* Move db_uri setting to DbSettings
* WIP: sqlite crate framework
* WIP: Migrations
* WIP: sqlite implementation
* Add sqlite3 to Docker image
* verified_at needed for user query
* chore(deps): bump debian (#2772)
Bumps debian from bookworm-20250428-slim to bookworm-20250520-slim.
---
updated-dependencies:
- dependency-name: debian
dependency-version: bookworm-20250520-slim
dependency-type: direct:production
...
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
* fix(doctor): mention the required ble.sh version (#2774)
References:
https://forum.atuin.sh/t/1047
* fix: Don't print errors in `zsh_autosuggest` helper (#2780)
Previously, this would result in long multi-line errors when typing,
making it hard to see the shell prompt:
```
$ Error: could not load client settings
Caused by:
0: could not create config file
1: failed to create file `/home/jyn/.config/atuin/config.toml`
2: Required key not available (os error 126)
Location:
atuin-client/src/settings.rs:675:54
fError: could not load client settings
Caused by:
0: could not create config file
1: failed to create file `/home/jyn/.config/atuin/config.toml`
2: Required key not available (os error 126)
Location:
atuin-client/src/settings.rs:675:54
faError: could not load client settings
```
Silence these in autosuggestions, such that they only show up when
explicitly invoking atuin.
* fix: `atuin.nu` enchancements (#2778)
* PR feedback
* Remove sqlite3 package
* fix(search): prevent panic on malformed format strings (#2776) (#2777)
* fix(search): prevent panic on malformed format strings (#2776)
- Wrap format operations in panic catcher for graceful error handling
- Improve error messages with context-aware guidance for common issues
- Let runtime-format parser handle validation to avoid blocking valid formats
Fixes crash when using malformed format strings by catching formatting
errors gracefully and providing actionable guidance without restricting
legitimate format patterns like {command} or {time}.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
* Satisfy cargo fmt
* test(search): add regression tests for format string panic (#2776)
- Add test for malformed JSON format strings that previously caused panics
- Add test to ensure valid format strings continue to work
- Prevent future regressions of the format string panic issue
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
---------
Co-authored-by: Claude <noreply@anthropic.com>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Koichi Murase <myoga.murase@gmail.com>
Co-authored-by: jyn <github@jyn.dev>
Co-authored-by: Tyarel8 <98483313+Tyarel8@users.noreply.github.com>
Co-authored-by: Brian Cosgrove <cosgroveb@gmail.com>
Co-authored-by: Claude <noreply@anthropic.com>
* chore: update rust toolchain to 1.85
* nix things
* make clippy happy
I've replaced a bunch of &Option<String> with Option<String>.
They were not in hot loops, so a single clone is really no big deal +
keeps things simpler.
* fmt
These dependencies are unused in actual code, and the test I've removed
is a remnant from a move to use an external library -- it was useful to
show that the mechanical transformation was correct, but it's only
testing that library nowadays.
* feat: allow advertising a fake version to clients
The server usually runs unstable Atuin, and is well monitored. But let's
not advertised the unstable version to clients, as they will notify
users there is an update available.
* fix test build
* add verified column to users table
* add database functions to check if verified, or to verify
* getting there
* verification check
* use base64 urlsafe no pad
* add verification client
* clippy
* correct docs
* fix integration tests
* init daemon crate
* wip
* minimal functioning daemon, needs cleanup for sure
* better errors
* add signal cleanup
* logging
* things
* add sync worker
* move daemon crate
* 30s -> 5mins
* make clippy happy
* fix stuff maybe?
* fmt
* trim packages
* rate limit fix
* more protoc huh
* this makes no sense, why linux why
* can it install literally just curl
* windows in ci is slow, and all the newer things will not work there. disable the daemon feature and it will build
* add daemon feature
* maybe this
* ok wut where is protoc
* try setting protoc
* hm
* try copying protoc
* remove optional
* add cross config
* idk nix
* does nix want this?
* some random pkg I found does this
* uh oh
* hack, be gone!
* update contributing
P99 is usually <100ms which is excellent, but occasionally has big
spikes to 1000ms. This is only on the record index.
I don't want this to get out of hand. I've ran a few test queries and
they all complete very fast, and are purely index scans.
Hopefully this helps figure out if it's a specific user with tonnes of
stores or something? Otherwise there could be something up with my db.
I should probably also figure out some proper log levels or tracing lol.
