botserver/src/email/vectordb.rs

use anyhow::Result;
use chrono::{DateTime, Utc};
use serde::{Deserialize, Serialize};
use std::path::PathBuf;
use std::sync::Arc;

use uuid::Uuid;

#[cfg(feature = "vectordb")]
use qdrant_client::{
    prelude::*,
    qdrant::{vectors_config::Config, CreateCollection, Distance, VectorParams, VectorsConfig},
};

/// Email metadata for vector DB indexing
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct EmailDocument {
    pub id: String,
    pub account_id: String,
    pub from_email: String,
    pub from_name: String,
    pub to_email: String,
    pub subject: String,
    pub body_text: String,
    pub date: DateTime<Utc>,
    pub folder: String,
    pub has_attachments: bool,
    pub thread_id: Option<String>,
}

/// Email search query
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct EmailSearchQuery {
    pub query_text: String,
    pub account_id: Option<String>,
    pub folder: Option<String>,
    pub date_from: Option<DateTime<Utc>>,
    pub date_to: Option<DateTime<Utc>>,
    pub limit: usize,
}

/// Email search result
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct EmailSearchResult {
    pub email: EmailDocument,
    pub score: f32,
    pub snippet: String,
}

/// Per-user email vector DB manager
pub struct UserEmailVectorDB {
    user_id: Uuid,
    bot_id: Uuid,
    collection_name: String,
    db_path: PathBuf,
    #[cfg(feature = "vectordb")]
    client: Option<Arc<QdrantClient>>,
}

impl UserEmailVectorDB {
    /// Create new user email vector DB instance
    pub fn new(user_id: Uuid, bot_id: Uuid, db_path: PathBuf) -> Self {
        let collection_name = format!("emails_{}_{}", bot_id, user_id);

        Self {
            user_id,
            bot_id,
            collection_name,
            db_path,
            #[cfg(feature = "vectordb")]
            client: None,
        }
    }

    /// Initialize vector DB collection
    #[cfg(feature = "vectordb")]
    pub async fn initialize(&mut self, qdrant_url: &str) -> Result<()> {
        let client = QdrantClient::from_url(qdrant_url).build()?;

        // Check if collection exists
        let collections = client.list_collections().await?;
        let exists = collections
            .collections
            .iter()
            .any(|c| c.name == self.collection_name);

        if !exists {
            // Create collection for email embeddings (1536 dimensions for OpenAI embeddings)
            client
                .create_collection(&CreateCollection {
                    collection_name: self.collection_name.clone(),
                    vectors_config: Some(VectorsConfig {
                        config: Some(Config::Params(VectorParams {
                            size: 1536,
                            distance: Distance::Cosine.into(),
                            ..Default::default()
                        })),
                    }),
                    ..Default::default()
                })
                .await?;

            log::info!("Created email vector collection: {}", self.collection_name);
        }

        self.client = Some(Arc::new(client));
        Ok(())
    }

    #[cfg(not(feature = "vectordb"))]
    pub async fn initialize(&mut self, _qdrant_url: &str) -> Result<()> {
        log::warn!("Vector DB feature not enabled, using fallback storage");
        Ok(())
    }

    /// Index a single email (on-demand)
    #[cfg(feature = "vectordb")]
    pub async fn index_email(&self, email: &EmailDocument, embedding: Vec<f32>) -> Result<()> {
        let client = self
            .client
            .as_ref()
            .ok_or_else(|| anyhow::anyhow!("Vector DB not initialized"))?;

        let point = PointStruct::new(email.id.clone(), embedding, serde_json::to_value(email)?);

        client
            .upsert_points_blocking(self.collection_name.clone(), vec![point], None)
            .await?;

        log::debug!("Indexed email: {} - {}", email.id, email.subject);
        Ok(())
    }

    #[cfg(not(feature = "vectordb"))]
    pub async fn index_email(&self, email: &EmailDocument, _embedding: Vec<f32>) -> Result<()> {
        // Fallback: Store in JSON file
        let file_path = self.db_path.join(format!("{}.json", email.id));
        let json = serde_json::to_string_pretty(email)?;
        fs::write(file_path, json).await?;
        Ok(())
    }

    /// Index multiple emails in batch
    pub async fn index_emails_batch(&self, emails: &[(EmailDocument, Vec<f32>)]) -> Result<()> {
        for (email, embedding) in emails {
            self.index_email(email, embedding.clone()).await?;
        }
        Ok(())
    }

    /// Search emails using vector similarity
    #[cfg(feature = "vectordb")]
    pub async fn search(
        &self,
        query: &EmailSearchQuery,
        query_embedding: Vec<f32>,
    ) -> Result<Vec<EmailSearchResult>> {
        let client = self
            .client
            .as_ref()
            .ok_or_else(|| anyhow::anyhow!("Vector DB not initialized"))?;

        // Build filter if specified
        let mut filter = None;
        if query.account_id.is_some() || query.folder.is_some() {
            let mut conditions = vec![];

            if let Some(account_id) = &query.account_id {
                conditions.push(qdrant_client::qdrant::Condition::matches(
                    "account_id",
                    account_id.clone(),
                ));
            }

            if let Some(folder) = &query.folder {
                conditions.push(qdrant_client::qdrant::Condition::matches(
                    "folder",
                    folder.clone(),
                ));
            }

            filter = Some(qdrant_client::qdrant::Filter::must(conditions));
        }

        let search_result = client
            .search_points(&qdrant_client::qdrant::SearchPoints {
                collection_name: self.collection_name.clone(),
                vector: query_embedding,
                limit: query.limit as u64,
                filter,
                with_payload: Some(true.into()),
                ..Default::default()
            })
            .await?;

        let mut results = Vec::new();
        for point in search_result.result {
            if let Some(payload) = point.payload {
                let email: EmailDocument = serde_json::from_value(serde_json::to_value(&payload)?)?;

                // Create snippet from body (first 200 chars)
                let snippet = if email.body_text.len() > 200 {
                    format!("{}...", &email.body_text[..200])
                } else {
                    email.body_text.clone()
                };

                results.push(EmailSearchResult {
                    email,
                    score: point.score,
                    snippet,
                });
            }
        }

        Ok(results)
    }

    #[cfg(not(feature = "vectordb"))]
    pub async fn search(
        &self,
        query: &EmailSearchQuery,
        _query_embedding: Vec<f32>,
    ) -> Result<Vec<EmailSearchResult>> {
        // Fallback: Simple text search in JSON files
        let mut results = Vec::new();
        let mut entries = fs::read_dir(&self.db_path).await?;

        while let Some(entry) = entries.next_entry().await? {
            if entry.path().extension().and_then(|s| s.to_str()) == Some("json") {
                let content = fs::read_to_string(entry.path()).await?;
                if let Ok(email) = serde_json::from_str::<EmailDocument>(&content) {
                    // Simple text matching
                    let query_lower = query.query_text.to_lowercase();
                    if email.subject.to_lowercase().contains(&query_lower)
                        || email.body_text.to_lowercase().contains(&query_lower)
                        || email.from_email.to_lowercase().contains(&query_lower)
                    {
                        let snippet = if email.body_text.len() > 200 {
                            format!("{}...", &email.body_text[..200])
                        } else {
                            email.body_text.clone()
                        };

                        results.push(EmailSearchResult {
                            email,
                            score: 1.0,
                            snippet,
                        });
                    }
                }

                if results.len() >= query.limit {
                    break;
                }
            }
        }

        Ok(results)
    }

    /// Delete email from index
    #[cfg(feature = "vectordb")]
    pub async fn delete_email(&self, email_id: &str) -> Result<()> {
        let client = self
            .client
            .as_ref()
            .ok_or_else(|| anyhow::anyhow!("Vector DB not initialized"))?;

        client
            .delete_points(
                self.collection_name.clone(),
                &vec![email_id.into()].into(),
                None,
            )
            .await?;

        log::debug!("Deleted email from index: {}", email_id);
        Ok(())
    }

    #[cfg(not(feature = "vectordb"))]
    pub async fn delete_email(&self, email_id: &str) -> Result<()> {
        let file_path = self.db_path.join(format!("{}.json", email_id));
        if file_path.exists() {
            fs::remove_file(file_path).await?;
        }
        Ok(())
    }

    /// Get indexed email count
    #[cfg(feature = "vectordb")]
    pub async fn get_count(&self) -> Result<u64> {
        let client = self
            .client
            .as_ref()
            .ok_or_else(|| anyhow::anyhow!("Vector DB not initialized"))?;

        let info = client.collection_info(self.collection_name.clone()).await?;

        Ok(info.result.unwrap().points_count.unwrap_or(0))
    }

    #[cfg(not(feature = "vectordb"))]
    pub async fn get_count(&self) -> Result<u64> {
        let mut count = 0;
        let mut entries = fs::read_dir(&self.db_path).await?;

        while let Some(entry) = entries.next_entry().await? {
            if entry.path().extension().and_then(|s| s.to_str()) == Some("json") {
                count += 1;
            }
        }

        Ok(count)
    }

    /// Clear all indexed emails
    #[cfg(feature = "vectordb")]
    pub async fn clear(&self) -> Result<()> {
        let client = self
            .client
            .as_ref()
            .ok_or_else(|| anyhow::anyhow!("Vector DB not initialized"))?;

        client
            .delete_collection(self.collection_name.clone())
            .await?;

        // Recreate empty collection
        client
            .create_collection(&CreateCollection {
                collection_name: self.collection_name.clone(),
                vectors_config: Some(VectorsConfig {
                    config: Some(Config::Params(VectorParams {
                        size: 1536,
                        distance: Distance::Cosine.into(),
                        ..Default::default()
                    })),
                }),
                ..Default::default()
            })
            .await?;

        log::info!("Cleared email vector collection: {}", self.collection_name);
        Ok(())
    }

    #[cfg(not(feature = "vectordb"))]
    pub async fn clear(&self) -> Result<()> {
        if self.db_path.exists() {
            fs::remove_dir_all(&self.db_path).await?;
            fs::create_dir_all(&self.db_path).await?;
        }
        Ok(())
    }
}

/// Email embedding generator using LLM
pub struct EmailEmbeddingGenerator {
    llm_endpoint: String,
}

impl EmailEmbeddingGenerator {
    pub fn new(llm_endpoint: String) -> Self {
        Self { llm_endpoint }
    }

    /// Generate embedding for email content
    pub async fn generate_embedding(&self, email: &EmailDocument) -> Result<Vec<f32>> {
        // Combine email fields for embedding
        let text = format!(
            "From: {} <{}>\nSubject: {}\n\n{}",
            email.from_name, email.from_email, email.subject, email.body_text
        );

        // Truncate if too long (max 8000 chars for most embedding models)
        let text = if text.len() > 8000 {
            &text[..8000]
        } else {
            &text
        };

        // Call LLM embedding endpoint
        // This is a placeholder - implement actual LLM call
        self.generate_text_embedding(text).await
    }

    /// Generate embedding from raw text
    pub async fn generate_text_embedding(&self, text: &str) -> Result<Vec<f32>> {
        // Try OpenAI embeddings first if API key is available
        if let Ok(api_key) = std::env::var("OPENAI_API_KEY") {
            return self.generate_openai_embedding(text, &api_key).await;
        }

        // Try local embedding service if configured
        if let Ok(embedding_url) = std::env::var("LOCAL_EMBEDDING_URL") {
            return self.generate_local_embedding(text, &embedding_url).await;
        }

        // Fall back to simple hash-based embedding for development
        self.generate_hash_embedding(text)
    }

    /// Generate embedding using OpenAI API
    async fn generate_openai_embedding(&self, text: &str, api_key: &str) -> Result<Vec<f32>> {
        use reqwest::header::{HeaderMap, HeaderValue, AUTHORIZATION, CONTENT_TYPE};
        use serde_json::json;

        let client = reqwest::Client::new();
        let mut headers = HeaderMap::new();
        headers.insert(CONTENT_TYPE, HeaderValue::from_static("application/json"));
        headers.insert(
            AUTHORIZATION,
            HeaderValue::from_str(&format!("Bearer {}", api_key))?,
        );

        let body = json!({
            "input": text,
            "model": "text-embedding-3-small"
        });

        let response = client
            .post("https://api.openai.com/v1/embeddings")
            .headers(headers)
            .json(&body)
            .send()
            .await?;

        if !response.status().is_success() {
            return Err(anyhow::anyhow!("OpenAI API error: {}", response.status()));
        }

        let result: serde_json::Value = response.json().await?;
        let embedding = result["data"][0]["embedding"]
            .as_array()
            .ok_or_else(|| anyhow::anyhow!("Invalid OpenAI response format"))?
            .iter()
            .map(|v| v.as_f64().unwrap_or(0.0) as f32)
            .collect();

        Ok(embedding)
    }

    /// Generate embedding using local embedding service
    async fn generate_local_embedding(&self, text: &str, embedding_url: &str) -> Result<Vec<f32>> {
        use serde_json::json;

        let client = reqwest::Client::new();
        let body = json!({
            "text": text,
            "model": "sentence-transformers/all-MiniLM-L6-v2"
        });

        let response = client.post(embedding_url).json(&body).send().await?;

        if !response.status().is_success() {
            return Err(anyhow::anyhow!(
                "Local embedding service error: {}",
                response.status()
            ));
        }

        let result: serde_json::Value = response.json().await?;
        let embedding = result["embedding"]
            .as_array()
            .ok_or_else(|| anyhow::anyhow!("Invalid embedding response format"))?
            .iter()
            .map(|v| v.as_f64().unwrap_or(0.0) as f32)
            .collect();

        Ok(embedding)
    }

    /// Generate deterministic hash-based embedding for development
    fn generate_hash_embedding(&self, text: &str) -> Result<Vec<f32>> {
        use std::collections::hash_map::DefaultHasher;
        use std::hash::{Hash, Hasher};

        const EMBEDDING_DIM: usize = 1536;
        let mut embedding = vec![0.0f32; EMBEDDING_DIM];

        // Create multiple hash values for different dimensions
        let words: Vec<&str> = text.split_whitespace().collect();

        for (i, chunk) in words.chunks(10).enumerate() {
            let mut hasher = DefaultHasher::new();
            chunk.join(" ").hash(&mut hasher);
            let hash = hasher.finish();

            // Distribute hash across embedding dimensions
            for j in 0..64 {
                let idx = (i * 64 + j) % EMBEDDING_DIM;
                let value = ((hash >> j) & 1) as f32;
                embedding[idx] += value;
            }
        }

        // Normalize the embedding
        let norm: f32 = embedding.iter().map(|x| x * x).sum::<f32>().sqrt();
        if norm > 0.0 {
            for val in &mut embedding {
                *val /= norm;
            }
        }

        Ok(embedding)
    }
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn test_email_document_creation() {
        let email = EmailDocument {
            id: "test-123".to_string(),
            account_id: "account-456".to_string(),
            from_email: "sender@example.com".to_string(),
            from_name: "Test Sender".to_string(),
            to_email: "receiver@example.com".to_string(),
            subject: "Test Subject".to_string(),
            body_text: "Test email body".to_string(),
            date: Utc::now(),
            folder: "INBOX".to_string(),
            has_attachments: false,
            thread_id: None,
        };

        assert_eq!(email.id, "test-123");
        assert_eq!(email.subject, "Test Subject");
    }

    #[tokio::test]
    async fn test_user_email_vectordb_creation() {
        let temp_dir = std::env::temp_dir().join("test_vectordb");
        let db = UserEmailVectorDB::new(Uuid::new_v4(), Uuid::new_v4(), temp_dir);

        assert!(db.collection_name.starts_with("emails_"));
    }
}