cloudillo-types 0.8.16

// SPDX-FileCopyrightText: Szilárd Hajba
// SPDX-License-Identifier: LGPL-3.0-or-later

//! Real-Time Database Adapter
//!
//! Trait and types for pluggable real-time database backends that store JSON documents
//! using hierarchical path-based access (e.g., `posts/abc123/comments/xyz789`).
//!
//! Read operations (query, get, subscribe) work directly on the adapter.
//! Write operations (create, update, delete) require a transaction for atomicity.
//!
//! Each adapter implementation provides its own constructor handling backend-specific
//! initialization (database path, connection settings, etc.).

use async_trait::async_trait;
use futures_core::Stream;
use serde::{Deserialize, Serialize};
use serde_json::Value;
use std::collections::HashMap;
use std::fmt::Debug;
use std::pin::Pin;

use crate::prelude::*;

/// Lock mode for document locking.
#[derive(Debug, Clone, PartialEq, Eq, Serialize, Deserialize)]
#[serde(rename_all = "camelCase")]
pub enum LockMode {
	Soft,
	Hard,
}

/// Information about an active lock on a document path.
#[derive(Debug, Clone)]
pub struct LockInfo {
	pub user_id: Box<str>,
	pub mode: LockMode,
	pub acquired_at: u64,
	pub ttl_secs: u64,
}

/// An aggregation operation to compute per group.
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(tag = "op", rename_all = "camelCase")]
pub enum AggregateOp {
	Sum { field: String },
	Avg { field: String },
	Min { field: String },
	Max { field: String },
}

/// Aggregation options: group by a field and compute statistics.
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(rename_all = "camelCase")]
pub struct AggregateOptions {
	/// Field to group by. For array fields, each element becomes a separate group.
	pub group_by: String,

	/// Additional operations per group (count is always included implicitly).
	#[serde(default, skip_serializing_if = "Vec::is_empty")]
	pub ops: Vec<AggregateOp>,
}

/// Query filter for selecting documents.
///
/// Supports multiple filter operations on JSON document fields.
/// A document matches if ALL specified conditions are satisfied (AND logic).
#[derive(Debug, Clone, Default, Serialize, Deserialize)]
pub struct QueryFilter {
	/// Field equality constraints: field_name -> expected_value
	#[serde(default, skip_serializing_if = "HashMap::is_empty")]
	pub equals: HashMap<String, Value>,

	/// Field not-equal constraints: field_name -> expected_value
	#[serde(default, skip_serializing_if = "HashMap::is_empty", rename = "notEquals")]
	pub not_equals: HashMap<String, Value>,

	/// Field greater-than constraints: field_name -> threshold_value
	#[serde(default, skip_serializing_if = "HashMap::is_empty", rename = "greaterThan")]
	pub greater_than: HashMap<String, Value>,

	/// Field greater-than-or-equal constraints: field_name -> threshold_value
	#[serde(default, skip_serializing_if = "HashMap::is_empty", rename = "greaterThanOrEqual")]
	pub greater_than_or_equal: HashMap<String, Value>,

	/// Field less-than constraints: field_name -> threshold_value
	#[serde(default, skip_serializing_if = "HashMap::is_empty", rename = "lessThan")]
	pub less_than: HashMap<String, Value>,

	/// Field less-than-or-equal constraints: field_name -> threshold_value
	#[serde(default, skip_serializing_if = "HashMap::is_empty", rename = "lessThanOrEqual")]
	pub less_than_or_equal: HashMap<String, Value>,

	/// Field in-array constraints: field_name -> array of allowed values
	#[serde(default, skip_serializing_if = "HashMap::is_empty", rename = "inArray")]
	pub in_array: HashMap<String, Vec<Value>>,

	/// Array-contains constraints: field_name -> value that must be in the array field
	#[serde(default, skip_serializing_if = "HashMap::is_empty", rename = "arrayContains")]
	pub array_contains: HashMap<String, Value>,

	/// Not-in-array constraints: field_name -> array of excluded values
	#[serde(default, skip_serializing_if = "HashMap::is_empty", rename = "notInArray")]
	pub not_in_array: HashMap<String, Vec<Value>>,

	/// Array-contains-any constraints: field_name -> array of values (at least one must be in the array field)
	#[serde(default, skip_serializing_if = "HashMap::is_empty", rename = "arrayContainsAny")]
	pub array_contains_any: HashMap<String, Vec<Value>>,

	/// Array-contains-all constraints: field_name -> array of values (all must be in the array field)
	#[serde(default, skip_serializing_if = "HashMap::is_empty", rename = "arrayContainsAll")]
	pub array_contains_all: HashMap<String, Vec<Value>>,
}

impl QueryFilter {
	/// Create a new empty filter (matches all documents).
	pub fn new() -> Self {
		Self::default()
	}

	/// Create a filter with a single equality constraint.
	pub fn equals_one(field: impl Into<String>, value: Value) -> Self {
		let mut equals = HashMap::new();
		equals.insert(field.into(), value);
		Self { equals, ..Default::default() }
	}

	/// Add an equality constraint to this filter (builder pattern).
	pub fn with_equals(mut self, field: impl Into<String>, value: Value) -> Self {
		self.equals.insert(field.into(), value);
		self
	}

	/// Add a not-equal constraint to this filter (builder pattern).
	pub fn with_not_equals(mut self, field: impl Into<String>, value: Value) -> Self {
		self.not_equals.insert(field.into(), value);
		self
	}

	/// Add a greater-than constraint to this filter (builder pattern).
	pub fn with_greater_than(mut self, field: impl Into<String>, value: Value) -> Self {
		self.greater_than.insert(field.into(), value);
		self
	}

	/// Add a greater-than-or-equal constraint to this filter (builder pattern).
	pub fn with_greater_than_or_equal(mut self, field: impl Into<String>, value: Value) -> Self {
		self.greater_than_or_equal.insert(field.into(), value);
		self
	}

	/// Add a less-than constraint to this filter (builder pattern).
	pub fn with_less_than(mut self, field: impl Into<String>, value: Value) -> Self {
		self.less_than.insert(field.into(), value);
		self
	}

	/// Add a less-than-or-equal constraint to this filter (builder pattern).
	pub fn with_less_than_or_equal(mut self, field: impl Into<String>, value: Value) -> Self {
		self.less_than_or_equal.insert(field.into(), value);
		self
	}

	/// Add an in-array constraint to this filter (builder pattern).
	pub fn with_in_array(mut self, field: impl Into<String>, values: Vec<Value>) -> Self {
		self.in_array.insert(field.into(), values);
		self
	}

	/// Add an array-contains constraint to this filter (builder pattern).
	pub fn with_array_contains(mut self, field: impl Into<String>, value: Value) -> Self {
		self.array_contains.insert(field.into(), value);
		self
	}

	/// Add a not-in-array constraint to this filter (builder pattern).
	pub fn with_not_in_array(mut self, field: impl Into<String>, values: Vec<Value>) -> Self {
		self.not_in_array.insert(field.into(), values);
		self
	}

	/// Add an array-contains-any constraint to this filter (builder pattern).
	pub fn with_array_contains_any(mut self, field: impl Into<String>, values: Vec<Value>) -> Self {
		self.array_contains_any.insert(field.into(), values);
		self
	}

	/// Add an array-contains-all constraint to this filter (builder pattern).
	pub fn with_array_contains_all(mut self, field: impl Into<String>, values: Vec<Value>) -> Self {
		self.array_contains_all.insert(field.into(), values);
		self
	}

	/// Check if a document matches this filter (all conditions must be satisfied).
	pub fn matches(&self, doc: &Value) -> bool {
		// Equality checks
		for (field, expected) in &self.equals {
			if doc.get(field) != Some(expected) {
				return false;
			}
		}

		// Not-equal checks (missing fields are inherently "not equal")
		for (field, expected) in &self.not_equals {
			if doc.get(field) == Some(expected) {
				return false;
			}
		}

		// Greater-than checks
		for (field, threshold) in &self.greater_than {
			match doc.get(field) {
				Some(actual)
					if compare_json_values(Some(actual), Some(threshold))
						== std::cmp::Ordering::Greater => {}
				_ => return false,
			}
		}

		// Greater-than-or-equal checks
		for (field, threshold) in &self.greater_than_or_equal {
			match doc.get(field) {
				Some(actual) => {
					let ord = compare_json_values(Some(actual), Some(threshold));
					if ord != std::cmp::Ordering::Greater && ord != std::cmp::Ordering::Equal {
						return false;
					}
				}
				_ => return false,
			}
		}

		// Less-than checks
		for (field, threshold) in &self.less_than {
			match doc.get(field) {
				Some(actual)
					if compare_json_values(Some(actual), Some(threshold))
						== std::cmp::Ordering::Less => {}
				_ => return false,
			}
		}

		// Less-than-or-equal checks
		for (field, threshold) in &self.less_than_or_equal {
			match doc.get(field) {
				Some(actual) => {
					let ord = compare_json_values(Some(actual), Some(threshold));
					if ord != std::cmp::Ordering::Less && ord != std::cmp::Ordering::Equal {
						return false;
					}
				}
				_ => return false,
			}
		}

		// In-array checks (field value must be in the provided array)
		for (field, allowed_values) in &self.in_array {
			match doc.get(field) {
				Some(actual) if allowed_values.contains(actual) => {}
				_ => return false,
			}
		}

		// Array-contains checks (field must be an array containing the value)
		for (field, required_value) in &self.array_contains {
			match doc.get(field) {
				Some(Value::Array(arr)) if arr.contains(required_value) => {}
				_ => return false,
			}
		}

		// Not-in-array checks (field value must NOT be in the provided array; missing fields pass)
		for (field, excluded_values) in &self.not_in_array {
			if let Some(actual) = doc.get(field)
				&& excluded_values.contains(actual)
			{
				return false;
			}
		}

		// Array-contains-any checks
		for (field, candidate_values) in &self.array_contains_any {
			match doc.get(field) {
				Some(Value::Array(arr)) if candidate_values.iter().any(|v| arr.contains(v)) => {}
				_ => return false,
			}
		}

		// Array-contains-all checks
		for (field, required_values) in &self.array_contains_all {
			match doc.get(field) {
				Some(Value::Array(arr)) if required_values.iter().all(|v| arr.contains(v)) => {}
				_ => return false,
			}
		}

		true
	}

	/// Check if this filter is empty (matches all documents).
	pub fn is_empty(&self) -> bool {
		self.equals.is_empty()
			&& self.not_equals.is_empty()
			&& self.greater_than.is_empty()
			&& self.greater_than_or_equal.is_empty()
			&& self.less_than.is_empty()
			&& self.less_than_or_equal.is_empty()
			&& self.in_array.is_empty()
			&& self.array_contains.is_empty()
			&& self.not_in_array.is_empty()
			&& self.array_contains_any.is_empty()
			&& self.array_contains_all.is_empty()
	}
}

/// Sort order for a field.
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct SortField {
	/// Field name to sort by
	pub field: String,

	/// Sort direction: true for ascending, false for descending
	pub ascending: bool,
}

impl SortField {
	/// Create ascending sort order.
	pub fn asc(field: impl Into<String>) -> Self {
		Self { field: field.into(), ascending: true }
	}

	/// Create descending sort order.
	pub fn desc(field: impl Into<String>) -> Self {
		Self { field: field.into(), ascending: false }
	}
}

/// Options for querying documents (filter, sort, limit, offset).
#[derive(Debug, Clone, Default)]
pub struct QueryOptions {
	/// Optional filter to select documents
	pub filter: Option<QueryFilter>,

	/// Optional sort order (multiple fields supported)
	pub sort: Option<Vec<SortField>>,

	/// Optional limit on number of results
	pub limit: Option<u32>,

	/// Optional offset for pagination
	pub offset: Option<u32>,

	/// When set, returns aggregated groups instead of documents.
	pub aggregate: Option<AggregateOptions>,
}

impl QueryOptions {
	/// Create new empty query options (no filter, sort, or limit).
	pub fn new() -> Self {
		Self::default()
	}

	/// Set the filter.
	pub fn with_filter(mut self, filter: QueryFilter) -> Self {
		self.filter = Some(filter);
		self
	}

	/// Set the sort order.
	pub fn with_sort(mut self, sort: Vec<SortField>) -> Self {
		self.sort = Some(sort);
		self
	}

	/// Set the limit.
	pub fn with_limit(mut self, limit: u32) -> Self {
		self.limit = Some(limit);
		self
	}

	/// Set the offset.
	pub fn with_offset(mut self, offset: u32) -> Self {
		self.offset = Some(offset);
		self
	}

	/// Set the aggregation options.
	pub fn with_aggregate(mut self, aggregate: AggregateOptions) -> Self {
		self.aggregate = Some(aggregate);
		self
	}
}

/// Options for subscribing to real-time changes.
#[derive(Debug, Clone)]
pub struct SubscriptionOptions {
	/// Path to subscribe to (e.g., "posts", "posts/abc123/comments")
	pub path: Box<str>,

	/// Optional filter (only matching changes are sent)
	pub filter: Option<QueryFilter>,
}

impl SubscriptionOptions {
	/// Create a subscription to all changes at a path.
	pub fn all(path: impl Into<Box<str>>) -> Self {
		Self { path: path.into(), filter: None }
	}

	/// Create a subscription with a filter.
	pub fn filtered(path: impl Into<Box<str>>, filter: QueryFilter) -> Self {
		Self { path: path.into(), filter: Some(filter) }
	}
}

/// Real-time change event emitted when a document is created, updated, or deleted.
#[derive(Debug, Clone, Serialize, Deserialize)]
#[serde(tag = "action", rename_all = "camelCase")]
pub enum ChangeEvent {
	/// A new document was created
	Create {
		/// Full path to the document (e.g., "posts/abc123" or "posts/abc123/comments/xyz789")
		path: Box<str>,
		/// Full document data
		data: Value,
	},

	/// An existing document was updated
	Update {
		/// Full path to the document
		path: Box<str>,
		/// Full updated document data
		data: Value,
		/// Previous document data (for incremental aggregate computation)
		#[serde(default, skip_serializing_if = "Option::is_none")]
		old_data: Option<Value>,
	},

	/// A document was deleted
	Delete {
		/// Full path to the document
		path: Box<str>,
		/// Document data before deletion (for incremental aggregate computation)
		#[serde(default, skip_serializing_if = "Option::is_none")]
		old_data: Option<Value>,
	},

	/// A lock was acquired on a document path
	Lock {
		/// Full path to the locked document
		path: Box<str>,
		/// Lock metadata (userId, mode)
		data: Value,
	},

	/// A lock was released on a document path
	Unlock {
		/// Full path to the unlocked document
		path: Box<str>,
		/// Unlock metadata (userId)
		data: Value,
	},

	/// Signals that all initial documents have been yielded for a subscription
	Ready {
		/// Subscription path
		path: Box<str>,
		/// Optional initial dataset
		#[serde(default, skip_serializing_if = "Option::is_none")]
		data: Option<Value>,
	},
}

impl ChangeEvent {
	/// Get the full path from this event.
	pub fn path(&self) -> &str {
		match self {
			ChangeEvent::Create { path, .. }
			| ChangeEvent::Update { path, .. }
			| ChangeEvent::Delete { path, .. }
			| ChangeEvent::Lock { path, .. }
			| ChangeEvent::Unlock { path, .. }
			| ChangeEvent::Ready { path, .. } => path,
		}
	}

	/// Get the document ID (last segment of the path).
	pub fn id(&self) -> Option<&str> {
		self.path().split('/').next_back()
	}

	/// Get the parent path (all segments except the last).
	pub fn parent_path(&self) -> Option<&str> {
		let path = self.path();
		path.rfind('/').map(|pos| &path[..pos])
	}

	/// Get the document data if this is a Create or Update event.
	pub fn data(&self) -> Option<&Value> {
		match self {
			ChangeEvent::Create { data, .. }
			| ChangeEvent::Update { data, .. }
			| ChangeEvent::Lock { data, .. }
			| ChangeEvent::Unlock { data, .. } => Some(data),
			ChangeEvent::Delete { .. } => None,
			ChangeEvent::Ready { data, .. } => data.as_ref(),
		}
	}

	/// Check if this is a Create event.
	pub fn is_create(&self) -> bool {
		matches!(self, ChangeEvent::Create { .. })
	}

	/// Check if this is an Update event.
	pub fn is_update(&self) -> bool {
		matches!(self, ChangeEvent::Update { .. })
	}

	/// Check if this is a Delete event.
	pub fn is_delete(&self) -> bool {
		matches!(self, ChangeEvent::Delete { .. })
	}
}

/// Compare two JSON values for ordering (used by filter range operators).
fn compare_json_values(a: Option<&Value>, b: Option<&Value>) -> std::cmp::Ordering {
	match (a, b) {
		(None, None) => std::cmp::Ordering::Equal,
		(None, Some(_)) => std::cmp::Ordering::Less,
		(Some(_), None) => std::cmp::Ordering::Greater,
		(Some(Value::Number(a)), Some(Value::Number(b))) => {
			a.as_f64().partial_cmp(&b.as_f64()).unwrap_or(std::cmp::Ordering::Equal)
		}
		(Some(Value::String(a)), Some(Value::String(b))) => a.cmp(b),
		(Some(Value::Bool(a)), Some(Value::Bool(b))) => a.cmp(b),
		(Some(a), Some(b)) => a.to_string().cmp(&b.to_string()),
	}
}

/// Convert a JSON value to a string key for aggregate group indexing.
pub fn value_to_group_string(value: &Value) -> String {
	match value {
		Value::String(s) => s.clone(),
		Value::Number(n) => n.to_string(),
		Value::Bool(b) => b.to_string(),
		Value::Null => "null".to_string(),
		_ => serde_json::to_string(value).unwrap_or_default(),
	}
}

/// Database statistics.
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct DbStats {
	/// Total size of database files in bytes
	pub size_bytes: u64,

	/// Total number of documents across all tables
	pub record_count: u64,

	/// Number of tables in the database
	pub table_count: u32,
}

/// Transaction for atomic write operations.
///
/// All write operations must be performed within a transaction to ensure atomicity.
#[async_trait]
pub trait Transaction: Send + Sync {
	/// Create a new document with auto-generated ID. Returns the generated ID.
	async fn create(&mut self, path: &str, data: Value) -> ClResult<Box<str>>;

	/// Update an existing document (stores the provided data as-is).
	///
	/// Note: This method performs a full document replacement at the storage level.
	/// Merge/PATCH semantics should be handled by the caller before invoking this method.
	async fn update(&mut self, path: &str, data: Value) -> ClResult<()>;

	/// Delete a document at a path.
	async fn delete(&mut self, path: &str) -> ClResult<()>;

	/// Read a document from the transaction's view.
	///
	/// This method provides transaction-local reads with "read-your-own-writes" semantics:
	/// - Returns uncommitted changes made by this transaction
	/// - Provides snapshot isolation from other concurrent transactions
	/// - Essential for atomic operations like increment, append, etc.
	///
	/// # Returns
	/// - `Ok(Some(value))` if document exists (either committed or written by this transaction)
	/// - `Ok(None)` if document doesn't exist or was deleted by this transaction
	/// - `Err` if read operation fails
	async fn get(&self, path: &str) -> ClResult<Option<Value>>;

	/// Commit the transaction, applying all changes atomically.
	async fn commit(&mut self) -> ClResult<()>;

	/// Rollback the transaction, discarding all changes.
	async fn rollback(&mut self) -> ClResult<()>;
}

/// Real-Time Database Adapter trait.
///
/// Unified interface for database backends. Provides transaction-based writes,
/// queries, and real-time subscriptions.
#[async_trait]
pub trait RtdbAdapter: Debug + Send + Sync {
	/// Begin a new transaction for write operations.
	async fn transaction(&self, tn_id: TnId, db_id: &str) -> ClResult<Box<dyn Transaction>>;

	/// Close a database instance, flushing pending changes to disk.
	async fn close_db(&self, tn_id: TnId, db_id: &str) -> ClResult<()>;

	/// Query documents at a path with optional filtering, sorting, and pagination.
	async fn query(
		&self,
		tn_id: TnId,
		db_id: &str,
		path: &str,
		opts: QueryOptions,
	) -> ClResult<Vec<Value>>;

	/// Get a document at a specific path. Returns None if not found.
	async fn get(&self, tn_id: TnId, db_id: &str, path: &str) -> ClResult<Option<Value>>;

	/// Subscribe to real-time changes at a path. Returns a stream of ChangeEvents.
	async fn subscribe(
		&self,
		tn_id: TnId,
		db_id: &str,
		opts: SubscriptionOptions,
	) -> ClResult<Pin<Box<dyn Stream<Item = ChangeEvent> + Send>>>;

	/// Create an index on a field to improve query performance.
	async fn create_index(&self, tn_id: TnId, db_id: &str, path: &str, field: &str)
	-> ClResult<()>;

	/// Get database statistics (size, record count, table count).
	async fn stats(&self, tn_id: TnId, db_id: &str) -> ClResult<DbStats>;

	/// Export all documents from a database.
	///
	/// Returns all `(path, document)` pairs. The path is relative to the db_id
	/// (e.g., `posts/abc123`). Used for duplicating RTDB files.
	async fn export_all(&self, tn_id: TnId, db_id: &str) -> ClResult<Vec<(Box<str>, Value)>>;

	/// Acquire a lock on a document path.
	///
	/// Returns `Ok(None)` if the lock was acquired successfully.
	/// Returns `Ok(Some(LockInfo))` if the path is already locked by another user (denied).
	async fn acquire_lock(
		&self,
		tn_id: TnId,
		db_id: &str,
		path: &str,
		user_id: &str,
		mode: LockMode,
		conn_id: &str,
	) -> ClResult<Option<LockInfo>>;

	/// Release a lock on a document path.
	async fn release_lock(
		&self,
		tn_id: TnId,
		db_id: &str,
		path: &str,
		user_id: &str,
		conn_id: &str,
	) -> ClResult<()>;

	/// Check if a path has an active lock. Returns the lock info if locked.
	async fn check_lock(&self, tn_id: TnId, db_id: &str, path: &str) -> ClResult<Option<LockInfo>>;

	/// Release all locks held by a specific user (called on disconnect).
	async fn release_all_locks(
		&self,
		tn_id: TnId,
		db_id: &str,
		user_id: &str,
		conn_id: &str,
	) -> ClResult<()>;

	/// Delete every RTDB database owned by the tenant.
	///
	/// Used by tenant purge orchestration. Implementations should treat a
	/// missing tenant store as success.
	async fn delete_tenant_databases(&self, tn_id: TnId) -> ClResult<()>;
}

// vim: ts=4