mirror of
https://github.com/zellij-org/zellij.git
synced 2024-11-27 14:34:32 +03:00
8504881e4e
13 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
Aidan Hobson Sayers
|
8fd7a67898
|
fix(terminal): improve reflow performance and general memory usage by reducing size of TerminalCharacter (#3043)
* refactor: Simplify transfer_rows_from_viewport_to_lines_above next_lines is always consolidated to a single Row, which immediately gets removed - we can remove some dead code as a result * perf: Batch remove rows from the viewport for performance Given a 1MB line catted into the terminal, a toggle-fullscreen + toggle-fullscreen + close-pane + `run true` goes from ~9s to ~3s * perf: Optimize Row::drain_until by splitting chars in one step Given a 10MB line catted into the terminal, a toggle-fullscreen + toggle-fullscreen + close-pane + `run true` goes from ~23s to ~20s * refactor: Simplify `if let` into a `.map` * refactor: There are only new saved coordinates when there were old ones * refactor: Unify viewport transfer: use common variable names * fix: Use same saved cursor logic in height resize as width See #2182 for original introduction that only added it in one branch, this fixes an issue where the saved cursor was incorrectly reset when the real cursor was * fix: Correct saved+real cursor calculations when reflowing long lines * fix: Don't create canonical lines if cursor ends on EOL after resize Previously if a 20 character line were split into two 10 character lines, the cursor would be placed on the line after the two lines. New characters would then be treated as a new canonical line. This commit fixes this by biasing cursors to the end of the previous line. * fix: for cursor index calculation in lines that are already wrapped * chore: test for real/saved cursor position being handled separately * chore: Apply cargo format * refactor: Unify viewport transfer: transfer + cursor update together * perf: Reduce size of TerminalCharacter from 72 to 60 bytes With a 10MB single line catted into a fresh terminal, VmRSS goes from 964092 kB to 874020 kB (as reported by /proc/<pid>/status) before/after this patch Given a 10MB line catted into the terminal, a toggle-fullscreen + toggle-fullscreen + close-pane + `run true` goes from ~15s to ~12.5s * fix(build): Don't unconditionally rebuild zellij-utils * refactor: Remove Copy impl on TerminalCharacter * perf: Rc styles to reduce TerminalCharacter from 60 to 24 bytes With a 10MB single line catted into a fresh terminal, VmRSS goes from 845156 kB to 478396 kB (as reported by /proc/<pid>/status) before/after this patch Given a 10MB line catted into the terminal, a toggle-fullscreen + toggle-fullscreen + close-pane + `run true` goes from ~12.5s to ~7s * perf: Remove RcCharacterStyles::Default, allow enum niche optimisation This reduces TerminalCharacter from 24 to 16 bytes With a 10MB single line catted into a fresh terminal, VmRSS goes from 478396 kB to 398108 kB (as reported by /proc/<pid>/status) before/after this patch Given a 10MB line catted into the terminal, a toggle-fullscreen + toggle-fullscreen + close-pane + `run true` goes from ~7s to ~4.75s * docs: link anchor omission from reset_all is deliberate reset_all is only used from ansi params, and ansi params don't control link anchor * fix: Remove no-op on variable that gets immediately dropped * refactor: Simplify replace_character_at logic The original condition checked absolute_x_index was in bounds, then used the index to manipulate it. This is equivalent to getting a ref to the character at that position and manipulating directly * chore: Run xtask format * chore(repo): update issue templates * Bump rust version to 1.75.0 (#3039) * rust-toolchain: Bump toolchain version to 1.69.0 which, compared to the previous 1.67.0, has the following impacts on `zellij`: - [Turn off debuginfo for build deps][2]: Increases build time (on my machine) from ~230 s in 1.67.0 to ~250 s now, *which is unexpected* This version also changes [handling of the `default-features` flag][3] when specifying dependencies in `Cargo.toml`. If a dependent crate requires `default-features = true` on a crate that is required as `default-features = false` further up the dependency tree, the `true` setting "wins". We only specify `default-features = false` for three crates total: - `names`: This is used only by us - `surf`: This is used only by us - `vte`: This is also required by `strip-ansi-escapes`, but that has `default-features = false` as well How this affects our transitive dependencies is unknown at this point. [2]: https://github.com/rust-lang/cargo/pull/11252/ [3]: https://github.com/rust-lang/cargo/pull/11409/ * rust-toolchain: Bump toolchain version to 1.70.0 which, compared to the previous 1.69.0, as the following impacts on `zellij`: 1. [Enable sparse registry checkout for crates.io by default][1] This drastically increases the time to first build on a fresh rust installation/a rust installation with a clean cargo registry cache. Previously it took about 75s to populate the deps/cache (with `cargo fetch --locked` and ~100 MBit/s network), whereas now the same process takes ~10 s. 2. [The `OnceCell` type is now part of std][2] In theory, this would allow us to cut a dependency from `zellij-utils`, but the `once_cell` crate is pulled in by another 16 deps, so there's no point in attempting it right now. Build times and binary sizes are unaffected by this change compared to the previous 1.69.0 toolchain. [1]: https://github.com/rust-lang/cargo/pull/11791/ [2]: https://doc.rust-lang.org/stable/std/cell/struct.OnceCell.html * rust-toolchain: Bump toolchain version to 1.75.0 which, compared to the previous 1.70.0, has the following impacts on `zellij`: 1. [cross-crate inlining][8] This should increase application performance, as functions can now be inlined across crates. 2. [`async fn` in traits][9] This would allow us to drop the `async_trait` dependency, but it is currently still required by 3 other dependencies. Build time in debug mode (on my own PC) is cut down from 256s to 189s (for a clean build). Build time in release mode is cut down from 473s to 391s (for a clean build). Binary sizes only change minimally (825 MB -> 807 MB in debug, 29 MB -> 30 MB in release). [8]: https://github.com/rust-lang/rust/pull/116505 [9]: https://github.com/rust-lang/rust/pull/115822/ * chore: Apply rustfmt. * CHANGELOG: Add PR #3039. * feat(plugins): introduce 'pipes', allowing users to pipe data to and control plugins from the command line (#3066) * prototype - working with message from the cli * prototype - pipe from the CLI to plugins * prototype - pipe from the CLI to plugins and back again * prototype - working with better cli interface * prototype - working after removing unused stuff * prototype - working with launching plugin if it is not launched, also fixed event ordering * refactor: change message to cli-message * prototype - allow plugins to send messages to each other * fix: allow cli messages to send plugin parameters (and implement backpressure) * fix: use input_pipe_id to identify cli pipes instead of their message name * fix: come cleanups and add skip_cache parameter * fix: pipe/client-server communication robustness * fix: leaking messages between plugins while loading * feat: allow plugins to specify how a new plugin instance is launched when sending messages * fix: add permissions * refactor: adjust cli api * fix: improve cli plugin loading error messages * docs: cli pipe * fix: take plugin configuration into account when messaging between plugins * refactor: pipe message protobuf interface * refactor: update(event) -> pipe * refactor - rename CliMessage to CliPipe * fix: add is_private to pipes and change some naming * refactor - cli client * refactor: various cleanups * style(fmt): rustfmt * fix(pipes): backpressure across multiple plugins * style: some cleanups * style(fmt): rustfmt * style: fix merge conflict mistake * style(wording): clarify pipe permission * docs(changelog): introduce pipes * xtask: Disable pusing during publish (#3040) * xtask: Add `--no-push` flag to `publish` which can be used when simulating releases to work without a writable git fork of the zellij code. * xtask: Fix borrow issues * xtask/pipe: Require lockfile in publish to avoid errors from invalid dependency versions. * CHANGELOG: Add PR #3040. * fix(terminal): some real/saved cursor bugs during resize (#3032) * refactor: Simplify transfer_rows_from_viewport_to_lines_above next_lines is always consolidated to a single Row, which immediately gets removed - we can remove some dead code as a result * perf: Batch remove rows from the viewport for performance Given a 1MB line catted into the terminal, a toggle-fullscreen + toggle-fullscreen + close-pane + `run true` goes from ~9s to ~3s * perf: Optimize Row::drain_until by splitting chars in one step Given a 10MB line catted into the terminal, a toggle-fullscreen + toggle-fullscreen + close-pane + `run true` goes from ~23s to ~20s * refactor: Simplify `if let` into a `.map` * refactor: There are only new saved coordinates when there were old ones * refactor: Unify viewport transfer: use common variable names * fix: Use same saved cursor logic in height resize as width See #2182 for original introduction that only added it in one branch, this fixes an issue where the saved cursor was incorrectly reset when the real cursor was * fix: Correct saved+real cursor calculations when reflowing long lines * fix: Don't create canonical lines if cursor ends on EOL after resize Previously if a 20 character line were split into two 10 character lines, the cursor would be placed on the line after the two lines. New characters would then be treated as a new canonical line. This commit fixes this by biasing cursors to the end of the previous line. * fix: for cursor index calculation in lines that are already wrapped * chore: test for real/saved cursor position being handled separately * chore: Apply cargo format * chore(repo): update issue templates * Bump rust version to 1.75.0 (#3039) * rust-toolchain: Bump toolchain version to 1.69.0 which, compared to the previous 1.67.0, has the following impacts on `zellij`: - [Turn off debuginfo for build deps][2]: Increases build time (on my machine) from ~230 s in 1.67.0 to ~250 s now, *which is unexpected* This version also changes [handling of the `default-features` flag][3] when specifying dependencies in `Cargo.toml`. If a dependent crate requires `default-features = true` on a crate that is required as `default-features = false` further up the dependency tree, the `true` setting "wins". We only specify `default-features = false` for three crates total: - `names`: This is used only by us - `surf`: This is used only by us - `vte`: This is also required by `strip-ansi-escapes`, but that has `default-features = false` as well How this affects our transitive dependencies is unknown at this point. [2]: https://github.com/rust-lang/cargo/pull/11252/ [3]: https://github.com/rust-lang/cargo/pull/11409/ * rust-toolchain: Bump toolchain version to 1.70.0 which, compared to the previous 1.69.0, as the following impacts on `zellij`: 1. [Enable sparse registry checkout for crates.io by default][1] This drastically increases the time to first build on a fresh rust installation/a rust installation with a clean cargo registry cache. Previously it took about 75s to populate the deps/cache (with `cargo fetch --locked` and ~100 MBit/s network), whereas now the same process takes ~10 s. 2. [The `OnceCell` type is now part of std][2] In theory, this would allow us to cut a dependency from `zellij-utils`, but the `once_cell` crate is pulled in by another 16 deps, so there's no point in attempting it right now. Build times and binary sizes are unaffected by this change compared to the previous 1.69.0 toolchain. [1]: https://github.com/rust-lang/cargo/pull/11791/ [2]: https://doc.rust-lang.org/stable/std/cell/struct.OnceCell.html * rust-toolchain: Bump toolchain version to 1.75.0 which, compared to the previous 1.70.0, has the following impacts on `zellij`: 1. [cross-crate inlining][8] This should increase application performance, as functions can now be inlined across crates. 2. [`async fn` in traits][9] This would allow us to drop the `async_trait` dependency, but it is currently still required by 3 other dependencies. Build time in debug mode (on my own PC) is cut down from 256s to 189s (for a clean build). Build time in release mode is cut down from 473s to 391s (for a clean build). Binary sizes only change minimally (825 MB -> 807 MB in debug, 29 MB -> 30 MB in release). [8]: https://github.com/rust-lang/rust/pull/116505 [9]: https://github.com/rust-lang/rust/pull/115822/ * chore: Apply rustfmt. * CHANGELOG: Add PR #3039. * feat(plugins): introduce 'pipes', allowing users to pipe data to and control plugins from the command line (#3066) * prototype - working with message from the cli * prototype - pipe from the CLI to plugins * prototype - pipe from the CLI to plugins and back again * prototype - working with better cli interface * prototype - working after removing unused stuff * prototype - working with launching plugin if it is not launched, also fixed event ordering * refactor: change message to cli-message * prototype - allow plugins to send messages to each other * fix: allow cli messages to send plugin parameters (and implement backpressure) * fix: use input_pipe_id to identify cli pipes instead of their message name * fix: come cleanups and add skip_cache parameter * fix: pipe/client-server communication robustness * fix: leaking messages between plugins while loading * feat: allow plugins to specify how a new plugin instance is launched when sending messages * fix: add permissions * refactor: adjust cli api * fix: improve cli plugin loading error messages * docs: cli pipe * fix: take plugin configuration into account when messaging between plugins * refactor: pipe message protobuf interface * refactor: update(event) -> pipe * refactor - rename CliMessage to CliPipe * fix: add is_private to pipes and change some naming * refactor - cli client * refactor: various cleanups * style(fmt): rustfmt * fix(pipes): backpressure across multiple plugins * style: some cleanups * style(fmt): rustfmt * style: fix merge conflict mistake * style(wording): clarify pipe permission * docs(changelog): introduce pipes * fix: add some robustness and future proofing * fix e2e tests --------- Co-authored-by: Aram Drevekenin <aram@poor.dev> Co-authored-by: har7an <99636919+har7an@users.noreply.github.com> * fix integer overflow again (oops) --------- Co-authored-by: Aram Drevekenin <aram@poor.dev> Co-authored-by: har7an <99636919+har7an@users.noreply.github.com> |
||
har7an
|
592cabeda8
|
xtask: Disable pusing during publish (#3040)
* xtask: Add `--no-push` flag to `publish` which can be used when simulating releases to work without a writable git fork of the zellij code. * xtask: Fix borrow issues * xtask/pipe: Require lockfile in publish to avoid errors from invalid dependency versions. * CHANGELOG: Add PR #3040. |
||
|
har7an
|
8031d6bf64
|
xtask/pipeline: Fix publish task (#2711)
* xtask/pipeline: Fix publish task which was previously stuck in an infinite loop after successfully publishing a crate. The error originated in the code only checking for error conditions but not breaking out of the inner infinite loop in case of success. * xtask: Improve publish failure UX by offering the user more actions to choose from when an error occured. * utils/assets: Add generated prost files to assets to make sure they're available at build time and are picked up by all components. It seems we hit some strange bug with the build script where, when running `cargo publish --dry-run` the build script **is not** run before regularly compiling zellij-utils. This shouldn't happen according to the docs, but I cannot explain what's causing it. So we're using this as a workaround for now to make a smooth release. * xtask: Prevent accidental git commit deletion when dry-running a publish. * utils: Add comments to protobuf-related code to explain why these changes were performed. The comments all include a link to an issue comment explaining the situation in greater detail. * xtask: Build protobuf definitions when building any part of the project, similar to how we build the plugins when required. This should ensure that all crates built through `cargo xtask` (which is the officially supported build method) will receive up-to-date protobuf definitions. |
||
Aram Drevekenin
|
bc628abc12
|
feat(sessions): add a session manager to switch between sessions, tabs and panes and create new ones (#2721)
* write/read session metadata to disk for all sessions * switch session client side * fix tests * various adjustments * fix full screen focus bug in tiled panes * fix tests * fix permission sorting issue * cleanups * add session manager * fix tests * various cleanups * style(fmt): rustfmt * clear screen before switching sessions * I hate you clippy * truncate controls line to width * version session cache * attempt to fix plugin tests * style(fmt): rustfmt * another attempt to fix the tests in the ci |
||
|
Aram Drevekenin
|
5fe4d60c22
|
feat(plugins): Plugin workers and strider (#2449)
* mvp of strider fuzzy find * improve search ui * various refactoringz * moar refactoring * even more refactoring * tests and more refactoring * refactor: remove unused stuff * style(fmt): rustfmt * debug ci * debug ci * correct path for plugin system tests * fix plugin system ci tests * remove debugging statements from test * fix plugin worker persistence * add test for plugin worker persistence * style(fmt): rustfmt * final cleanups * remove outdated comments |
||
|
har7an
|
30d0cffa42
|
Use rust 1.67 (#2375)
* rust: Update toolchain version to 1.67 * xtask/pipeline/publish: Drop manual "wait" for crates.io to catch up, which is obsolete with rust 1.66 and up. Cargo does that on its own now. See https://github.com/rust-lang/cargo/pull/11062 * xtask: Add function to obtain asset_dir instead of assembling it on demand throughout the codebase. * xtask/run: Add '--quick-run' flag as a convenient shorthand for `cargo xtask run --data-dir $PROJECT_ROOT/zellij-utils/assets`. * cargo: Add 'q' command alias as a shorthand for 'cargo xtask run --quick-run' * cargo: Update thiserror to 1.0.40 * cargo: Update anyhow to 1.0.70 and specify dependency only once inside `zellij-utils`, not inside the zellij root crate. * cargo: Update names to 0.14.0 * cargo: Update miette to 5.7.0 and re-export the dependency from zellij-utils, to avoid duplicate (incompatible) includes from inside zellij-utils and the root crate. * cargo: Update dialoguer to 0.10.4 * fix formatting * changelog: Add PR #2375 |
||
Thomas Linford
|
f598ca3738
|
improve build/ci times (#2396)
- avoid building all workspace crates with `cargo x build` (only plugins and main binary) - only set the target triple in tests for plugins - add new profile for `cargo x run` to build with optimized dependencies => FAST plugins when developing (thanks [Bevy Book](https://bevyengine.org/learn/book/getting-started/setup/#compile-with-performance-optimizations) for the idea) - use https://github.com/Swatinem/rust-cache to avoid rebuilding dependencies every time in ci - split `Build & Test` job into two so they run in parallel - hopefully improve the flaky tests situation, this also makes the e2e tests run much faster (some tests produced correct snapshots but had some logic errors causing them to loop for much longer than necessary). Add some output to the tests so it is easier to see if something goes wrong. - remove verbose build output from e2e test build |
||
|
Aram Drevekenin
|
4a32547fd4 | hotfix: fix assets | ||
|
har7an
|
46b9bc755e
|
Feature: simulate publishing (#2194)
* xtask: Add arguments to `publish` that allow specifying a custom git remote to push to and a custom cargo registry to publish packages to. * xtask/publish: Don't release `xtask` subcrate because it's not meant to be released at all. * xtask/publish: Add status messages to publish so we see what crate is currently being published, too. * xtask/publish: Disable default features on `zellij` because otherwise it tries to pick up the debug builds of the plugins, which aren't part of released version of `zellij utils`. * xtask/publish: Fix handling of custom registry * docs: Add `RELEASE.md` which explains how to simulate a zellij release. * xtask: Apply rustfmt * xtask: Remove `wasm-opt` from build steps because recent versions cause havoc in the release process in GitHub pipelines and it's primary goal is to only reduce binary size. Current rust versions seem to produce very compact wasm binaries themselves, though. * .github: Don't install wasm-opt in workflows |
||
m-lima
|
d1b458e37b
|
chore(build): make the singlepass compiler opt-in through flags (#2146) | ||
|
m-lima
|
5b3e1ecacd
|
chore(development): use singlepass in debug mode (#2134)
* Add new feature flags * Use singlepass in debug mode * Use Cow to avoid unnecessary copies - Instead of removing and reinserting into the memory cache, use Cow to model both owned an borrowed data - Log at debug-level the time to compile/load a wasm module - A little clippy drive-by on touched files * Satisfy the assumption from zellij-utils/src/consts.rs for target-dir * Allow forcing cranlift in debug mode * Remove deprecated comments * PR comment: typo * Remove extras |
||
|
har7an
|
beddfb77a8
|
Improve client disconnect handling (#2068)
* xtask/run: Use varargs when run with `-data-dir` Previously any additional arguments passed on the command line were ignored. Now they are appended to `cargo run ...` as documented. * server/os_i_o: Improve error message when IPC dies and display the last send/recv error to the user instead of a generic "Buffer full" message. * server/lib: Log error in `send_to_client!` so we will know when an error occured while trying to send a message to the client. The most likely cause for this is that the client buffer filled up and hence we cannot send any new messages. While we still disconnect the client as before, now we also write a log message that explains the situation. * utils/channel: Apply rustfmt * server/lib: Detect when client responds too slow and log a message before disconnecting it. * server/os_i_o: Add retry queue to client senders that is dynamically allocated on-demand and stores `ServerToClientMsg` in case the regular IPC channel is currently full. This acts as a dynamic buffer to hold and buffer messages for a while until the client hopefully catches up. Also write a message to the log to indicate when the client is recognized to be too slow in handling server messages. * server: apply rustfmt * utils/ipc: Add session name to "Disconnect" error * utils/ipc: Fix error message indent * server/os_i_o: Undo IPC channel extension via `Vec` and drastically increase the IPC message queue size instead. Measurements didn't discover a drastic increase in RAM caused by this, and it is a much easier fix for the problem at hand. * CHANGELOG: Add PR #2068 |
||
|
har7an
|
d1f50150f6
|
WIP: Use xtask as build system (#2012)
* xtask: Implement a new build system xtask is a cargo alias that is used to extend the cargo build system with custom commands. For an introduction to xtask, see here: https://github.com/matklad/cargo-xtask/ The idea is that instead of writing makefiles, xtask requires no additional dependencies except `cargo` and `rustc`, which must be available to build the project anyway. This commit provides a basic implementation of the `build` and `test` subcommands. * xtask/deps: Add 'which' * xtask/test: Handle error when cargo not found * xtask/flags: Add more commands to perform different useful tasks. Includes: - clippy - format - "make" (composite) - "install" (composite) Also add more options to `build` to selectively compile plugins or leave them out entirely. * xtask/main: Return error when cargo not found * xtask/build: Add more subtasks - `wasm_opt_plugins` and - `manpage` that perform other build commands. Add thorough documentation on what each of these does and also handle the new `build` cli flags appropriately. * xtask/clippy: Add job to run clippy * xtask/format: Add job to run rustfmt * xtask/pipeline: Add composite commands that perform multiple atomic xtask commands sequentially in a pipeline sort of fashion. * xtask/deps: Pin dependencies * xtask/main: Integrate new jobs and add documentation. * xtask: Implement 'dist' which performs an 'install' and copies the resulting zellij binary along with some other assets to a `target/dist` folder. * cargo: Update xflags version * xtask: Measure task time, update tty title * xtask: Update various tasks * xtask: wasm-opt plugins in release builds automatically. * xtask/build: Copy debug plugins to assets folder * xtask: Add 'run' subcommand * xtask: Add arbitrary args to test and run * xtask: Rearrange CLI commands in help * xtask: Add deprecation notice * docs: Replace `cargo make` with `xtask` * github: Use `xtask` in workflows. * xtask: Add support for CI commands * xtask: Streamline error handling * github: Use new xtask commands in CI * xtask: Add 'publish' job * xtask/publish: Add retry when publish fails * xtask: Apply rustfmt * xtask: Refine 'make' deprecation warning * xtask: add task to build manpage * contributing: Fix e2e commands * xtask/run: Add missing `--` to pass all arguments following `xtask run` directly to the zellij binary being run. * xtask: Stay in invocation dir and make all tasks that need it change to the project root dir themselves. * xtask/run: Add `--data-dir` flag which will allow very quick iterations when not changing the plugins between builds. * xtask/ci: Install dependencies without asking * utils: Allow including plugins from target folder * utils/assets: Reduce asset map complexity * utils/consts: Update asset map docs * xtask: Fix plugin includes * xtask/test: Build plugins first because the zellij binary needs to include the plugins. * xtask/test: Fix formatting * xtask: Add notice on how to disable it |