sal/buildah_debug_implementation_plan.md
2025-04-05 11:03:58 +02:00

207 lines
6.7 KiB
Markdown

# Buildah Debug Implementation Plan
## Current State
1. The `Builder` struct already has a `debug` field and methods to get and set it (`debug()` and `set_debug()`).
2. There's a thread-local `DEBUG` variable with functions to get and set it.
3. The `execute_buildah_command` function checks the thread-local debug flag and outputs some debug information.
4. There's an unused `execute_buildah_command_with_debug` function that takes a `Builder` reference.
## Requirements
1. Use only the Builder's debug flag, not the thread-local debug flag.
2. When debug is true, output stdout/stderr regardless of whether the command succeeds or fails.
3. If debug is false but there's an error, still output all information.
## Implementation Plan
### 1. Keep the Existing Thread-Local DEBUG Variable
We'll keep the existing thread-local DEBUG variable that's already in the code:
```rust
// Thread-local storage for debug flag
thread_local! {
static DEBUG: std::cell::RefCell<bool> = std::cell::RefCell::new(false);
}
```
### 2. Modify the Builder Methods to Set/Clear the Thread-Local Debug Flag
We'll modify each Builder method to set the thread-local debug flag from the Builder's debug flag before calling `execute_buildah_command` and restore it afterward:
```rust
pub fn run(&self, command: &str) -> Result<CommandResult, BuildahError> {
if let Some(container_id) = &self.container_id {
// Save the current debug flag
let previous_debug = thread_local_debug();
// Set the thread-local debug flag from the Builder's debug flag
set_thread_local_debug(self.debug);
// Execute the command
let result = execute_buildah_command(&["run", container_id, "sh", "-c", command]);
// Restore the previous debug flag
set_thread_local_debug(previous_debug);
result
} else {
Err(BuildahError::Other("No container ID available".to_string()))
}
}
```
### 3. Keep the Existing execute_buildah_command Function
The existing `execute_buildah_command` function already checks the thread-local debug flag, so we don't need to modify it:
```rust
pub fn execute_buildah_command(args: &[&str]) -> Result<CommandResult, BuildahError> {
// Get the debug flag from thread-local storage
let debug = thread_local_debug();
if debug {
println!("Executing buildah command: buildah {}", args.join(" "));
}
// ... rest of the function ...
}
```
### 4. Update the execute_buildah_command Function to Output stdout/stderr
We need to modify the `execute_buildah_command` function to output stdout/stderr when debug is true, regardless of success/failure:
```rust
pub fn execute_buildah_command(args: &[&str]) -> Result<CommandResult, BuildahError> {
// Get the debug flag from thread-local storage
let debug = thread_local_debug();
if debug {
println!("Executing buildah command: buildah {}", args.join(" "));
}
let output = Command::new("buildah")
.args(args)
.output();
match output {
Ok(output) => {
let stdout = String::from_utf8_lossy(&output.stdout).to_string();
let stderr = String::from_utf8_lossy(&output.stderr).to_string();
let result = CommandResult {
stdout,
stderr,
success: output.status.success(),
code: output.status.code().unwrap_or(-1),
};
// Always output stdout/stderr when debug is true
if debug {
if !result.stdout.is_empty() {
println!("Command stdout: {}", result.stdout);
}
if !result.stderr.is_empty() {
println!("Command stderr: {}", result.stderr);
}
if result.success {
println!("Command succeeded with code {}", result.code);
} else {
println!("Command failed with code {}", result.code);
}
}
if result.success {
Ok(result)
} else {
// If command failed and debug is false, output stderr
if !debug {
println!("Command failed with code {}: {}", result.code, result.stderr.trim());
}
Err(BuildahError::CommandFailed(format!("Command failed with code {}: {}",
result.code, result.stderr.trim())))
}
},
Err(e) => {
// Always output error information
println!("Command execution failed: {}", e);
Err(BuildahError::CommandExecutionFailed(e))
}
}
}
```
### 5. Handle Static Methods
For static methods, we'll just call `execute_buildah_command` directly, which will use the thread-local debug flag:
```rust
pub fn images() -> Result<Vec<Image>, BuildahError> {
let result = execute_buildah_command(&["images", "--json"])?;
// Rest of the method...
}
```
If we want to support debugging in static methods, we could add an optional debug parameter:
```rust
pub fn images(debug: bool) -> Result<Vec<Image>, BuildahError> {
// Save the current debug flag
let previous_debug = thread_local_debug();
// Set the thread-local debug flag
set_thread_local_debug(debug);
// Execute the command
let result = execute_buildah_command(&["images", "--json"]);
// Restore the previous debug flag
set_thread_local_debug(previous_debug);
// Process the result
match result {
Ok(cmd_result) => {
// Parse JSON and return images...
},
Err(e) => Err(e),
}
}
// Backward compatibility method
pub fn images() -> Result<Vec<Image>, BuildahError> {
Self::images(false)
}
```
### 6. Update Rhai Bindings
We still need to update the Rhai bindings to expose the debug functionality:
```rust
// Add a debug getter and setter
engine.register_get("debug", |builder: &mut Builder| builder.debug());
engine.register_set("debug", |builder: &mut Builder, debug: bool| { builder.set_debug(debug); });
```
### 7. Remove Unused Code
We can remove the unused `execute_buildah_command_with_debug` function.
## Implementation Flow
1. Modify the `execute_buildah_command` function to output stdout/stderr when debug is true
2. Update all Builder methods to set/restore the thread-local debug flag
3. Update static methods to optionally accept a debug parameter
4. Update Rhai bindings to expose the debug functionality
5. Remove unused code
## Benefits
1. Minimal changes to the existing codebase
2. No changes to function signatures
3. Backward compatibility with existing code
4. Improved debugging capabilities