Commit Graph

32256 Commits

Author SHA1 Message Date
Cole Miller
f5f7eb0c70 wip it renders 2025-11-14 16:48:23 -05:00
cameron
b15c7be652 some doc comments 2025-11-14 15:13:09 +00:00
Cole Miller
6cde3b2502 get the fold map compiling 2025-11-14 02:15:58 -05:00
Cole Miller
c12c220d3e get the inlay map compiling 2025-11-14 00:50:31 -05:00
Cole Miller
744de9cefa finish implementing chunks 2025-11-13 22:51:06 -05:00
Cole Miller
a75230c63f implement some filtermap apis for the inlaymap to consume
Co-authored-by: cameron <cameron.studdstreet@gmail.com>
2025-11-13 19:40:00 -05:00
Cole Miller
9ecc2445e2 wip 2025-11-12 20:34:22 -05:00
Cole Miller
d3c1a5fe20 candidate fix for insertion halves of diff hunks going missing 2025-11-12 01:48:43 -05:00
Cole Miller
8dfe7ebdb6 wip 2025-11-11 00:43:08 -05:00
Cole Miller
a1a4e8ebe5 fix wrong diff hunk row ranges with adjacent hunks 2025-11-10 23:47:13 -05:00
Cole Miller
fc448d7d44 simplify and fix up edit expansion 2025-11-10 20:06:28 -05:00
Cole Miller
56c70e9d1c fix diff hunks disagreeing with transforms when an excerpt boundary
falls in the middle of an added line

Co-authored-by: Cameron <cameron@zed.dev>
2025-11-10 19:06:47 -05:00
Cole Miller
52cdc838f5 more 2025-11-09 16:33:04 -05:00
Cole Miller
6f3c81ab8c fix wrong changed range computation in buffer_diff 2025-11-09 15:28:53 -05:00
Cole Miller
6ab75d3d4b expand edit new ranges to the whole line 2025-11-09 15:24:37 -05:00
Cole Miller
34d2063456 progress on seed = 266 2025-11-08 16:25:40 -05:00
Cole Miller
bb56a17798 correct overshoot calculation for edit old range 2025-11-08 15:47:23 -05:00
Cole Miller
8c008cb9ae fix output edit overshoot calculation
Co-authored-by: cameron <cameron.studdstreet@gmail.com>
2025-11-07 09:54:06 -05:00
cameron
b2a8c08cef slightly improved debug outputs 2025-11-07 14:03:55 +00:00
Cole Miller
e080b3a75a add some assertions about the output edits 2025-11-07 00:09:18 -05:00
Cole Miller
f5557a1867 actually fix the coalescing logic 2025-11-06 22:29:08 -05:00
Cole Miller
11a356b6e6 fix coalescing logic for additional edits 2025-11-06 22:01:24 -05:00
cameron
3614b623b8 extend test to check case where next edit exhibits same wonky
intersection behaviour
2025-11-06 19:36:23 +00:00
Cole Miller
4e1b674e39 fix an issue with the multibuffer emitting incorrect edits when a diff
hunk is invalidated

Co-authored-by: cameron <cameron.studdstreet@gmail.com>
2025-11-06 13:29:50 -05:00
Cole Miller
532dce97f1 rework expected text/summary calculation to use row infos, first fails
for seed = 111
2025-11-04 20:14:20 -05:00
Cole Miller
97ea848ac0 checkpoint, passing up to seed = 13
Co-authored-by: cameron <cameron.studdstreet@gmail.com>
2025-11-04 17:04:30 -05:00
Cole Miller
49ebf59d05 wip turns out the test wasn't working
Co-authored-by: cameron <cameron.studdstreet@gmail.com>
2025-11-03 18:48:31 -05:00
Cole Miller
20b457de2a get randomized test passing for many iterations 🎉
Co-authored-by: cameron <cameron.studdstreet@gmail.com>
2025-11-03 16:38:10 -05:00
Cole Miller
a63fada5f5 expand all diff hunks and implement todo 2025-11-02 23:49:46 -05:00
Cole Miller
23adeab8d2 wip
Co-authored-by: Cameron <cameron@zed.dev>
2025-11-02 23:38:10 -05:00
Cole Miller
6e0f65e05a sketch in split editor and start working on new logic in block map
Co-authored-by: cameron <cameron.studdstreet@gmail.com>
2025-10-29 21:30:16 -04:00
Adam Richardson
370d4ce200 rope: Micro optimize the creation of masks (#41132)
Using compiler explorer I saw that the compiler wasn't clever enough to
optimise away the branches in the masking code. I thought the compiler
would have a better chance if we always branched, which [turned out to
be the case](https://godbolt.org/z/PM594Pz18).

Running the benchmarks the biggest benefit I saw was:
```
push/65536              time:   [2.9067 ms 2.9243 ms 2.9417 ms]
                        thrpt:  [21.246 MiB/s 21.373 MiB/s 21.502 MiB/s]
                 change:
                        time:   [-8.3452% -7.2617% -6.2009%] (p = 0.00 < 0.05)
                        thrpt:  [+6.6108% +7.8303% +9.1050%]
                        Performance has improved.
```
But I did also see some regressions:
```
slice/4096              time:   [66.195 µs 66.815 µs 67.448 µs]
                        thrpt:  [57.915 MiB/s 58.464 MiB/s 59.012 MiB/s]
                 change:
                        time:   [+3.7131% +5.1698% +6.6971%] (p = 0.00 < 0.05)
                        thrpt:  [-6.2768% -4.9157% -3.5802%]
                        Performance has regressed.
```

Release Notes:

- N/A
2025-10-27 16:16:16 +01:00
Richard Feldman
2284131bfc Add @rtfeldman to reviewers list (#41127)
Closes #ISSUE

Release Notes:

- N/A *or* Added/Fixed/Improved ...
2025-10-27 11:13:12 -04:00
Ben Kunkle
2f7045f724 settings_ui: Show migration banner (#41112)
Closes #ISSUE

Release Notes:

- N/A *or* Added/Fixed/Improved ...

---------

Co-authored-by: Danilo <danilo@zed.dev>
2025-10-27 10:13:26 -04:00
Jakub Konka
5d359ea2f2 remote: Fall back to SCP if SFTP fails (#41255)
Fixes https://github.com/zed-industries/zed/issues/41260

After experimenting and reading through the implementation of OpenSSH
stack on Windows, it looks like batch mode precludes use of passwords.
In the listing
b8c08ef9da/sshconnect2.c (L417),
the last field of each `Authmode` struct is a pointer to the config
value that *disables* that particular mode. In this case, `keyboard`
(interactive) and `password` modes are both disabled if batch mode is
used. We should therefore fall back to `scp` if `sftp` fails rather than
to fail outright.

Release Notes:

- N/A
2025-10-27 15:01:13 +01:00
Ben Kunkle
72c6a74505 keymap_editor: Fix updating empty keymap (#40909)
Closes #40898

Release Notes:

- Fixed an issue where attempting to add or update a key binding in the
keymap editor with an empty `keymap.json` file would fail
2025-10-27 09:56:54 -04:00
Alvaro Parker
941033e373 settings_ui: Add vim motions on navigation menu (#39988)
Closes #ISSUE

Release Notes:

- Added vim motions on settings navigation menu
2025-10-27 09:53:37 -04:00
Piotr Osiewicz
83884ca36f lsp: Support tracking multiple registrations of diagnostic providers (#41096)
Closes #40966
Closes #41195
Closes #40980 

Release Notes:

- Fixed diagnostics not working with basedpyright/pyright beyond an
initial version of the document

---------

Co-authored-by: Smit Barmase <heysmitbarmase@gmail.com>
Co-authored-by: Anthony Eid <hello@anthonyeid.me>
Co-authored-by: dino <dinojoaocosta@gmail.com>
2025-10-27 14:52:23 +01:00
Kirill Bulatov
f503c65924 Make "About Zed" menu entry to work when all Zed windows are closed (#41272)
Closes https://github.com/zed-industries/zed/issues/41267

The only downside would be the fact that a Zed window will appear behing
the version modal, but this is a limitation for all "window-less"
actions currently.

Release Notes:

- Made "About Zed" menu entry to work when all Zed windows are closed
2025-10-27 11:33:50 +00:00
Lukas Wirth
c1cd371786 fix: Edit predictions using stale snapshot (#41271)
Release Notes:

- N/A *or* Added/Fixed/Improved ...

Co-authored-by: David Kleingeld <davidsk@zed.dev>
2025-10-27 11:00:37 +00:00
Ben Brandt
7cb2d83608 acp: Start sending Client Info to the Agent (#41265)
Updates to acp crate 0.7, which allows us to send information about the
client to the Agent.
In the future, we can also use the AgentInfo on the response for
internal metrics.

Release Notes:

- N/A
2025-10-27 10:05:50 +00:00
Lukas Wirth
ae3abf50d8 editor: Fix panics in CursorPosition::update_position (#41237)
Fixes a regression introduced in
https://github.com/zed-industries/zed/pull/39857. As for the exact
reason this causes this issue I am not yet sure will investigate (as per
the todos in code)

Fixes ZED-23R

Release Notes:

- N/A *or* Added/Fixed/Improved ...
2025-10-27 08:26:39 +00:00
deltamaya
edf2ec7d4c editor: Make hover popover delay strictly respect hover_popover_delay setting (#41149)
Previously, the hover popover delay was implemented using two
overlapping timers, which caused the minimum delay to always be at least
HOVER_REQUEST_DELAY_MILLIS, regardless of the hover_popover_delay
setting.
This change updates the logic to wait for hover_popover_delay only,
ensuring the total delay is always equals to hover_popover_delay . As a
result, the hover popover now appears after the intended delay, matching
the user's configuration more accurately.

Release Notes:

- Improved hover popover respecting settings delay correctly
2025-10-27 08:01:43 +00:00
Marshall Bowers
2919e1976a docs: Display action names in backticks instead of quotes (#41248)
This PR fixes some instances where we were displaying action names in
quotes instead of in backticks.

Also fixed some mentions of using an action to open the settings file,
as this has changed after the release of the settings UI.

Release Notes:

- N/A
2025-10-27 00:50:31 +00:00
Lukas Wirth
fd3ca0303f workspace: Handle non-cloneable items better (#41215)
When trying to split and clone a non clone-able workspace item we now
attempt split and move instead of doing nothing. Additionally we disable
the split menu buttons if we can't split the active item at all.

Release Notes:

- Improved handling of unsplittable panes
2025-10-26 13:24:26 +00:00
Lukas Wirth
61e4a1d16a settings: Fix out of bounds index (#41227)
Fixes ZED-2J8

Release Notes:

- N/A *or* Added/Fixed/Improved ...
2025-10-26 12:57:44 +00:00
Lukas Wirth
2471ae451c Pre-initialize global rayon threadpool (#41226)
We only use it a handful of times and the default amount of threads
(logical cpu core number) its spawns is overkill for this. This also
gives the threads names oppose to being labeled `<unknown>`

Release Notes:

- N/A *or* Added/Fixed/Improved ...
2025-10-26 12:44:34 +00:00
Lukas Wirth
d6b31d8932 gpui: Fix TextLayout::layout producing invalid text runs (#41224)
The issues is that the closure supplied to `request_measured_layout`
could be run multiple times with differing `known_dimensions`. This in
turn will mean we truncate the font runs once, inserting a multibyte
char at the end but then in the next iteration use those truncated runs
for possible the original untruncated string which has multibyte
characters overlapping the now truncated run end resulting in a faulty
utf8 boundary index.

Solution to this is simple, truncate a clone of the runs when needed
instead of modifying the original.

Fixes https://github.com/zed-industries/zed/issues/36925
Fixed ZED-2FF
Fixes ZED-2KM
Fixes ZED-2KK
Fixes ZED-1FF
Fixes ZED-255
Fixes ZED-2JD
Fixes ZED-2FX
Fixes ZED-2K2
Fixes ZED-2JX
Fixes ZED-2GE
Fixes ZED-2FC
Fixes ZED-2GD
Fixes ZED-2HY
Fixes ZED-2HR
Fixes ZED-2FN
Fixes ZED-2GT
Fixes ZED-2FK
Fixes ZED-2EY
Fixes ZED-27E
Fixes ZED-272
Fixes ZED-2EM
Fixes ZED-2CC
Fixes ZED-29V
Fixes ZED-25B
Fixes ZED-257
Fixes ZED-24R
Fixes ZED-24Q
Fixes ZED-23Z
Fixes ZED-227

Release Notes:

- Fixed a crash in text shaping when truncating rendered text
2025-10-26 13:00:52 +01:00
Lukas Wirth
95ad7a6cae gpui: Drop unnecessary use of StringIndexConverter (#41221)
Might help with ZED-1FF

Release Notes:

- N/A *or* Added/Fixed/Improved ...
2025-10-26 10:35:33 +00:00
Kirill Bulatov
54a7da364c Fix task terminal split (#41218)
Re-lands https://github.com/zed-industries/zed/pull/40824

Release Notes:

- N/A
2025-10-26 10:01:40 +00:00