Merged
Conversation
📝 WalkthroughWalkthroughAdds an advanced multi-step wizard to the TUI: Architecture, Inject, Init, and Compression screens, new WizardMode (Quick/Advanced), BuildConfig extensions (injections, init_mode), screen sync/validation, updated build/CLI generation to include injections and custom init, and per-screen event handlers. Changes
sequenceDiagram
participant User
participant App
participant Screens as Arch/Inject/Init/Compress
participant Config as BuildConfig
participant Builder as InitramfsBuilder
User->>App: Start TUI
App->>App: Default Quick flow (Language → Image → Summary)
User->>App: Press 'a' (enter advanced)
App->>App: enter_advanced_mode()
App->>Screens: Show Architecture
User->>Screens: Configure Arch, Inject, Init, Compression
Screens->>Config: sync fields (arch, injections, init_mode, compression)
User->>App: Proceed to Summary / Build
App->>Config: validate config
App->>Builder: start_build(image, arch, injections, init_mode, compression)
Builder-->>App: build complete / result
Estimated code review effort🎯 4 (Complex) | ⏱️ ~45 minutes Poem
🚥 Pre-merge checks | ❌ 3❌ Failed checks (1 warning, 2 inconclusive)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches
🧪 Generate unit tests (beta)
📝 Coding Plan
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 6
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (2)
src/tui/app.rs (2)
127-149:⚠️ Potential issue | 🟠 MajorPersist screen state on every non-forward transition.
Line 127 and Line 145 switch screens without the
sync_screen_on_exit()step thatnext_screen()already uses. In the advanced flow that can silently drop edits: a custom image entered beforeenter_advanced_mode()stays stale inconfig, and an architecture change can be reset after Back → forward becausesync_screen_on_enter()reloads it from the old config value.Suggested fix
pub fn prev_screen(&mut self) { self.validation_error = None; + self.sync_screen_on_exit(); self.screen = match self.screen { Screen::Language => Screen::Language, Screen::Image => Screen::Language, Screen::Architecture => Screen::Image, Screen::Inject => Screen::Architecture, @@ } self.sync_screen_on_enter(); } pub fn enter_advanced_mode(&mut self) { + self.sync_screen_on_exit(); self.mode = WizardMode::Advanced; self.screen = Screen::Architecture; self.sync_screen_on_enter(); }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/tui/app.rs` around lines 127 - 149, The prev_screen and enter_advanced_mode flows are switching screens without persisting current edits; call self.sync_screen_on_exit() before modifying self.screen (and before changing self.mode in enter_advanced_mode) so the current UI state is saved to config like next_screen() does; update prev_screen to invoke self.sync_screen_on_exit() immediately before the match that assigns self.screen, and update enter_advanced_mode to call self.sync_screen_on_exit() before setting self.mode = WizardMode::Advanced and self.screen = Screen::Architecture.
1-297:⚠️ Potential issue | 🟡 MinorRun
cargo fmton this file before merging.CI is already failing on
src/tui/app.rsfor rustfmt, so this needs formatting cleanup before the PR can land.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/tui/app.rs` around lines 1 - 297, The file fails rustfmt check; run "cargo fmt" (or rustfmt) on the repository to reformat src/tui/app.rs (which contains App, BuildConfig, detect_host_arch, start_build, generate_cli_command, etc.) so imports, spacing, and line-wrapping match rustfmt rules, then commit the formatted changes; no logic changes are needed—just apply rustfmt to this file (or the whole workspace) before merging.
🧹 Nitpick comments (1)
src/tui/screens/arch.rs (1)
12-28: Consider extracting the arch-to-index mapping to reduce duplication.
new_with_defaultandsync_from_configcontain identical logic for mapping architecture strings to indices.Proposed refactor
impl ArchScreen { + fn arch_to_index(arch: &str) -> usize { + match arch { + "arm64" => 1, + _ => 0, + } + } + pub fn new_with_default(arch: &str) -> Self { - let selected = match arch { - "arm64" => 1, - _ => 0, - }; Self { - selected, + selected: Self::arch_to_index(arch), options: ["amd64", "arm64"], } } pub fn sync_from_config(&mut self, arch: &str) { - self.selected = match arch { - "arm64" => 1, - _ => 0, - }; + self.selected = Self::arch_to_index(arch); }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/tui/screens/arch.rs` around lines 12 - 28, Both new_with_default and sync_from_config duplicate the same arch-to-index matching logic; extract that into a single private helper (e.g., fn arch_to_index(arch: &str) -> usize) and call it from new_with_default and sync_from_config to compute selected. Update new_with_default to set selected = Self::arch_to_index(arch) and sync_from_config to set self.selected = Self::arch_to_index(arch), keeping the options = ["amd64", "arm64"] initialization unchanged.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@src/tui/app.rs`:
- Around line 280-295: The generated CLI string in generate_cli_command
currently concatenates raw path/argument values (from self.config.injections ->
inj.src/inj.dest, self.config.init_mode when InitMode::CustomFile(path), and
self.config.output) which will break if they contain spaces or shell-sensitive
characters; update generate_cli_command to properly shell-quote or escape those
values before appending (e.g., wrap with a safe quoting function or escape
spaces/quotes/backslashes), ensuring --inject arguments, the --init path and the
-o output are quoted/escaped consistently while preserving the existing
formatting and line continuations.
- Around line 43-47: The default output filename is hardcoded to
"initramfs.cpio.gz" which can mismatch config.compression; update the
initializer that sets output (where the struct with arch, injections, init_mode,
compression, output is created) to derive the extension from Compression (e.g.,
Compression::Gzip -> ".cpio.gz", None -> ".cpio"), and also update the code path
that changes config.compression (the place noted as "line 167" where
config.compression is set) to synchronize config.output accordingly so the
filename extension always reflects the current config.compression value.
- Around line 234-236: The InitMode::CustomFile branch currently passes any
PathBuf into builder.init_script(path) and lets the builder later fail during
the async task; instead validate the path synchronously before spawning: check
that when InitMode::CustomFile(path) is matched the path.exists() &&
path.is_file() (or use equivalent std::fs metadata checks) and return or surface
an immediate error to the UI if the check fails, otherwise call
builder.init_script(path) and then proceed to tokio::spawn; update the code
around the InitMode::CustomFile match and the call sites that handle
builder.init_script to propagate a clear error to the init screen rather than
deferring to the builder.
In `@src/tui/screens/init.rs`:
- Around line 1-4: This file has formatting inconsistencies in the imports
(e.g., the use of ratatui::{ prelude::*, widgets::{Block, Borders, List,
ListItem, ListState, Paragraph}, };) — run `cargo fmt` to reformat the file so
the import blocks and spacing conform to rustfmt rules and CI will pass; commit
the resulting changes.
- Around line 30-36: get_init_mode currently constructs
InitMode::CustomFile(PathBuf::from(&self.path_input)) even when path_input is
empty; change get_init_mode to return Option<InitMode> (or Result if you prefer
validation errors) and return None when selected == 1 and
self.path_input.trim().is_empty(), otherwise return
Some(InitMode::CustomFile(PathBuf::from(&self.path_input))) for the custom case
and Some(InitMode::Default) when selected == 0; update callers of get_init_mode
to handle the Option/Result accordingly.
In `@src/tui/ui.rs`:
- Line 1: This file fails rustfmt formatting; run rustfmt (e.g., cargo fmt) to
reformat the file so imports and code conform to project style — specifically
reformat the top-level use line that references App, Screen, and WizardMode in
src/tui/ui.rs (symbols: App, Screen, WizardMode) so the file passes CI
formatting checks.
---
Outside diff comments:
In `@src/tui/app.rs`:
- Around line 127-149: The prev_screen and enter_advanced_mode flows are
switching screens without persisting current edits; call
self.sync_screen_on_exit() before modifying self.screen (and before changing
self.mode in enter_advanced_mode) so the current UI state is saved to config
like next_screen() does; update prev_screen to invoke self.sync_screen_on_exit()
immediately before the match that assigns self.screen, and update
enter_advanced_mode to call self.sync_screen_on_exit() before setting self.mode
= WizardMode::Advanced and self.screen = Screen::Architecture.
- Around line 1-297: The file fails rustfmt check; run "cargo fmt" (or rustfmt)
on the repository to reformat src/tui/app.rs (which contains App, BuildConfig,
detect_host_arch, start_build, generate_cli_command, etc.) so imports, spacing,
and line-wrapping match rustfmt rules, then commit the formatted changes; no
logic changes are needed—just apply rustfmt to this file (or the whole
workspace) before merging.
---
Nitpick comments:
In `@src/tui/screens/arch.rs`:
- Around line 12-28: Both new_with_default and sync_from_config duplicate the
same arch-to-index matching logic; extract that into a single private helper
(e.g., fn arch_to_index(arch: &str) -> usize) and call it from new_with_default
and sync_from_config to compute selected. Update new_with_default to set
selected = Self::arch_to_index(arch) and sync_from_config to set self.selected =
Self::arch_to_index(arch), keeping the options = ["amd64", "arm64"]
initialization unchanged.
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: fda7699c-f6d7-4021-a1df-54022a5e16a9
📒 Files selected for processing (9)
src/tui/app.rssrc/tui/events.rssrc/tui/screens/arch.rssrc/tui/screens/compress.rssrc/tui/screens/init.rssrc/tui/screens/inject.rssrc/tui/screens/mod.rssrc/tui/screens/summary.rssrc/tui/ui.rs
Comment on lines
+1
to
+4
| use ratatui::{ | ||
| prelude::*, | ||
| widgets::{Block, Borders, List, ListItem, ListState, Paragraph}, | ||
| }; |
There was a problem hiding this comment.
Run cargo fmt to fix formatting.
The CI pipeline reports formatting issues in this file. Please run cargo fmt to ensure consistent formatting.
🧰 Tools
🪛 GitHub Actions: CI
[error] 1-1: Rust fmt check failed. Detected formatting changes in src/tui/screens/init.rs; please run 'cargo fmt' to format the file.
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@src/tui/screens/init.rs` around lines 1 - 4, This file has formatting
inconsistencies in the imports (e.g., the use of ratatui::{ prelude::*,
widgets::{Block, Borders, List, ListItem, ListState, Paragraph}, };) — run
`cargo fmt` to reformat the file so the import blocks and spacing conform to
rustfmt rules and CI will pass; commit the resulting changes.
src/tui/screens/init.rs
Outdated
Comment on lines
+30
to
+36
| pub fn get_init_mode(&self) -> InitMode { | ||
| if self.selected == 0 { | ||
| InitMode::Default | ||
| } else { | ||
| InitMode::CustomFile(PathBuf::from(&self.path_input)) | ||
| } | ||
| } |
There was a problem hiding this comment.
Empty path accepted when CustomFile is selected.
When selected == 1 (CustomFile), get_init_mode() creates a PathBuf from path_input even if it's empty. This could cause issues downstream if the caller expects a valid file path.
Consider adding validation or returning an error/option when the path is empty:
Proposed validation approach
pub fn get_init_mode(&self) -> InitMode {
if self.selected == 0 {
InitMode::Default
} else {
- InitMode::CustomFile(PathBuf::from(&self.path_input))
+ InitMode::CustomFile(PathBuf::from(self.path_input.trim()))
}
}Alternatively, consider returning Option<InitMode> or validating non-empty paths at the screen transition boundary.
📝 Committable suggestion
‼️ IMPORTANT
Carefully review the code before committing. Ensure that it accurately replaces the highlighted code, contains no missing lines, and has no issues with indentation. Thoroughly test & benchmark the code to ensure it meets the requirements.
Suggested change
| pub fn get_init_mode(&self) -> InitMode { | |
| if self.selected == 0 { | |
| InitMode::Default | |
| } else { | |
| InitMode::CustomFile(PathBuf::from(&self.path_input)) | |
| } | |
| } | |
| pub fn get_init_mode(&self) -> InitMode { | |
| if self.selected == 0 { | |
| InitMode::Default | |
| } else { | |
| InitMode::CustomFile(PathBuf::from(self.path_input.trim())) | |
| } | |
| } |
🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed.
In `@src/tui/screens/init.rs` around lines 30 - 36, get_init_mode currently
constructs InitMode::CustomFile(PathBuf::from(&self.path_input)) even when
path_input is empty; change get_init_mode to return Option<InitMode> (or Result
if you prefer validation errors) and return None when selected == 1 and
self.path_input.trim().is_empty(), otherwise return
Some(InitMode::CustomFile(PathBuf::from(&self.path_input))) for the custom case
and Some(InitMode::Default) when selected == 0; update callers of get_init_mode
to handle the Option/Result accordingly.
There was a problem hiding this comment.
Actionable comments posted: 1
♻️ Duplicate comments (4)
src/tui/app.rs (3)
159-165:⚠️ Potential issue | 🟠 MajorKeep
outputaligned with the selected compression.This branch updates
config.compressiononly.config.outputstill defaults toinitramfs.cpio.gz, so switching to another compression silently produces a misnamed archive. Please derive the filename fromCompressionboth inBuildConfig::default()and when leaving the compression screen.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/tui/app.rs` around lines 159 - 165, The compression branch in sync_screen_on_exit only updates config.compression, leaving config.output as the default filename; update sync_screen_on_exit so that when Screen::Compression is handled (where compress_screen.get_selected() is used) you also set config.output to a filename derived from that Compression (e.g., use Compression to pick extension like .cpio, .cpio.gz, etc.). Also update BuildConfig::default() so the default config.output is computed from the default Compression value instead of hardcoding "initramfs.cpio.gz", ensuring both the initial default and any changes via sync_screen_on_exit remain consistent with Compression.
211-229:⚠️ Potential issue | 🟡 MinorFail fast on invalid custom init scripts.
Any
InitMode::CustomFile(path)is passed straight into the async build. Missing files, directories, and blank paths only surface aftertokio::spawn(), which makes the wizard look like it accepted the value. Validatepath.as_os_str().is_empty()/path.is_file()before switching toScreen::Buildand surface avalidation_errorimmediately.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/tui/app.rs` around lines 211 - 229, Validate custom init script paths synchronously before spawning the async build: when handling InitMode::CustomFile(path) (the variable init_mode and enum InitMode::CustomFile), check path.as_os_str().is_empty() and path.is_file() and if invalid set the wizard state to show validation_error and avoid switching to Screen::Build; only pass a proven-valid path into the InitramfsBuilder.init_script(...) inside the tokio::spawn block so missing/blank/non-file paths fail fast in the UI rather than during the background build.
277-287:⚠️ Potential issue | 🟡 MinorQuote filesystem arguments in the exported CLI command.
--inject,--init, and-oare emitted raw here. Paths with spaces or shell metacharacters will not round-trip when users paste the command back into a shell. Quote each path-bearing argument before appending it.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/tui/app.rs` around lines 277 - 287, The CLI builder appends raw filesystem paths for injections (inj.src, inj.dest), custom init files (InitMode::CustomFile -> path.display()), and output (self.config.output) which breaks when paths contain spaces or shell metacharacters; update the code that formats these flags in the loop and the init/output appends to properly quote or shell-escape each path-bearing value (e.g., wrap in single quotes and escape embedded single quotes, or use a shell-escaping utility) before pushing to cmd so the exported command round-trips correctly.src/tui/screens/init.rs (1)
30-35:⚠️ Potential issue | 🟡 MinorReject blank custom init paths before returning
InitMode.This still turns
""/ whitespace-only input intoInitMode::CustomFile(...), so the wizard can proceed with an unusable init script and only fail later. Trim and reject the empty case here, or returnOption<InitMode>/Result<InitMode, _>and make the caller block the transition.Suggested direction
- pub fn get_init_mode(&self) -> InitMode { - if self.selected == 0 { - InitMode::Default - } else { - InitMode::CustomFile(PathBuf::from(&self.path_input)) - } + pub fn get_init_mode(&self) -> Option<InitMode> { + match self.selected { + 0 => Some(InitMode::Default), + 1 if !self.path_input.trim().is_empty() => { + Some(InitMode::CustomFile(PathBuf::from(self.path_input.trim()))) + } + _ => None, + } }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/tui/screens/init.rs` around lines 30 - 35, get_init_mode currently treats whitespace/empty self.path_input as a valid InitMode::CustomFile; change its contract to return Option<InitMode> and reject blank paths: update the signature of get_init_mode to -> Option<InitMode>, trim self.path_input (e.g., let p = self.path_input.trim()), return Some(InitMode::Default) when self.selected == 0, return None when self.selected != 0 and p.is_empty(), otherwise return Some(InitMode::CustomFile(PathBuf::from(p))); update callers to handle the None case (block the transition or show validation error).
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@src/tui/app.rs`:
- Around line 145-149: The enter_advanced_mode path skips syncing the active
screen, so any edits on the current Screen (e.g., Screen::Image) remain
unstored; update enter_advanced_mode to first call the same screen-exit sync
used elsewhere (call sync_screen_on_exit() or otherwise sync the current screen
state) before changing self.mode and self.screen, then call
sync_screen_on_enter() as before so config.image and other fields are up-to-date
when switching to WizardMode::Advanced.
---
Duplicate comments:
In `@src/tui/app.rs`:
- Around line 159-165: The compression branch in sync_screen_on_exit only
updates config.compression, leaving config.output as the default filename;
update sync_screen_on_exit so that when Screen::Compression is handled (where
compress_screen.get_selected() is used) you also set config.output to a filename
derived from that Compression (e.g., use Compression to pick extension like
.cpio, .cpio.gz, etc.). Also update BuildConfig::default() so the default
config.output is computed from the default Compression value instead of
hardcoding "initramfs.cpio.gz", ensuring both the initial default and any
changes via sync_screen_on_exit remain consistent with Compression.
- Around line 211-229: Validate custom init script paths synchronously before
spawning the async build: when handling InitMode::CustomFile(path) (the variable
init_mode and enum InitMode::CustomFile), check path.as_os_str().is_empty() and
path.is_file() and if invalid set the wizard state to show validation_error and
avoid switching to Screen::Build; only pass a proven-valid path into the
InitramfsBuilder.init_script(...) inside the tokio::spawn block so
missing/blank/non-file paths fail fast in the UI rather than during the
background build.
- Around line 277-287: The CLI builder appends raw filesystem paths for
injections (inj.src, inj.dest), custom init files (InitMode::CustomFile ->
path.display()), and output (self.config.output) which breaks when paths contain
spaces or shell metacharacters; update the code that formats these flags in the
loop and the init/output appends to properly quote or shell-escape each
path-bearing value (e.g., wrap in single quotes and escape embedded single
quotes, or use a shell-escaping utility) before pushing to cmd so the exported
command round-trips correctly.
In `@src/tui/screens/init.rs`:
- Around line 30-35: get_init_mode currently treats whitespace/empty
self.path_input as a valid InitMode::CustomFile; change its contract to return
Option<InitMode> and reject blank paths: update the signature of get_init_mode
to -> Option<InitMode>, trim self.path_input (e.g., let p =
self.path_input.trim()), return Some(InitMode::Default) when self.selected == 0,
return None when self.selected != 0 and p.is_empty(), otherwise return
Some(InitMode::CustomFile(PathBuf::from(p))); update callers to handle the None
case (block the transition or show validation error).
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: a5ff242b-f254-4f32-948e-918fd59ce0b3
📒 Files selected for processing (3)
src/tui/app.rssrc/tui/screens/init.rssrc/tui/ui.rs
🚧 Files skipped from review as they are similar to previous changes (1)
- src/tui/ui.rs
axelfrache
approved these changes
Mar 16, 2026
There was a problem hiding this comment.
🧹 Nitpick comments (3)
src/tui/app.rs (2)
343-362: Quote the image argument for consistency.The
imagevalue on line 344 is not shell-quoted, but it could contain characters that need escaping (e.g., spaces in a local path, or registry URLs with special characters). For consistency with other path arguments, consider quoting it.♻️ Proposed fix
pub fn generate_cli_command(&self) -> String { - let mut cmd = format!("initramfs-builder build {}", self.config.image); + let mut cmd = format!("initramfs-builder build {}", shell_quote(&self.config.image));🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/tui/app.rs` around lines 343 - 362, The CLI builder currently inserts self.config.image unquoted in generate_cli_command, which can break for spaces/special chars; change the initial command construction to shell-quote the image (use shell_quote(&self.config.image) or equivalent) so the image argument is consistently escaped with the other path/injection arguments; update the format call in generate_cli_command that builds "initramfs-builder build {}" to use the quoted image string.
192-196: Consider preserving InitScreen state even when path is empty.When
get_init_mode()returnsNone(CustomFile selected with empty path),config.init_modeis not updated, which can cause a mismatch between UI state (init_screen.selected == 1) and config state (InitMode::Default). While validation prevents proceeding with an empty path, this silent skip could be surprising.Consider storing the screen's selection intent even when invalid, or explicitly resetting
init_screen.selectedto matchconfig.init_mode.🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/tui/app.rs` around lines 192 - 196, When handling Screen::Init, preserve the InitScreen selection even when init_screen.get_init_mode() returns None: either persist the screen's intent (e.g., record selected/index on the screen) so config.init_mode reflects the user's selection attempt, or explicitly reset init_screen.selected to reflect the current self.config.init_mode to avoid UI/config mismatch; update the code around Screen::Init (where get_init_mode() is called) to implement one of these approaches so that self.config.init_mode and init_screen.selected remain consistent (reference InitScreen::get_init_mode, self.config.init_mode, init_screen.selected, and InitMode::Default).src/tui/screens/init.rs (1)
30-38: Trim whitespace from path before creating PathBuf.The empty check uses
trim()on line 33, but when creating thePathBufon line 36, the rawpath_inputis used. This allows paths with leading/trailing whitespace to pass validation but potentially cause issues downstream.♻️ Proposed fix
pub fn get_init_mode(&self) -> Option<InitMode> { if self.selected == 0 { Some(InitMode::Default) } else if self.path_input.trim().is_empty() { None } else { - Some(InitMode::CustomFile(PathBuf::from(&self.path_input))) + Some(InitMode::CustomFile(PathBuf::from(self.path_input.trim()))) } }🤖 Prompt for AI Agents
Verify each finding against the current code and only fix it if needed. In `@src/tui/screens/init.rs` around lines 30 - 38, In get_init_mode, you check the path with self.path_input.trim() but then construct the PathBuf from the raw self.path_input, allowing leading/trailing whitespace to slip through; fix by computing a trimmed variable (e.g. let trimmed = self.path_input.trim();) and use that trimmed value when constructing the InitMode::CustomFile PathBuf (PathBuf::from(trimmed)) so the validation and created PathBuf are consistent.
🤖 Prompt for all review comments with AI agents
Verify each finding against the current code and only fix it if needed.
Nitpick comments:
In `@src/tui/app.rs`:
- Around line 343-362: The CLI builder currently inserts self.config.image
unquoted in generate_cli_command, which can break for spaces/special chars;
change the initial command construction to shell-quote the image (use
shell_quote(&self.config.image) or equivalent) so the image argument is
consistently escaped with the other path/injection arguments; update the format
call in generate_cli_command that builds "initramfs-builder build {}" to use the
quoted image string.
- Around line 192-196: When handling Screen::Init, preserve the InitScreen
selection even when init_screen.get_init_mode() returns None: either persist the
screen's intent (e.g., record selected/index on the screen) so config.init_mode
reflects the user's selection attempt, or explicitly reset init_screen.selected
to reflect the current self.config.init_mode to avoid UI/config mismatch; update
the code around Screen::Init (where get_init_mode() is called) to implement one
of these approaches so that self.config.init_mode and init_screen.selected
remain consistent (reference InitScreen::get_init_mode, self.config.init_mode,
init_screen.selected, and InitMode::Default).
In `@src/tui/screens/init.rs`:
- Around line 30-38: In get_init_mode, you check the path with
self.path_input.trim() but then construct the PathBuf from the raw
self.path_input, allowing leading/trailing whitespace to slip through; fix by
computing a trimmed variable (e.g. let trimmed = self.path_input.trim();) and
use that trimmed value when constructing the InitMode::CustomFile PathBuf
(PathBuf::from(trimmed)) so the validation and created PathBuf are consistent.
ℹ️ Review info
⚙️ Run configuration
Configuration used: defaults
Review profile: CHILL
Plan: Pro
Run ID: 2fa07376-dbbf-46ed-b9a8-af5309232cb3
📒 Files selected for processing (3)
src/tui/app.rssrc/tui/events.rssrc/tui/screens/init.rs
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Add advanced TUI mode
Description
Add advanced mode flow and persist TUI state across screen transitions.
Checklist:
Summary by CodeRabbit