try to use local ai to generate commit message

src-tauri/crates/commit_ai/src/lib.rs

@@ -0,0 +1,157 @@
+use std::sync::Arc;
+
+use mistralrs::{GgufModelBuilder, Model, TextMessageRole, TextMessages};
+use tokio::sync::RwLock;
+
+// Small instruct model, quantized to keep the one-time download reasonable (~1 GB) while
+// still being fast enough for short, structured generations like a commit message on CPU.
+const HF_REPO: &str = "Qwen/Qwen2.5-1.5B-Instruct-GGUF";
+const GGUF_FILE: &str = "qwen2.5-1.5b-instruct-q4_k_m.gguf";
+const TOKENIZER_MODEL_ID: &str = "Qwen/Qwen2.5-1.5B-Instruct";
+
+#[derive(Debug, Clone, Copy, PartialEq, Eq, serde::Serialize)]
+#[serde(rename_all = "lowercase")]
+pub enum CommitAiPhase {
+    /// Nothing has been requested yet.
+    Idle,
+    /// Downloading (first run only, then cached by hf-hub) and/or loading into memory.
+    Loading,
+    Ready,
+    Error,
+}
+
+#[derive(Debug, Clone, serde::Serialize)]
+pub struct CommitAiStatus {
+    pub phase: CommitAiPhase,
+    pub error: Option<String>,
+}
+
+struct Inner {
+    phase: CommitAiPhase,
+    error: Option<String>,
+    model: Option<Arc<Model>>,
+}
+
+/// Cheap to clone: shares one model instance across the app via an inner `Arc`.
+#[derive(Clone)]
+pub struct CommitAiEngine {
+    inner: Arc<RwLock<Inner>>,
+}
+
+impl Default for CommitAiEngine {
+    fn default() -> Self {
+        Self {
+            inner: Arc::new(RwLock::new(Inner {
+                phase: CommitAiPhase::Idle,
+                error: None,
+                model: None,
+            })),
+        }
+    }
+}
+
+impl CommitAiEngine {
+    pub fn new() -> Self {
+        Self::default()
+    }
+
+    pub async fn status(&self) -> CommitAiStatus {
+        let guard = self.inner.read().await;
+        CommitAiStatus {
+            phase: guard.phase,
+            error: guard.error.clone(),
+        }
+    }
+
+    /// Downloads (first run only; hf-hub caches the files afterwards) and loads the model.
+    /// Safe to call multiple times — only the first caller actually triggers a load, later
+    /// callers just return once the in-flight or previous attempt is done.
+    pub async fn ensure_loaded(&self) {
+        {
+            let mut guard = self.inner.write().await;
+            if guard.phase != CommitAiPhase::Idle {
+                return;
+            }
+            guard.phase = CommitAiPhase::Loading;
+            guard.error = None;
+        }
+
+        let result = GgufModelBuilder::new(HF_REPO, vec![GGUF_FILE])
+            .with_tok_model_id(TOKENIZER_MODEL_ID)
+            .with_logging()
+            .build()
+            .await;
+
+        let mut guard = self.inner.write().await;
+        match result {
+            Ok(model) => {
+                guard.model = Some(Arc::new(model));
+                guard.phase = CommitAiPhase::Ready;
+                guard.error = None;
+            }
+            Err(err) => {
+                guard.phase = CommitAiPhase::Error;
+                guard.error = Some(err.to_string());
+            }
+        }
+    }
+
+    pub async fn generate_commit_message(
+        &self,
+        diff: &str,
+        notes: Option<&str>,
+    ) -> Result<String, String> {
+        let model = {
+            let guard = self.inner.read().await;
+            match (guard.phase, &guard.model) {
+                (CommitAiPhase::Ready, Some(model)) => model.clone(),
+                _ => return Err("Das KI-Modell ist noch nicht bereit.".to_string()),
+            }
+        };
+
+        if diff.trim().is_empty() {
+            return Err("Keine gestagten Änderungen für eine Commit-Message vorhanden.".to_string());
+        }
+
+        let (system, user) = build_messages(diff, notes);
+        let messages = TextMessages::new()
+            .add_message(TextMessageRole::System, system)
+            .add_message(TextMessageRole::User, user);
+
+        let response = model
+            .send_chat_request(messages)
+            .await
+            .map_err(|err| err.to_string())?;
+
+        let content = response
+            .choices
+            .first()
+            .and_then(|choice| choice.message.content.clone())
+            .ok_or_else(|| "Das Modell hat keine Antwort geliefert.".to_string())?;
+
+        Ok(content.trim().to_string())
+    }
+}
+
+fn build_messages(diff: &str, notes: Option<&str>) -> (String, String) {
+    // grobe Token-Schätzung, kleine Modelle haben oft 8–32k Kontext
+    const MAX_CHARS: usize = 24_000;
+    let diff = if diff.len() > MAX_CHARS {
+        format!("{}\n\n[... Diff gekürzt ...]", &diff[..MAX_CHARS])
+    } else {
+        diff.to_string()
+    };
+
+    let system = "Du bist ein Werkzeug, das Git-Commit-Messages erzeugt. \
+Antworte ausschließlich mit der Commit-Message im Conventional-Commits-Format \
+(<type>(<scope>): <subject>), optional gefolgt von einem Body nach einer Leerzeile. \
+Subject imperativ, max. 72 Zeichen. Kein Vorspann, keine Erklärung, keine Code-Fences, in Englisch antworten"
+        .to_string();
+
+    let mut user = String::new();
+    if let Some(n) = notes.filter(|n| !n.trim().is_empty()) {
+        user.push_str(&format!("Anmerkungen des Entwicklers:\n{n}\n\n"));
+    }
+    user.push_str(&format!("Staged diff:\n{diff}"));
+    (system, user)
+}