON-blocking.
We don't explicitly do that on *startup* (libuv does it internally),
but on "continue" it seems wise to restore libuv's expectations.

Closes #5459
2016-11-02 01:20:16 +01:00
James McCoy
d37c2e5dc2 doc: Remove features from vim_diff which have been merged upstream (#5554)
- TabNew/TabClosed: Introduced in 7.4.2075 and 7.4.2077
- hl-EndOfBuffer: Introduced in 7.4.2213
2016-10-31 17:26:32 +01:00
James McCoy
859e9fa65f Merge pull request #5527 from brcolow/vim-7.4.1730
vim-patch:7.4.[1730,1734,1741,1742,1779,1782]
2016-10-31 10:13:41 -04:00
Justin M. Keyes
ed198737fd Merge #5546 from justinmk/global-ctrlc
ex_global: Catch CTRL-C even if it is mapped.
2016-10-30 14:20:40 +01:00
Justin M. Keyes
49cf0b6ac4 lint 2016-10-30 13:10:29 +01:00
Justin M. Keyes
7cd204dbfa ex_global: Catch CTRL-C even if it is mapped.
Problem: If CTRL-C is mapped, it does not interrupt :global output.
Solution: clear `mapped_ctrl_c` during :global.
2016-10-30 13:10:29 +01:00
KillTheMule
e8c0f90962 Linted 2016-10-30 10:52:05 +01:00
KillTheMule
561c1e39e0 Tests for incsubstitution feature 2016-10-30 10:52:05 +01:00
KillTheMule
13841a56b4 Incsubsitution feature
Originally implemented by

* Clement0
* DesbyP
* aym7
* Adrey06
* Robinhola

in #4811. Major reworkings and bug
fixes by

* bfredl

Most tests suggested by ZyX-l, suggestions for improvements by oni-link.
2016-10-30 10:21:18 +01:00
James McCoy
6143cfdeec Merge pull request #5543 from jamessan/fix-shada-marks-test
test: Fix shada/marks_spec.lua failure
2016-10-29 12:49:50 -04:00
J Phani Mahesh
3d49fc1f50 CheckHealth: fix check for tmux escape-time (#5545) 2016-10-29 14:48:19 +02:00
James McCoy
93f3896300
test: Fix shada/marks_spec.lua failure
The 'dump and read back mark " from a closed tab' test needs to actually
create a second tab.  Since it wasn't doing so, the 'q!' command caused
nvim to exit and the subsequent 'qall' command fails.
2016-10-28 22:32:14 -04:00
Justin M. Keyes
242c3d5f0b version bump 2016-10-28 15:16:59 +02:00
Justin M. Keyes
cc1ec959f1 NVIM v0.1.6
FEATURES:
0b5a7e4ad5 #4432 API: external UIs can render custom popupmenu
c6ac4f84b1 #4934 API: call any API method from vimscript
31df051ed9 #4568 API: nvim_call_atomic(): multiple calls in a single request
b268ba353a #5424 API: nvim_win_get_number(), nvim_tabpage_get_number()
e7e2844d46 has("nvim-1.2.3") checks for a specific Nvim version
522b885a0d #5295, #5493 `:CheckHealth` checks tmux, terminfo, performance
719dae2e01 #5384 events: allow event processing in getchar()
f25797f869 #5386 API: metadata: Nvim version & API level
22dfe6925d #5389 API: metadata: "since", "deprecated_since"
605e74327a Added QuickFixLine highlight group

CHANGES:
4af6ec746c #5253 perf: Disable clipboard in do_cmdline()
6e9f329d05 #5299 perf: Skip foldUpdate() in insert-mode.
9d4fcec7c6 #5426 perf: Do not auto-update folds for some foldmethods.
eeec0cab58 #5419 tui: Default to normal-mode cursor shape.

FIXES:
e83845285c #5436 tui: Fix "weird characters" / "bleeding termcodes"
10a54ad12e #5243 signal_init: Always unblock SIGCHLD.
bccb49bedb #5316 eval.c: Fix memory leak for detached pty job
626065d385 #5227 tchdir: New tab should inherit CWD.
cd321b7d0f #5292 getcwd(): Return empty string if CWD is invalid.
6127eaef05 shada: Fix non-writeable ShaDa directory handling
ca65514a24 #2789 system(): Respect shellxescape, shellxquote
2daf54ee8d #4874 Restore vim-like tab dragging
0c536b5d8a #5319 syntax.c: Support bg/fg special color-names.
3c53371b0c #4972 from justinmk/schedule-ui_refresh
68bcb32ec4 #4789 tui.c: Do not wait for tui loop on teardown.
c8b6ec2e6a #5409 v:count broken in command-line window
6bc3bcefc6 #5461 fix emoji display
51937e1322 #5470 fix :terminal with :argadd, :argu
79d77da8a0 #5481 external UIs: opening multiple files from command-line
657ba62a84 #5501 rplugin: resolve paths in manifest file
6a6f188d2a #5502 system('foo &', 'bar'): Show error, don't crash.
1ff162c0d9 #5515 os_nodetype: open fd with O_NONBLOCK
2a6c5bb0c4 #5450 modeline: Handle version number overflow.
0ade1bb706 #5225 CI tests now run against Windows!
2016-10-28 15:16:45 +02:00