Struct Subscription 

pub struct Subscription {

    pub id: SubscriptionId,
    pub query: String,
    pub tables: HashSet<String>,
    pub last_result_hash: u64,
    pub last_result: Option<Vec<Row>>,
    pub notify_tx: Sender<SubscriptionUpdate>,
    pub retry_policy: SubscriptionRetryPolicy,
    pub retry_count: u32,
    pub updates_sent: u64,
    pub updates_dropped: u64,
    pub channel_buffer_size: usize,
    pub slow_consumer_threshold_percent: u8,
    pub connection_id: Option<String>,
    pub wire_subscription_id: Option<[u8; 16]>,
    pub filter: Option<String>,
    pub pk_columns: Vec<usize>,
    pub selective_eligible: bool,
    pub selective_updates_override: Option<SelectiveColumnConfig>,
}

A single query subscription

Tracks the query, its table dependencies, and the channel for sending updates.

Fields

id: SubscriptionId

Unique identifier for this subscription

query: String

The SQL query being monitored

tables: HashSet<String>

Tables this query depends on (extracted from AST)

last_result_hash: u64

Hash of the last result set (for change detection)

last_result: Option<Vec<Row>>

Last result set (for delta computation) This stores the previous result to enable computing deltas on change.

notify_tx: Sender<SubscriptionUpdate>

Channel to send updates to the subscriber

retry_policy: SubscriptionRetryPolicy

Retry policy for handling transient errors

retry_count: u32

Current retry attempt count (resets on successful execution)

updates_sent: u64

Total updates sent to this subscription

updates_dropped: u64

Total updates dropped due to channel being full

channel_buffer_size: usize

Buffer size for the subscription channel

slow_consumer_threshold_percent: u8

Slow consumer threshold percentage

connection_id: Option<String>

Optional connection/session ID that owns this subscription Used for connection-level subscription tracking and cleanup

wire_subscription_id: Option<[u8; 16]>

Optional wire protocol subscription ID (UUID bytes) Used to bridge between wire protocol IDs and internal SubscriptionId

filter: Option<String>

Optional filter expression (SQL WHERE clause) to apply to updates Only rows matching the filter will be included in subscription updates

pk_columns: Vec<usize>

Primary key column indices in the result set Used for selective column updates to always include PK columns Default: [0] (assumes first column is PK if not detected)

selective_eligible: bool

Whether this subscription is eligible for selective column updates True when PK columns were confidently detected

selective_updates_override: Option<SelectiveColumnConfig>

Per-subscription override for selective column update configuration If set, this overrides the server-level selective_updates config for this subscription

Implementations

impl Subscription

pub fn new( query: String, tables: HashSet<String>, notify_tx: Sender<SubscriptionUpdate>, ) -> Self

Create a new subscription

pub fn with_policy( query: String, tables: HashSet<String>, notify_tx: Sender<SubscriptionUpdate>, retry_policy: SubscriptionRetryPolicy, ) -> Self

Create a new subscription with a custom retry policy

pub fn with_config( query: String, tables: HashSet<String>, notify_tx: Sender<SubscriptionUpdate>, config: &SubscriptionConfig, ) -> Self

Create a new subscription with custom configuration

pub fn for_connection( query: String, tables: HashSet<String>, notify_tx: Sender<SubscriptionUpdate>, connection_id: String, wire_subscription_id: [u8; 16], filter: Option<String>, config: &SubscriptionConfig, ) -> Self

Create a new subscription for a specific connection (wire protocol)

This associates the subscription with a connection ID for tracking and cleanup when the connection closes.

pub fn for_connection_with_pk( query: String, tables: HashSet<String>, notify_tx: Sender<SubscriptionUpdate>, connection_id: String, wire_subscription_id: [u8; 16], filter: Option<String>, config: &SubscriptionConfig, pk_columns: Vec<usize>, ) -> Self

Create a new subscription for a specific connection with custom PK columns

This associates the subscription with a connection ID for tracking and cleanup when the connection closes. It also allows specifying which columns are primary keys for selective column updates.

pub fn set_pk_columns(&mut self, pk_columns: Vec<usize>)

Set the primary key columns for this subscription

Used after detection to update the subscription with actual PK columns.

pub fn set_pk_columns_with_eligibility( &mut self, pk_columns: Vec<usize>, confident: bool, ) -> bool

Set both PK columns and selective eligibility

Used after PK detection to update the subscription. Returns true if the subscription is newly marked as selective-eligible.

pub fn set_selective_updates_override(&mut self, config: SelectiveColumnConfig)

Set per-subscription selective updates override

Allows clients to override server-level selective update thresholds on a per-subscription basis.

pub fn clear_selective_updates_override(&mut self)

Clear the selective updates override (use server defaults)

pub fn get_effective_selective_config( &self, server_config: &SelectiveColumnConfig, ) -> SelectiveColumnConfig

Get the effective selective column update config for this subscription

Returns the per-subscription override if set, otherwise creates a config from the server-level config with this subscription’s PK columns.

Trait Implementations

impl Debug for Subscription

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.