Error Classification
Module: src-tauri/src/error_classification.rs
Classifies terminal error messages and calculates exponential backoff delays for retry logic.
Tauri Commands
| Command | Signature | Description |
|---|---|---|
classify_error_message | (message: String) -> String | Classify error type |
calculate_backoff_delay_cmd | (retry_count, base_delay_ms, max_delay_ms, backoff_multiplier) -> f64 | Calculate backoff delay |
Error Categories
classify_error(message) returns one of:
| Category | Description | Examples |
|---|---|---|
"transient" | Temporary, retry-safe | Network timeout, connection reset, rate limit |
"server" | API server-side error | 5xx responses, service unavailable, auth failures from providers |
"permanent" | Will not resolve with retry | Auth failure, not found, invalid input |
"unknown" | Unclassified | Default for unrecognized patterns |
Backoff Calculation
#![allow(unused)]
fn main() {
pub fn calculate_backoff_delay(
retry_count: u32,
base_delay_ms: f64,
max_delay_ms: f64,
backoff_multiplier: f64,
) -> f64
}Formula: min(base_delay_ms * backoff_multiplier^retry_count, max_delay_ms)
Typical usage:
base_delay_ms: 1000 (1 second)max_delay_ms: 30000 (30 seconds)backoff_multiplier: 2.0 (double each retry)
Result sequence: 1s → 2s → 4s → 8s → 16s → 30s → 30s → …
Frontend Integration
The errorHandlingStore calls these Rust functions to classify errors detected in terminal output and calculate retry delays. The store manages active retries and respects the user’s configured strategy (retry, ignore, or manual).